Introducing Item Pool Visualization: A method for investigation of concepts in self-reports and psychometric tests

Abstract

In this article, we introduce Item Pool Visualization. It is an illustration system that locates items and item pools (scales) from multiple psychological instruments regarding their commonality and distinguishability along several dimensions of nested radar charts. The application of Item Pool Visualization creates illustrations that represent different item pools by different circles that do not overlap. Item Pool Visualization illustrates a comparison of different structural equation models that are estimated with the same data. It combines the advantages of general and correlated factor models when evaluating psychological instruments. Furthermore, in contrast to other visualization methods, Item Pool Visualization provides an empirically driven categorization of psychological constructs and their subconcepts (facets) that is suited to provide professionals with help in comparing psychometric constructs, questionnaires, and selecting tests.

Are some interoceptive sensibility components more central than others? Using item pool visualisation to understand the psychometric representation of interoception

“…IPV analyses were computed using the IPV package [ 46 ] in R [ 41 ], and structural equation models (SEMs) were calculated using the lavaan package [ 43 ]. The planned analytic strategy for the IPV analyses followed the procedural steps outlined by Dantlgraber and colleagues [ 31 ] and Swami and colleagues [ 47 ]. Briefly, we planned to first produce a general factor model of interoceptive sensibility using SEM (i.e., a single factor extracted from the overall item pool).…”

“…Finally, centre distances were calculated. Centre distances represent the proportional increase of the explained item variance when the items are allocated to smaller sub-pools compared to the larger common pool (see Section 1; see also [ 31 , 47 ]).…”

“…Here, we suggest that using Item Pool Visualisation (IPV [ 31 ]) could be a useful approach for investigating construct commonality and distinguishability, by illustrating comparisons of facets both within and between measures that aim to assess interoceptive sensibility. IPV is a visual tool that uses different structural equation modelling (SEM) estimations to locate items and item pools from within the same dataset.…”

Illustrating psychometric tests, scales, and constructs: An R package for Item Pool Visualization

“…To offer a systematic and detailed portrayal of the actual content and internal balance of competing measures, Item Pool Visualization (IPV) was developed (Dantlgraber et al, 2019). In IPV, an item pool is a large set of correlated items (manifest indicator variables) that are assessed together.…”

“…For example, "agreeableness" in the Big Five model is conceptually and empirically distinct from "agreeableness" in the HEXACO model (Ashton & Lee, 2019;Ashton et al, 2014). In another example, the "Sports Mental Toughness Questionnaire" turned out to be empirically more similar to two "Self-Esteem" measures than these were similar to each other (Dantlgraber et al, 2019). Secondly, the integrated content of multidimensional measures is usually unbalanced (e.g., Hulleman et al, 2010).…”