Date of Award


Document Type


Degree Name

Doctor of Philosophy


School of Education


Educational Psychology, Ph.D.

First Advisor

Conrad A. Reichert

Second Advisor

Peter Blitchington

Third Advisor

Bernard M. Lall


Problem. There are very few objective instruments available for use in the psychological diagnosis of emotional impairment at the elementary-school level. The determination of emotional impairment has often been based on clinical judgment and subjectivity. In addition, psychological instruments have been attacked from the standpoint of reliability and validity. Major concerns relate to the seriousness of misclassification and the limited success-rate of rehabilitation efforts. The sentence-completion method is popularly used by clinicians. However, the available incomplete-sentence blanks used at the elementary-school level rely almost entirely on face validity. Public Legislation 94-142 and the Rehabilitation Act of 1973 specifically require that psychological instruments be validated for the purposes for which they are used. Therefore the purpose of this investigation was to develop and validate an objective sentence-completion method. The ability of theinstrument to discriminate among special education groups, the sensitivity of theinstrument to long-term treatment effects, and the consistency of scores produced by school psychologists using the objective scoring device were the important factors analyzed.

Method. Based on a review of the previous research done on this method of personality study, an initial list of sentence-completion stems was developed. A more refined list ofstems was developed through two field tests using the Chi-square method of analysis to determine which stems exhibited a tendency to discriminate between the emotionally impaired and regular-education youngsters. Feedback from test administrators pertaining to language comprehension difficulty, as well as analysis for disproportionate numbers ofresponses were also useful guides to final stem selection. The final form consisted oftwenty stems. A personality adjustment scale of Negative, Denial, Neutral, Acceptance, and Positive was devised and the quantitative values 1,2,3,4, and 5, respectively, were assigned. The final administration to determine the ability of the instrument to segregate groups was conducted on emotionally impaired, educable mentally impaired, learning impaired, and regular-education youngsters and analyzed by a nonparametric sign test of differences between medians. Tests of inter-scorer reliability using school psychologists trained on thescoring system were conducted. In addition, the changes in performance on the instrument were compared to personality-adjustment changes over a one-year period as evaluated by an experienced teacher of the emotionally impaired.

Results. Inner-city special-education groups were placed on the Personality Adjustment Continuum in the order: emotionally impaired--2.3, educable mentally impaired--2.6, learning impaired--2.7, regular education--3.2. A statistically significant difference in performance on the instrument was seen between all special-education groups and regular-education youngsters (p < .05). Six correlation coefficients of inter-scorer reliability, ranging from .96 to 1.00, were obtained by school psychologists using the scoring device, suggesting that theinstrument can be scored with relatively high consistency. The instrument was able to measure significant (t = 5.08, dt = 7, p < .05), long-term, personality-adjustment changes which correctly corresponded to teacher evaluation.

Conclusion. The instrument produced responds to personality-adjustment changes and can be consistently scored. The instrument was able to distinguish special-education from regular-education pupils. However, special-education youngsters in this population appear to be similar in emotional status even though they are statistically significantly different from regular-education youngsters. The Bacchus Sentence-Completion Method, produced from this investigation, represents a major tool that can be used in the non-biased assessment of emotional impairment at the elementary-school level.

Subject Area

Personality tests, Emotions in children

Files over 3MB may be slow to open. For best results, right-click and select "save as..."