skip to main content skip to footer

New Findings in the Development of a Test of Intellectual Ability

Author(s):
Findley, Warren G.
Publication Year:
1954
Report Number:
RM-54-03
Source:
ETS Research Memorandum
Document Type:
Report
Page Count:
7
Subject/Key Words:
Cognitive Ability, Intelligence Tests, Power Tests, Test Construction

Abstract

This paper presents findings incidental to the development of a test of intellectual ability that 1) yields separate scores for verbal and quantitative reasoning and 2) depends on power rather than speed. Preliminary studies of the reliability and validity of various item-types were undertaken in 25 different school systems chosen to reflect differences in geographical location, size of school and community and level of expenditure per pupil. The test scores were compared with high school grades in grades 9 and 12. Results and possible interpretations of them are presented. Conclusions include: 1) A 30-minute verbal aptitude test, a 40- minute quantitative aptitude test, or a 35-minute composite test can be built to yield within-grade reliabilities of .90; 2) a combination of sentence completion and synonyms exercises would form the most efficient measure of verbal competence and/or aptitude at the secondary and early college levels; 3) a combination of routine computation and arithmetic problem-solving exercises would form the most appropriate measure of quantitative reasoning ability at the secondary and early college levels.; 4) a combination of tests of these two components would from the most efficient measure of general scholastic aptitude for a short battery at the secondary and early college levels; and 5) the data on the validity of the routine computation test in predicting concurrent secondary school achievement, taken in conjunction with other data and studies, suggest the presence of a "character" factor in performance on mastery tests that is predictive of achievement generally. It is also possible that this "character" factor is overrepresented in teachers grades at the ninth-grade level and hence is overvalued by this study with its emphasis on concurrent validity. Paper read at Joint Meeting of AAAS and AERA December 29, 1953.

Read More