Search

The aim of this study was threefold: First, to examine the dimensionality of the construct of General Academic Ability (GAA) at the subscale level providing additional insights over and above on the conceptualization of the construct. Second, to explore different degrees of measurement invariance of the GAA across gender using more recent advancements in the examination of Measurement Invariance (i.e., Bias-Corrected bootstrap Confidence Intervals). Third, to examine gender differences across the different facets of the GAA at the latent mean level. The sample consisted of 1,800 high school graduates who applied for higher education in Saudi Arabia. The results from the analysis indicated that the hierarchical model with one higher-order factor (i.e., general academic ability) and four lower-order cognitive factors (i.e., verbal ability, quantitative ability, scholastic aptitude, and GPA) exhibited an excellent fit to the data. In terms of the measurement invariance hypothesis, it was found that the hierarchical model exhibits full configural and metric invariance and partial scalar invariance. Finally, using the Latent Mean Difference procedure, the results showed gender differences in the Verbal and GPA domains. Although significant differences were also found in the Scholastic aptitude domain, this finding is not stable due to several non-invariant items within the domain. In both cases, females scored higher than males. Finally, regarding the higher-order factor (GAA), the results showed that females scored higher than males. There were no significant differences in the Quantitative domain. [ABSTRACT FROM AUTHOR]

The main aim of the present study was to investigate the presence of Differential Item Functioning (DIF) using a latent class (LC) analysis approach. Particularly, we examined potential sources of DIF in relation to gender. Data came from 6,265 Saudi Arabia students, who completed a high-stakes standardized admission test for university entrance. The results from a Latent Class Analysis (LCA) revealed a three-class solution (i.e., high, average, and low scorers). Then, to better understand the nature of the emerging classes and the characteristics of the people who comprise them, we applied a new stepwise approach, using the Multiple Indicator Multiple Causes (MIMIC) model. The model identified both uniform and non-uniform DIF effects for several items across all scales of the test, although, for the majority of them, the DIF effect sizes were negligible. Findings from this study have important implications for both measurement quality and interpretation of the results. Particularly, results showed that gender is a potential source of DIF for latent class indicators; thus, it is important to include those direct effects in the latent class regression model, to obtain unbiased estimates not only for the measurement parameters but also of the structural parameters. Ignoring these effects might lead to misspecification of the latent classes in terms of both the size and the characteristics of each class, which in turn, could lead to misinterpretations of the obtained latent class results. Implications of the results for practice are discussed. [ABSTRACT FROM AUTHOR]

The purpose of the present studies was to evaluate and predict academic cheating with regard to a national examination in a Middle East country. In Study 1, 4,024 students took part and potential cheaters were classified as those having discrepant scores in multiple administrations that exceeded 1SDin absolute terms. A latent class mixture analysis suggested two pathways for potential cheating: (a) The first path involved students—most male—who changed city or region of examination during test taking, and (b) the second path described students—most male—who did not change city, region, or center of administration. Study 2 profiled cheaters using a sample of examinees who were actually caught cheating. Participants were 545 students, 253 of whom were caught cheating between 2002 and 2012. Both samples were selected from a pool of 319,219 testees using random sampling procedures. Results indicated that a 4-class solution best fitted the data as in Study 1. Furthermore, a predictive model was tested with an independent cross-validation sample of 112 examinees (56 cheaters, 56 noncheaters). Results indicated that the model classified correctly 78.57 of the new cheating cases (sensitivity) and 94.64% of noncheaters (specificity). [ABSTRACT FROM AUTHOR]