Evaluating Adolescents' Popularity with the Draw-A-Friend Test

Abstract

While projective drawings have been widely used to evaluate children in schools (Goh & Fuller. 1983) and aggressive traits o n clinical referral (Handler & Mclntosh, 1971) , techniques and methods are not available for the evaluation of popularity of adolescents among peers. T o investigate the potential of such an instrument five college-age smdents ( 2 men, 3 women of M age 19 yr., range 17 to 22 yr.) involved in a life skills program were administered the Draw-a-Friend Test, specifically designed for this inquiry. Clients were asked to respond to the following statement: a picture of you and a friend doing something. Draw the best picture that you can. Clients were then asked to respond indicating the nature of the activity, identity of the friend, age, position in the picmre, and future events. The life-skills counselor ranked the clients on popularity based on his observation of interactions of acceptance among peers during group sessions. Three doctoral students in school psychology next ranked the clients, based on their evaluation of the drawings and descriptions. All three judges had extensive experience as school psychologists and training in personality assessment. They were asked to use a holistic approach or over-all impression in ranking subjects from least to most popular. Judges' and counselors' rankings were then correlated by the Spearman rank-order procedure: for Judge 1 x Counselor p 1.0, Judge 2 X Counselor p 1.0, Judge 3 x Counselor p .26 ( p > .05). While interjudge reliability (p3.r 1.0, p .05, p3.3 .26, p > .05) indicates disagreement, the values suggest that the procedure may be useful in an initial evaluation of social desirability among adolescents. Further research must delineate the specific psychometric properties of this instrument.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 2 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.