IRT Information Function of Polytomously Scored Items under the Graded Response Model

Zhaosheng, Luo; Xue-Lian, Ouyang; Shu-Qing, QI; Haiqi, Dai; Ding, Shuliang

doi:10.3724/sp.j.1041.2008.01212

Cited by 2 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The formula for calculating the reliability coefficient is r xx (θ) = 1 − 1 I(θ) , and the formula for deviation is SE(θ) = 1 √ I(θ) . "I(θ)" refers to the amount of information that a subject whose psychological properties level is θ contributes to the test (Luo et al, 2009). The reliability and deviation of the two scales were calculated under the GRM model, and the results are shown in Figures 1, 2, respectively.…”

Section: Test Reliability Coefficient and Deviation Curvementioning

confidence: 99%

“…First, the category response threshold (parameter b) of the item and the trust property of the subject use the same metric system. Second, the results of different experiments with the same psychological traits can be compared (Luo, 2012). Third, through item parameter estimation, IRT can directly and accurately reflect the experimental characteristics of each item.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Comparison of Trust Assessment Scales Based on Item Response Theory

et al. 2020

View full text Add to dashboard Cite

Three widely used interpersonal trust measurement scales [Interpersonal Trust Scale (ITS), Philosophies of Human Nature Scale (RPHNS), Company Trust Scale (CTS)] have seldom been applied in non-Western contexts. Different social environments may lead to variation in the level or structure of trust. Therefore, it is necessary to compare the applicability of these scales to different levels of trust-related traits in Eastern cultures so that researchers can choose appropriate scales for relevant studies. This study attempted to conduct a comparative analysis of the ITS, RPHNS, and CTS. A sample of 725 Chinese college students was analyzed. Total score correlations and latent factor correlations estimated by confirmatory factor analysis (CFA) for a first-order three-factor model were assessed, and then the quality of the item parameters, test reliability and standard errors, and test information were assessed. The results are as follows: (1) the ITS and the RPHNS assessed almost the same trust traits; therefore, only the ITS and the RPHNS are compared in the next sections; (2) the original structure of only the RPHNS is verified; (3) some items on the ITS do not work well, while the RPHNS has higher overall test reliability; and (4) the average item information provided by the RPHNS is higher across all trait levels. In most cases, the RPHNS is the better choice in the Chinese cultural context.

show abstract

Section: Test Reliability Coefficient and Deviation Curvementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%