Fadia Nasser

A Monte Carlo Study Investigating the Impact of Item Parceling on Measures of Fit in Confirmatory Factor Analysis

A simulation study was conducted to examine the effect of item parceling on goodness-of-fit indices at different levels of sample size, number of indicators per factor, factor structure/pattern coefficients, interfactor correlations, and item-level data distribution. Results revealed that the use of item parcels yielded more nonconverged solutions and Heywood cases than individual items. The likelihood of nonconverged solutions and Heywood cases increased as the number of indicators per factor (more items per parcel) decreased. Meanwhile, parcel solutions as compared with item solutions resulted in better fit as measured by the chi-square to degrees-of-freedom ratio, Goodness-of-Fit Index (GFI), Expected Cross-Validation Index (ECVI), and root mean square error of approximation (RMSEA), as well as two incremental fit indices, the Non-Normed Fit Index (NNFI) and Comparative Fit Index (CFI). The same pattern of results was found with data that varied in terms of skewness and kurtosis at the item level. However, the likelihood of nonconverged solutions and Heywood cases was more pronounced when data were extremely skewed/kurtotic at the item level.

Predicting Teacher Commitment.

Abstract An empirical investigation was made of a model for predicting commitment to teaching, as measured by the extent to which teachers expressed an unwillingness to change careers. Predictor variables included personal variables as well as job-related factors. Data are reported from 175 teachers who had completed their preservice training at an Israeli teachers college over a ten-year period. Results indicated that only job satisfaction could directly predict commitment. Other factors, such as professional self-image, abilities, gender, job advancement, and pupil grade level were indirectly related, generally through their relationship with satisfaction. Teaching experience was unrelated to other variables in the model.

How Well Do Item Parcels Represent Conceptually Defined Latent Constructs? A Two-Facet Approach

This article presents a methodology for examining the content and nature of item parcels as indicators of a conceptually defined latent construct. An essential component of this methodology is the 2-facet measurement model, which includes items and parcels as facets of construct indicators. The 2-facet model tests assumptions required for accepting parcels as aggregates of item covariation in representing the latent construct. According to this methodology, parcels are acceptable indicators of the latent construct if the 2-facet model meets parametric assumptions for unidimensionality and if items and parcels have content validity as measures of the latent construct. The proposed methodology is illustrated using a 1-factor model of the Worry construct in the test anxiety measurement tradition

Modeling Mathematics Achievement of Jewish and Arab Eighth Graders in Israel: The Effects of Learner-Related Variables.

This study examined a structural model of mathematics achievement of 2 culturally different groups of Jewish and Arab 8th graders in terms of 5 learner-related variables, namely, gender, epistemological beliefs, self-efficacy, attitudes, and mathematics anxiety. Multigroup structural modeling analysis indicated that the goodness of fit of the hypothesized structural model and the total effects of mathematics self-efficacy and epistemological beliefs were comparable in both groups. The 2 groups differed in the effects that gender, attitudes toward mathematics, and mathematics anxiety exerted on mathematics achievement. They also diverged in terms of the amount of variance in mathematics achievement that the 5 learner-related variables accounted for.

The Performance of Regression-Based Variations of the Visual Scree for Determining the Number of Common Factors

The accuracy and consistency of four regression-based variations of the visual scree (Cattell-Nelson-Gorsuch, multiple regression, t-value index, and standard error scree [SEscree]) were examined using artificial data with known characteristics. The results indicated that in general, the SEscree was more accurate than the other three procedures. The observation-to-variable ratio was most influential on the accuracy and the consistency of the number of factors yielded by the SEscree followed by the magnitude of factor structure and pattern coefficients. Unlike the other regression variations of the scree, the SEscree performed equally well whether the factors were correlated or uncorrelated.

Predicting student ratings: the relationship between actual student ratings and instructors’ predictions

This study examined the relationship between student ratings and instructors’ predictions of these ratings, taking into account other instructor, student, and course characteristics. Participants in the study were 198 instructors in the School of Education at a major teacher training college in Israel. Data corresponding to one randomly selected course per instructor were collected using student and instructor questionnaires and college records. Results indicate a systematic positive relationship between instructors’ predictions and actual student ratings with respect to overall ratings and the ratings of three dimensions of teaching. Results also demonstrate a systematic trend whereby low‐rated instructors tend to overestimate their student ratings, high‐rated instructors underestimated ratings, and moderately rated instructors gave accurate predictions. Results have implications for using predictions to motivate teaching improvement.

Effects of Gender and Ethnicity on Fourth Graders' Knowledge States in Mathematics.

This study sheds light on the achievement gap between two culturally diverse populations in Israel by employing a diagnostic model for analysing responses of a representative sample of Jewish and Arab fourth graders on a national mathematics test. The results indicated large significant differences, in favour of the Jewish group, on most attributes underlying the test, and relatively small significant gender effects only in the Jewish group, where boys outscored girls on higher-order thinking attributes. These results were discussed in light of cultural differences between the two populations, educational resources, and prevalent instruction–learning–assessment cultures in their respective schools.

Modeling the Observation-to-Variable Ratio Necessary for Determining the Number of Factors by the Standard Error Scree Procedure Using Logistic Regression.

Logistic regression was used for modeling the observation-to-variable (n/v) ratio required for the standard error scree (SEscree) procedure to correctly identify the number of factors in simulated data. The correlation matrices were generated to possess known characteristics: number of factors (f), number of variables (v), sample size (n), magnitude of pattern coefficients (p), and degree of interfactor correlations (r). The results indicated that under all conditions, the n/v ratio required for the SEscree procedure to correctly identify the true number of factors with high probability exceeded the minimum of 5:1 recommended in some of the related literature. This study demonstrated the ability of the logistic regression to simplify summarizing and reporting findings from simulation studies that involve a large number of conditions.

Large-scale diagnostic assessment: Mathematics performance in two educational systems

A diagnostic methodology for large-scale assessment was employed to compare performance on a national test in mathematics of representative samples of Jewish and Arab 8th graders in Israel in order to shed light on a previously identified large achievement gap between these 2 populations. The results revealed significant differences between the 2 groups in patterns of strengths and weaknesses with respect to content, process, and skill/item-type attributes, indicating different paths for remedial interventions.