Comparison of two random-effects methods of meta-analysis.

Hall, Steven M.; Brannick, Michael Τ.

doi:10.1037/0021-9010.87.2.377

Cited by 103 publications

(127 citation statements)

References 37 publications

(83 reference statements)

Supporting

Mentioning

117

Contrasting

Unclassified

Order By: Relevance

“…As such the Hunter and Schmidt method is preferable to alternative approaches such as the fixed-effects Hedges and Olkin (1985) method that explicitly assumes that all variation in observed effect sizes is a function of study artifacts rather than possibly reflecting substantive variation across samples, and hence assumes that SD ρ is equal to zero. The Hunter and Schmidt method is also preferable to the random-effects Hedges and Olkin method in that it does not convert correlations to Fisher's z; a method that attempts to eliminate the negative bias inherent in the use of sample correlations but that has been shown to result in an even larger positive bias (Field 2001;Hall and Brannick 2002). That is, the bias produced by the Hunter and Schmidt method is negligibly negative and substantially smaller than the positive bias produced by alternate methods.…”

Section: Analytic Proceduresmentioning

confidence: 91%

Adjustment to College as Measured by the Student Adaptation to College Questionnaire: A Quantitative Review of its Structure and Relationships with Correlates and Consequences

2011

View full text Add to dashboard Cite

This paper presents a meta-analytic review (k=237, N=44,668) of the adjustment to college literature. The review, based on studies using the Student Adaptation to College Questionnaire, is organized around three primary themes: (1) the structure of students' adjustment to college, (2) the relationship of adjustment to college constructs with possible antecedents and correlates, and (3) the relationship of adjustment to college constructs with college grades and college retention. Meta-analytic results indicate that adjustment to college is multidimensional, predictive of college grades, and an unusually good predictor of college retention. Adjustment to college is also shown to be moderately related to individual traits, social support, and students' relationships with their parents. Weaker relationships are evident with demographic variables, prior achievement, coping approaches, and variables that reflect students' psychological independence from their parents. Theoretical and practical implications for the study of students' adjustment to college, academic performance, and retention are discussed.Keywords Adjustment to college . College retention . College GPA . Meta-analysis Entry into college presents students with numerous novel challenges that extend beyond the greater academic demands, greater autonomy, and lower level of academic structure that characterizes life at many colleges and universities. First-year students are also required to navigate a new social environment, develop an orientation toward the institution of which they are now a member, become productive members of the university community, adapt to new roles and responsibilities (e.g., managing own finances), manage the separation from friends and family, and engage in the process of making career decisions. Concern about the manner in which students navigate these myriad challenges is reflected both in the orientation programs offered by many colleges and universities as well as the substantial body of educational research that has focused on students' adjustment to college. Many educational researchers have viewed Educ Psychol Rev (2012) 24:133-165

show abstract

Section: Analytic Proceduresmentioning

confidence: 91%

Adjustment to College as Measured by the Student Adaptation to College Questionnaire: A Quantitative Review of its Structure and Relationships with Correlates and Consequences

2011

View full text Add to dashboard Cite

show abstract

“…Next, the distribution of r pb was corrected for sampling error to get r c . Note that the correction was done using frequency-weighted average, not Fisher's z transformation, since the latter has been shown to produce upwardly biased correlation estimates (Hall and Brannick 2002). Finally, standard deviations of the corrected observed correlation were calculated to compute 99% confidence intervals around r c .…”

Section: Computation and Analysis Of Effect Sizesmentioning

confidence: 99%

Expertise Differences in the Comprehension of Visualizations: a Meta-Analysis of Eye-Tracking Research in Professional Domains

2011

View full text Add to dashboard Cite

This meta-analysis integrates 296 effect sizes reported in eye-tracking research on expertise differences in the comprehension of visualizations. Three theories were evaluated: Ericsson and Kintsch's (Psychol Rev 102:211-245, 1995) theory of long-term working memory, Haider and Frensch's (J Exp Psychol Learn Mem Cognit 25:172-190, 1999) information-reduction hypothesis, and the holistic model of image perception of Kundel et al. (Radiology 242:396-402, 2007). Eye movement and performance data were cumulated from 819 experts, 187 intermediates, and 893 novices. In support of the evaluated theories, experts, when compared with non-experts, had shorter fixation durations, more fixations on task-relevant areas, and fewer fixations on task-redundant areas; experts also had longer saccades and shorter times to first fixate relevant information, owing to superiority in parafoveal processing and selective attention allocation. Eye movements, reaction time, and performance accuracy were moderated by characteristics of visualization (dynamics, realism, dimensionality, modality, and text annotation), task (complexity, time-on-task, and task control), and domain (sports, medicine, transportation, other). These findings are discussed in terms of their implications for theories of visual expertise in professional domains and their significance for the design of learning environments.Expertise in the comprehension of visualizations has gained growing attention over the past years (de Groot and

show abstract

“…To calculate the true mean effect size, which represents the true effects of the autonomysupportive interventions, the researchers adopted the meta-analysis method in Hunter and Schmidt (2004) because this approach has been recommended following comparative analysis of the various meta-analytic approaches (Hall and Brannick 2002;Schulze 2004) and because it is the only procedure that corrects not only for sampling error but also for measurement error (Schmidt 2010). Moreover, the random effect method was preferred because it assumes that the population parameter may be different from study to study and allows for any possible value of SD d , rather than the fixed effect model which assumes homogeneity of effect parameters (Hedges and Vevea 1998;Hunter and Schmidt 2004) that could result both in a higher type I error and in narrower confidence intervals (Hunter and Schmidt 2000).…”

Section: Effect Size Calculationsmentioning

confidence: 99%

A Meta-analysis of the Effectiveness of Intervention Programs Designed to Support Autonomy

2010

View full text Add to dashboard Cite

The twofold purpose of the present study was, first, to determine whether training intervention programs designed to help people support the autonomy of others are effective and, second, to identify the set of conditions that allowed these interventions to be most effective. A meta-analysis of the findings from 19 studies with 20 effect sizes showed that the training programs were, overall, effective with a weighted effect size of 0.63. Moderator analyses of the overall effect size showed that the relatively more effective intervention programs were structured in ways that trained multiple elements of autonomy support and were presented in relatively brief (1-3 h) sessions in a laboratory training setting that focused on skill-based activities and utilized multiple types of media to deliver its content. Furthermore, relatively effective intervention programs were offered to teachers (rather than to other professionals), trainees (rather than to experienced professionals), and individuals with an autonomy (rather than a control) causality orientation. Though the small number of included studies warrants caution, results generally affirmed the effectiveness of autonomy-supportive training programs and identified the conditions under which future programs can be designed to be highly effective.According to self-determination theory (SDT; Ryan and Deci 2000), people function positively when others support their autonomy rather than control their behavior. Because the benefits from autonomy support have been found to be both widespread and predictive of important outcomes (discussed below), researchers and practitioners alike have asked whether or not people such as

show abstract

Comparison of two random-effects methods of meta-analysis.

Cited by 103 publications

References 37 publications

Adjustment to College as Measured by the Student Adaptation to College Questionnaire: A Quantitative Review of its Structure and Relationships with Correlates and Consequences

Adjustment to College as Measured by the Student Adaptation to College Questionnaire: A Quantitative Review of its Structure and Relationships with Correlates and Consequences

Expertise Differences in the Comprehension of Visualizations: a Meta-Analysis of Eye-Tracking Research in Professional Domains

A Meta-analysis of the Effectiveness of Intervention Programs Designed to Support Autonomy

Contact Info

Product

Resources

About