Validation of the short parallel and extra-short form of the Heidelberg figural matrices test (HeiQ)

• Verfasst von: Pallentin, Vanessa S. [VerfasserIn]
• Verfasst von: Danner, Daniel [VerfasserIn]
• Verfasst von: Lesche, Sven [VerfasserIn]
• Verfasst von: Rummel, Jan [VerfasserIn]
• Titel: Validation of the short parallel and extra-short form of the Heidelberg figural matrices test (HeiQ)
• Verf.angabe: Vanessa S. Pallentin, Daniel Danner, Sven Lesche and Jan Rummel
• E-Jahr: 2024
• Jahr: 14 October 2024
• Umfang: 13 S.
• Illustrationen: Illustrationen
• Fussnoten: Gesehen am 24.03.2025
• Titel Quelle: Enthalten in: Journal of Intelligence
• Ort Quelle: Basel : MDPI, 2013
• Jahr Quelle: 2024
• Band/Heft Quelle: 12(2024), 10, Artikel-ID 100, Seite 1-13
• ISSN Quelle: 2079-3200
• DOI: doi:10.3390/jintelligence12100100
• Datenträger: Online-Ressource
• Sprache: eng
• Sach-SW: figural matrices test
• Sach-SW: intelligence
• Sach-SW: item response theory
• Sach-SW: parallel test forms
• Sach-SW: reasoning
• K10plus-PPN: 1920379347

Abstract

Figural matrices tests are frequently used to measure fluid intelligence. The HeiQ—an operation-oriented figural matrices test—was developed to tackle limitations of previous matrices tests, mainly the possibility of excluding distractors based on superficial features instead of actively solving the items. However, allowing for a total administration time of 60 min for the assessment of one construct is not feasible in many study designs. Thus, the goal of this study was to develop three short forms of the existing HeiQ. Two parallel 20-item short forms (the HeiQ-S A and HeiQ-S B) that are comparable in content as well as on a psychometric basis and a 6-item short form (the HeiQ-XS) were generated. All tests showed good internal consistency (Cronbach’s Alpha ranging from α = 0.82 to α = 0.86) and good criterion-related validity (correlations with high school grade (Abitur) ranging from r = −0.34 to r = −0.38); construct validity (correlations with the global intelligence scores of the Intelligence Structure Test 2000R were between r = 0.58 and r = 0.71). Further, all test versions showed to be Rasch-scalable, implying a uniform underlying ability. Thus, we conclude that all three newly developed short versions are valid tools for assessing fluid intelligence.