Construct validation of self-concept in the Rorschach inkblot test

Date of Publication


Document Type


Degree Name

Doctor of Philosophy in Counseling Psychology

Subject Categories

Cognition and Perception


Br. Andrew Gonzalez FSC College of Education


Counseling and Educational Psychology

Thesis Adviser

Gundelina A. Velazco

Defense Panel Chair

Naomi R. Ruiz

Defense Panel Member

Rose Marie Salazar-Clemeña
Natividad Dayan
Imelda V.G. Villar
Salud P. Evangelista


This study conducts a construct validation of self-concept in the Rorschach Test.The descriptive quantitative and qualitative design employed in the study enabled the investigator to describe the occurrence of the variables, the underlying dimensions in the set of variables or relationship between or among the variables.The subjects involved in the study were 100 Filipino males and females belonging to the middle adulthood stage and ranging from 34-60 years. At random, these 100 individuals who belonged to contrasted groups and were willing to participate in the study, were administered the Tennessee Self-Concept Scale. Fifty of them belonged to white collar professions (administrators, deans, etc.) and 50 came from the blue collar jobs (drivers, janitors, etc.). With reference to the study conducted by Gary (1973), 10 respondents with high self-concept were chosen from the former group based on the Tennessee Self-Concept Scale. Ten respondents with low self-concept were also chosen from the latter group based on the same instrument.The Rorschach test was administered to the 20 subjects belonging to the contrasted groups.Two instruments were used to gather the needed data: the Rorschach Test (RT) and the Tennessee Self-Concept Scale (TSCS).

Two types of statistical treatment were used in the study. First was descriptive statistics where the mean and standard deviation was computed to describe the level of self-concept in the RT and in the TSCS. Second was Pair-wise Correlation used to differentiate the Rorschach scores from scores of the TSCS and to determine the significant relationship in the self-concept scores. The F-ratio was computed to find the difference in self-concept between the 2 contrasted (white collar and blue collar) groups.The findings were: The first hypotheses that there is a significant relationship between over-all self-concept scores of the respondents on the RT and their scores on the TSCS was accepted, r=.652,p.002. This proves that self-concept in the RT has concurrent criterion-related validity.The second hypothesis that there is a significant difference between the over-all self-concept scores between the contrasted groups on the RT was accepted, F(1,22)=17.105,p.001, thus proving that self-concept in the RT has construct validity.The white collar respondents had higher self-concept than the blue collar respondents.The content analysis showed a clear difference in the quality of content between the white collar and blue collar respondents.Based on the findings it was concluded that the RT has concurrent criteria-related validity. Therefore it is effective in counteracting mono-operational bias in research and counseling. The variables on self-concept in the RT at Level 1 have construct validity. Hence, they can be reliably used to measure self-concept.

Abstract Format






Accession Number


Shelf Location

Archives, The Learning Commons, 12F Henry Sy Sr. Hall

Physical Description

89 numb. leaves ; computer print-out


Self-perception--Testing; Psychometrics; Rorschach test; Personality tests; Tests and scales--Validity; Inkblot test, Rorschach; Projective techniques

This document is currently not available here.