What does the Detroit tests of learning abilities, fifth edition measure? Revelations from a hierarchical exploratory factor analysis

Abstract

This study examined the factor structure of the Detroit Tests of Learning Abilities, Fifth Edition (DTLA‐5) using principal axis factoring, multiple factor extraction criteria, and the Schmid–Leiman orthogonalization procedures not utilized by test publishers. Results suggest that the publisher's six‐factor structure model was over factored. Although there is support for five group factors and a hierarchical general factor that correspond with some of the test's most reliable composites, there was no evidence warranting a distinction between the test's hypothesized Reasoning Abilities and Processing Abilities domains or separation of Acquired Knowledge and Verbal Comprehension subdomains. Some test users may prefer to interpret the four DTLA‐5 subdomains that correspond to some of our basic findings in a profile indicating intra‐individual strengths and weaknesses. Based on our hierarchical factor analysis and omega statistics, we believe that the DTLA‐5 should be interpreted as a measure of psychometric g and not used to examine intra‐individual strengths and weaknesses for most test‐takers.

Structural Models of Organizing Cognitive Learning Skills

“…Как и его предшественники, DTLA-5 был разработан для оценки различных когнитивных способностей у детей и подростков в возрасте от 6 до 17 лет. Он состоит из 12 подтестов, дающих девять составных баллов, в том числе те, которые представляют шесть когнитивных субдоменов, два домена и один глобальный показатель (общие когнитивные способности) [29]. В тесте «Оценка основных способностей к обучению» (ABLA) используются стандартные процедуры подсказки и подкрепления для оценки легкости или трудности, с которой испытуемый может выучить простую имитацию и пять различений с двумя вариантами ответов.…”