A preliminary study examined the interrater reliability, factor structure, and convergent validity of the Haak Sentence Completion (in which the respondent is given a sentence stem and asked to provide the remainder of the sentence) using a three-point scoring system. The subjects were 100 boys ranging in age from 6 to 11 years who were randomly selected from a roster of students who had been referred for evaluation of emotional disturbance. All the boys came from three elementary schools within a suburban school district; all the boys were white and upper-middle class. Subjects were administered the Haak Sentence Completion while their parents were given the Child Behavior Checklist (CBCL) to fill out and return. The scores from the parent CBCL (the parents' report of a child's problem behaviors) were then correlated with the item total score and factors of the Haak (the child's self-report of perceptions). Results revealed that the quantification of sentence completion tests is feasible and useful. The findings suggest that the quantification of such measures makes them useful as research tools to generate self-report data and to investigate hypotheses related to factors that contribute to emotional disturbance. Data generated from such measures can be used for screening, placement, and intervention with children suspected of being emotionally disturbed. Further psychometric investigation of sentence completion techniques in support of this preliminary study appear to be called for. (Two tables of data are included; 19 references are attached.) (KEH)