Cognitive Component Analysis Comparing Three Chinese Remote Associates Tests: Linear Logistic Latent Trait Model Approach

Su Pin Hung, Ching Lin Wu*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

7 Citations (Scopus)

Abstract

The Remote Associates Test, generally used in creativity research, has Chinese versions for the three levels of “radical-word-vocabulary.” However, research has not been conducted on the influence of the item components on the difficulties among these Chinese Remote Associates Tests (CRATs). The present study selected six item components from the construction rules for three versions of CRATs, collected 555 participants’ responses to them, and analyzed the differences and similarities in the predictions of item difficulty among the six item components. The results show that item difficulty decreased if the targets are high-frequency or the stimuli are low-frequency words, but item difficulty increases if the stimulus can connect to more words, while the concreteness of the target does not significantly predict item difficulty. Moreover, the concreteness of the stimulus negatively predicts the difficulty of the CRRAT but positively predicts the difficulty of the CCRAT High association transfer increased the difficulty of the CWRAT but had no influence on the other two versions of CRAT. Therefore, these findings indicate the different influences of item components on item difficulty, strengthen the structural validity of the test, and also provide a reference for test development of the CRAT.

Original languageEnglish
Pages (from-to)224-234
Number of pages11
JournalCreativity Research Journal
Volume33
Issue number3
DOIs
Publication statusPublished - 2021

Keywords

  • creativity, remote association, item response theory, item component

ASJC Scopus subject areas

  • Developmental and Educational Psychology
  • Visual Arts and Performing Arts
  • Psychology (miscellaneous)

Fingerprint

Dive into the research topics of 'Cognitive Component Analysis Comparing Three Chinese Remote Associates Tests: Linear Logistic Latent Trait Model Approach'. Together they form a unique fingerprint.

Cite this