Evaluating the Bender Visual Motor Gestalt Test II as a Diagnostic

Graduate Theses, Dissertations, and Problem Reports

2010

Evaluating the Bender Visual Motor Gestalt Test II as a Diagnostic Evaluating the Bender Visual Motor Gestalt Test II as a Diagnostic

Screening Instrument Among Clinically Referred Children and Screening Instrument Among Clinically Referred Children and

Adolescents Adolescents

Linda R. Marnic

West Virginia University

Follow this and additional works at: https://researchrepository.wvu.edu/etd

Recommended Citation Recommended Citation

Marnic, Linda R., "Evaluating the Bender Visual Motor Gestalt Test II as a Diagnostic Screening Instrument

Among Clinically Referred Children and Adolescents" (2010).

Graduate Theses, Dissertations, and

Problem Reports

. 3162.

https://researchrepository.wvu.edu/etd/3162

This Dissertation is protected by copyright and/or related rights. It has been brought to you by the The Research

Repository @ WVU with permission from the rights-holder(s). You are free to use this Dissertation in any way that is

permitted by the copyright and related rights legislation that applies to your use. For other uses you must obtain

permission from the rights-holder(s) directly, unless additional rights are indicated by a Creative Commons license

in the record and/ or on the work itself. This Dissertation has been accepted for inclusion in WVU Graduate Theses,

Dissertations, and Problem Reports collection by an authorized administrator of The Research Repository @ WVU.

For more information, please contact researchreposit[email protected].

Evaluating the Bender Visual Motor Gestalt Test II as a Diagnostic Screening

Instrument Among Clinically Referred Children and Adolescents

Linda R. Marnic

Dissertation submitted to the

College of Human Resources and Education

at West Virginia University

in partial fulfillment of the requirements

for the degree of

DOCTOR OF PHILOSOPHY

Counseling Psychology

James Bartee, Ph.D., Chair

Jeffrey Daniels, Ph.D

Margaret Glenn, Ed.D.

Richard Walls, Ph.D.

Eric Youngstrom, Ph.D.

Department of Counseling, Rehabilitation Counseling,

and Counseling Psychology

Morgantown, West Virginia

2010

Key Words: Bender Gestalt II, Koppitz 2, KSADS-PL+, CBCL, Children

and Adolescents, Clinical Utility.

ABSTRACT

EVALUATING THE BENDER VISUAL MOTOR GESTALT TEST II AS A

DIAGNOSTIC SCREENING INSTRUMENT AMONG CLINICALLY

REFERRED CHILDREN AND ADOLESCENTS

LINDA R. MARNIC

This research was designed to investigate the diagnostic utility of the Bender Gestalt II

(BGII) test using the Bender Global Scoring System (BGSS) and the Koppitz 2 scoring

systems. The scores from these two systems were correlated with scores derived from the

Kiddie Schedule for Affective Disorders and Schizophrenia for School Aged Children,

Present and Lifetime Edition (KSADS-PL+), a semistructured interview along with

Longitudinal Evaluation of all Available Data (LEAD) and the Child Behavior Checklist

(CBCL). Of the 115 children and adolescents who initially participated in the study to

assess the validity of the Bender Visual Motor Gestalt II test as a screening instrument in

psychological decision making, 75 completed all protocols and the relevant data were

entered into the subsequent analysis. A correlational design was employed, and a post

hoc test was used to incorporate Receiver Operating Characteristic (ROC) analysis into

the findings. Results from both the Bender Gestalt II Global Scoring System and Koppitz

2 scoring systems showed moderate correlations with results from the CBCL on the

symptom categories of aggressive behaviors, depressed behaviors, and attention deficit

hyperactivity disorder (ADHD) in children and adolescents. However, the Koppitz 2

Emotional Indicators scoring measure did not accurately discriminate for the presence or

absence of psychopathology. The Koppitz 2 Total Error score was found to be modestly

correlated with receiving a diagnosis of ADHD. None of the other diagnoses based on

results from KSADS-PL+ with LEAD showed any significant correlations with the

Koppitz 2 Total Error Score. Adding Receiver Operating Characteristic analysis for

sensitivity and specificity, improved the diagnostic likelihood ratio from 50% to 66% for

ADHD diagnosis using the Koppitz 2 Emotional Indicators. The main hypothesis that the

Bender Gestalt II would improve diagnostic accuracy of psychopathology was not

supported. The unexpected finding that the BGII is useful in diagnosing ADHD indicates

a possible direction for future research.

iii

Acknowledgments

I would like to thank all the people who have been instrumental in the completion

of this dissertation. I would like to thank Eric Youngstrom, who believes in promoting

the field and cultivating the next generation of psychologists. His positive demeanor and

endless energy is a benefit not only to the areas of research and academia but to each

student under his tutelage. I cannot thank enough, Margalit Persing, Andrew Freeman,

and Dr. James Bartee for their editorial work and numerous reviews that made this paper

readable. Also, I am grateful to the whole Applewood family who made my years in

Cleveland truly wonderful. My inspiration has been the various married working mothers

in the field of psychology. They have shown me that it’s not impossible to achieve your

dreams, impact the lives of others, and still have enough left to be active, loving mothers

to wonderful children and partners to the men in their lives. I have found from their

examples, the more you give the more you have to give. Thank you.

I would like to acknowledge my family and friends. Stephen, my partner in this

life, your love and patience is my strength. You have given the most valuable gift in my

life: our children. My life is chaotic and difficult with work, kids, activities, and finding

time to write: however, you have stood by me, honored me, and that means so much.

Thank you.

I would like to thank all the clients I have seen over the years. Without them I

would have never started this endeavor and maintained the belief that I make a

difference. I know that I have learned from them as much as they from me. I realize that

this paper is truly a gift from God, and I could not achieve anything without His help.

When I thought all was lost, and I couldn’t take another disappointment, I found strength

through prayer. I also thank the many who encouraged me to finish and were there for

me. Thank you. I am a better person for all the relationships I have had or will have in

my life. They have shown me the person I can be and what I am capable of doing. I thank

you all. I am truly blessed. If it were not for them then I would not have become me. I

am grateful and indebted to all of you.

Table of Contents

List of Tables....................................................................................................................viii

List of Figures.....................................................................................................................ix

Chapter 1: Introduction and Overview...............................................................................1

Statement of the Problem............................................................................................................9

Research Questions...................................................................................................................12

Research Significance................................................................................................................12

Definition of Terms...................................................................................................................14

Chapter 2: Review of Selected Literature.........................................................................21

Research Focus.................................................................................................................28

The Bender Visual Motor Gestalt Test.............................................................................30

Current Use of the Bender Visual Motor Gestalt–Second Edition............................................33

The Koppitz 2 and the Koppitz Developmental Scoring System..............................................35

Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children...........37

The Child Behavior Checklist...................................................................................................41

Chapter 3: Method............................................................................................................45

Participants................................................................................................................................45



Measures....................................................................................................................................46

The Bender Visual Motor Gestalt II- Global Scoring System.............................................................46

The Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children (K-SADS).

.............................................................................................................................................................51



The Child Behavior Checklists, (CBCL) and Youth Self Report Form (YSR)...................................53

Examiners..................................................................................................................................54

Procedure...................................................................................................................................54

Data Analyses............................................................................................................................55

Chapter 4: Results.............................................................................................................59

Research Question 1..................................................................................................................59

Research Question 2..................................................................................................................68

Research Question 3..................................................................................................................71

Research Question 4..................................................................................................................75

Chapter 5: Discussion.......................................................................................................76

Discussion of Hypotheses..........................................................................................................76

Limitations of the Study............................................................................................................82

Implications for Practice............................................................................................................84

Conclusions...............................................................................................................................87

References.........................................................................................................................90

Appendix A: IRB Protocol..............................................................................................103

Appendix B: CREC Program Notice of Certification.....................................................108

vii

Appendix C: Administration Manual for ABACAB......................................................109

Appendix D: Demographic Form...................................................................................111

Appendix E: Consent Forms...........................................................................................115

Appendix F: Statistical Findings.....................................................................................120

Author’s Note..................................................................................................................121

viii

List of Tables

Table 1 Pearson Product-Moment Correlation Matrix Between the BGSS Copy and

Recall Scores and the CBCL Main Scores .............................................................. 62

Table 2 Pearson Correlational Matrix of the BGSS Scores and the CBCL Syndrome

Scores ....................................................................................................................... 64



Table 3 Pearson Correlational Matrix Comparing BGSS and CBCL Diagnostic

Scores ....................................................................................................................... 65

Table 4 Pearson Correlational Matrix Comparing CBCL and Koppitz 2 Scores ............ 66

Table 5 Pearson Correlation Matrix Comparing Koppitz 2 Total Score, Total Emotional

Indicators, and Visual Motor Index to the CBCL Syndrome Scales ....................... 67

Table 6 Pearson Correlation Matrix Comparing the BGII, BGSS and Koppitz2 Scores

and KSADS-PL+ Symptomology. ........................................................................... 70

Table 7 Pearson Product-Moment Correlations of the Koppitz 2 Scores and KSADS-PL+

Diagnoses ................................................................................................................. 70

List of Figures

Figure 1. ROC of errors made using the Koppitz 2 scoring system to predict................. 74

UTILITY OF THE BGII 1

Chapter 1: Introduction and Overview

In an era of increased emphasis on accountability and outcomes, psychologists are

continually pushed to become more accurate in diagnosing and treating clients while

balancing the cost of providing services. The task of developing more efficacious

diagnostic and treatment protocols is indeed a challenging one in this time of managed

care and limited client contact in part due to insurance considerations and heavy

caseloads. Finding valid diagnostic measures that are inexpensive, easy to administer,

reliable, and that serve multiple purposes would be beneficial to all. This research was

used to investigate the use of the Bender Visual Motor Gestalt II (BGII) as a possible

diagnostic tool to add to clinical protocols that assist in initial diagnostic decision making

for children and adolescents.

The field of psychology is faced with many challenges. Mental-health statistics

for children in the United States are very disturbing. More than 400,000 children are in

therapy for treatment of a diagnosed mental illness (Kamphaus, Petoskey, & Rowe,

2000). Additionally, over 5 million children have received psychoeducational evaluations

in public schools to assess for learning, behavior, and information processing problems

(Kamphaus et al., 2000). Astoundingly, this number does not include those children who

were evaluated outside of the school system for various mental-health or neurological

problems. It has been reported by Tolan and Dodge (2005) that over two thirds of the

children and adolescents who receive mental-health services have been previously

treated, and over three quarters of adult clients in treatment report that their problems

UTILITY OF THE BGII 2

began in their childhood years. These staggering numbers further highlight the need for

increased accuracy and more efficient diagnostic protocols in practice. The importance of

early identification of mental-health disorders and the need for effective forms of

treatment have been the focus of the American Psychological Association for several

years (Tolan & Dodge, 2005). However, it is important to note that effective treatment

relies on the ability of the treating clinician to provide an accurate diagnostic profile of

the individual.

Diagnosis is a cornerstone of the practice of counseling psychology. Nowhere is

this more evident than in the work with children and adolescents. Not only is the

formulation of a diagnosis the first step in treatment planning, but it also lays the

foundation upon which all future therapeutic work is built. The Diagnostic and Statistical

Manual, 4th edition Text Revision (DSM-IV-TR) is the current manual on which all

clinical diagnostic decisions are based and is considered to be the most comprehensive

psychological diagnostic manual to date (American Psychiatric Association, 2000). Use

of the DSM-IV-TR has been credited with an increase in appropriate diagnoses and early

intervention in significant childhood mental illnesses. This in turn has led to improved

childhood outcomes in autism spectrum disorders, attention deficit hyperactivity disorder,

childhood depression, and bipolar disorder (American Psychiatric Association, 2000;

Charman & Baird, 2002; Lipovsky, Finch, & Belter, 1989; Valderhaug & Ivansson,

2005). However, one of the criticisms of reliance on the DSM-IV-TR is that the manual

allows too much overlap among different diagnoses (DeClercq, DeFruyt, Van Leewen, &

UTILITY OF THE BGII 3

Mervielde, 2006). Questions have also been raised about the meaning of disorders

categorized as adult disorders with consideration of a childhood onset, leading to further

overlap and difficulty among professionals with regard to treatment venues (Sourander et

al., 2005).

Psychologists have long sought quick, inexpensive, and empirically sound

measures that would assess pathology while giving maximum information with minimum

time spent scoring. The original Bender Visual Motor Gestalt (OBG) Test has always

been one such quick and direct measure of a child’s ability to perform visual motor

integration tasks. The OBG has a history of meeting these needs by being fast, easy to

administer, and capable of assessing multiple areas of client ability. It has been used by

clinicians as a standard measure in psychological batteries for more than 60 years. Before

any further detailed discussion of the revised Bender and its use in this research, a brief

discussion of the history of this significant test is in order.

The studies of children using direct measures, rather than the use of

questionnaires, have been debated for many years. German psychologists Brentano

(1838-1917) and Stumpf (1848-1936) supported the study of children and advocated

experimental studies based on internal cognitive processes rather than the questionnaire

approach pioneered by Hall (1844-1924) and early functionalists in the United States

(Hothersall, 1995).

The use of drawing as a medium for the measure of visual- motor integration

ability has been. The original Bender-Gestalt Test (OBG) titled The Bender Visual Motor

UTILITY OF THE BGII 4

Gestalt Test (Bender, 1938), was composed of drawings to be visually perceived and

reproduced by patients and was developed by Wertheimer (1880-1943) in his perceptual

psychology experiments and later formalized by Bender (1897-1987). Wertheimer was

one of the original Gestalt theorists whose theoretical works were cut short by World War

II and his flight from Nazi Germany. Although he came to the United States from

Germany, he did little to advance Gestalt theory in the United States in the years after the

war (Hothersall, 1995). Bender (1938), a psychologist working in New York, further

experimented with the figures by using the drawings with both her child and adult

patients. The resulting cards which Bender developed were slightly different than the

visual motor patterns of the originals by Wertheimer. The resultant client drawings were

considered an integration of the discrete internal processes of the drawer (Bender, 1938).

Based on Gestalt theory, the resultant drawing represents more than visual and motor

associations within the physical body. The drawings reflect association and cognitive

complexity within the individual. Individual variables that may affect the accuracy of a

drawing include age, physical and emotional development, as well as individual mental

and emotional states (Bender, 1938).

According to Bender (1938), developmental maturation of drawing is an ongoing

process that follows sequential stages, incorporating motor development from gross to

fine motor, visual imagery, and perceptual awareness. Thus, a beginning scribble

becomes more circular. Circles become loops and a tendency from vertical to horizontal

movements occurs, and then finally, dimensional awareness emerges in the drawing.

UTILITY OF THE BGII 5

These factors are all components of the maturational process that leads to the more

intricate representations found within mature representational drawings which then

represent a completed integration of internal processes. A deviation within this

developmental or maturational process would obviously lead to a disintegration of the

original representation. Psychology has found this process helpful in identifying those

individuals who have not yet matured, were delayed in visual motor perception, or who

were once matured, yet for various reasons may be losing such integration faculties.

The original Bender test (OBG) consists of nine figures on separate 3 x 5 cards.

There have been many research articles that have utilized the Bender as a criterion in

developmental processing, perceptual motor skills, and neurological intactness

(Brannigan & Decker, 2003). Horn and O’Donnell (1984) researched the early

identification of learning disabilities and found that the OBG was effective in identifying

those children classified as learning disabled and those with low achievement. Based on

original theory, the cards were also used as a direct measure of the underlying emotional

state in those thought to be of normal development and as a personality assessment to

assess internal motives (Hutt, 1985). The OBG has also been used as a test for emotional

problems and a personality test. The OBG was found to be useful in comparing

impulsive adolescents to those without impulse difficulties (Oas, 1984). Oas (1984)

found that adolescents with impulsivity disorders were significantly different from those

designated as nonimpulsive on the Matching Familiar Figures Test, a behavior rating

scale.

UTILITY OF THE BGII 6

The OBG monograph (Bender, 1938) was also used for patients within psychiatric

hospitals to differentiate between the functionally mentally ill and malingerers. However,

later research by Pascal and Suttell (1952) found this function to be invalid.

Subsequently, Mehlman and Vatovec (1956), Bowland and Deabler (1956), and Stewart

(1957) found that the Bender was reliable in differentiating between psychiatric and

nonpsychiatric patients being admitted to the hospital. Regardless of differing findings,

professionals continue to find the Bender to be of value in their evaluations and even

prior to revision it remained a favored test in use (Piotrowski, 1995; Piotrowski & Keller,

1989). Bender’s scoring system evaluated the overall quality of each design on a scale

that ranged from 1 to 5 on one design and from 1 to 7 on the other eight designs. Her

scoring system is based on accuracy toward perfection of the design in a completion of a

gestalt (Bender, 1938).

In an attempt to validate a direct measure of the emotional state of the individual

or the projective use of the Bender, different psychologists developed specific scoring

procedures The psychologists included Pascal and Suttell in 1952, Hutt and Briskin in

1960, Koppitz for children in 1963, Keogh and Smith in 1961 and Canter in 1976. These

specific scoring procedures have been joined by more recent scoring systems such as the

Advanced Psychodiagnostic Interpretation Scale by Rosenberg and Raphael in 2000

(Canter, 1968; Keogh, 1965; Pascal & Sutter, 1952; Piotrowski, 1995). These diverse

scoring systems have been criticized for insufficient reliability. Hutt’s popular scale for

determining psychopathology, developed in 1977, was found in 1983 to have

UTILITY OF THE BGII 7

questionable reliability and validity (Rossini, 1983). Keogh and Smith’s scoring system

did not provide normative data (Brannigan & Decker, 2006). The result of the

questionable validity and the numerous scoring systems has led to a slight decrease in the

use of the OBG cards among professionals (Archer et al., 1991; Wilson & Reschly,

1996). However, this situation has further led those in favor of the OBG to explore its

uses and revise the test with the goal of standardization as an empirically valid measure

(Brannigan & Decker, 2003, 2006; Brannigan, Decker, & Madsen, 2004; Reynolds,

2007).

The revision of the OBG was in process for many years, first at the American

Orthopsychiatric Association then later at Riverside Publishing, which held the

copyrights, with the work of many advisors and more than 25 years of collaboration

(Brannigan & Decker, 2006). Tolor and Brannigan in 1980 stressed the need for research

on the OBG, not only to assess personality dynamics and psychopathology, but also to

diagnose organic pathology, and predict school learning problems (Brannigan & Decker,

2006; Brannigan, Decker, & Madsen, 2004). Four main components were of importance

during the revision process: (a) keeping the original nine designs, but increasing the

number of designs; (b) inclusion of a memory procedure; (c) comparison of both the

deviation and quality based scoring systems; and (d) obtaining results from a nationally

representative sample for validity (Brannigan & Decker, 2003, 2006). The Bender Gestalt

II (BGII) was released for publication in 2003 by the Riverside Publishing Company.

UTILITY OF THE BGII 8

Along with the release of the BGII, the Global Scoring System (BGSS) was released as a

recommended standardized measure by Riverside Publishing Company.

Revisions were also begun on the original Koppitz scoring system. Pro-Ed

Publishing obtained the rights to the Koppitz scoring system and retained Cecil Reynolds

to revise its scoring version. It was released in 2007 as the Koppitz 2: the Koppitz

Developmental Scoring System for the Bender- Gestalt Test II (Reynolds, 2007). The

Bender Gestalt II Global Scoring System (BGSS) and the newly revised Koppitz 2

developmental scoring system are now used as the preferred empirically validated

measures for visual motor integration. However, they have not been researched beyond

this use and thus may not be a valid measure for personality assessment. It is worth

noting that the Koppitz 2 (Reynolds, 2007) has included the Emotional Indicators as an

additional measure to assess personality. But this component was not included in the

original norming process.

Only an unpublished dissertation from 2004 used the Bender Visual Motor

Gestalt Test II (BGII) with the scoring of the Koppitz 2 Emotional Indicators. Fidal in

2004 examined the BGII, using the Koppitz 2 Developmental Scoring System and the

Emotional Indicators Scoring systems on adolescents who had incidents of abuse versus

those who reported no such history. Fidal found that the independent t tests could not

significantly differentiate between the groups (Fidal, 2004). More research is needed to

compare the two methods to determine whether to support or retire the use of the

Emotional Indicators in order to end the controversy about the original Koppitz

UTILITY OF THE BGII 9

Emotional Indicators. A literature search for the use of the BGII as an assessment in

evidenced-based practice led to no findings. Thus this current research may be the first to

address the practical use of the BGII with the BGSS and the Koppitz 2 Scoring System in

clinical decision making using likelihood ratios (Koppitz, 1968, 1971, 1975).

Statement of the Problem

This study attempts to address one of the main deficits that can occur in the

diagnosis and treatment of children and adolescents with mental-health issues: that is the

difficulty in finding a direct measure of the child’s functioning and internal state without

either asking overly obvious questions or relying on information from significant others.

Thus, this study will combine the task of the practitioner with that of the clinical

researcher to determine the practicality for the more frequent use of the BGII. The OBG

has been widely used as a projective screening device, as previously defined. However,

most of the research was limited to adult pathology. Little research was conducted in

regard to childhood pathology and less was directed toward the adolescent population

(Belter, McIntosh, Finch, Williams, & Edwards, 1989; Rossini & Kasper, 1987). In

children, the OBG was used mainly as a test of visual motor development and the

research at the time supported this conclusion (Decker, Allen, & Choca, 2006; Koppitz,

1971). However, the original Bender test, now the BGII has been revised to include more

drawing items that increases the baseline and ceiling of the test, added the recall

procedure and developed an empirically normed scoring system called the Global

Scoring System (BGSS) for standard use among clinicians (Brannigan & Decker, 2003,

UTILITY OF THE BGII 10

2006). It would be of benefit to clinicians if the revisions and new scoring systems

contributed to the psychologist’s ability to accurately diagnose. Furthermore, it would be

beneficial to compare the newly revised BGII forms with another highly valid behavioral

report measure that identifies childhood and adolescent diagnostic areas of concern in an

attempt to revive the original purpose of the Bender as a psychodiagnostic technique. It

also might be useful to compare the BGII findings with the newer clinical assessments

available to psychologists such as the semistructured clinical interview.

In this study the results of the BGSS and the Koppitz 2 scoring systems for the

newly revised Bender Visual Motor Gestalt II (BGII) test were compared with two

clinical screening measures of mental disorders: the Child Behavior Checklist, known as

the CBCL (Achenbach, 2004), and the Kiddie Schedule for Affective Disorders and

Schizophrenia-Present and Lifetime Edition, known by the acronym KSADS-PL+

(Findling et al., 2001; Youngstrom & Duaz, 2005; Youngstrom, Findling, Danielson, &

Calabrese, 2001). The CBCL and KSADS-PL+ have been previously compared to each

other in research (Wassenberg, Max, Koele, & Firme, 2004; Youngstrom et al., 2001).

The current research is exploratory in nature as it will attempt to find relationships

between the revised Bender Gestalt II with the CBCL and KSADS-PL+ used in the

diagnosis of childhood psychopathology. In order to assess the clinical utility of these

results, the BGII using the BGSS and Koppitz 2 data were compared to results derived

from the previously mentioned valid measures of pediatric diagnostic decision making,

the KSADS-PL+ and the CBCL. In the current study, two updated scoring procedures for

UTILITY OF THE BGII 11

the BGII test were compared, the BGSS and the Koppitz 2, to see if either provided

clinicians with a valid, cost-effective, projective-screening measure that would aid in

diagnosis of mental-health issues in clinical practice with children.

UTILITY OF THE BGII 12

Research Questions

RQ1. Is the BG II, using the BGSS and Koppitz 2 scoring measures, an effective

psychometric screening tool to use with a clinically referred population of children aged

5 to 18 when compared to a commonly used screening instrument such as the Child

Behavior Checklist (Achenbach System of Empirical Based Assessment [ASEBA],

2006)?

RQ2. Are the scores derived from the BGSS and the Koppitz 2 scoring systems

effective measures for the diagnosis of the presence or absence of pathology in subject

children and adolescents when compared to the Washington University version of the

Kiddie Schedule for Affective Disorders and Schizophrenia (WASH-U-KSADS-PL+), a

research diagnostic instrument?

RQ3. Are there significant relationships among the results derived from the BGII,

using the BGSS and Koppitz 2 scoring systems, with final pediatric DSM-IV-TR

diagnoses of LEAD consensus results following the KSADS-PL+?

RQ4. Are there any significant relationships between the BGSS and Koppitz 2

scores (including the Koppitz 2 Emotional Indicators) derived from the BGII when it is

used with clinically referred children from 5 to 18 years of age?

Research Significance

The significance of this study lies in the joining of the BGII, BGSS scoring

system with the KSADS PL+, a semistructured diagnostic interview, and was termed

“groundbreaking” research according to Gary G. Brannigan (personal communication,

UTILITY OF THE BGII 13

11/30/2006). Brannigan is the co-author of the Bender Visual Motor Gestalt Test –

Second Edition (BGII). Comparisons of the BGII with several behavioral rating scales

(Belter et al., 1987; McCormick &Brannigan, 1984) and pathology groups (Field, Bolton,

& Dana, 1982; Rossini & Kaspar, 1987; Shapiro, & Simpson, 1995) yielded varying

findings. Shapiro and Simpson (1995) found that primary psychiatric diagnosis was not

related to Bender performance when using the earlier Koppitz scoring system. However,

these studies were conducted with the OBG and are now considered outdated. This study

will examine individual patient drawings and emotional factors to determine the

diagnostic value of assessing childhood pathology in relation to the BGII cards and the

two scoring methods as previously mentioned. From this study, it is hoped that a simple

test will eventually emerge providing a bridge between an extensive and laborious

clinical evaluation and a parental checklist of subjective problems, thus providing another

source of valid psychometric data to be used in psychological batteries. Such a test may

prove to be an effective, additional diagnostic instrument in standard intake protocols.

This study was open to all children 5 to 18 years of age who sought mental-health

services at a large community mental-health center in Cleveland, Ohio. It was part of a

larger study which included profiles of 825 children. The July 2006 population of

Cleveland was approximately 444,313, with an estimated median household income of

$24,105 (2006). The percentage of residents living in poverty was 32.4% in 2005 (20.4 %

for White non-Hispanic residents, 39.3% for African American residents, and 38.5% for

Hispanic residents). The racial makeup of Cleveland in 2006 was 51% African American,

UTILITY OF THE BGII 14

38.8 % White non- Hispanic, 7.3% Hispanic, 0.9% American Indian, and 3.6% other

races, with 2.2% of the population listed as two or more races. The Cleveland Municipal

School District is the largest school district in Ohio (City Data.com, 2009).

Definition of Terms

Base rate. This is a quantitative representation of a particular event occurring in a

population or setting. Base rates in mental-health settings are usually based upon

demographic information of the clients, and can be influenced by referral sources,

geographical locations, previous diagnoses given by different clinicians, and the market

area of the clinician. Base rate is the starting point in diagnostic decision making when

using Evidence Based Practice assessment (Youngstrom & Duax, 2005).

The Bender Gestalt II Global Scoring System. The BGSS was specifically

designed to assess visual motor integration across a lifespan and aid in discriminating

various types of learning, psychological, and neurological problems using the revised

BGII (Brannigan & Decker, 2003). It has been chosen because it represents the first

standardization of the revised drawings of the OBG and is preferred by the BGII

publishers, Riverside Publishing Company. The BGSS allows two phases of

administering the BGII, the copy phase and the recall (memory) phase with a total score

developed for each phase (Brannigan & Decker, 2003).

Bender Visual Motor Gestalt Test II. The BGII is the test based on the original

nine Bender drawings (OBG) plus seven additional drawings developed by Brannigan

and Decker for the revised edition. These researchers added four cards to the beginning

UTILITY OF THE BGII 15

of the protocol to be used for children 4 to 7 years of age resulting in a total of 13 cards.

Three additional cards were added to the end of the original protocol for the subjects over

eight years of age resulting in a total of 12 cards.

The Child Behavior Checklist. The CBCL, developed by Achenbach and

Rescorla (2001), is a rating scale commonly used to assist in making a pediatric

psychological diagnosis. It assesses a broad range of behavioral symptoms found in

children with emotional difficulties (Achenbach, 1991). The CBCL was revised in 2001

when two different forms were created: the CBCL for ages 1 to 5 years (CBCL/1 -5) and

the CBCL for ages 6 to 18 years (CBCL/6-18; ASEBA, 2006). The CBCL uses three

different forms for reporting: parent, teacher, and self-report. Rater diagnosis is based on

how often each item has occurred currently or within the past 6 months for specific

questions using a forced 3-point scale response form. There are two open-ended items for

the individual to report additional problems. However, the open-ended items are

qualitative and are not included in standard scoring (ASEBA, 2006). The CBCL

provides T scores and percentile scores for different behavioral areas. These areas are

divided into three competency scales (Activities, Social, School), a total competency

scale, eight syndrome scales, two broad problem scales, and a total problem scale. The

eight syndromes are aggressive behavior, anxious/depressed, attention problems, rule-

breaking behaviors, social problems, somatic complaints, attention-deficit hyperactivity

problems, oppositional-defiant problems and conduct problems (ASEBA, 2006).The

broad problem scales are comprised of internalizing and externalizing problem scales.

UTILITY OF THE BGII 16

High T scores are scores above 70 on the problem scales and are indicative of pathology

in that particular scale. However, resultant high T scores on the competency scales are

indicative of mental resiliency and internal strengths of the individual (ASEBA, 2006).

Diagnostic Likelihood Ratio (DLR). The ratio of two proportions: the sample of

people with a particular test result among all those who have a specific condition divided

by the sample of people with the same test result among all those without the condition

(Hamza, 2008). A DLR is the ratio of the posttest odds of having a particular diagnosis

to the pretest odds of having that diagnosis among the general population based on a

specific assessment (CHOI, 1998). This is calculated by obtaining the sensitivity and

specificity of a particular test. Sensitivity is the probability of obtaining a positive test

result among those with a true diagnosis. Specificity is the probability of obtaining a

negative test result among those individuals without the particular diagnosis (CHOI,

1998).

Evidence-based practice. An effective treatment approach that is used for

specific disorders based on systematic empirical research. Most areas of medicine,

psychology, and sociology encourage the promotion of treatments based on empirical

evidence (Luebbe, Radcliffe, Callands, Green, & Thorn, 2007; Youngstrom & Duax,

2005). Evidence-based practice incorporates the best research evidence available, patient

preference and clinical judgment in an effort to combine clinical practice and research

advancements in psychology to advance the field (Luebbe et al., 2007).

UTILITY OF THE BGII 17

KSADS-PL+ with LEAD percentages. The Kiddie Schedule for Affective

Disorders and Schizophrenia for School-Age Children (K-SADS) is a semistructured

interview for the psychiatric assessment of children and adolescents aged 6 to 18. It was

adapted from the adult version, Schedule for Affective Disorders and Schizophrenia

(SADS) developed in 1978 by Endicott and Spitzer (Ghanizadeh, Mohammadi, &

Yazdanshenas, 2006). There are currently three accepted versions of the KSADS. The

Kiddie Schedule for Affective Disorders and Schizophrenia- Present and Lifetime Edition

(KSADS-PL+) was originally developed by Puig-Antich and is the version used in this

study (Ghanizadeh et al., 2006). The KSADS-PL+ version is a further modification that

includes additional items sensitive to symptoms of depression and mania and has been

used to assess the presence of 32 DSM-IV-TR diagnoses. Scoring allows the researcher to

question for the presence or absence of various symptoms within essential DSM-IV-TR

categories with further questioning based on the clinical judgment of the interviewer

(Ghanizadeh et al., 2006). The current version includes items from the Young Mania

Rating Scale as well as the mood disorders module originated at Washington University,

(WASH-U-KSADS-PL+; Geller et al., 2001; Youngstrom, 2005).

The Koppitz—2 Visual Motor Index. This instrument is a measure of overall

visual-motor integration skill and is defined as the ability to relate visual stimuli to motor

responses in an accurate and appropriate manner based on the developmental scoring

system of Elizabeth Koppitz and adapted for the Bender Gestalt II (BGII; Reynolds,

2007). The Koppitz 2 raw score is based on yes/no questions specific to common errors.

UTILITY OF THE BGII 18

On the Koppitz 2, each drawing may have several questions regarding errors, such as on

Design 7, the columns are all slanted left to right, No = 0, Yes = 1; the more errors

produced by the individual, the lower the score. The Koppitz -2 Visual Motor Index was

an age-corrected deviation-scaled score set (M= 100, SD= 15). This error- based

approach is different than the BGSS which focuses on drawing accuracy in copying the

original stimulus card. The categorical descriptive ratings were identical to the BGSS as

previously described: 80-89 is below average, 90-109 is average, and 110-119 is high

average (Reynolds, 2007). The time to complete the drawings has been standardized for

age and results are interpreted according to the number of errors (Reynolds, 2007). The

older version of the Koppitz Developmental Scoring System (1975) was the most

commonly used procedure to score the OBG designs produced by children ages 5 to 12

(Shapiro & Simpson, 1995).

The Koppitz 2 adds an additional score to the BGII, which is the Koppitz 2

Emotional Indicators (EI). The EI score is based on the OBG and original research

regarding emotional pathology. Koppitz (1975) developed the EI test when used as a

projective measure to identify children with emotional problems (Reynolds, 2007).

Unlike the Koppitz 2 Visual Motor Scores, the Emotional Indicators are added based on

errors drawn. The Koppitz 2 Visual Motor Index scores and Emotional Indicators scores

have been included in this study as they are popular for use among clinicians treating

children and adolescents.

UTILITY OF THE BGII 19

Longitudinal Expert Evaluation of all Available Data (LEAD; Spitzer, 1983).

The LEAD assessment involves a formal review of all the data by an expert or team of

experts (Klein, Ouimette, Kelly, Ferro, & Riso, 1994; Pilkonis, Heape, Ruddy, & Serrao,

1991). LEAD was held post interview and was a final review of all presenting data under

the direction of a licensed child psychologist. LEAD was held in person, by telephone, or

video conferencing with all raters reviewing the results of the KASDS-PL+, along with

the child’s supplemental information such as the child’s history, family history, and

school behaviors. Final diagnoses were made by consensus based on a degree of

certainty by the assigned raters.

Pediatric Diagnostic Categories. The categories use the DSM-IV-TR as the basis

for the diagnosis and categorization of mental disorders for both research and clinical

practitioners. Affective Disorders include both the unipolar and bipolar types of affective

disorders. Behavioral disorders include Attention Deficit Hyperactivity Disorder

(ADHD) and Disruptive Behavior Disorders. The Residual Disorder category includes

Anxiety Disorders which further includes Panic Disorders, Obsessive Compulsive

Disorder, Posttraumatic Stress Disorder, Phobias, or any other diagnosed anxiety disorder

within the DSM-IV-TR. The Residual Disorders category may also include those

individuals not diagnosed with any disorders from those major categories but who were

still seeking mental-health interventions. These may include those with Adjustment

Disorders, Psychotic Disorders, Learning Disorders or those individuals that have

symptoms that do not meet any DSM-IV-TR Axis I criteria (Youngstrom et al., 2001).

UTILITY OF THE BGII 20

Psychometric screening tool. This is an instrument developed to determine

minimum criteria for inclusion in a specified group without the loss of the reliability or

accuracy of the instrument (Hildreth, 1945). A screening instrument is often meant to be

a brief assessment within a larger more detailed examination.

Receiver Operating Characteristic (ROC). ROC is a statistical method which

graphically plots the sensitivity or true positive rate against the false positives rate of the

occurrence of a specific condition. This is done to illustrate the accuracy of a particular

diagnosis using a specific measure versus the probability of an inaccurate diagnosis using

that same measure (Zweig & Campbell, 1993). The ROC analysis and the ROC curve are

useful in the selection of optimal tests or measures and the dismissal of nonuseful ones

based on a statistically derived discrimination threshold. The ideal prediction measure

results in a 1.0 diagnostic likelihood ratio, which represents 100% sensitivity and 100%

specificity regarding a particular diagnosis. This outcome is highly unlikely, however, as

there are usually a number of false negative and false positive outcomes in diagnostic

testing (Zweig & Campbell, 1993).

Semistructured interview. This is a diagnostic assessment technique with

specific standardized questioning areas. These include suggested questions to cover

specific DSM areas while using the clinical knowledge and judgment of the rater to

decide on appropriate interventions for each individual client (Geller et al., 2001).

UTILITY OF THE BGII 21

Chapter 2: Review of Selected Literature

The psychological interview and the psychological assessment battery are two

methods used by clinicians in making psychological and psychiatric diagnoses. Once a

diagnosis has been formulated the results of the interview and assessment battery can be

further used in developing an effective treatment plan. There are three basic methods for

diagnosing individuals, the more traditional, less structured, open-ended interview, the

structured interview and the semistructured interview. The more unstructured clinical

interview has given way to more formalized interviewing techniques in order to increase

accuracy and accountability in diagnosis and treatment. Each method has its advantages

and disadvantages. The unstructured interview is qualitatively based, whereas the

structured interview is quantitative and symptom based, while the semistructured

interview attempts to balance diagnostic accuracy with clinical meaning by combining

aspects of both the unstructured and structured approaches (Jellinek & McDermott,

2004). Additionally, the intervention of managed health-care panels and insurance policy

constraints has changed the level of accountability for psychologists with the requirement

to evaluate and diagnose based on the DSM-IV-TR (Cashel, 2002; Jellinek & McDermott,

2004).

Cashel (2002) found that the training of clinicians to conduct psychological

assessments has changed very little over the years even with changes in psychological

training objectives which now focus more on evidenced-based practices (Luebbe et al.,

2007). Practicing clinicians working with children have, in recent years, increased their

UTILITY OF THE BGII 22

use of structured observations, behavior rating scales, and shorter formats of intelligence

measures (Archer et al., 1991; Cashel, 2002). These abbreviated measures and screening

instruments also serve as a form of cost containment for psychologists (Archer et al.,

1991). Cashel reported that many of the clinicians surveyed reported significant

limitations placed upon them by outside sources. He further believed that this ultimately

restricted effective diagnostic decision making. Clinicians find themselves bound by

increased pressure to be more outcome and diagnosis based and less focused on process

in order to quickly arrive at a diagnostic label.

Contrary to this tendency is the reality that appropriate diagnosis and early

intervention in mental illness are not only necessary but also have led to improved

outcomes in children with different clinical diagnoses, such as autistic spectrum

disorders, attention deficit hyperactivity disorder, depression, and bipolar disorder

(American Psychiatric Association, 2000; Charman & Baird, 2002; Lipovsky et al., 1989;

Valderhaug, & Ivansson, 2005). For example, Charman and Baird (2002) found that

increased recognition of symptoms by primary health practitioners, more frequent use of

screening instruments by professionals, and evidence that intervention improves

outcomes, have all contributed to the earlier diagnosis of autism from a mean age of 12

years to four years. Kamphaus et al. (2000) found that the expansion of school-based

psychological services has led to an increased awareness of childhood mental-health

issues and earlier identification of those in need of instructional and developmental

services.

UTILITY OF THE BGII 23

Psychologists continue to have an important role in striving to maintain a high

standard of care while adhering to the goals of achieving diagnostic accuracy and dealing

with the demands of an ever-changing profession. Therefore, the initial diagnostic

assessment becomes a pivotal point in the treatment process (Krueger & Finger, 2001). A

thorough diagnostic assessment should include detailed information on development,

family and social history, parental description of everyday behavior, activities of the

child, direct assessment of the child’s communicative, intellectual and adaptive

functioning, as well as the child’s self-perception (Charman & Baird, 2002; Elbert, &

Holden, 1987; Sourander et al., 2005). How psychologists obtain that information in the

clinical interview has been the prerogative of the clinician. Originally, a subjective

clinical interview process was employed, but this utilized open-ended questions and

projective assessment measures to evaluate and diagnose (Jellinek & McDermott, 2004).

Under certain conditions, clinical judgment came less into play as the use of behavioral

assessment tools, both formal and informal, became available to assist in information

gathering and assess client functioning (Aklin & Turner, 2006; Dryden, 1986). However,

it still remains largely an area of individual preference for each clinician to decide exactly

how to gather data, evaluate, and report findings.

Evaluated data and findings are then summarized into a clinical picture of the

individual using language common to professionals and treatment specialists regardless

of their theoretical orientations (Aklin & Turner, 2006; Dryden, 1986). The DSM-IV- TR

(American Psychiatric Association, 2000) is the current clinical resource manual that

UTILITY OF THE BGII 24

guides diagnostic decision making, and that allows professionals to have a common basis

or language in which to initiate treatment, coordinate care, and discuss prognostic

concerns (Morgan, Olson, Krueger, Schellenberg, & Jackson, 2000). The authors of the

DSM-IV-TR do not claim to adhere to any particular theoretical framework (American

Psychiatric Association, 2000). The DSM-IV-TR is a working medically based manual

used not only by psychologists but also by many other health care practitioners, and

remains important in this constantly changing field of science. Therefore, it has been

under constant revision since its inception, attempting to alleviate weaknesses in previous

editions and to facilitate diagnosis of individuals based on new research findings (Morgan

et al., 2000).

Achieving individual diagnostic accuracy especially with children and adolescents

has been challenging and is also well documented (Aklin & Turner, 2006; Gerber,

Appleton, Dykeman, Sampson, & Toews, 1994; Smith, Muir, & Blackwood, 2004).

Although children and adolescents present with their own particular mental-health issues,

until recently research and treatment have relied heavily on adult research and treatment

guidelines for diagnosis and treatment of these specific groups. This situation has lead to

the likelihood of misdiagnosis or under-diagnosis within the field (Aklin & Turner, 2006;

Gerber et al., 1994, Tolan & Dodge, 2005). For example, misdiagnosis based on adult

guidelines can result in children receiving medications that do not produce the desired

clinical effects and which may also result in serious negative outcomes. Misdiagnosis has

also been related to longer and more costly treatment regimens. These negative aspects

UTILITY OF THE BGII 25

have resulted in a general lack of trust in the profession by both clients and their families

(Aklin & Turner, 2006; Smith et al., 2004). In an effort to lessen the likelihood of

misdiagnosis and eliminate the need for costly testing, psychologists have begun to look

at the use of clinical base rates and evidence-based practice, as previously discussed,

when diagnosing children and adolescents (Youngstrom & Duax, 2005).

To assist in diagnostic decision making, psychologists have often relied on the use

of lengthy multimethod assessments to insure a more global perspective on the individual

(Archer et al., 1991; Dryden, 1986). Multimethod assessments within a psychological test

battery involve several informants’ reports of the child’s behavior, as well as direct

measures of the child’s intelligence and abilities. For many reasons, including possible

denial, mistrust of professionals, or limited insight, children may not be the best reporters

of their own problems. Therefore, professionals have resorted to obtaining information

regarding children from other sources, often in the form of checklists (Achenbach, 1991).

It is understandable that reports of children and adolescent behaviors elicited at school

may differ from behaviors seen within the home and vice versa. Therefore, information

regarding the parent-child relationship and the teacher-child relationship is thought to add

to the global picture in an assessment (Achenbach, 1991; Achenbach & Rescorla, 2001).

Along with reported views of the child, a multimethod assessment would include

psycho-educational testing instruments, personality profiles, behavioral checklists and

mental status examinations. Direct assessment of a child’s skills or functioning is

obtained through intellectual evaluation and cognitive assessment. Performance results

UTILITY OF THE BGII 26

are standardized and compared to the results of a reference group, normed either for age

or grade. Some of the more common tests include the Wechsler Intelligence Scales and

the Woodcock Johnson Scales (Archer et al., 1991). Psycho-educational testing alone

does not specifically address mental-health and pathological concerns and tends to focus

only on intellect and ability. Thus, personality or psychopathology is often mainly

assessed in the form of behavioral checklists as the increase in the use of empirical

methods has caused psychologists to be more cautious in the use of projective measures

(Cashel, 2002; Elbert & Holden, 1987; Kamphaus et al., 2000).

Predoctoral internship programs offering psychological training for child

psychologists have found that generally teaching methods have remained relatively stable

over the past two decades (Archer et al., 1991, Cashel, 2002). However, changes within

general clinical practice have occurred. These include as noted, school- based assessment,

growth of abbreviated IQ measures, and popularity of behavioral measures over

projective measures (Cashel, 2002; Elbert & Holden, 1987; Kamphaus et al., 2000).

Clinical researchers have also become aware of the need to develop a thorough

and efficient intake protocol geared toward accurate diagnosis of childhood mental-health

disorders (Youngstrom, 2001). As previously discussed, psychology has traditionally

utilized both objective and subjective measures to aid the clinician in reaching accurate

diagnoses. Since the release of the original Binet intelligence test in 1908, assessment

measures have expanded to include not only tests for intelligence, but also for

achievement, vocational aptitude, personality, neuropsychological, development, and

UTILITY OF THE BGII 27

individual behavioral. Currently clinicians also use many different subjective, objective,

self-report, checklist, or observational techniques. The development of the empirically

based DSM-IV-TR classification system of diagnosis has resulted in more reliance by

clinicians on formal data collection of symptoms, identification of problem behaviors,

and determination of functioning (First et al., 2004).

Therefore, the initial interview and diagnostic formulation become pivotal points

in the treatment process. The interview or assessment method used to obtain pertinent and

valid information should include the following: development, family and social history,

parental description of everyday behavior, activities of the child, direct assessment of the

child’s functioning, and the child’s self-perception (Charman & Baird, 2002). A continual

quandary for many practitioners is the need to get as much information as possible within

a limited time frame and to balance the rigorous task of information gathering with that

of cost effectiveness.

One evidenced-based method developed to help obtain accurate diagnoses and

reduce clinician diagnostic error allows psychologists to use the statistically based

method of probabilities used by medical professionals. The Diagnostic Likelihood Ratio

(DLR) method was adapted for use by psychologists and incorporated the inclusive

symptom approach of the DSM-IV-TR with quantitative analyses (Youngstrom & Duax,

2005). This evidence-based approach to assessment allows the clinician to weigh the use

of various costly testing measures against the likelihood of supporting or negating

possible diagnoses. Evidenced-based practice if used in conjunction with accurate base

UTILITY OF THE BGII 28

rates and likelihood ratios can reduce the previously discussed conflict between costly

batteries of tests and limited time for diagnosis. However, there is a twofold problem in

incorporating this approach to assessment. The practicing clinician needs to know his

client base rate which is based on the likelihood of anyone in his market area having the

specific diagnosis as well as the best assessment measures to evaluate for that specific

disorder. More simply stated, the question is whether a particular assessment is going to

add to the clinician’s fund of knowledge about a specific diagnosis and will aid in

reaching an accurate diagnosis in a timely and cost effective manner.

Research Focus

The purpose of this study was to compare the BGII drawings, using the BGSS and

Koppitz 2 scoring methods, with the diagnostic accuracy of the WASH-U-KSADS-PL+

interview and the popular CBCL measure to determine the diagnostic utility of the BGII

as a decision making tool. In the current study, the BGSS and the Koppitz 2 scoring

system for the BGII test were evaluated in a research setting to assess their contribution

as projective screening measures that would aid diagnosis and treatment planning in

clinical practice with children. It has been the hope of this researcher to begin a process

leading to an increase in the clinical utility of the BGII not only as a developmental visual

motor performance test but also as an additional effective assessment tool in a

psychological test battery. Since this is exploratory research, the BGII, BGSS and the

Koppitz 2 scoring results have been used. It was predicted that the BGSS and the Koppitz

2 would do equally well in screening children for visual motor integration errors. It

UTILITY OF THE BGII 29

remains uncertain whether they can effectively reach a diagnosis when compared with the

WASH-U-KSADS-PL+ and the CBCL in a clinical sample of children. Both scoring

measures of the Bender Gestalt II Test, although similarly normed by current research

and measurement techniques, may yield different results.

The WASH-U-KSADS-PL+ (Youngstrom et al., 2005) is currently being used in

an ongoing study “Assessing Bipolar Disorder: A Community Blend (ABACB)” to

assess accuracy of childhood diagnosis with regard to childhood bipolar disorder in

Cleveland, Ohio, through Case Western Reserve University. Permission to conduct the

larger study was originally given by Case Western University Institutional Review Board

in 2003 as a 5-year study comparing different instruments in diagnostic assessment for

juvenile bipolar disorder. Addenda to this study have been routinely added during the

yearly review process to compare various measures, both research and clinically driven.

The University of North Carolina Institutional Review Board further reviewed this study

in 2006 for data collection at an additional site. Once the IRB addendum was approved

through Case Western Reserve University, this study was presented to the West Virginia

University Institutional Review Board (Youngstrom, 2005). West Virginia University

determined the data collection was archival and the present study was given exempt

status. All Institutional Review Board releases can be viewed in Appendix A.

The major purpose of the Case Western Reserve University study with sites at

Applewood Centers Inc., Case Western University, and University of North Carolina was

to clarify the characteristic features of childhood bipolar disorder in children, to cross

UTILITY OF THE BGII 30

validate a childhood bipolar screening protocol for use in clinical settings, and investigate

the developmental changes in symptoms across the age span using both cross sectional

and longitudinal approaches (Youngstrom, 2005). Because one of the main purposes of

the research study was the validation of potential screening measures, these screening

tools were segregated from the LEAD assessment and reviewed and analyzed separately

(Youngstrom et al., 2005). The BGII was one screening tool that was added in the fifth

year of the study and was also segregated from the LEAD process (see Appendix B). If

any of the Bender results had been found to be an adequate predictor of any specific

diagnostic criteria then further analyses would have been conducted to determine its

future use by developing a diagnostic likelihood ratio. In this study, a DLR was only

performed for children with the diagnosis of ADHD based on the KSADS-PL+ with

LEAD results (Frazier, 2006; Jaeschke, Guyatt, & Sackett, 1994; Youngstrom, 2006).

The Bender Visual Motor Gestalt Test

The OBG has been commonly used as a quick, simple, direct measure of the

individual’s internal psychological state (Archer et. al., 1991; Bender, 1938; Piotrowski

& Keller, 1989). The originally titled OBG, or the Bender Visual Motor Gestalt Test

(Bender, 1938), was composed of simple geometric drawings developed by Max

Wertheimer (1880-1943) in his perceptual psychology experiments and later formalized

by Bender (Archer et al., 1991; Bender, 1938; Koppitz, 1975).

According to Bender, developmental maturation of drawing ability is an ongoing

process that follows sequential stages, incorporating gross to fine-motor skills, visual

UTILITY OF THE BGII 31

imagery, and perceptual awareness and a developmental drawing tendency from vertical

to horizontal movements, and finally from two dimensional to three dimensional

awareness. These stages were all considered part of the maturational process leading to

the more intricate representations found within a mature representational drawing that

become a complete gestalt or perceptual integration (Bender, 1938). Bender theorized

that a deviation within this maturational process would obviously lead to a disintegration

of the original representation or errors between the drawing and its stimulus. As

previously noted, psychologists have found this concept helpful in identifying those

individuals who have not yet developmentally matured, or are delayed in their visual

motor perception, as well as those who were once psychologically mature yet for various

possible reasons may have lost such integration (Bender, 1938; Brannigan & Decker,

2003, 2006; Koppitz, 1975 ; Reynolds, 2007).

The relationship of visual-spatial skills and working memory in areas of

intelligence when correlated with tasks of executive functioning has been more recently

addressed by Miyake, Friedman, Rettinger, Shah, and Hegarty, (2001). These researchers

found that people who are good at complex visual-spatial tasks also perform better on

executive function tasks which are crucial in regulating and controlling behavior (Miyake

et al., 2001). Shapiro and Simpson (1995) found the Koppitz scoring system in disturbed

adolescents with emotional and behavioral disorders did not measure intelligence, but

was able to interpret maturational development, which correlates to the use of the BGII.

The researchers evaluated a clinical group of adolescents 12 to 17 years of age and

UTILITY OF THE BGII 32

determined that visual motor skills continue to develop beyond the age of 11 and the

scoring results were not found to be related to initial diagnosis, gender, or intelligence

level of the adolescent (Shapiro & Simpson, 1995). Their work with the original Koppitz

scoring system provided useful information with regard to adolescents for this research

study.

As previously stated, the OBG was one of the most popular assessment measures

for more than 60 years and it was consistently listed as one of the top used instruments by

psychologists when queried over those years (Archer et al., 1991; Brannigan & Decker,

2003; Piotrowski & Keller, 1989). There have been many researchers who cited the OBG

as the criterion measure in their research (Brannigan & Decker, 2003; Reynolds, 2007;

Hutt, 1985). Horn and O’Donnell (1984) studied the early identification of learning

disabilities and found that the OBG was effective in identifying those children classified

as learning disabled as well as those with low achievement. The OBG was found to be

useful in separating those adolescents diagnosed with impulsivity from those without

impulse difficulties (Oas, 1984). The Bender Gestalt–Recall technique was found to be a

valid measure of short-term visual memory in children and adolescents and comparable

to the Coding recall measure on the WISC-III (Imm, Kim, Belter, & Finch, 1991). Since

the OBG has been used in over 1,300 published articles and 60 years of research, its

strengths and weaknesses are well documented (Brannigan & Decker, 2003).

The use of the OBG became popular and widespread even with cautions on its use

as a projective and neurological test. It continues to be a preferred measure possibly

UTILITY OF THE BGII 33

because of its simplicity and short administration time (Bigler & Ehrfurth, 1981). The

popularity of the Bender as a clinical tool has led to the development of a plethora of

scoring measures and interpretive manuals over the years, although some exhibit highly

questionable reliability and validity (Brannigan, & Decker, 2003; Hutt, 1985; Perticone,

1998).

As a result the broad utility of the OBG has come into question over time. Many

psychologists have cautioned against wide use of the OBG for diagnostic evaluations

(Bigler & Ehrfurth, 1981) because of low reliability with regard to its use as a projective

test (Naglieri & Pfeiffer, 1992). However, it still remains one of the top 10 instruments

used by psychologists (Piotrowski & Keller, 1989). Thus, the OBG has had two preferred

uses in a childhood population: first, as a visual motor development instrument and then

as a projective measure for identification of certain psychological conditions in both

children and adults (Rossini & Kaspar, 1987).

Current Use of the Bender Visual Motor Gestalt—Second Edition

The Bender Gestalt II (BGII; Brannigan & Decker, 2003) is a modification of the

OBG and maintains the importance of the quality of the drawing as its basis of scoring.

However, the number of designs was increased from 9 to a total of 16 (Brannigan &

Decker, 2003). Four of the new designs are to be given to children younger than 8 years

old and precede the original set of drawings. When giving the Bender Gestalt II drawings

to anyone 8 years old or above, there are three additional drawings that follow the

original administration (Brannigan & Decker, 2003).

UTILITY OF THE BGII 34

Once developed and refined, the BGII was renormed using a national

representative sample obtained through the 2000 US census as well as clinical samples of

selected diagnoses (Brannigan & Decker, 2003). Another change in the BGII is that it

now includes perception and motor subtests to detect specific problems separate from

integrative processes for those individuals who perform below expectation (Brannigan &

Decker, 2003). In order to measure visual-motor integration skills in children and adults

from four to 85 years of age, the BGII is administrated in two stages. The first consists of

a Copy and Recall phase, followed by two supplementary tests (the Motor Test and the

Perception Test).

On the BGSS, scoring is performed by conversion of raw scores to Standard

Scores ranging from 40 to 160 (M = 100, SD = 15). Scores can also be converted to

percentile ranks, t scores, z scores, and age equivalents. Individuals scoring 1.3

standard deviations below the mean Visual Motor Index of 100 were considered to be

“mildly impaired” and those scoring more than 2 standard deviations below the mean

were considered “significantly impaired” (Brannigan & Decker, 2003). Above average

performance on the BGSS and Koppitz 2 presupposes normal levels of both visual-

perceptual skills and fine-motor coordination. Below-average performance may be due

to problems in the domain, the integrative process, motivation, attention, or related

concerns (Reynolds, 2007). A recent study examined the use of the BGII with children

diagnosed with ADHD. Results indicated that those with ADHD tended to do more

UTILITY OF THE BGII 35

poorly than a normal child group. However, these differences disappeared when

intellectual level was statistically controlled (Allen, 2005).

Despite the 2003 revision of the OBG, it is still regarded as controversial to many

in the field, as will be discussed in greater detail in subsequent sections. However, as

noted in the purpose of this research there are those who hope to rectify this situation and

establish the BGII as a clinically useful tool.

The Koppitz 2 and the Koppitz Developmental Scoring System

The Koppitz 2 scoring system for the BGII is a revision of the OBG Koppitz

developmental scoring system developed in 1963 by Koppitz, a child psychologist

(Reynolds, 2007). Koppitz (1975) developed her own specific instructions for

administration and scoring using a developmental approach with scores based on errors

found within the design construction (Perticone, 1998). Scoring was based on total

number of errors throughout the OBG. The original Koppitz was revised by Reynolds, a

colleague and personal friend of Koppitz. In her early research, Koppitz had concluded

that both developmental and emotional results had an overall diagnostic value and

represented a possible indication of emotional disturbance within an individual

(Perticone, 1998). Koppitz (1975) was guarded about generalizing individual indicators

as personality characteristics. She saw them as more of a guide for further inquiry

(Perticone, 1998). Following the introduction of the BGII, the need for revisions in the

Koppitz Developmental scoring system became clear, resulting in Reynolds developing

the Koppitz 2 Developmental Scoring System.

UTILITY OF THE BGII 36

Along with the developmental scoring, Koppitz included a scoring manual for

Emotional Indicators, to differentiate children having emotional difficulties from those

appearing well adjusted (Perticone, 1998; Reynolds, 2007). The Koppitz Emotional

Indicators were developed using empirical methods based on psychodynamic theory to

distinguish children with serious emotional problems from those without problems.

Koppitz (1975) reported that emotional indicators were clinical symptoms or signs that

should be individually evaluated by the rater. A single indicator by itself may not indicate

serious pathology; however, a single indicator may show a manifestation or tendency

toward some disturbance. Therefore, indicators could occur separately or in combination.

The emotional indicators identified for scoring on the OBG were confused order, wavy

lines, dashes for circles, increasing size, large size of drawing, or overly small size of the

drawing, fine line, overworked or reinforced lines, second attempt of a drawing, and use

of two or more sheets of paper to complete the test. However, these remain controversial

due to their projective nature (Koppitz, 1975; Perticone, 1998; Reynolds, 2007).

Additionally, Koppitz (1975) cautioned that the indicators were not comparable or

correlated to the developmental test score.

The Koppitz 2 authors reviewed earlier research supporting the emotional

indicators based on over 500 studies from Koppitz and others. They removed those that

were demonstrated to be age related, and then included two indicators that were of

considerable empirical significance based on previous research (Reynolds, 2007). The

Koppitz 2 emotional indicators added drawing a box around one or more designs and

UTILITY OF THE BGII 37

spontaneous elaborations or changes in the overall gestalt of the design to the 10

emotional indicators reported by Koppitz in 1975. This resulted in a final total of 12

emotional indicators that when found in a profile were to be taken into consideration for

concern by the clinician (Koppitz, 1975; Reynolds, 2007; Rossini & Kaspar, 1987). The

original research for the Koppitz Emotional Indicators reported that three or more errors

were indicative of possible pathology (Koppitz, 1975). However, the current Koppitz 2

Emotional Indicators scoring is quantitative with more than four considered to be high

risk or of concern for emotional difficulties (Reynolds, 2007).

Archer et al. (1991) researched psychological test usage among psychologists and

found that the OBG rated third among all assessment measures for use with adolescents.

According to this research, the OBG was primarily reported as diagnostically useful in

learning disability assessments. This leads to a larger question: Exactly what resultant

data have been derived from the clinical use of the OBG and now the BGII? Is it a viable

screening method that can add information to the diagnostic assessment interview? Or

should the BGII be replaced by more structured diagnostic techniques to arrive at the

same necessary information? Finally, is the BGII of diagnostic benefit to include as an

assessment measure in children and adolescents?

Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age Children

The Kiddie Schedule for Affective Disorders and Schizophrenia for Children (K-

SADS) is a semistructured interview. Methods of clinical interviewing have changed

since 1938 when the Bender cards were first introduced as a screening instrument (Aklin

UTILITY OF THE BGII 38

& Turner, 2006). In an attempt to reduce the high error rate, structured, and

semistructured interviews have been developed to replace the open-ended, more

unstructured interview process (Aklin & Turner, 2006). The structured interview process

provides systematized ratings, outlining specific behaviors and symptoms that need to be

addressed using a standardized format. These standardized formats, however, have

limitations. The structured interviews have been criticized for their inflexibility and lack

of depth in obtaining information (Aklin & Turner, 2006). Semistructured interviews

offer more latitude to the clinician for the diagnosis of patient symptoms. Although the

semistructured interviews alone may be less reliable due to the individual style of the

clinician, the qualitative and quantitative analysis found that the combination of the

semistructured interview and the use of empirical screening measures, may yield the

deepest understanding of the client’s situation (Aklin & Turner, 2006; Findling et al.,

2001; Findling et al., 2005).

The K-SADS is a semistructured diagnostic interview for children and

adolescents (6-18 years) designed to assess current and past episodes of psychopathology

according to DSM-IV-TR criteria. The K-SADS is administered by interviewing the

parent(s) and the child, and finally developing summary ratings which include all sources

of information (parent, child, school, chart, and other collateral sources).

The KSADS-PL+ and the CBCL have undergone considerable review and found

to have acceptable validity and reliability (Achenbach System of Empirically Based

Assessment [ABESA], 2006; Achenbach, 1991; Ambrosini, 2000; Ghanizadeh et al.,

UTILITY OF THE BGII 39

2006). However, they are very different instruments (ABESA, 2006; Aklin & Turner,

2006; Chambers et al., 1985; Findling et al., 2001; Findling et al., 2002). The KSADS is

a research derived instrument with limited use in clinical practice due to the cost and

administrative time demands (Chambers et al., 1985). The CBCL, on the other hand, is

both time and cost efficient. It consists of a behavioral checklist of distressing behaviors

to the responding individual. It is generally used in clinical practice as a screening

instrument (ASEBA, 2006).

The K-SADS is a semistructured interview that was developed in the late 1970s

by Drs. Puig-Antich and Chambers (Ambrosini, 2000). The original version of the

KSADS has gone through several modifications. The K-SADS PL+ includes the present

and lifetime edition of the KSADS along with the Youth Mania Rating Scale questions to

cover all possible symptoms of bipolar disorder (Findling et al., 2005; Gracious,

Youngstrom, Findling, & Calabrese, 2002). The K-SADS PL+ is DSM III-R (1987) and

DSM-IV (1994) compatible and includes interviews with both the parent and child. The

administration of the KSADS-PL+ takes approximately 90 minutes for each interview

and can take longer depending on the number of supplemental diagnostic areas that are

deemed significant (Ambrosini, 2000; Findling et al., 2001; Findling et al., 2005). The

KSADS, as well as the KSADS-PL+, allows for clinical judgment based on the research

diagnostic criteria and the placement of a specific symptom within the diagnostic

criterion (Ambrosini, 2000, Findling et al., 2001, Findling et al., 2005).

UTILITY OF THE BGII 40

The Wash-U-KSADS-PL+ interview is based on DSM –IV-TR diagnostic criteria,

and as previously discussed in regard to the DSM-IV-TR, there can be considerable

overlap across diagnoses and different clinician interpretation of a symptom. Using the

KSADS, KSADS PL+, or the WASH-U-KSADS-PL+ versions, can all be an arduous

process for the clinician, because training for test administration focuses on a clear

understanding and interpretation of various symptoms, administering the test must adhere

to a strict protocol, and learning to interpret the interviewee’s responses are all difficult

tasks (Findling et al., 2001, 2005; Youngstrom et al., 2002). WASH-U-KSADS PL+

training requires both a didactic and interview-observer component in order to teach the

interview process and insure scoring reliability (Ambrosini, 2000; Geller et al., 2001).

The K-SADS PL+ is administered by interviewing the parent and the child and finally

developing summary ratings, which include all sources of information. At some current

research sites, all the raters then meet for a consensus of the different summary ratings to

determine clinical diagnoses; this is known as the longitudinal expert evaluation of all

available data or LEAD (Spitzer, 1983).

The LEAD assessment involves a formal review of all the data by an expert or

team of experts (Klein et al., 1994; Pilkonis et al., 1991). LEAD is implemented

following interviews and includes a final review of all presenting data under the direction

of a licensed child psychologist at one of the current sites (Youngstrom et al., 2005).

LEAD may be implemented in person, by telephone, or video conferencing with various

raters reviewing the results of the KASDS-PL+. The final diagnoses are determined

UTILITY OF THE BGII 41

based on percentage of certainty by trained raters (Youngstrom, 2005). Although raters

are highly trained, which enhances scoring reliability, the interviews are very time

consuming, so the KSADS PL+ with LEAD, is not likely to be used in most clinical

practices. It is, however, used for research purposes and compared with commonly used

assessment measures (Findling et al., 2002; Findling et al., 2005; Youngstrom et al.,

2005). The WASH-U- KSADS PL+ with LEAD is a thorough and lengthy interview. It

can be useful in achieving accurate diagnoses using the DSM-IV-TR criteria (Ghanizadeh

et al., 2006). As previously stated, it is, however, impractical for the practicing clinician

due to its labor intensive administration and low cost efficiency.

The Child Behavior Checklist

The Child Behavior Checklist (CBCL) by Achenbach and Rescorla (2001) is a

commonly administered checklist given to both parent and child to disclose current

symptomology (Achenbach & Rescorla, 2001, ASEBA, 2006). It assesses 120 emotional,

a behavioral, social and psychological problem frequently reported by parents; and is

meant to be part of multi-informant instruments of empirically based assessments

originally developed by Achenbach (1991) and later revised by Achenbach and Rescorla

(2001). It assesses a broad range of behavioral symptoms found among children with

emotional difficulties (Achenbach, 1991). The revised CBCL resulted in the development

of two different formats: The CBCL for ages 1 to 5 years (CBCL/1 -5) and CBCL for

ages 6 to 18 years (CBCL/6-18) (Achenbach & Rescorla, 2001, ASEBA, 2006). The

Achenbach behavioral checklists are multiinformant instruments and have three different

UTILITY OF THE BGII 42

formats which include: parent, teacher, and self report forms. The Achenbach Youth Self

Report Form is a standardized forced response test designed for 11 to 18 year olds to

report their own strengths and areas of difficulties. Rater answers are based on accuracy

of each item currently or within the previous 6-month period (ASEBA, 2006). The CBCL

provides t scores and percentiles for three competency scales (Activities, Social, School),

total competency in these areas, internalizing problems scale, externalizing problems

scale, and total problems scale. There are also eight syndromes that can be differentiated

on the CBCL: aggressive behavior, anxiety /depression, attention problems, rule-

breaking behaviors, social problems, somatic complaints, attention deficit hyperactivity

problems, oppositional defiant problems and conduct problems (ASEBA, 2006). The

CBCL was intended to serve as one component of a multimethod empirically based

assessment and record children’s competencies and deficiencies as reported by their

parents, parent surrogates, or teachers. The CBCL provides clinical information relating

to strengths and competencies as well as problems within the individual (internalizing or

externalizing problems).

The CBCL had been found to be useful in assessing child symptomology based

on clinical levels of internalizing problems and/or externalizing problems, and

distinguishing attention deficit disorder from other disorders (Ivanova et al., 2007,

Lengua, Sadowski, Friedrich, & Fisher, 2001). In research by Ivanova and others, the

CBCL was found to useful be across cultures using the eight-syndrome structure across

30 diverse societies throughout the world (Ivanova et al., 2007). However the CBCL,

UTILITY OF THE BGII 43

similar to the DSM-IV-TR, has been criticized for its overlapping of diagnoses and a

tendency to over-estimate the co-occurrence of diagnoses (Lengua et al., 2001).

One study found the CBCL to be the most commonly utilized behavioral checklist

among clinical and school psychologists for use with children (Cashel, 2002). In recent

years, behavioral checklists and rating scales have replaced personality and projective

measures as the instrument of choice among psychologists (Kamphaus et al., 2000). The

popularity of behavioral checklists has been found to be related to their time efficiency,

straightforwardness and ease in quantifying the results (Kamphaus et al., 2000). However

as with any testing method, there are limitations. The self-report measures may raise

questions of honesty and defensiveness of the reporter (La Fiosca & Loyd, 1986).

Validity of self-reports has been found to improve when there are different sources of

measurement. The reports often include parent, child, and teacher versions when

screening (Achenbach & Rescorla, 2001). This researcher supports the use of behavioral

checklists to aid in diagnostic decision making and finds the CBCL is an adequate

assessment tool to use for children within this current study. Although the Youth Self

Report (YSR) by Achenbach (1991) would have been an ideal comparison assessment,

because it is geared for children older than 11, it was not used as a comparison measure

due to the younger aged children assessed.

As described in this section the OBG has been used for many years and previous

research into the BGII lays the ground work for the current study examining the utility of

the BGII and the previously discussed scoring methods. This current study also supports

UTILITY OF THE BGII 44

the assessment tools of the WASH-U-KSADS-PL+ and the CBCL as diagnostic decision

making tools. How these instruments are used in the current study will be further

addressed within the procedure section of the next chapter.

UTILITY OF THE BGII 45

Chapter 3: Method

This study is comparative research using archival data from a 1-year subset of a

larger 5-year study that included over 600 participants. This parent study aimed at

investigating the effectiveness of screening measures in diagnosing bipolar disorder in

children and adolescents. For my study, a total of 115 children were initially evaluated

and 75 completed all the procedures of the research protocol comprising the KSADS-

PL+, the CBCL, and the Bender Visual Motor Gestalt Test II. The data for all subsequent

analyses are derived from these 75 participants, as will be clarified below.

Participants

Inclusion criteria for this research as well as the larger general study were (a)

youth between the ages of 5 years and 18 years, (b) written consent and assent from both

caregiver and client, (c) both caregiver and youth presented for the assessment, and (d)

both caregiver and youth were functional English speakers. Participants for the larger

study were 620 caregivers and youth invited from the intakes of a community mental-

health center in Cleveland, Ohio, following a consecutive case-series design. A

consecutive case series design reports the outcomes of a group of individuals or clients

with a similar condition treated in the same manner (McKeon, Medina, & Hertel, 2006).

The parent study enrolled participants between July 2003 and March 2008, while my

research, which focuses on the BGII as a screening instrument, covered a period of time

from May 2007 through March 2008.

As noted my study reports findings for the 75 youth, ages 5 to 18, with complete

data for all measures. The most common reason for missing data was no Koppitz scoring

UTILITY OF THE BGII 46

for the youth (n = 26) followed by a missing CBCL (n = 7). There were seven other

individuals whose protocols were missing a variety of data points. Of the 75 youth

completing all elements of the study the average age was 11.92 (SD = 2.56), with a

median of three diagnoses, 64% male, 84% African American, 11% Caucasian, and 5%

other. Participants enrolled in the larger study over the 5-year time period between ages

5-18 were on average 12.00 years old (SD = 2.67), with a median of two diagnoses, 89%

African American, 7% Caucasian, and 4% other. All subsequent data points are derived

using an n = 75. Independent sample t test using a t test for the equality of means indicate

that youth in the current study were not significantly different in age or comorbidity than

other youth in the larger study t (74) =.817, p = .076. Chi-square analyses indicated no

significant difference between those who received the BGII and those who did not in

ethnicity, x² (3) = 2.86, p = .41. Chi-square analysis indicated no significant difference

between those who received the BGII and those who did not in gender, x² (1) = 1.67, p =

.20.

Measures

The Bender Visual Motor Gestalt II- Global Scoring System. This measure

was designed to assess visual motor integration across the lifespan and aid in

differentiating learning problems as well as psychological and neurological problems

(Brannigan & Decker, 2003). The BGII consists of the original Bender drawings plus

additional drawings added by Brannigan and Decker (2003) to expand testable age range

and provide more discrimination for levels of visual motor perception. Compared to the

OBG, the BGII includes new items (k =16) on the BGII compared to (k = 9) on the OBG,

UTILITY OF THE BGII 47

a memory recall phase, national norms, clinical validity studies, time estimates,

quantitative and qualitative scoring, test observation forms, and co-norming with the

Stanford-Binet (Brannigan & Decker,2003). Both the original and new figures were

based upon laws of perception postulated by Wertheimer (1923). Figures appear

individually on numbered cards. Administration of the BGII consists of two phases: copy

and recall. The copy phase measures visual motor integration. During the copy phase,

cards with a single figure are shown to the individual client one at a time. Participants

then copy each figure onto a blank paper with a pencil. Immediately after the copy phase,

the recall phase occurs. The recall phase measures short-term memory. The individual

draws as many of the figures as he or she can remember on a new blank piece of paper

immediately following the copy phase.

For the BGII, the Global Scoring System (BGSS) replaced the plethora of

alternate scoring systems that had been used for the OBG (Brannigan & Decker, 2003).

The BGSS is a measure to assess the accuracy of the BGII drawings. Scoring is based on

clinician comparison of examinee drawings to examples in the manual (Brannigan,

Decker, & Madsen, 2004). Each figure is scored on a 5-point Likert-type rating scale (0

to 4). A 0 means the drawn figure did not represent the stimulus at all. A score of 4

indicates the drawn figure matches the stimulus nearly perfectly. Drawings are rated from

both the copy and recall phases of administration using the BGSS. The BGSS produces a

sum raw score of all items administered. The raw scores are then converted to standard

scores based on similarly aged individuals within the norming sample (Brannigan &

Decker, 2003). Standard scores range from 40 to 160 (M = 100, SD = 15) (Brannigan &

UTILITY OF THE BGII 48

Decker, 2003). Classification ranges based on standard scores are represented by the

following: 145 to160, extremely high or extremely advanced; 130 to144, very high or

very advanced; 120 to 129, high or advanced; 110 to 119, high average; 90 to 109,

average; 80-89, low average; 70 to 79, borderline delayed; 55 to 69, mildly delayed or

low; and 40 to 54, moderately delayed or extremely low (Brannigan & Decker, 2003).

Validity and reliability of the BGSS were originally discussed by Brannigan and

Decker (2003) during the test revision process. Construct validity of the BGSS is

assessed by comparison with the Beery-Buktenica Developmental Test of Visual Motor

Integration, Fourth Edition (VMI-IV). The BGSS standard scores demonstrate strong,

positive correlations with the VMI-IV on both copy (r = .65) and recall (r = .44)

(Brannigan, Decker, & Madsen, 2004). Additionally, the BGSS was compared to tests of

achievement and cognitive ability. The BGSS has demonstrated modest correlations with

the Woodcock –Johnson II Tests of Achievement (Woodcock, McGrew, & Mather,

2001). The BGSS shows modest correlational scores ranging from .27 to .53 for the Copy

Phase, and .25 to .49 in the Recall phase (Brannigan & Decker, 2003). A study examining

the relationship between the BGSS and the Stanford Binet Intelligence Scales, Fifth

Edition (Roid, 2003) corrected correlations between the BGSS and the SB-5 IQ scores

ranged from .50 to .54 for the Copy phase and .45 to .48 for the Recall phase. Interrater

reliability of the BGSS was extremely high, yielding a Kappa = 0.90, p < 0.05. Another

measure of reliability of the BGSS was found using the split-half method for internal

consistency; Spearman-Brown prophecy formula was > 0.91, at all age ranges indicating

a consistent and stable measurement (Brannigan & Decker, 2003). As reported in the

UTILITY OF THE BGII 49

manual, the BGSS demonstrates adequate validity and reliability for measuring visual-

motor integration (Brannigan & Decker, 2003).

In addition to the BGSS, other researchers have used the revised Koppitz 2 by

Reynolds (2007) to score the figures of the BGII. For this research, Koppitz 2 scoring

occurs after administration of the BGII and BGSS scoring. The Koppitz 2 Emotional

Indicators score was incorporated and scored along with the Koppitz 2 Visual Motor

Integration scoring. The Koppitz 2 is a scoring system used to derive assessment data

from the Bender Gestalt II (2003). It provides an overall visual-motor integration score

that shows the ability to relate visual stimuli to motor responses in an accurate and

appropriate manner (Reynolds, 2007). The Koppitz 2 Visual Motor Index uses an age

corrected deviation scaled score set (M = 100, SD = 15). As previously mentioned, the

categorical descriptive ratings are identical to those of the Bender Visual Motor Gestalt

Test II- Global Scoring system results. Koppitz found significant differences between

children with learning or behavioral difficulties and those who demonstrated average

performance (Brannigan & Decker, 2003). Koppitz recognized excessive time taken for

test completion as a possible indicator of psychopathology. The time to normally

complete the drawings has been examined in previous research and found to be between a

mean of 9 minutes to 14 minutes depending on age for the copy phase with standard

deviations of 4 to 7 minutes on both the BGSS and the Koppitz 2 (Brannigan & Decker,

2003; Reynolds, 2007). Time is often a reported variable in the testing interpretation and

has been associated with number of errors and willingness to perform the test accurately.

UTILITY OF THE BGII 50

Brannigan and Decker (2003) conducted studies during their research

standardization of the BGSS to investigate the relationship between that system and the

Koppitz 2. Correlations with that system and the BGSS were .80 for the copy phase and

.51 for the recall phase. The lowest Cronbach’s alpha of the Koppitz 2 as compared to the

BGSS based on age ranges was found in the 5-year age group at 0.77. In other aspects of

the norming study of the BGSS, age, gender and race variables were considered. When

divided into gender and racial groups, the Cronbach’s coefficients on the BGSS were

found to be approximately 0.90 for ages 8 and above (Brannigan & Decker, 2003). These

were calculated as internal measures of the BGSS during the norming process.

Test- retest method of reliability over time shows an average correlational

coefficient, r = 0 .77 with a range of 0.73 to 0.85 in the normative sample. Interrater

reliability was found to be Cohen’s kappa = 0.91 and 0.93 for the two different protocols

of the Koppitz 2 (Reynolds, 2007). The correlation coefficients for the original Koppitz

Scoring System and BGSS were 0.80 for the Copy phase and 0.51 for the Recall phase

(Brannigan & Decker, 2003). It appears that the scoring systems are stable and measuring

similar constructs, but not to the point of redundancy.

Koppitz 2 scoring is based on specific aspects in the different designs that are

considered indicators for the presence or absence of perceptual difficulties. Tests are

scored on a point system based on examiner affirmative responses to the representation of

a specific item. (For example, for Design 5, 1 point = yes, the two items touch or nearly

touch). There is a possibility of 45 correct in the Koppitz 2 raw score. Raw scores are

then converted to standard scores and are the basis for the Visual Motor Integration score.

UTILITY OF THE BGII 51

The Koppitz 2 extracts a further score from the Bender drawings which is known

as the Koppitz 2 Emotional Indicators (EI). This score is based on the use of the BGII as

a projective indicator to identify children with severe emotional problems. This usage

derived from research data collected by Koppitz and others over many years (Reynolds,

2007). Koppitz identified 12 emotional indicators that she believed were of considerable

significance to those children with severe emotional problems and not a reflection of

intelligence alone (Reynolds, 2007). Those 12 emotional indicators (EI) are (a) confused

order, (b) wavy line, (c) dashes for circles, (d) progressive increase in drawing size, (e)

large size of drawings, (f) small size of drawings, (g) fine lines, (h) overworked or

reinforced lines, (i) second attempts at drawing a design, (j) expansion, (k) box around a

design, and (l) spontaneous elaboration or additions to the design (Reynolds, 2007). In

developing the Koppitz 2, Reynolds added the EI as a supplemental test based on the

original Koppitz research and results (Koppitz , 1963, 1971, 1975).

The Kiddie Schedule for Affective Disorders and Schizophrenia for School-Age

Children (K-SADS). This measure is a semistructured interview that was developed by

Puig-Antich and Chambers (1978). The KSADS allows the examiner to systematically

inquire about symptoms of psychopathology in children and adolescents. Kaufman and

others modified the original KSADS so that it also inquires about present and past

episodes of psychopathology, becoming the KSADS-PL (Kaufman, Birmaher, & Brent,

2003). This version demonstrates adequate reliability and validity as a clinical diagnostic

measure (Kaufman et al., 2003). The Washington University KSADS (WASH-U-

KSADS) version was further modified to include additional symptoms and associated

UTILITY OF THE BGII 52

features of depression and mania (Geller et al 2001). The WASH-U-KSADS-PL +

combines the KSADS–PL and the WASH-U-KSADS. Additionally the WASH-U-

KSADS-PL+ includes the Child Depression Rating Scale (Geller et al., 2001) and the

Young Mania Rating Scale (Gracious et al., 2002). The WASH-U-KSADS-PL+ was the

modification used in the current study and considered the reference standard in childhood

bipolar research. In the current study, interrater reliability was established by training all

research assistants to attain a Kappa coefficient > 0.85 at the symptom level and Kappa =

1.0 at the diagnosis level. To do so, new research assistants were initially required to

score five administrations with a certified rater followed by five interview

administrations leading and scoring the interview with participation of a certified rater.

Passing was scored at an item level Kappa > 0.85.

Final diagnoses were given via the Longitudinal Evaluation of all Available Data.

(LEAD) process (Spitzer, 1983). The process involves a formal review of all data by an

expert or team of experts. Raters were blind to the final diagnoses for study the research

questionnaires and BGSS results, but were later integrated with family history, prior

treatment history, prior testing history, and other clinical observations under the

supervision of licensed clinical psychologists. LEAD was conducted in person, by

telephone conferencing, or video conferencing with all raters reviewing the results of the

WASH-U-KASDS-PL+, along with child history, family history, supplemental

information, and other screening results. Final diagnoses were then determined based on

consensus of certainty by the raters.

UTILITY OF THE BGII 53

The Child Behavior Checklists, (CBCL) and Youth Self Report Form (YSR).

These measures by Achenbach and Rescorla (2001) are rating scales of common Axis I

psychopathology as rated by caregivers and self-report. The CBCL and the YSR assess a

broad range of behavioral symptoms found among children with emotional difficulties

(Achenbach, 1991). The CBCL and YSR were developed to serve as components of a

multiinformant empirically based assessment. The CBCL was revised in 2001 and

resulted in two different forms, the CBCL for ages 1 to 5 (CBCL/1 -5), and the CBCL

for ages 6 to 18 (CBCL/6-18; ASEBA, 2006). Caregivers and youth respond to each item

using the past 6 months as the time frame. Responses consist of a 3-point Likert-type

scale (0 to 2), with 0 indicating never true, 1 indicating sometimes true, and 2 indicating

always true. Raw sum scores are transformed to T scores (M = 50, SD = 10) (Rescorla &

Wagner, 2001). The CBCL and YSR yield eight syndrome scales and three general

scales. The eight syndrome scales are aggressive behavior, anxiety/depression, attention

problems, rule-breaking behaviors, social problems, somatic complaints, attention-deficit

hyperactivity problems, oppositional-defiant problems and conduct problems (ASEBA,

2006). The three general scales are internalizing, externalizing, and total problems. High

T scores on the problem scales are indicative of pathology. T scores higher than 70 are in

the clinical range, T scores of 64 to 70 are in the borderline clinical range, while less than

64 are in the healthy range (Achenbach & Rescorla, 2001). While the CBCL and YSR are

widely used and considered valid measures of psychopathology, some critique their use

for having limited response choices and the inability of the clinician to query responses,

thus limiting the amount of clinical information gathered (Barker & Pistrang, 2004).

UTILITY OF THE BGII 54

Examiners

The researchers were predominantly predoctoral interns participating as part of

American Psychiatric Association predoctoral internship research requirements. Five

research assistants were full-time employees who had undergone extensive training. All

research assistants were under the direct supervision of the principal investigator and

coinvestigators. The raters administered the protocols independently once deemed

reliably trained. The reliability of raters was established over 10 passed trials with a

veteran examiner until a 0.85 coefficient Kappa interrater reliability was obtained

(Findling et al., 2001, 2002, 2005; Youngstrom et al., 2001; Youngstrom et al., 2005).

Procedure

All youth and caregivers presenting for a mental-health intake were invited to

participate in a more detailed assessment during the clinical intake session for the general

clinic. If youth and caregivers agreed to participate (62% agreed; primary reason for not

participating was duration of the study), they were then scheduled to meet with a research

assistant (57% attended). Institutional review of this study was conducted under board

approval from the University Hospitals of Cleveland and the Case Western Reserve

University Institutional Review Board. The West Virginia University Institutional

Review Board approval was sought; the study was granted exempt status because of the

use of archival data (Appendix A). Both caregivers and youth provided written and verbal

consent.

The WASH-U-KSADS-PL+ was conducted individually and separately with the

youth and caregivers. The youth completed all other measures with a separate research

UTILITY OF THE BGII 55

assistant while the caregiver was being interviewed. The caregiver completed all other

measures with another research assistant while the participant was being interviewed. The

BGII drawings were scored according to the BGSS and Koppitz 2 by both the research

assistant who administered the measure and by a research assistant blind to the interview

administration and testing results. No significant difficulties were reported by research

assistants during administration of the BGII. Youth completed the copy phase of the BGII

in 8 minutes and 32 seconds on average (SD = 4.3). As previously reported, the normal

time to complete the drawings in previous research has been found to be between a mean

of 9 to 14 minutes depending on age for the copy phase with standard deviations of 4 to 7

minutes on both the BGSS and the Koppitz 2 for interpretation of results in relation to

errors (Brannigan & Decker, 2003; Reynolds, 2007).

Data Analyses

Data from the ABACB study (Youngstrom, 2006) were analyzed in SPSS v. 16.0.

Data analyses were performed on the BGII GSS scores, BGII Koppitz 2 scores, the

CBCL syndrome and general scales, the WASH-U-KSADS+PL and LEAD results. The

CBCL syndrome and general scales were correlated using Pearson correlations with the

BGSS scores and Koppitz 2 scores to examine whether similar constructs were being

measured. The BGSS scores and the Koppitz 2 scores were compared to diagnoses using

independent sample t tests, ANOVAs, and Receiver Operating Characteristic (ROC).

Diagnoses were categorized into four hierarchical groups: Bipolar Spectrum Disorders,

Depressive Disorder, Behavior Disorders, and Other Diagnoses. For example, if a youth

UTILITY OF THE BGII 56

was diagnosed with Bipolar II and ADHD- Combined type, then he/she would be placed

in the Bipolar Spectrum Disorders group (Youngstrom et al., 2001).

Significant ANOVA results were further explored using post hoc tests; Tukey’s

HSD was used if variances among groups could be considered equal. Games-Howell was

used if variances among groups could not be considered equal. Holm’s step-down

correction procedure was used to compare independent t tests with hierarchical diagnosis

groups.

The evaluator could then assess the BGSS and Koppitz 2 level of specificity and

sensitivity with regard to specific measurement results as well as common childhood

diagnoses by testing different cut-off scores related to those significant ANOVA findings.

If any of the hypothesized results had been found, they could have demonstrated the

utility of the Bender in diagnosis in evidence-based practice (Youngstrom & Duax,

2005). The standard of using significant ANOVA findings in evidenced-based practice

had already resulted in the development of Receiver Operating Characteristic (ROC)

analyses to examine the diagnostic sensitivity and specificity of the assessment measures

(Altman & Bland, 1994b; Frazier, 2006; Jaeschke et al., 1994; Youngstrom, 2006).

The ROC analysis plots sensitivity and false alarm rate (1-specificity) to aid in the

interpretation of scores and can separate cases from those identified as noncases.

Sensitivity is the proportion of individuals identified by the test as positive for the

diagnosis that had been previously positively classified for that diagnosis (Altman &

Bland, 1994a; Choi, 1998). A false alarm is the belief that something is positive but it is

not; therefore an incorrect diagnosis may be assumed when the child does not actually

UTILITY OF THE BGII 57

have the disorder. Specificity is the proportion of individuals identified correctly without

the diagnosis (Altman & Bland, 1994a; Choi, 1998). An ROC plots the sensitivity/false

alarm rate for each score thus producing a curve. A straight line indicates a 50/50 chance

of being given the specific diagnosis and is considered random. Tests that do not

discriminate above the level of random chance are of no use to clinicians (Zweig &

Campbell, 1993). ROC curves examined diagnostic efficiency by comparing the

sensitivity and false alarm rate (1 – specificity) for each score (Altman & Bland, 1994a).

ROC results are determined by nonparametric methods that result in a decimal fraction

that represents the area under the curve (AUROC). An AUROC of .50 would indicate

chance performance.

In evidence-based practice, clinical decision making uses the diagnostic

likelihood ratios derived from ROC analysis to improve the likelihood of a clinically

significant diagnosis. Mathematically, this is a result of the Bayes’ Theorem. Visually a

nomogram can be used to simplify the process in clinical practice (Frazier, 2006;

Jaeschke et al., 1994; Youngstrom, 2006). The use of diagnostic likelihood ratios, ROC,

and nomograms are relatively new to the field of psychology yet have been used in

determining the utility of medical procedures and tests for years (Frazier, 2006).

Diagnostic likelihood ratios were calculated to aid in understanding the clinical

significance. Using Bayes’ Theorem, the diagnostic likelihood ratio is multiplied by

pretest odds (or base rate) to determine the posttest odds. For ease of clinical use, the

odds are presented as probabilities in percent form. If one uses a binary cut point (e.g.,

UTILITY OF THE BGII 58

clustering all clients below a score and all others above a score), then the positive

diagnostic likelihood ratio is sensitivity divided by 1- specificity for the cut point score.

In the current research and prior to statistical analysis, a power analysis was

conducted using G Power (Erdfelder, Lang, & Buchner, 1996). Power of 80% was

calculated a priori for the correlational results. A moderate correctional effect size was

used (r = .30). Results indicated that a sample of 67 clients was required for statistical

utility. However, power of 80% was calculated a priori for the difference between group

results using independent sample t tests with a moderate effect size (Cohen’s d = 0.5).

Results indicated a sample of more than 67 clients was sufficient and therefore an N of 75

or larger was sufficient.

This chapter has detailed the methods and procedures used in this study along

with the rationale for these methods. As stated the data were gathered for approximately

one-year in the duration of a larger 5-year study. The study was designed to fit into the

framework of the larger study and the data collected was archival.

UTILITY OF THE BGII 59

Chapter 4: Results

The present study was conducted to examine the clinical utility of the Bender

Gestalt Visual Motor Test II (BGII) as a diagnostic discriminator for young children and

adolescents. Two different methods for scoring the BGII, the Global Scoring System

(BGSS) and the Koppitz 2 Scoring System, were compared to determine whether the

BGII is an adequate measure for identifying pathology in children at initial intake.

Participants in the present study were administered the BGII during an interview process

to assess clinical diagnostic accuracy among a variety of screening instruments. All data

were converted to T scores with a mean of 50 and standard deviation of 10 for analysis.

Data from the current study were analyzed in SPSS v. 16.0. Data analyses were

performed on the BGII GSS scores, BGII Koppitz 2 scores, the CBCL syndrome and

general scales, the WASH-U-KSADS+PL and LEAD results. The CBCL syndrome and

general scales were compared with the BGSS scores and Koppitz 2 scores using Pearson

correlations to determine whether similar constructs were being measured. The BGSS

scores and the Koppitz 2 scores were then compared to diagnoses using independent

samples t tests, to determine whether similar constructs were being measured. The BGSS

scores and the Koppitz 2 scores were then compared to diagnoses using independent

samples t tests, ANOVAs, and Receiver Operating Characteristic (ROC).

Research Question 1

When compared to a common screening instrument such as the CBCL, the BGII

using the BGSS and Koppitz 2 scoring methods will be useful in the diagnostic screening

UTILITY OF THE BGII 60

of a clinically referred population of children. Results of the BGSS scores are given first

followed by results of the Koppitz 2 scoring. The BGSS scoring of the BGII was

compared to the CBCL using Pearson correlations. The BGSS Copy T scores were not

found to be significantly related to the CBCL Internalizing, Externalizing, and Total

Problems T scores. The Bender GSS Recall T scores, (r = -0.26, p < 0.01, r² = .07) have

a negative relationship to the CBCL Total Problems T scores. Therefore, an increase in

the total number of behavior problems was associated with a decrease in the recall scores.

The BGSS Recall T scores demonstrated a significantly negative relationship to the

CBCL Externalizing T scores (r = -0.23, p = 0.01, r² = 0.05). This indicates that as scores

on externalizing or acting-out behaviors increased child recall scores decreased. The

CBCL Internalizing T scores were found not to be significantly related to the BGSS Copy

T scores (r = 0.002, p = 0.98) or to the BGSS Recall T scores (r = -0.08, p = 0.36). The

Internalizing and Externalizing CBCL scores were significantly correlated to each other(r

= 0.44, p > 0.01, r² = 0.19). Not surprisingly, both the Internalizing and Externalizing

scores were highly correlated with the Total Problems T scores (r = 0.76, r = 0.89, p >

0.01, respectively). Table 1 displays the complete correlation matrix for the CBCL scores

and the BGII using the BGSS scoring system.

Results of the CBCL on the larger study indicate that the clinically referred

groups of children were in the clinically borderline range for emotional problems: CBCL

Total Problems (M = 68.74, SD = 8.82, N = 782), CBCL Externalizing Behaviors (M =

69.63, SD = 9.72, N = 782), and CBCL Internalizing Behaviors (M = 63.48, SD = 10.34,

N = 782, in the larger study). This is between one and two standard deviations above the

UTILITY OF THE BGII 61

mean on the CBCL based on the larger 5-year study results. This is indicative of high

number of symptoms as reported by the parent or primary caregiver at the time of initial

intake.

As the sample group, the BGII, Copy T scores, when adjusted for age were (M =

52.10, SD = 10.85, n=75), and the Recall T scores were (M = 50.98, SD = 11.83, n =75).

These results are in the average range for the BGSS and are not considered clinically

significant. The overall mean differences between the CBCL and BGII are indicative of a

significant difference between the measures. The mean scores between the CBCL and

BGSS scores were compared using a dependent sample t test as they both provide T

scores and are comparable one to the other. The results suggest that the BGSS indicates

on average less visual motor problems in the sample than the CBCL indicates

psychopathology in the sample. However, it is important to note that the BGII was

completed by the child (client) and the CBCL was completed by parent or primary

caretaker.

Pearson correlation coefficients were computed between the BGSS Copy and

BGSS Recall T scores and the CBCL. The CBCL scores and results are presented in

Table 1. The BGSS Copy T score has a moderately high positive relationship with the

BGSS Recall T score, (r = 0.49, p < .01). In Table 1, BGSS Recall scores were

moderately negatively related to the CBCL Externalizing scores, (r = -0.23, p < .01) and

Total Problems scores, (r = -0.26, p < .01). These results suggest a negative relationship

between client’s ability to recall drawings and number of externalizing behaviors

reported by their caregivers as well as total number of symptoms endorsed.

UTILITY OF THE BGII 62

Overall the sample group and the time to complete the BGII show a moderate

positive association with BGII GSS Copy T scores, r =.34, p <.05. Time to complete the

BGII was not significantly related to any of the CBCL T Scores, all p < .05.

Table 1

Pearson Product-Moment Correlation Matrix Between the BGSS Copy and Recall Scores

and the CBCL Main Scores

Copy

Recall

CBCL

internalize

CBCL

externalize

CBCL

total problems

BGSS copy -

BGSS recall

.49

---

CBCL

internalizing

.01 -.09

CBCL

externalizing

-.16 -.23* .44**

CBCL Total

problems

-.16 -.26** .76** .86** ---

*p > .05, **p > .01

Further analyses of the BGSS and CBCL were performed using the syndrome

scales of the CBCL; Table 2 displays the complete correlation matrix for the CBCL

syndrome scales and the BGSS scores. The CBCL syndrome categories were (a)

aggressive behavior, (b) anxiety/depression, (c) attention problems, (d) rule-breaking

behaviors, (e) social problems, (f) somatic complaints, (g) thought problems, and (h)

withdrawn depressed.

UTILITY OF THE BGII 63

The CBCL attention problems scale was found to be negatively related to the

BGSS copy scores, (r = -0.14) and negatively related to BGSS recall scores (r = -.24, p =

.01). The strength of the association of the relationship between CBCL attention

problems and Bender Copy scores was r² = 0.42. The BGSS recall score and the CBCL

thought problems syndrome category were positively related, (r = -.26, p =. 01);

suggesting a relationship between the ability of the child to recall the drawings and

reporting thinking difficulties or reality based behaviors.

The Bender Recall scores were found to correlate to several of the subcategories

of the CBCL. The Bender Recall scores were found to have a weak negative correlation

with the subcategories of anxiety-depression, social problems, thought problems,

attention problems, and rule breaking, all p < .05. Again this suggests a closer connection

between BGSS recall scores and symptom scores than the copy phase of the BGSS.

Further analyses of the BGSS method of scoring with the CBCL were conducted

on the clinical diagnoses scales. Results revealed that BGSS copy scores were not found

to be significant for affective disorder, anxiety disorder, ADHD or oppositional defiant

disorder. Again however, the BGSS recall scores were found to be significant in those

children who met the diagnoses of ADHD using the CBCL findings, (r = .19, p = .05).

This suggests that as BGSS scores increase so do ADHD subscale scores. In other words,

poor performance on the recall is associated with higher ADHD scores. Findings of

BGSS compared to the CBCL clinical diagnoses are found in Table 3.

UTILITY OF THE BGII 64

Table 2

Pearson Correlational Matrix of the BGSS Scores and the CBCL Syndrome Scores

Subcategory

GSS

copy

GSS

recall

CBCL

internalize

CBCL

externalize

CBCL

total

problems

Anxiety/

depression

.07

-.13

.85**

.40**

.68**

Attention

problems

-.14

-.24**

.47**

.55**

.71**

Aggressive

behaviors

-.20*

-.18

.45**

.93

.82**

Social problems -.07

-.17

.61**

.55**

.75**

Rule breaking

.01

-.15

.33**

.86**

.71**

Thought problems

-.13

.26**

-.60**

.53**

-.74**

Withdrawn

behaviors

.09

.13

.77**

.31**

.56**

Somatic

concerns

.04

-.10

.69**

.26**

.50**

*p > .05, **p > .01

UTILITY OF THE BGII 65

Table 3

Pearson Correlational Matrix Comparing BGSS and CBCL Diagnostic Scores

Affective

disorder

Anxiety

disorder

ADHD

ODD

GSS copy

scores

-.09

-.17

-.18

-.07

GSS recall

scores

-.16

-.08

-.19*

.11

*p > .05

To determine whether the Koppitz 2 scoring of the BGII was associated with

psychopathology, Pearson correlational coefficients were computed between the Koppitz

2 and the CBCL results using their respective T scores to determine if there were any

relationships. The initial analysis compared the Koppitz 2 total VMI and total EI scores

with the CBCL scores for internalizing, externalizing and total problems (Table 4). These

findings indicated a weak but significant relationship between CBCL Total Problem and

Koppitz EI scores. No other significant relationships were found between the different

Koppitz and CBCL testing measures.

UTILITY OF THE BGII 66

Table 4

Pearson Correlational Matrix Comparing CBCL and Koppitz 2 Scores

CBCL

internalizing

CBCL

externalizing

CBCL

total problems

Koppitz EI .15

.06

.18*

Koppitz VMI -.08

-.16

-.19

*p > .05

Further analyses comparing the Koppitz 2 findings with the CBCL subcategories

resulted in the following findings as shown in Table 5. First, the Koppitz EI scores were

found to be related to the Koppitz Total Score (r = 0.48, p ≤ .05) and the Koppitz VMI

Scores (r = .31, p ≤ .05). Koppitz Total Scores were found to be correlated to the Koppitz

VMI Scores (r = 0.88, p ≤ .05).

Koppitz 2 Total Scores were also found to be correlated to social problems, (r = -

.23, p ≤ .05), and aggressive problems on the CBCL, (r = -.24, p ≤ .05), Koppitz 2 EI

scores were correlated with social problems, (r = 0.28, p ≤ .05), and attention problems,

(r = .28, p ≤ .05). The Koppitz 2 scoring methods were not found to be significantly

correlated with anxiety-depression, withdrawal, somatic complaints, thought problems,

rule-breaking group, internalizers, externalizers, or total problems groups. This suggests a

connection between the number of Koppitz 2 EI exhibited by the child and the number of

social and attention problems described by the caregiver on the CBCL. Koppitz 2 Total

Scores were strongly related (r = 0.88, p ≤ .05) as previously suspected.

UTILITY OF THE BGII 67

Table 5

Pearson Correlation Matrix Comparing Koppitz 2 Total Score, Total Emotional

Indicators, and Visual Motor Index to the CBCL Syndrome Scales

1 2 3 4 5 6 7 8 9 10 11 12 13

1.Koppitz

total score

2.Emotional

indicators

-.48* -

3. VMI .88

-.31* -

4.Anxious-

depressed

-.07 .19 .01 -

Withdrawn

.22 -.10 .19 .50* -

6.Somatic

complaints

.11 -.01 .18 .43* .41* -

7.Social

problems

.23

.28* -.12 .60* .17 .35* -

8.Thought

problems

-.10 .08 -.09 .55* .20 .49* .61* -

Attention

problems

-.22 .28* -.11 .47* .16 .31* .67* .62* -

10.Rule

breaking

-.09 .09 -.09 .37* .21 .34* .51* .51* .46* -

11.

Aggressive

.24

.19 -.20 .53* .11 .24* .75* .63* .65* .69* -

12.

Internalizing

.00 .09 .06 .83* .75* .69* .51* .51* .40* .38* .42* -

13. External-

izing

-.21 .15 -.17 .51* .16 .27* .69* .64* .63* .84* .95* .42

14.Total

problems

-.22 .22 -.14 .73* .37* .51* .79* .78* .74* .73* .85* .71

.89*

* p < .05

UTILITY OF THE BGII 68

Research Question 2

When compared to the research instrument of the WASH-U-KSADS-PL+, scores

derived from the BGII BGSS and the Koppitz- 2 scoring systems will be found to be

valid predictors of pathological symptom severity among the children studied. Again the

results section is ordered by the two systems used, BGSS first and Koppitz 2 second.

Results of the BGII BGSS were compared to the diagnostic categories of the

KSADS-PL+ using Pearson Product-Moment correlations. The current study compared

the different scoring results with specific pediatric diagnostic categories on the KSADS-

PL+ to determine if the BGII could be useful in identification of pathology. Although

when compared with the KSADS-PL+ diagnoses the BGSS results showed no significant

correlations (Table 6), additional analyses found that within the WASH-U-KSADS-PL+

semiinterview, the ADHD diagnosis was positively correlated with the ODD diagnosis (r

= .27, p = .01, r² = .07).

The Koppitz 2 scoring results were found to be associated with the diagnostic

findings on the KSADS-PL+. The KSADS-PL+ diagnoses were further compared to the

Koppitz 2 Subtotals, VMI and EI using the Pearson Correlational Coefficient analyses as

presented in Table 7. When comparing the Koppitz 2 VMI scoring results with that of the

KSADS-PL+ diagnoses, there were no significant correlations found at the p = .05 level.

However, when comparing the Koppitz 2 Emotional Indicators with different KSADS-

UTILITY OF THE BGII 69

PL+ diagnoses, there was a positive correlation between ADHD diagnoses and the EI

scores (r = 0.30, p <0.01) with a strength of association of r² = .08.

UTILITY OF THE BGII 70

Table 6

Pearson Correlation Matrix Comparing the BGII, BGSS and Koppitz2 Scores and

KSADS-PL+ Symptomology

________________________________________________________________________

1 2 3 4 5 6 7 8

1. Time to complete -

2. BGSS score .34* -

3. BGSS recall .20 .53* -

4. Koppitz2 total score .16 .72* .50 -

5. Emotional indicators .04 -.23* -.33* -.48 -

6. Koppitz2 VMI .25* .82* .51* .88* -.31* -

7. KSADS depression -.10 .02 .14 .14 -.19 .01 -

8. KSADS mania -.06 -.10 -.05 .04 -.18 -.07 .50* -

9. KSADS ODD -.15 -.04 -.21 .04 -.09 .04 -.03 .32

________________________________________________________________________

* p > .05, n= 75.

Table 7

Pearson Product-Moment Correlations of the Koppitz 2 Scores and KSADS-PL+

Diagnoses.

Bipolar /

KSADS

ADHD/

KSADS

ODD/

KSADS

CD disorder/

KSADS

Koppitz VMI

T score

-.05

-.12

-.11

.08

Koppitz EI

T score

-.09

.30**

.03

-.18

*p > .05, **p > .01

UTILITY OF THE BGII 71

Research Question 3

There will be significant relationships among the scoring results of the Bender

Visual Motor Gestalt II and Koppitz- 2 depending on diagnostic categories when related

to final LEAD results. Children with bipolar disorders will differ from those with

unipolar disorder, anxiety disorders, behavior disorders, and ADHD scores derived from

the BGSS or the Koppitz 2 scoring systems.

Using the Holm’s step-down correction procedure, independent samples t tests

were conducted to determine whether there was a significant difference on each of the

separate scales. The scales are (a) BGSS copy scores, (b) BGSS recall scores, (c) Koppitz

2 total score, (d)Total emotional indicators and (e) Koppitz 2 VMI between KSADS-PL+

diagnoses categories. The use of Holm’s step-down correction was preferred because it

maintains alpha at close to .05 overall and protects statistical power more than the

common Bonferroni Correction method. When using this approach, there are four

comparisons being made on each scale based on the following categories. The categories

are (a) bipolar disorder versus all others, (b) ADHD versus all others, (c) conduct

disorder versus all others and (d) ODD versus all others.

The BGSS Copy scores (t (73) = 1.05, p = .29), BGSS Recall Scores (t (73) =

0.07, p = .94), Koppitz 2 Total Score (t (73) = 0.32, p = .78), Koppitz 2 Emotional

Indicators (t (73) = 0.14, p = .10), and Koppitz 2 VMI (t (73) =1.09, p = .42) did not

significantly differentiate between those diagnosed with bipolar (n = 9) and all other

disorders (n = 66). ROC indicated the BGSS Copy score (AUROC = .40), BGSS Recall

UTILITY OF THE BGII 72

score (AUROC = .49), Koppitz 2 Total Score (AUROC = .47), Koppitz 2 Emotional

Indicators (AUROC = .36), and the Koppitz 2 VMI (AUROC = .41) did not predict

bipolar disorder significantly better than chance, at p > .05. The BGSS Copy score (t (73)

= 1.66, p = .10), BGSS Recall Scores (t (73) = 1.69, p = .10), Koppitz 2 Total EI (t (73) =

1.79, p = .07), and Koppitz 2 VMI (t (73) = 1.34, p = .19) did not significantly

differentiate between ADHD (n = 47) and all other disorders (n = 28). On the Koppitz 2

Total Score children and adolescents diagnosed with ADHD (M = 19.47, SD = 8.44) had

a significantly lower scores than all others (M = 24.93, SD = 9.49), (t (73) = 2.56, p =

.01). An error score was created for the Koppitz 2 Total Score. An error score was an

actual score subtracted from the total possible score. Receiver Operating Characteristic

(ROC) analyses were also conducted to determine whether any aspects of the researched

scoring systems could predict any specific diagnoses compared to those not having the

diagnosis in the sample group. This score was created to aid in the interpretation of the

ROC analysis because it moved the ROC curve to the conventional top left instead of the

bottom right of the diagram. Figure 1 displays the ROC curve that demonstrates that the

number of errors significantly predicted an ADHD diagnosis over all other diagnoses

(AUROC = .67) 95% CI = [.54, .80].

Using diagnostic likelihood ratios, sample participants who made more than 23

errors have a positive diagnostic likelihood ratio (DLR) of 2:1. Expressed as a

probability, an odds ratio of 2:1 is 66%. This is calculated by the following equation:

probability = odds/ (1 + odds); using the formula shown with the current data: 2/ (2 + 1)

UTILITY OF THE BGII 73

= .66. This indicates that in a sample in which 50% of the youth have a diagnosis of

ADHD, a positive test result increases the probability of the child truly having an ADHD

diagnosis to 66%, and represents the increased likelihood of an accurate diagnostic

finding when the prior odds are modified by incorporating the diagnostic likelihood ratio

as indicated by Bayes’ theorem to calculate posterior odds. Negative diagnostic

likelihood ratios can also be calculated by taking (1-sensitivity) divided by the specificity

for the cut point score. Clients making less than 23 errors have a negative diagnostic

likelihood ratio of 0.7. Again using the Bayes’ Theorem, prior or base rate odds (1.0) are

multiplied by the DLR (0.7) which equal posterior odds of 0.7, which convert

mathematically to 41% using the formula above .7/ (1 + .7) = .41. This indicates that in

the same sample of which 50% of the youth could have an ADHD diagnosis, a test result

of less than 23 decreases the probability of a child having the diagnosis of ADHD to

41%.

UTILITY OF THE BGII 74

ADHD. 

*Note: The solid diagonal line is chance (AUROC=.5).

Figure 1. ROC of errors made using the Koppitz 2 scoring system to predict.

The BGSS Copy score (t (73) = .99, p = .99), BGSS Recall Score (t (73) =.06, p =

.47), Koppitz 2 Total Score (t (73) = -1.77, p = .08), Koppitz 2 Emotional Indicators (t

(73) = 1.82, p = .08) and Koppitz 2 VMI (t (73) = -.62, p = .53) did not significantly

differentiate between conduct disorder (n = 30) and all other disorders (n = 45). An

ANOVA was performed using the mean of the BGSS Recall Scores across the KSADS-

PL+ groupings of diagnoses found during LEAD. BGSS Recall Scores were found not to

be significantly different with F (3,104) = 0.76, p = 0.97. Due to a nonsignificant

ANOVA and low shared variance among the different measures, additional analyses were

not performed on this relationship.

UTILITY OF THE BGII 75

Using subscores from the Koppitz- 2 scoring systems of the Bender Visual Motor

Gestalt Test II were compared with the final LEAD diagnoses groups to determine if any

significant relationships exist. The ANOVA of the Koppitz -2 VMI -Scores were not

significant. An ANOVA was performed with the Koppitz 2 Total EI scores and were

found to be not significant. As previously shown in Table 5, the unipolar group had

reported fewer emotional indicators than the disruptive behavior group and, though not

significant, showed a difference between the cyclothymic/bipolar group and residual

group. The average difference is only one emotional indicator. However, this difference

of one emotional indicator is the difference between the normal range and the range of

concern (Reynolds, 2006).

Research Question 4

Using hierarchical groups of diagnoses, four one-way ANOVAs indicated that the

BGSS Copy Scores (F (3,71) = 0.56), BGSS Recall Scores (F (3,71) = 0.76), Koppitz 2

Total Score (F (3,71) = 0.46), Koppitz 2 Emotional Indicators (F (3,71) = 1.07) and

Koppitz 2 VMI (F (3,71) = 0.46) did not significantly differentiate sample participants

with (a) bipolar, (b) unipolar depression, (c) behavior disorders, or (d) any other

disorders, at p > .05. Due to nonsignificant ANOVAs and low shared variance among the

different measures, additional analyses were not performed on this relationship to

determine the level of specificity and sensitivity in relationship to clinical decision

making when using the Bender Gestalt II together with the BGSS for ROC. 



UTILITY OF THE BGII 76

Chapter 5: Discussion

In this study, the relationship between two scoring systems of the Bender Visual

Motor Gestalt Test II and two empirically validated diagnostic instruments: the Child

Behavior Checklist and the KSADS-PL+ were evaluated. The primary goal of this study

was to determine the clinical utility of the Bender Gestalt II as a screening instrument in

the initial intake process for identification of psychopathology in children and

adolescents in a clinical setting. Specifically, is this nonverbal measure of visual motor

integration useful as a projective indicator of individual psychopathology in children?

Examination of the data indicates that aspects of both scoring systems for the Bender

Gestalt II, the BGSS and the Koppitz 2 were found to have some limited utility in

identifying psychopathology in children.

Discussion of Hypotheses

The initial hypothesis that the BGII, when compared with the CBCL, would be

useful as a psychometric clinical screening instrument with children and adolescents was

weakly validated by the current findings. These results indicate that when compared to

the CBCL, the Bender Recall scores showed a significant negative correlation with the

CBCL Externalizing scores. That is, children who exhibit a greater number or frequency

of externalizing or acting out behaviors were found to have lower scores on the recall

subtest. Bender Recall scores and CBCL Internalizing scores, reflective of such

disorders as anxiety, depression and other dysfunctions that are not socially disruptive,

UTILITY OF THE BGII 77

lacked statistical correlation. This is expected considering previous studies have

demonstrated that children who exhibit externalizing disruptive behaviors tend to have

more difficulty with short-term memory than those with internalizing disorders

(Kooistra, Crawford, Dewey, Cantell, & Kaplan, 2005; Raggi & Chronis, 2006).

When comparing the CBCL with the BGSS and the Koppitz 2 results, several

potentially useful findings did appear. Although the mean of the Bender results was

within the normal range upon initial review, the individual CBCL mean was in the

borderline clinical range. One possible reason for the difference between means is that

the children completed the Bender Gestalt II themselves, whereas the caregiver was

responsible for the CBCL results. In reviewing a series of reports, Youngstrom (2006)

reported previously that parents are the most accurate reporters of symptoms. It is thus

possible that the Bender is not assessing the same type of symptoms as children report

differently than their caregivers. However, when subcategories of the test results were

evaluated, a level of correlation began to emerge. Bender Recall T Scores were found to

be associated with those of CBCL Externalizing and Total Problem scores. Results

indicate that those with higher Externalizing scores and Total Problem scores found it

more difficult to recall the presented Bender items. These recall items on the Bender

Gestalt II are related to short-term memory and accuracy of drawing (Brannigan &

Decker, 2003). This is likely due to the relationship between attention and short-term

memory, which is a difficulty commonly reported by caregivers and further supported by

the Bender performance on recall tasks.

UTILITY OF THE BGII 78

On the CBCL syndrome scales, several further meaningful relationships can be

identified. Bender Recall scores tended to be negatively related to those scoring high on

ADHD symptoms, as well as those reporting with aggressive behavior problems. This

was similar to the findings of Allen (2005) who found a relationship between youth with

ADHD and low recall memory. Allen further found lower Bender Recall scores were

positively associated with increased scores for rule-breaking behaviors. However, this

was not supported in the current study when compared with the CBCL rule-breaking

subcategory. In the current research, higher BGSS copy and recall scores were found to

be negatively associated with CBCL attention scores. It is not surprising that those with

difficulty paying attention would have a difficult time attending to accuracy of the

drawings and later recalling the drawing to reproduce. Previous researchers found similar

relationships with regard to the BGSS and diagnosis of ADHD and the specific symptom

of impulsivity in children (Allen, 2005; Oas, 1984). Allen found that ADHD-diagnosed

children performed more poorly overall on the Bender Gestalt II than normal children.

Oas had previously found that results for adolescents with impulse disorders were

significantly different from those designated as nonimpulsive on the Matching Familiar

Figures Test, a behavior rating scale. It is important to report that the associated strength

of most of these findings was considered weak but in the hypothesized direction. Since I

did not have a nonclinical group as control and comparison, and this was a consecutive

case series design, the current research findings are preliminary and not meant to be

conclusive.

UTILITY OF THE BGII 79

A finding not previously reported but found in the current research is the strong

correlation of reported increase in thought problems on the CBCL and high BGSS recall

scores. The CBCL defines thought problems as obsessive thoughts, hallucinations,

strange behaviors and atypical sleep patterns. One would think thought problems would

impede the ability to maintain attend and recall. One possible reason for this finding may

be that these children are highly aware of their surroundings or hyper vigilant, but are not

outwardly expressing their thoughts and perceptions, thus resulting in their caregivers

misinterpreting their behavior as odd or thought disordered.

A further hypothesis of this study postulated that when compared to a research

instrument such as the KSADS-PL+, The Bender Gestalt II and the Koppitz 2 scoring

methods results would be found to be valid measures of childhood pathological symptom

severity. The KSADS-PL+ semistructured interview process represents both a qualitative

and quantitative approach to data collection which affords clinicians a wider depth and

scope of diagnoses. However, results from the Bender Gestalt II Global Scoring system

were not found to be correlated with scores derived from the WASH-U-KSADS-PL+.

There were no statistically significant findings on the BGSS when compared with the

WASH-U-KSADS-PL+ for symptoms within the childhood diagnoses groups. As

previously discussed, the KSADS-PL+ allows for more parental input than the forced

choice standardized method of the CBCL and may have contributed to the lack of

relationship between these two variables, as well as the varying results among these

measures when further compared to the BGII scoring systems. Research comparing a

UTILITY OF THE BGII 80

testing measure such as the Child Behavior Checklist and the General Behavior Inventory

with an interview method has been previously reported and the findings have been found

useful in comparing a more subjective clinical impression to an empirically valid

instrument (Youngstrom et al., 2000; Youngstrom et al., 2001, Youngstrom et al., 2005).

As previously discussed, the Koppitz 2 scoring system resulted in several

significant findings regarding individuals diagnosed with ADHD using the WASH-U-

KSADS-PL+ and the CBCL. As a screener for children with possible ADHD, the

Koppitz 2 may aid in clinician decision making and thus be clinically useful. However,

the Koppitz 2 Emotional Indicators (EI) were not found to be of benefit for diagnostic

purposes in my research. As previously noted, the Emotional Indicators are those items

that Koppitz recognized as highly indicative of pathology. Therefore, despite the weak

results in my initial research, the Bender Gestalt II as a screener for ADHD diagnostic

purposes in accompaniment with the WASH-U-KSADS-PL+ semistructured interview

could be an avenue for future research.

A third hypothesis suggested that significant relationships would be found

between the scoring systems of BGSS and the Koppitz 2 in relation to LEAD diagnostic

categories. The final diagnostic categories were reviewed and determined during the

formal LEAD process. These categories were defined previously as part of the bipolar

study and therefore delineated mood disorders into two separate categories, Unipolar/

Depression and Cyclothymic/ Bipolar Spectrum Disorders. The Koppitz 2 and BGSS

were unable to differentiate between the two mood disordered groups presented, although

UTILITY OF THE BGII 81

originally this was hypothesized. In previous research Shapiro and Simpson (1995)

found that clients’ primary psychiatric diagnoses were unrelated to their Bender

performance. Yet when examining results obtained using the Koppitz 2 scoring method,

they found Koppitz error rates to be weakly related to concentration. A more recent study

by Allen (2005) indicated that ADHD-diagnosed children tended to do more poorly on

the Bender Gestalt II than a normal group of children. However, these differences

disappeared when a control variable for intellectual level was introduced. My current

research findings also show that those diagnosed with ADHD have lower scores and thus

could further support Allen’s findings (Shapiro & Simpson, 1995).

The final hypothesis of this study investigated expected differences between the

BGSS and Koppitz 2 scores in a sample of 75 clinically referred children aged 5 to 18.

These two scoring systems were found to be statistically significantly related to each

other on most items, subcategories, and main scoring results with the exception of the

BGSS Recall. Recall is more a measure of memory than drawing accuracy, and therefore

the lack of statistically significant correlation is not surprising in comparison to the other

measures within the scoring systems.

Research by Brannigan, Decker, and Madsen (2004) found a significant

difference in the scoring approach of the Koppitz 2 versus the BGSS, leading them to

describe the Koppitz 2 as “more lenient” in scoring than the BGSS. They noted that

Koppitz 2 focused on specific aspects within a drawing whereas the BGSS scores the

whole gestalt of the drawing. This finding was supported in the current research and may

UTILITY OF THE BGII 82

have resulted in some of the positive findings with regard to the Koppitz 2 and not the

BGSS. However, it is not clear if the scoring systems’ differences were the only factor in

false positive findings or that other factors, such as a small sample size (75 out of 100 or

more), may have contributed to the increase of false positives.

Additional research showed that when compared to the Beery-Buktenica

Developmental Test of Visual Motor Integration- Fourth Edition (VMI-IV), the Koppitz

2 scoring system was less reliable than the BGSS. Nonetheless, the Koppitz 2 remains the

preferred system for analysis of visual motor perception. In my research, I found that

both systems when combined added the most clinical information in regard to visual-

motor perception, as well as clinical understanding of the child’s test taking behaviors

during the administration of the BGII measure.

Limitations of the Study

The sample of this study initially consisted of a clinical group of 115 children.

Several items from other assessments were found to be missing in the 115 individuals

who completed the Bender Gestalt II test resulting in the lower final total sample of 75.

Additionally, there was no control group of children without problems for comparison in

this study. The majority of the subjects completing the protocol were 8 years and above,

limiting the number of younger children for my research. This may have been a further

limitation given the already small sample size. There may also have been an

unwillingness of parents to answer certain items, such as those related to legal issues and

the child’s conduct on the day of the interview.

UTILITY OF THE BGII 83

There are the limitations related to the demographics of the sample. Participants

in the present study were children living in urban Cleveland, Ohio. This may limit the

possibility of generalization of findings to other geographic locations and populations.

The sample was further limited by a lack of racial diversity. Participants were primarily

African American (n = 63), consistent with previous referral patterns at the research

facility (Youngstrom et al., 2000; Youngstrom et al., 2001; Youngstrom et al., 2005).

Results may differ in larger metropolitan areas, more rural settings, or areas with a

different demographic profile. For example, the data raised some questions about the

impact of demographics, but the sample sizes were too small to come to any conclusions

and further study may be warranted in this area. A further limitation is the failure to

include a measure of intelligence within the study, which leaves the contributions of a

potentially important variable unexplored.

Other limitations of this study are related to the clinical nature of the sample. The

present sample consisted of children who were referred for treatment, and met the criteria

for a DSM-IV-TR diagnosis prior to interview. Upon agreement to be in the study,

children and caregivers were both assessed according to the previously reported protocol.

Data were not collected from anyone in the community who did not have a clinical

relationship with the center. This suggests the possibility of a self-selection bias in

addition to the restriction based on clinical status. Additionally, participants were

grouped according to previously established clinical categories based on identifying

children with bipolar spectrum disorders. Although a strict interview protocol was

UTILITY OF THE BGII 84

maintained, the fact that the larger study was geared toward the diagnosis of bipolar

disorder in children makes it impossible to completely eliminate the possibility of

examiner bias.

Another concern in the present study is that the scores on the CBCL were found

to be two standard deviations above the normal average, suggesting that these children

were exhibiting behaviors in the clinical range at the time of interview. Although this

supports the previously stated limitation as a clinical group, it also should serve as a

caution to the reader not to compare these findings to normal children and adolescents.

In addressing the CBCL assessment for use in this study, previous research has found that

the parent report is a good predictor of diagnostic concerns (Youngstrom et al., 2000).

However, clinicians cannot always be certain about parental motivation or accuracy with

regard to the identification of pathology. That is, a parent may exaggerate certain

symptoms or frequency of behaviors while underreporting their perception of the child’s

internal thoughts and feelings. One more way to address this limitation may be to collect

more self-report data from the child clients themselves. However, some studies suggest

that child self-reporting also tends to be lacking in insight and underestimates problem

behaviors (Youngstrom, 2006). Another avenue to address this limitation may be to have

a second reporter complete the CBCL for each child.

Implications for Practice

The BGSS and the Koppitz 2, previously normed for developmental visual motor

integration (Brannigan & Decker, 2003, Reynolds, 2007), were found to be clinically

UTILITY OF THE BGII 85

useful when used with other assessments in diagnosing ADHD in children and

adolescents. However, to be most useful, clinicians should consider the base rate of a

possible ADHD diagnosis within their practices (Frazier, 2006). The use of the BGSS to

evaluate more than visual motor difficulties could be of benefit to clinicians who have

been trained in the use of the original Bender Gestalt (OBG) as a projective measure.

Although this is initial research, the results of comparing the Bender Gestalt II with the

CBCL and a semistructured interview, do suggest that the Bender has some clinical

utility. The findings support the use of the Bender Gestalt II in children and adolescents

as a screening instrument for visual perceptual difficulty, impulsivity, short-term memory

recall, and organizational ability of the individual. The findings in the study do not

support using the BGII as a purely projective measure. The BGII may be able to

distinguish healthy children and adolescents from those with psychopathology as the

OBG did, as previously reported by Bender (1938) and Koppitz (1975) along with other

proponents of the Bender. However, that assertion cannot be made in this study because

all children evaluated in this study were already from a clinical population.

Results of this study indicate that the Bender Gestalt II measures aspects other

than simple visual motor perception and possibly begins to provide insight into the

differentiation between types of psychopathology. Incremental validity improves when

this test is combined with other standardized interview techniques. Therefore, if the

clinician has difficulty in clearly making an ADHD diagnosis, the Bender Gestalt II may

provide further clinical evidence. It was found that those with ADHD had several

UTILITY OF THE BGII 86

significant correlations with the CBCL and WASH-U-KSADS-PL+. Children diagnosed

with ADHD tended to have poorer recall and lower overall Koppitz 2 scores than others.

Although my study is exploratory in nature it hopefully represents the beginning

of more empirically driven research into the utility of the BGII. This investigation,

conducted in the framework of a clinical setting, is intended to have applied research

implications. The clinical setting as a condition of the research may have created some

further limitations. Evidence supporting the use of the Bender for clinical practice as a

diagnostic instrument was not fully achieved within this study, as had been hypothesized.

This study employed two scoring systems of the BGII: the BGSS published in 2003 and

the Koppitz 2 published in 2007. Future researchers might wish to review other scoring

methods that were developed to identify psychopathology using the older version of the

Bender. Further research is needed to support the work of Brannigan (2003), who applied

the previously-developed Hutt scales to the BGII and found it reliable and valid for

personality assessment.

This study only reviewed the BGSS with the Koppitz 2 items for overall

correlation but did not address scoring on individual items within those measures. The

BGII test was selected for this study as a projective measure because it is time efficient

and simple for clinicians to administer. It is not the only projective measure available to

clinicians and it would be beneficial to explore other projective measures as well for

possible increased utility as they relate to empirically standardized measures, such as the

KSADS-PL+ and the CBCL.

UTILITY OF THE BGII 87

It would also be of benefit for follow up research to address incremental validity

of the Bender Visual Motor Gestalt II in relation to other screening tests utilized in the

larger ABACB study (Youngstrom, 2005), such as the General Behavior Inventory or the

Wechsler Intelligence Scale for Children-IV. This would add to previous research by

identifying strengths of the BGII test as a useful measure of client psychopathology. It

also could further address the current hypothesis of the BGII as a useful projective

measure in clinical diagnostic decision making. As the ABACB study has a longitudinal

component with a 5-year follow-up, the Bender Gestalt II could be readministered and

compared within an adult population (Garb, 2003). Therefore, further research could

address situational changes and consistency among the drawings in an individual for this

study over a 5-year time period. These findings could be very valuable in determining the

long term usefulness of the BGII test as a clinical measure.

Conclusions

The field of psychology has evolved and grown significantly over the last 50

years and currently there are many more clinical assessment tools that are attractive to

psychologists because of their perceived clinical usefulness. However, in some cases

these are being pushed aside for measures and checklists that, while useful, may be too

transparent in their questioning style. The publication of a revised Bender Gestalt II test

in 2003 provided this researcher with the opportunity to look at the current clinical utility

of one such historical diagnostic measure still in use.

UTILITY OF THE BGII 88

The use of the Bender Visual Motor Gestalt Test II as a direct measure of internal

mental state was explored as a possible addition to the tools used in reaching a clinical

diagnosis. It may provide an important “missing link” in the current evaluation process

by helping to bridge the gap between parent and child symptom reporting. The current

results suggest that the Bender Visual Motor Gestalt Test II is a less than adequate

screening tool for diagnosing clients with unipolar depression versus bipolar disorders,

but is somewhat more useful for identifying childhood disruptive behavioral disorders.

With the publication of the Bender Gestalt II, the instrument has been standardized with

specific instructions for administration and scoring. The test possesses adequate validity

and reliability as a test of visual motor integration.

As discussed, findings regarding the use of the Bender II as a projective

instrument were mixed depending on which scoring method was used. It appears that

internalizing children versus those considered externalizers or those with high levels of

total problems draw differently in overall quality, resulting in significantly different

scores derived from the Bender Visual Motor Gestalt Test II Global Scoring System and

Koppitz- 2 scoring systems. Thus it may be that the Bender drawings do assess some

internal states of the individual and may reflect some individual Gestalt processes as

originally proposed by Wertheimer and Bender (Bender, 1938). However results did not

support the use of Koppitz Emotional Indicators as a measure of psychopathology.

The BGSS and Koppitz 2 results in regard to significant findings of the KSADS

diagnoses were disappointing and limited. It was found that the only diagnosis

UTILITY OF THE BGII 89

consistently related to BGII performance was the diagnosis of ADHD. It was also found

that Koppitz 2 scores based on error rates were more sensitive than the Bender Global

Scoring system in clinician decision making regarding ADHD. The Global Scoring

System identified Recall Scores to be related to ADHD, but was not found to be as

predictive of that as the Koppitz 2.

In conclusion, the findings of the present study indicate that the Bender Gestalt II

may have use beyond its traditional value as a measure of visual motor ability. Despite

the mixed results in supporting the main hypotheses, the current findings are useful. This

is an initial investigation that suggests that further research on the Bender Gestalt II as a

screening tool for childhood pathology would be valuable.

UTILITY OF THE BGII 90

References

Achenbach, T. M. (1991). Manual for the Child Behavior Checklist/4-18 and 1991

Profile. Burlington: University of Vermont, Department of Psychiatry.

Achenbach System of Empirical Based Assessment. (2006). Achenbach Child Behavior

Checklist. Retrieved from http://www.aseba.org/

Achenbach T. M., & Rescorla, L. A. (2001). Manual for ASEBA School-Age Forms &

Profiles. Burlington, VT: University of Vermont, Research Center for Children,

Youth, & Families.

Aklin, W. M., & Turner, S. M. (2006). Toward understanding ethnic and cultural factors

in the interviewing process. Psychotherapy: Theory, Research, Practice,

Training, 43(1), 50-64.

Allen, R. A. (2005). Utility of the Bender Gestalt –Second edition in the assessment of

attention- deficit/hyperactivity disorder. Dissertation Abstracts International:

Section B: The Sciences and Engineering, 65(11-B) 6035.

Altman, D, G., & Bland, J. M. (1994a). Diagnostic tests 1: Sensitivity and specificity.

British Medical Journal, 308, 1552-1561.

Altman, D. G., & Bland, J. M. (1994b). Diagnostic tests 3: Receiver operating

characteristic plots. British Medical Journal, 309, 188-194.

Ambrosini, P. J. (2000). Historical development and present status of the schedule for

affective disorders and schizophrenia for school age children (K-SADS). Journal

of the American Academy of Child and Adolescent Psychiatry, 39(1), 49-58.

UTILITY OF THE BGII 91

American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental

Disorders (4

ed., text revision). Washington, DC: Author.

Archer, R. P., Maruish, M., Imhof, E. A., & Piotrowski, C. (1991). Psychological test

usage with adolescent clients: 1990 survey findings. Professional Psychology:

Research and Practice, 22, 247-252.

Barker, C., & Pistrang, N. (2004). Quality criteria under methodological pluralism:

Implications for conducting and evaluating research. Retrieved from

http://www.ucl.ac.uk/publications/papers.

Belter, R.W., McIntosh, J. A., Finch, A. J., & Williams, L. D. (1989). The Bender Gestalt

as a method of personality assessment with adolescents. Journal of Clinical

Psychology, 45, 414-423.

Bender, L. (1938). A Visual Motor Gestalt Test and its clinical use: Research

monographs No. 3. New York, NY: The American Orthopsychiatric Association.

(Original work published 1938).

Bigler, E. D., & Ehrfurth, J. W. (1981). The continued inappropriate singular use of the

Bender visual motor gestalt test. Professional Psychology: Research and

Practice, 12, 562-569.

Bowland, J. A., & Deabler, H. L. (1956). A Bender-Gestalt diagnostic validity study.

Journal of Clinical Psychology, 12 (1), 82-84.

Brannigan, G. G., & Decker, S. L. (2003). The Bender Visual- Motor Gestalt Test (2nd

ed.). Itasca, IL: Riverside.

UTILITY OF THE BGII 92

Brannigan, G. G., & Decker, S. L. (2006). The Bender –Gestalt II. American Journal of

Orthopsychiatry, 76 (1), 10-12.

Brannigan, G. G., Decker, S. L., & Madsen, D. H. (2004). Innovative features of the

Bender Gestalt II and expanded guidelines for the use of the global scoring

system. Itasca, IL: Riverside.

Canter, A. U. (1968). BIP Bender test for the detection of organic brain disorder:

Modified scoring method and replication. Journal of Consulting and Clinical

Psychology, 32, 522-526.

Cashel, M. L. (2002). Child and adolescent psychological assessment: Current clinical

practices and the impact of managed care. Professional Psychology: Research

and Practice, 33, 446-453. doi:10.1037/0735-7028.33.5.44

Charman, T., & Baird, G. (2002). Practitioner review: Diagnosis of autism spectrum

disorder with 2-and 3-year old children. Journal of Child Psychology and

Psychiatry, 43, 289-305.

Choi, B. C. K. (1998). Slopes of a receiver operating curve and likelihood ratios for a

diagnostic test. American Journal of Epidemiology, 148, 1127-1132.

Decker, S. L., Allen, R., & Choca, J. P. (2006). Construct validity of the Bender Gestalt

II: Comparison with Wechsler Intelligence Scale for Children -III. Perceptual

Motor Skills, 102(1), 133-141.

DeClercq, B., DeFruyt, F., Van Leewen, K., & Mervielde, I. (2006). The structure of

maladaptive personality traits in childhood: A step toward an integrative

UTILITY OF THE BGII 93

developmental perspective for DSM-V. Journal of Abnormal Psychology, 115,

639-657.

Dryden, W. (1986) Eclectic Psychotherapies: A critique of leading approaches. In J.C.

Norcross (Ed.) Handbook of Eclectic Psychotherapy (pp. 353-375). New York,

NY: Brunner/Mazel.

Elbert, J. C., & Holden, E. W. (1987) Child diagnostic assessment: Current trends in

clinical psychology internships. Professional Psychology: Research and Practice,

18, 587-596.

Erdfelder, E., Faul, F., & Buchner, A. (1996). GPOWER: A general power analysis

program. Behavior Research Method, Instruments & Computers, 28, 1-11.

Fidal, C. A. (2004). Examining abuse indicators on the Bender Gestalt Test. Dissertation

Abstracts International: Section B: The Sciences and Engineering, 64 (8-B),

4003.

Field, K., Bolton, B., & Dana, R. H. (1982). An evaluation of three bender-gestalt scoring

systems as indicators of pathology. Journal of Clinical Psychology, 38, 838-842.

Findling R. L., Gracious, B. L., McNamara, N. K., Youngstrom, E. A., Demeter, A. C.

A., Branicky, L. A., & Calabrese, J. R. (2001). Rapid, continuous cycling and

psychiatric co-morbidity in pediatric bipolar I disorder. Bipolar Disorders, 3, 202-

210.

UTILITY OF THE BGII 94

Findling, R. L., Youngstrom, E. A., Danielson, C. K., DelPorto-Bedoya, D., Papish-

David, R., & Townsend, L. (2002). Clinical decision-making using the General

Behavior Inventory in juvenile bipolarity. Bipolar Disorders, 4, 34-42.

Findling, R. L., Youngstrom, E. A., McNamara, N. K., Stansbrey, R. J., Demeter, C. A.,

Bedoya, D., … Calabrese, J. R. (2005). Early symptoms of mania and the role of

parental risk. Bipolar Disorders, 7, 623-634.

First, M. B., Pincus, H. A., Levine, J. B., Williams, J. B., Ustun, B., & Peele, R. (2004).

Clinical utility as a criterion for revising psychiatric diagnoses. American Journal

of Psychiatry, 161, 946-954.

Frazier, T. (2006). Evidence based assessment of Attention Deficit Hyperactivity

Disorder. Cleveland, OH: Applewood Centers.

Garb, H. N. (2003). Incremental validity and the assessment of psychopathology in

adults. Psychological Assessment, 15, 508-520. doi:10.1037/1040-3590.15.4.508

Geller, B., Zimmerman, B., Williams, M., Bolhofner, K., Craney, J. L., Delbello, M. P.,

& Soutullo, C. (2001). Reliability of the Washington University in St. Louis

Kiddie Schedule for affective disorders and schizophrenia (WASH-U-KSADS)

mania and rapid cycling sections. Journal of Academy of Child and Adolescent

Psychiatry, 40, 450-455.

Gerber, S., Appleton, V., Dykeman, J.C., Sampson, D., & Toews, J. (1994). The vital

balance revisited or the resolution of the Counseling Profession’s identity split.

UTILITY OF THE BGII 95

Journal of Counseling Psychology, 37, 2-14. (ERIC Document Reproduction

Service No. ED373279)

Ghanizadeh, A., Mohammadi, M. R., & Yazdanshenas, A. (2006) Psychometric

properties of the kiddie schedule for affective disorders and schizophrenia-present

and lifetime version. Biomedcentral Psychiatry, 6(10). Retrieved from http://

www.biomedcentral.com/1471-244X/6/10

Gracious, B. L. Youngstrom, E. A., Findling, R. L., & Calabrese, J. R. (2002).

Discriminative validity of the parent version of the young mania rating scale.

Journal of the American Academy of Child and Adolescent Psychiatry, 41, 1350-

1359.

Hamza, T. H. (2008). Meta analyses of diagnostic test evaluation data: Random effects

approaches. Annals of Internal Medicine, 149, 889-897. doi: 978-90-9023002-3.

Horn, W.F., & O’Donnell, J.P. (1984). Early identification of learning disabilities: A

comparison of two methods. Journal of Education Psychology, 76, 1106-1110.

Hothersall, D. (1995). History of psychology (3rd ed.). New York, NY: McGraw-Hill.

Hutt, M. L. (1985). The Hutt adaptation of the Bender-gestalt test (4th ed.). Orlando, FL:

Harcourt Brace Jovanovich.

Imm, P.S., Kim, Y.F., Belter, R.W., & Finch, A.W. (May 1991). Assessment of short

term visual memory in child and adolescent psychiatric inpatients. Journal of

Clinical Psychology, 47, 441-443.

UTILITY OF THE BGII 96

Ivanova, M. Y., Achenbach, T. M., Dumenci, L., Rescorla, L.A., Almqvist, F.,

Weintraub, S., … Frigerio, A. (2007) Testing the eight syndrome structure of the

Child Behavior Checklist in 30 societies. Journal of Clinical Child and

Adolescent Psychology, 36, 405-417.

Jaeschke, R., Guyatt, G. H., & Sackett, D. L. (1994). Users’ guides to the medical

literature: III. How to use an article about a diagnostic test: Section B: What are

the results and would they help me in caring for my patients? Journal of the

American Medical Association, 271, 389-391.

Jellinek, M.S., & McDermott, J.F. (2004). Formulation: Putting the diagnosis into a

therapeutic context and treatment plan. Journal of the American Academy of Child

and Adolescent Psychiatry, 43, 913-917.

Kamphaus, R.W., Petoskey, M. D., & Rowe, E. W. (2000). Current trends in

psychological testing of children. Professional Psychology: Research and

Practice, 31(2), 155-164. doi: 10.1037/0735-7028..31.2.155

Kaufman, J., Birmaher, B., & Brent, D. (2003). Schedule for affective disorders and

schizophrenia for school-age children- Present and lifetime version (K-SADS-

PL): initial reliability and validity data. Journal of American Academy of Child

Adolescent Psychiatry, 36, 980-988.

Keogh, B. (1965). School achievement associated with successful performance on the

bender gestalt test. Journal of School Psychology 3(3), 37-40.

UTILITY OF THE BGII 97

Klein, D. N., Ouimette, P. C., Kelly, H. S., Ferro, T., & Riso, L. P. (1994). Test-retest

reliability of team consensus best-estimate diagnoses of Axis I and II disorders in

a family study. The American Journal of Psychiatry, 151, 1043-1047.

Kooistra, L., Crawford, S., Dewey, D., Cantell, M., & Kaplan, B. J. (2005). Motor

correlates of ADHD: contribution of reading disability and oppositional defiant

disorder. Journal of Learning Disabilities, 38(3), 195-206.

Koppitz, E. M. (1968). Psychological Evaluation of children’s human figure drawings.

New York, NY: Grune & Stratton.

Koppitz, E. M. (1971). Children with learning disabilities: A five year follow-up study.

New York, NY: Grune & Stratton.

Koppitz, E. M. (1975) The Bender Gestalt Test for young children, Volume II: Research

and application, 1963-1973. New York, NY: Grune & Stratton.

Krueger, R., & Finger, M. (2001). Using item response theory to understand comorbidity

among anxiety and unipolar mood disorders. Psychological Assessment, 13(1),

140- 151.

La Fiosca, T., & Loyd, B. (1986). Defensiveness and the assessment of parent stress and

anxiety in parents. Journal of Clinical Psychology, 15, 254-59.

Lengua, L. J., Sadowski, C. A., Friedrich, W. N., & Fisher, J. (2001). Rationally and

empirically derived dimension of children’s symptomology expert ratings and

confirmatory factor analyses of the CBCL. Journal of Consulting and Clinical

Psychology, 69, 683-698.

UTILITY OF THE BGII 98

Lipovsky, J. A., Finch, A. J., & Belter, R.W. (1989). Assessment of depression in

adolescence: Objective and projective measures. Journal of Personality

Assessment, 53, 449-458.

Luebbe, A. M., Radcliffe, A. M., Callands, T. A., Green, D. & Thorn, B. E. (2007).

Evidence-based practice in psychology: Perceptions of graduate students in

scientist–practitioner programs. Journal of Clinical Psychology, 63, 643-655.

McCormick, T. T., & Brannigan, G. G. (1984). Bender Gestalt Signs as indicants of

anxiety, withdrawal, and acting-out behaviors in adolescents. Journal of

Psychology, 118(1), 71-74.

McKeon, P. O., Medina, J. M., & Hertel, J. (2006). Hierarchy of research design in

evidenced- based sports medicine. Athletic Therapy Today, 11(4), 41-45.

Mehlman, B., & Vatovec, E. (1956). A validation study of the Bender-Gestalt. Journal of

Consulting Psychology, 20(1), 71-74.

Miyake, A., Friedman, N. P., Rettinger, D. A., Shah, P., & Hegarty, M. (2001). How are

visuospatial working memory, executive functioning, and spatial abilities related?

A latent variable analysis. Journal of Experimental Psychology: General, 130,

621-640.

Morgan, R. D., Olson, K. R., Krueger, R. M., Schellenberg, R. P., & Jackson, T. (2000).

Do the DSM decision trees improve diagnostic ability? Journal of Clinical

Psychology 56 (1), 73–88. doi: 10.1002/ (SICI) 1097-4679(200001)56:1<73:

AID-JCLP7>3.0.CO; 2-I

UTILITY OF THE BGII 99

Naglieri, J. A., & Pfeiffer, S. I. (1992). Performance of disruptive behavior disordered

and normal samples on the draw a person: Screening procedure for emotional

disturbance. Psychological Assessment, 4(2), 156-159.

Oas, P. (1984). Validity of a Draw-A-Person and Bender-Gestalt as measures of

impulsivity with adolescents. Journal of Consulting and Clinical Psychology, 52,

1011-1019.

Pascal, B. J., & Suttell, G. R. (1952)."Regression" in schizophrenia as determined by

performance on the Bender-Gestalt test. The Journal of Abnormal and Social

Psychology, 47, 653-657.

Perticone, E. X. (1998). The clinical and projective use of the Bender-Gestalt test.

Springfield, IL: Charles C Thomas.

Pilkonis, P. A., Heape, C. L., Ruddy, J., & Serrao, P. (1991). Validity in the diagnosis of

personality disorders: The use of the LEAD standard. Psychological Assessment,

3(1), 46-54.

Piotrowski, C. (1995). A review of the clinical and research use of the Bender-Gestalt

Test. Perceptual and Motor Skills, 81, 1272-1274.

Piotrowski, C., & Keller, J.W. (1989). Psychological testing in outpatient mental health

facilities. Professional Psychology: Research and Practice, 20, 423-425.

Raggi, V. L., & Chronis, A. M. (2006). Interventions to address the academic impairment

of children and adolescents with ADHD. Clinical Child and Family Psychology

Review, 9(2), 85-111. doi:10.2007/s10567-006-0006-0

UTILITY OF THE BGII 100

Reynolds, C. R. (2007). Koppitz developmental scoring system for the Bender-gestalt

test-2nd edition (Koppitz 2) Rater’s manual. Austin, TX: Pro-Ed Inc.

Roid, G. H. (2003). Stanford-Binet intelligence scales (5th ed.) Itasca, IL: Riverside.

Rossini, E. D. (1983). The Bender-Gestalt psychopathology scale: Failure to infer

validity in a school-aged sample. Journal of Personality Assessment, 51, 254-261.

Rossini, E. D. & Kaspar, J. C. (1987). The validity of the Bender- Gestalt emotional

indicators. Journal of Personality Assessment, 51, 254-261.

Shapiro, S. K., & Simpson, R. G. (1995). Koppitz Scoring System as a measure of

Bender –Gestalt performance in behaviorally and emotionally disturbed

adolescents. Journal of Clinical Psychology, 51(1), 108-112.

Smith, D. J., Muir, W. J., & Blackwood, D. H. (2006). Neurocognitive impairment in

euthymic young adults with bipolar spectrum disorder and recurrent major

depressive disorder. Bipolar Disorder, 8(1), 40-46.

Sourander, A., Haavisto, A., Ronning, J.A., Multimaki, P. Parkkola, K., Santalahti, P., …

Almqvist, F. (2005). Recognition of psychiatric disorders, and self perceived

problems. A follow up study from age 8 to age 18. Journal of Child Psychology

and Psychiatry, 46, 1124-1134.

Spitzer, R. L. (1983). Psychiatric diagnosis: Are clinicians still necessary?

Comprehensive Psychiatry, 24, 399-411.

Stewart, H. F. (1957). A note on recall patterns using the Bender Gestalt with psychotic

and non-psychotic patients. Journal of Clinical Psychology, 13, 95-97.

UTILITY OF THE BGII 101

Tolan, P. H., & Dodge, K. A. (2005). Children’s mental health as a primary care and

concern: A system for comprehensive support and service. American

Psychologist, 60, 601-614.

Valderhaug, R., & Ivansson, T. (2005). Functional impairment in clinical samples of

Norwegian and Swedish children and adolescents with obsessive-compulsive

disorder. European Child & Adolescent Psychiatry, 14(3), 164-173.

Wassenberg, R., Max, J. E., Koele, S. L., & Firme, K. (2004). Classifying psychiatric

disorders after traumatic brain injury and orthopedic injury in children; adequacy

of KSADS versus CBCL. Brain Injury, 18, 377-390. doi:10.1080/

02699250310001617325

Wertheimer, M. (1923). Laws of organization in perceptual forms. First published as

Untersuchungen zur Lehre von der Gestalt II, in Psycologische Forschung, 4,

301-350. Translation published in Ellis, W. (1938). A source book of Gestalt

psychology (pp. 71-88). London: Routledge & Kegan Paul.

Wilson, M. S., & Reschly, D. J. (1996). Assessment in school psychology training and

practice. School Psychology Review, 25(1), 9-23.

Woodcock, R. W., McGrew, K. S., & Mather, N. (2001). Woodcock Johnson III Tests of

Achievement. Itasca, IL: Riverside.

Youngstrom, E. (2005). Improving the assessment of juvenile bipolar disorder.

Unpublished Training Manual. (NIH R01 5MH066647)

UTILITY OF THE BGII 102

Youngstrom, E. A. (2006) Training manual for ABACB. Unpublished work for

Applewood Centers, Cleveland, OH.

Youngstrom, E. A., & Duaz, B.A. (2005). Evidenced based assessment of Pediatric

Bipolar Disorder, Part I: Base rate and family history. Journal of American

Academy of Adolescent Psychiatry, 44, 712-717.

Youngstrom, E. A., Findling, R. L., Danielson, C. K., & Calabrese, J. R. (2001).

Discriminative validity of parent report of hypomania and depressive symptoms

of the General Behavior Inventory for juvenile bipolarity. Psychological

Assessment, 13, 267-276.

Youngstrom, E. A., Loeber, R., & Stouthamer-Loeber, M. (2000). Patterns and correlates

of agreement between parent, teacher, and male adolescent ratings of

externalizing and internalizing problems. Journal of Consulting and Clinical

Psychology, 68, 1038-1050.

Youngstrom, E. A., Meyers, O., Demeter, C., Youngstrom, J., Morello, L., Piiparinen,

R..., & Findling, R. L. (2005). Comparing diagnostic checklists for pediatric

bipolar disorder in academic and community mental health setting. Bipolar

Disorders 7, 507-517.

Zweig, M. H., & Campbell, G. (1993). Receiver operating characteristic (ROC) plots: a

fundamental evaluation tool in clinical medicine. Clinical Chemistry 39, 561-577.

UTILITY OF THE BGII 103

Appendix A: IRB Protocol

University Hospitals of Cleveland

Case Western Reserve University IRB Number 01-02-39

Title: Improving the Assessment Process of Children

UTILITY OF THE BGII 104

Proposed IRB Addendum (2007)

Overview:

The proposed amendment would add one measure of developmental visual motor

ability to the ongoing study. The addition of this measure would not increase the length

of the visit for the participating families at all. This was because the parent interview

portion takes substantially longer than the youth interview component. As a result the

addition of the measure filled a gap when the youth would otherwise be waiting for the

parent to complete their portion of the assessment. The introduction of this measure

would provide valuable information regarding the individual functioning of the child and

would form the basis of a doctoral dissertation comparing the KSADS-PL with the

Bender Gestalt Visual Motor Integration Test using two newly reintroduced results. Thus

the addition would add greatly to the validation of clinically driven measurements,

contribute to a limited body of knowledge regarding these measures, and as well have the

potential to fulfill important educational goals.

Rationale:

Using a group of participants already being measured using approved research

methods and deidentified data, a secondary analysis would be performed incorporating

the Bender Visual Motor Gestalt Test II (Brannigan & Decker, 2006). The Bender Gestalt

II is a revision of the original Bender gestalt test developed by Bender in 1938 (Bender,

UTILITY OF THE BGII 105

1938), which had been one of the most used measures in psychological assessments until

the 1990’s (Archer, Maruish, Imhof, & Piotrowski, 1991; Brannigan & Decker, 2006). Its

clinical attractiveness was that it was quick to give, taking less than 10 minutes, was used

in an ice breaker in most psychological evaluations because of its ease on the subject,

found to be a cross cultural, nonverbal measure that had a scoring capabilities for a large

age range (Brannigan & Decker, 2006). The Bender fell out of favor with many

psychologists due to research cautioning the broad use of the Bender for diagnostic

evaluations (Bigler & Ehrfurth, 1981) and low reliability with regard to its use for

projective means (Naglieri & Pfeiffer, 1992).

The reality is that clinicians were in want of effective diagnostic instruments like

the KSADS-PL yet have to manage the time/ cost balance in practice, this study allows

the unique opportunity of comparing the utility of the Bender Gestalt II as a screening

instrument with regard to diagnostic criteria. The study allows the unique opportunity of

a comparison of the clinically efficient measure of the Bender with the diagnostic

reliability of the KSADS-PL. Completing the Bender Gestalt-II measure is not difficult

for the client and typically takes less than 15 minutes total time for the participant, and

would not add any time to the total amount that families spend participating in the current

study. Individual functioning of children is often evaluated to determine level of current

functioning/ability in paper and pencil tasks such as the BVMGT-II. Over the years, such

testing instruments have been compared to various constructs such as achievement levels,

perceptual ability, visual motor ability, developmental capability and various emotional

UTILITY OF THE BGII 106

symptoms related to various diagnostic categories (Bender, 1938; Decker, Allen, &

Choca, 2006; Hutt, 1985; Valderhaug & Ivansson, 2005).

TILITY O

To:

Fro

Date

Subj

Trac

Title:

com

Than

[boar

sub

proje

2004

Invol

cont

Furth

In di

deid

nam

there

cont

Than

Boar

Lette

THE BGII

Expedite

: W

ct: N

ing #: H-

The Clini

ared with t

k you for y

dname] In

ission was

t does no

Office for

ing Code

ct the IRB

ermore, it

cussions

ntified bef

to the dat

ore not hu

ct the IRB

k you.

Designe

Sent By:

- IRB Pr

nick, Roy

U Office

nday, Ma

action re

20822

al utility o

he KSADS

ur submis

titutional

reviewed.

constitute

uman Re

Private In

office for a

as deter

ith the PI,

re the PI r

and ther

man subje

at (304) 2

: Ast, Lilo

st, Lilo, 3/

IRB Proto

tocol - N

f Researc

ch 24, 200

uired

the Bend

-PL+ in chi

ion to the

eview Boa

As describ

human su

earch Pro

ormation

copy of th

ined that:

it is clear t

ceives it.

fore no id

t research

3-7073.

24/2008 6:

col

n Human

Complian

r Gestalt II

ldren 5 to

West Virgi

rd. On [Re

ed, it was

ject resea

ections G

r Biologica

guidance

at the dat

here is n

ntifiable in

. If you ha

03 PM

ubjects

and the K

ia Univer

iew Date

etermined

ch, per th

idance on

l Specime

to be ana

way for th

ividual. T

e any que

107

esearch

ppitz 2

ity

ot Found]

that your

10 Augus

Research

s. Please

yzed will b

e PI to ma

e study is

tions, ple

, the

ch a

UTILITY OF THE BGII 108

Appendix B: CREC Program Notice of Certification

Linda Marnic,

Congratulations! You are Core Certified in the Continuing Research Education

Credit (CREC) Program managed by Case Western Reserve University (Case).

To obtain Core certification you passed the online CITI Core Training course.

Core Certification means that you would have the ability to request review of

proposed human subject research proposals by Institutional Review Boards at

the following institutions: University Hospitals of Cleveland, The Metro Health

System, and Case Western Reserve University.

Certification in the CREC Program also means that you have met the NIH

educational requirements for the involvement of human participants in research

for Key Personnel. This certification is valid for 3 years and would expire on

11/15/2009. To be Re-Certified after this date you must obtain 12 CRECs before

your Core Certification expires. The following is the URL to our educational web

site outlining the training options available for continuing certification as well as

provide instructions for obtaining information on your current CREC status by

looking up your account information on SPRIDERWEB, Sponsored Projects

Information and Data Entry and Retrieval Website.

http://ora.ra.cwru.edu/orc_education.asp

You may view a printable “Certificate of Achievement” by going here. Please

maintain this document for your records. Copies of this document may also be

submitted to sponsors to indicate compliance with human subject education

requirements.

I hope this information is helpful. Please feel free to contact me if I can be of

further assistance.

The CREC Program

Sears 657, CWRU

216.368.6925

[email protected]

UTILITY OF THE BGII 109

Appendix C: Administration Manual for ABACAB

Bender Gestalt II Administration Manual for ABACAB

I. Who Administers the BG II?

Either the Rater 1 or the Rater 2 with appropriate training would administer the test. It

would be decided on a team basis by the raters.

Suggested time of testing: If Rater 2 administers the standard protocol, and then the

BG II should be administered prior to the questionnaires. If Rater 1 administers the

protocol, then the BG II should be administered at the end the KSADS if time

permits.

If the family is finished before 3:30 pm, the BG II must be administered before

finalizing with the family.

II. What is needed to administer the BG II?

The BG II stimulus cards, BG II observation form (ABACAB version), a motor test, a

perception test, 2 pencils with erasers, 10 sheets of blank printer paper, and a time

keeping device (stopwatch or a watch or clock that had seconds).

III. What is the administration order?

The official and only acceptable administration order is: copy, recall, motor, and

perception. Each phase occurs immediately after the previous phase. There are no

breaks between phases, unless absolutely necessary (this should be marked).

IV. How do I administer the Copy test?

Place the cards in order; design side down. Place a blank sheet of paper vertically and

a pencil in front of the child. Read the following directions:

I have a number of cards here. Each card has a different drawing on it. I would

show you the cards one at a time. Use this pencil to copy the drawing from each

card into this sheet of paper. Try to make your drawings look just like the

drawings on the cards. There are no time limits, so take as much time as you

need. Do you have any questions? Here is the first card.

Show the child the first card and mark the starting time (include seconds). Administer

cards 1 through 13 in order to children below age 8. Administer cards 5 through 16 in

order to children above age 8. Children may erase and use more than 1 sheet of paper.

Do not allow the children to touch the stimulus cards or draw/doodle nontest figures

on the paper.

If the examinee becomes discouraged say:

UTILITY OF THE BGII 110

Do the best you can.

If the examinee asks where to start drawing any figure, say:

Begin wherever you like.

When complete, record the time finished (include seconds). Label the child’s

sheet: COPY SHEET and TOP on the top of the page as started on by the child.

Things to be marked on the Observation sheet by the administrator:

1) Direction and order of drawing

2) Describe any counting that occurs

3) If any of the test-taking observations occur on more than 2 items, check the

box.

4) Mark the tilt of the paper.

V. How do I administer the Recall Phase?

Immediately after the Copy phase, give the child a new sheet of paper placed

vertically. Read:

Now, I want you to draw as many of the designs that I just showed you as you

can remember. Draw them on this new sheet of paper. Try to make your

drawings just like the ones on the cards that you saw earlier. There are no time

limits, so take as much as you need. Do you have any questions? Begin.

Begin timing (include seconds). Stop timing when child finishes all figures or does

not recall any more designs after 2 minutes. Record end time (include seconds). Label

the top of the sheet: Recall Sheet and Top (for top of sheet from child’s beginning).

On the Observation sheet, mark the order in which the items were recalled.

VI. How do I administer the Motor Test?

Immediately after the recall phase, hand the child the BG II Motor Test. Say:

For each item, start with the largest figure. For each figure draw a

line connecting the dots without touching the borders. Do not lift the pencil,

erase, or tilt the paper while drawing. Try the sample item. Do you have any

questions? Now you try it.

Make sure the child completes all items. Repeat directions as needed.

UTILITY OF THE BGII 111

VII. How do I administer the Perception Test?

Immediately after the Motor Test, hand the child the BG II Perception Test. Say:

Look at this picture (point to the design in the first box). There is another

picture that looks just like it in this row (run finger across the first row). Circle or

point to the picture that looks just like this one (point at design in first box).

If the child needs assistance for any item say:

Which one of these pictures looks like this one? (Point at

picture in first box of row)

If it takes the child more than 30 seconds, then say:

Let’s try the next one. Write an S next to any skipped items.

VIII. How do I score the BG II Copy and Recall

Phases?

The scoring system is 0 – 4 (no resemblance to nearly perfect). Please

score according to the following pictures for each item. If uncertain refer to

Bender Gestalt Manual for further pictures.

**Instructions for this manual were from the Bender Gestalt II Manual (2006).

Appendix D: Demographic Form

UTILITY OF THE BGII 112

Demographic Forms

CONTACT INFORMATION Subject #:______________

Child’s Name_______________________________________________

First Middle Last

Nickname: ___________________ Child’s Date of Birth: -

_____/_____/_______

Address:

________________________________________________________

_______________________________________________

Parent/Guardian’s Name:

________________________________________________

Relationship of guardian to child: ____________________________________

Address:

________________________________________________________________

(if different from children) Street/Apt# City State

Zip

Phone: ( ) ________ - __________ Work Phone: ( ) ________ -

____________

Cell Phone: ( ) _______ - ___________ User of Cell:

______________________

E-mail address: _______________________________

If applicable, other parent/guardian’s name:

___________________________________

Relationship of other parent/ guardian to child:

_________________________________

Other parent/guardian’s home address:

_______________________________________

_________________________________________________________

City State Zip

Home telephone (if different): ( ) ________ - ____________

Work Phone: ( ) ________ - ____________

Cell Phone: ( ) _________ - ____________ User of Cell:

_____________________

Emergency contact person: ______________________________

UTILITY OF THE BGII 113

Relation (if known): __________________________

Phone: ( ) ________ - ___________

Alt. Emergency contact person: ______________________________

Relation (if known): __________________________

Phone: ( ) ________ - ___________

Date of Assessment: ________________________

Rater 2 Interview with Primary Caregiver

ACI ID____________

Date of Review ________________

Employ these codes on the pages that follow:

UTILITY OF THE BGII 114

Race/Ethnicity

0 Native American or Alaskan native

1 Asian/Pacific Islander

2 Black/African American, not of Hispanic origin

3 Latino/Hispanic

4 White/Caucasian, not of Hispanic origin

5 Other

Highest Level of Education Completed:

1 Elementary school (less than 7 years of school)

2 Junior high school (7-9 years)

3 Partial high school (10-11 years)

4 high school graduates (includes G.E.D.)

5 1 to 3 years of college, business or trade school

6 College or university graduate (four year college

graduate)

7 Graduate school or professional training (at least 1 year)

8 Completed graduate school

Occupation:

0 Unemployed

1 Menial Service Worker, Farm Laborer

2 Unskilled Worker

3 Machine Operator, Semiskilled Worker

4 Smaller Business Owner, Skilled Manual Worker,

Craftsmen, Tenant Farmers

5 Clerical and Sales Workers, Small Farm and Business

Owners

6 Technicians, Semiprofessionals, Small Business

Owners

7 Smaller Business Owners, Farm Owners, Managers,

Minor Professionals

8 Administrators, Lesser Professionals, Proprietors of

Medium Businesses

9 Higher Execs, Proprietors of Large Businesses, Major

Professionals

Annual Income

1 $0 – 4, 999 7 $40,000 - $49, 000

2 $5,000 - $9, 999 8 $50,000 - $74, 999

3 $10,000 - $14, 999 9 $75,000 - $99,999

4 $15,000 - $19,999 10 $100,000 - $149,999

5 $20,000 - $29, 999 11 $150, 000 - $200,000

6 $30,000 - $39,999 12 More than $200,000

11 Mother

12 Maternal grandmother

13 Maternal grandfather

14 Aunt - mother’s sister

15 Uncle - mother’s brother

16 Sister - shared mother

17 Brother - shared mother

19 Other - biological mother’s relative

21 Step-mother

22 Foster mother

23 Adoptive mother

24 Maternal GM’s sister (Great Aunt)

25 Maternal GM’s brother (Great Uncle)

26 Maternal GF’s sister (Great Aunt)

27 Maternal GF’s brother (Great Uncle)

31 Father

32 Paternal grandmother

33 Paternal grandfather

34 Aunt - father’s sister

35 Uncle - father’s brother

36 Sister - shared father

37 Brother - shared father

39 Other - biological father’s relative

41 Step-father

42 Foster father

43 Adoptive father

44 Paternal GM’s sister (Great Aunt)

45 Paternal GM’s brother (Great Uncle)

46 Paternal GF’s sister (Great Aunt)

47 Paternal GF’s brother (Great Aunt)

51 Female significant other => 1 yr with family

52 Male significant other => 1 yr with family

53 Female SO < 1 yr with family

54 Male SO < 1yr with family

55 Full bio sister

56 Full bio brother

57 Step-sister (unrelated to either of proband’s BIO

parents)

58 Step-brother (unrelated to either of proband’s

BIO parents)

59 Random relative NOS

78 Female nonrelative NOS

79 Male non0relative NOS

96 Missing

97 N/A

98 Unknown

99 Refused

UTILITY OF THE BGII 115

Appendix E: Consent Forms

TILITY O

THE BGII

116

TILITY O

THE BGII

117

TILITY O

THE BGII

118

TILITY O

THE BGII

119

UTILITY OF THE BGII 120

Appendix F: Statistical Findings

Statistical findings /SPSS 16.0

Sample Demographics.

Age, years

Median

Mean

Minimum-Maximum

11.92

2.56

8-17

Number of Diagnoses

Median

Mean

Minimum-Maximum

2.76

1.42

0-8

Gender, n (%)

Boy

Girl

48 (64)

27 (36)

Ethnicity, n (%)

White

Black

Other

8 (11)

63 (84)

4 (5)



UTILITY OF THE BGII 121

Author’s Note

This work is in completion of a Ph.D. dissertation for Linda R. Marnic, Department of

Counseling Psychology at West Virginia University. Professional colleagues that were

integral to this work include Eric Youngstrom, Department of Psychology, University of

North Carolina at Chapel Hill; Oren Meyers, Department of Psychiatry, Case Western

Reserve University, Cleveland, Ohio; Andrew Freeman, University of North Carolina at

Chapel Hill; Heather Marcinek, Case Western Reserve University; Frank Ezzo,

Department of Psychology, Applewood Centers, Cleveland, Ohio.

The research was supported in part by a grant funded by the National Institute of Mental

Health in coordination with a study of Bipolar Disorder in Children.

Correspondence regarding this research may be sent to Linda R. Marnic, who is now in

private practice at Family Matters, P.O. Box 490, Lost Creek, West Virginia 26385

E-mail: lm[email protected]

John H.

Hagen

Digitally signed by John H. Hagen

DN: cn=John H. Hagen, o=West

Virginia University Libraries,

ou=Acquisitions Department,

[email protected],

c=US

Date: 2010.12.08 15:41:13 -05'00'