A Factor Analysis of Raven's Advanced Progressive Matrices Freed of Difficulty Factors

Abstract

The study presents a factor analysis of the 1962 revision of the Advanced Progressive Matrices (APM). The analysis was conducted such that substantive factor structure interpretations were freed of the effects of differences in item difficulty. The APM test was given to 237 examinees, 16–18 years old. The data were subjected to a Guttman scale analysis to determine whether the APM could be interpreted as a one factor instrument. Then the phi/phi max inter-item correlation matrix was factored. A principal components analysis, followed by a series of varimax rotations of the principal components, was performed. The Guttman coefficients of scalability were too small to support a one factor theory of the APM. The 2-factor solution provided the most interpretable factor structure. Factor I was composed of items in which the solution was obtained by adding or subtracting patterns. Factor II was composed of items in which the solution was based on the ability to perceive the progression of a pattern. Results are discussed in terms of representative cognitive tests and tasks believed to embody the logical operations responsible for successful performance on items loading on each factor. The possibility of forming subtests of items to enhance the predictive validity of the matrices also is discussed.

The quest for item types based on information processing: An analysis of Raven's Advanced Progressive Matrices, with a consideration of gender differences

“…These low reliabilities, combined with the relatively high correlations between composite scores (between the visual and analytic composites based on DeShon et al (1995): r = 0.50; between the pattern-progression and addition/subtraction composites based on Dillon et al (1981): r = 0.48) mean that the differential predictive power of any of the subscores considered here, relative to gender, was bound to be low. Also, all correlations with gender were weak (r ≤ 0.17, or d ≤ 0.34).…”

“…These models were a singlefactor model, a correlated two-factor model based on Dillon et al (1981), a correlated two-factor model based on DeShon et al (1995), and a correlated two-factor model based on item skewnesses. The four-factor model based on Mackintosh and Bennett (2005) could not be tested due to the limited number of items in some of the categories (in particular the quantitative-pairwise-progression category: only one item).…”

“…Barring any experimental manipulation, these three potential sources of item separation remain confounded. Considering that Dillon et al (1981) did not obtain factors based on item position in the test and used an untimed administration of the APM, whereas Vigneau and Bors (2005) did obtain such a pattern under timed conditions, it is also possible that item-position-based factor solutions reflect the common use of a time limit when administering the APM. Mackintosh and Bennett (2005), using both published data and a specially designed experiment, reported two sets of results showing that gender differences on the APM were mainly associated with item clusters characterized as requiring visuospatial processing-namely, in Carpenter et al's (1990) terminology, the figure-addition/subtraction rule and the distribution-of-two-values rule.…”

“…Mackintosh and Bennett's (2005) results are consistent with a number of dimensionality studies of Raven's Matrices Intelligence 36 (2008) reporting multifactorial solutions. Most of these studies pertained to the Standard version of the Matrices (SPM, Raven, 1956;e.g., Lynn, Allik & Irwing, 2004;van der Ven & Ellis, 2000), but there have been reports indicating that the advanced version of the Matrices (APM) could also be multidimensional (Dillon, Pohlmann & Lohman, 1981). The notion that Raven's Matrices items can be classified in terms of the information processing rules required for their solution was most notably developed by Carpenter et al (1990).…”

“…Dillon et al (1981) obtained results that indicated that the APM was dominated by two factors: one factor was interpreted as reflecting the ability to add and subtract patterns (addition/subtraction), and one factor was interpreted as the ability to detect a progression in a pattern (pattern progression) (see Appendix A). As mentioned above, however, Arthur and Woehr (1993) and DeShon et al (1995) both failed to replicate Dillon et al's (1981) results.…”