Glen E. Roberts, Ph.D.

Test Developed:
Roberts Apperception Test for Children

roberts.gif (5107 bytes)

I was drawn to psychological testing in graduate school and particularly to projectives in attempting to understand how the child feels and how feelings are expressed in behavior. It seems basic that the therapeutic process begins with the assessment in determining the dynamics of the child, as related to the presenting problem.

During my internship in a community child clinic, I became aware of the need to evaluate the attitudes and relationships with family members in assessing the personality functioning of children. Although we used the TAT, CAT, and the Rorschach, the nature of the family interactions were not clearly and thoroughly evaluated. The therapeutic work with the children helped to identify the significance of involving the family rather than focusing on the child alone. I decided to develop a thematic test to assess the family relationships for my dissertation upon my return to graduate school. As the enormity of such a task became apparent, I compared different types of stimulus figures for my dissertation as the first step in developing such a test. Four types of stimulus figures—human children, human adults, animals, and stick figures—were presented to children of three age groups in four different thematic situations. It was concluded that human children and human adults represented the figures that elicited the most projection. It was further concluded that situations were a significant variable in eliciting projection and should be thoroughly researched for the basis of a thematic test for children.

Influences on Definition of Construct

Several years after graduation, I began the process of collecting pictures that pulled for the family attitudes and relationships and also for a variety of emotional reactions and situations involving school and peers. In my practice in two clinic settings, I explored the responses my child clients produced to hundreds of pictures from magazines, newspapers, drawings, and so on. Twenty-three situations were identified as representing the important areas in a child’s life and were drawn in line drawings for consistency of the stimulus. Boys and girls in an elementary school in four age groups—six, eight, ten, and twelve—were shown the twenty-three pictures. Each picture was evaluated by a combination of scores and ratings. Sixteen pictures with the highest ratings were chosen to represent all of the interpersonal arenas and emotional states. The final test consisted of eleven cards depicting either boy or girl figures and five cards that were administered to both boys and girls.

The test was administered for several years to children at our child clinic and interpreted primarily by identifying

significant themes and unusual responses. Our clinic began training psychology interns, and significant contributions were made by several interns. Dorothea McArthur compared RATC cards with TAT and CAT cards with boys in two age groups and concluded that the RATC was superior in eliciting nonstereotyped responses. Later she co-authored the manual for the test and helped in the process of training interns to score the protocols of well-adjusted children and children who were clients of the agency for a validation study. Tom Muha contributed to the development of the scoring system in his study of comparing the responses to cards by nonclinic families and clinic families. He found that the nonclinic families projected more positive and helpful responses and that they tended to end the stories with positive outcomes. These findings provided the basis for the scores of Support and Resolution and were significant in developing the adaptive scores.

Test Development Process

The most interesting and significant aspect of developing the test was the process of constructing the norms. The scoring system was still in a developmental stage, and I experienced the same difficulties in creating a psychometric scoring system as had existed for the TAT and other thematic tests. Well-adjusted children ages 6 to 15 were identified by teachers in three school systems to serve as the population for the norms. When the test was first administered to 6- and 7-year-olds, it became obvious that these children produced stories quite different from the clinic children. In addition to the differences in behavior in terms of maintaining focus, staying on task, and controlling inappropriate behavior, the children tended to end the stories with positive outcomes and include a great number of supportive and positive interactions among the figures. The scores of Support and Resolution that had been identified by Muha as characteristics of nonclinic families were occurring in the stories of well-adjusted children of all ages in contrast to the projections of clinic children. Children of ages 8 and older included some description of process in resolving their stories instead of just positive closure utilized by the 6- and 7-year-old children. The idea of a developmental factor in the resolution of stories became obvious. The outcomes to the stories produced by the adolescents included more elaborate description of the process and often involved a generalization for coping in the future. The norming procedure identified the need for norms for different age groups and provided some surprises in how differently the well-adjusted children process and organize their stories.

Aspects of Publication Process

The publication process forced me to learn patience because it was a year from the time the manual was accepted to the time it actually appeared in print. In the process I had to compromise and learn to be nondefensive and open to feedback by the editors and research staff at the publishers, but I realized later that the publisher’s staff contributed greatly to the final product. The expertise of the staff improved all aspects of the test better than we could have achieved alone. The review and input by colleagues has been invaluable and constitutes a necessity in such a process.

Other Comments

The RATC seems to be relatively culture-free since children from different ethnic groups who are well-adjusted appear to demonstrate the same characteristics as identified in the norms, although content may vary from culture to culture. The children of different ethnic groups are able to identify the family relationships and the emotional and situational pull of the cards. How the child reacts to the emotional state or the situation provides us with important assessment data regardless of the influence of the culture. Since the publication of the RATC and the manual, the need for more detailed guidelines for scoring and interpretation has resulted in the publication of a handbook. This handbook was developed after years of clinical application and research and provides a thorough description of the scoring system with numerous samples of themes for each card, each scale, and each clinical indicator. Case studies are included with scoring and interpretation described in detail. Some clinicians have reported to me that they include the RATC in their assessment battery but do not utilize the scoring system. I strongly recommend learning the scoring system because the norms can be employed to provide a foundation to make a valid judgment in comparing the production of a clinical child to the standards of well-adjusted children of the same age group.

Perspective Pros and Cons

One of the pros in being a test developer has been the opportunity to apply what I learned in graduate school using the scientific procedure in research as it related to my clinical practice. The door has been opened to continue research with the test by graduate students, and the possibilities seem endless. Another pro is the opportunity to continue learning about assessment and personality functioning with the RATC in teaching and supervising graduate students as they test hundreds of children in our child clinic. There is great satisfaction in my role in teaching the test to mental health professionals and knowing that the instrument will contribute to a more comprehensive understanding of the child and to a meaningful therapy plan. There are few cons, but the prominent con is the position of being vulnerable in publishing and facing the criticisms of the psychology community of why I didn’t do something differently. Another con is the limitation in time and finances to expand the research and pursue idealistic studies such as testing groups of children in every major culture in the world.


The relationship and interaction between clinical practice and research has been one of the most rewarding aspects of the test development process. My advice to prospective test developers is to work from observations and needs in clinical practice to develop instruments that have meaning and applicability. Since an armchair conceptualization has serious limitations in determining how people really react and function, subject your ideas to rigorous research to provide guidelines for valid development. Several of the adaptive scales for the RATC were determined by research comparing the responses of well-adjusted children with those of clinic children and identified characteristics that were opposite to what I had previously theorized. Be adaptive to incorporate the research results, even if unexpected, into your theories and provide groundwork for new theories. Be open and nondefensive to the editing and input of the publisher’s staff. Be willing to allocate unlimited time to perfect your work.


McArthur, D. S. (1976). A comparison of the stimulus influence of three thematic projective techniques with children. Unpublished doctoral dissertation, California School of Professional Psychology.

McArthur, D. S., and Roberts, G. E. (1989). Roberts Apperception Test for Children: Manual. Los Angeles: Western Psychological Services.

Muha, T. W. (1977). A validation study of the Roberts Apperception Test as a measure of psychological dysfunction in families. Unpublished doctoral dissertation, California School of Professional Psychology.

Roberts, G. E. (1958). The effect of stimulus variation on the projection of children. Unpublished doctoral dissertation, Purdue University.

Roberts, G. E. (1994). Interpretive handbook for the Roberts Apperception Test for Children. Los Angeles: Western Psychological Services.

Return to Index 

[About this Book]  -  [Developer Profiles]  -  [Web Links]  - [Measurement Forum] -  [Further Information]   -  [Home]

Copyright ©2001 The McGraw-Hill Companies. Any use is subject to the Terms of Use and Privacy Policy. McGraw-Hill Higher Education is one of the many fine businesses of
The McGraw-Hill Companies, Inc.
Corporate Link