Éric Frenette

Maternal self-efficacy and hostile-reactive parenting from infancy to toddlerhood

A longitudinal study of maternal self-efficacy (SE) and hostile-reactive parenting (HRP) was conducted with a community sample of 1836 mothers. Mothers completed questionnaires when their child was 4.5, 16.6 and 28.5 months of age. Maternal SE showed little change, whereas HRP sharply increased from 4.5 to 28.5 months. Structural equation models suggest these initially correlated variables did not influence each other over time, but rather became increasingly independent. Three distinct developmental trajectories were fitted for both maternal SE and HRP. In contrast to the overall portrait of stability, 12.6% of mothers followed a trajectory of declining SE. Likewise, 12.1% of mothers showed initially high and sharply increasing HRP. Few mothers (3.1%) simultaneously followed both of these trajectories, yet overall, SE and HRP trajectories were modestly associated (tau(b)=-.23, p<.0001). Failure to support the expected reciprocal influences between SE and HRP over time, as well as maternal and child contributions to early manifestations of maternal HRP and their evolution are discussed.

Leadership: Validation of a Self-Report Scale

The aim of this paper was to propose and test the factor structure of a new self-report questionnaire on leadership. A sample of 373 school principals in the Province of Quebec, Canada completed the initial 46-item version of the questionnaire. In order to obtain a questionnaire of minimal length, a four-step procedure was retained. First, items analysis was performed using Classical Test Theory. Second, Rasch analysis was used to identify non-fitting or overlapping items. Third, a confirmatory factor analysis (CFA) using structural equation modelling was performed on the 21 remaining items to verify the factor structure of the scale. Results show that the model with a single third-order dimension (leadership), two second-order dimensions (transactional and transformational leadership), and one first-order dimension (laissez-faire leadership) provides a good fit to the data. Finally, invariance of factor structure was assessed with a second sample of 222 vice-principals in the Province of Quebec, Canada. This model is in agreement with the theoretical model developed by Bass (1985), upon which the questionnaire is based.

Nonparametric item analysis and confirmatory factorial validity of the computer attitude scale for secondary students

Abstract The use of computers in secondary schools allows a better preparation of students for their future careers and offers opportunities for refining critical thinking skills. Thus, the development of a valid attitude scale towards computers for students is very important because it could lead to the development or modification of computer fluency curricula. The aim of the study was (a) to examine the factorial structure (affect, behavior, and cognition) of the Computer Attitude Scale for Secondary Students (CASSS) through confirmatory factor analysis, and (b) to determine how well the test and its items discriminate among individuals differing in the strength of their attitude toward computers through nonparametric item analyses. The analyses were conducted on two samples of high school students (n1=421; n2=422) who completed the CASSS, a 40-item questionnaire. Results showed that (a) the option characteristic curves (OCC) and item characteristic curve (ICC) of each item revealed certains problems for some items, (b) the reliability of the CASSS is excellent at all levels of the attitude scale, being around 0.96 for students having the least favorable attitude toward computer and 0.88 for students having a more favorable attitude toward computers, and (c) a higher-order or hierarchical model underlies the theoretical structure of the CASSS.

Multiple tutorial-based assessments: a generalizability study

BackgroundTutorial-based assessment commonly used in problem-based learning (PBL) is thought to provide information about students which is different from that gathered with traditional assessment strategies such as multiple-choice questions or short-answer questions. Although multiple-observations within units in an undergraduate medical education curriculum foster more reliable scores, that evaluation design is not always practically feasible. Thus, this study investigated the overall reliability of a tutorial-based program of assessment, namely the Tutotest-Lite.MethodsMore specifically, scores from multiple units were used to profile clinical domains for the first two years of a system-based PBL curriculum.ResultsG-Study analysis revealed an acceptable level of generalizability, with g-coefficients of 0.84 and 0.83 for Years 1 and 2, respectively. Interestingly, D-Studies suggested that as few as five observations over one year would yield sufficiently reliable scores.ConclusionsOverall, the results from this study support the use of the Tutotest-Lite to judge clinical domains over different PBL units.