Emerging Scholars Methods Workshop
Meta-analysis: Why, When, and How?
Presenter: Elisabetta Crocetti
Utrecht University, the Netherlands
Thursday, October 10th from 2pm to 5:15pm
Location: River C
Registration: the workshop is FREE for all conference participants.
Registration is not required. However, those who are interested in participating can communicate their intention to participate to Elisabetta Crocetti sending an email to firstname.lastname@example.org with the subject "EA Meta-Analysis Workshop" by October 5. Doing so, they will receive a print copy of all workshop materials. Those who do not notify their participation can attend the workshop without receiving the print copy of the workshop materials.
The meta-analysis is a methodological approach to systematically review available literature and provide a statistical synthesis of it, testing for factors that can moderate differences in study findings. The workshop will address three questions.
The first part will address why is worthwhile conducting a meta-analysis, covering advantages of this approach in the context of the statistics reform in the behavioral sciences.
The second part will explain when is particularly important to conduct a meta-analysis considering issues related to both the state-of-the-art of a field and the career of a scholar.
The third part will address how to conduct a meta-analysis. Specifically, the following steps will be explained: 1) defining the research question and formulating hypotheses; 2) specifying inclusion and exclusion criteria; 3) searching the literature (e.g., computerized database search, searching indexes of journals, searching in reference lists, research strategies for the grey literature); 4) selecting primary studies; 5) coding and preparing the meta-analytic database; 6) computing effect sizes for each study (e.g., Cohen's d, Hedges's g, correlations, risk ratios, odds ratios, hazard ratios); 7) combining effect sizes (fixed-effect model vs. random-effects model); 8) assessing heterogeneity; 9) explaining heterogeneity: test of categorical (subgroup analysis) and continuous (meta-regression) moderators; 10) managing complex databases (e.g., multiple subgroups, outcomes, comparisons, and/or waves within primary studies); 11) assessing publication bias (e.g., Failsafe N, Funnel plot, Begg and Mazumdar's rank correlation method, Egger's linear regression method, Trim and Fill method); 12) publishing a meta-analysis (e.g., PRISMA and MARS guidelines).
For each step, practical examples and software applications will be shown. At the end of the workshop participants will be able to a) critically understand a published meta-analysis and b) planning their own meta-analytic review.
The first part of the workshop will have a lecture format. Ample time will be provided throughout for participant questions and comments. In the second part of the workshop participants will work in groups on interpreting a published meta-analysis and on planning a meta-analytic review.
Elisabetta Crocetti, PhD is a researcher at the Utrecht University, the Netherlands. She received a Master in Psychology from the University of Bologna and a PhD from the University of Macerata (Italy).
Her primary scholarly interest concerns the processes of identity formation in adolescence and emerging adulthood. In particular, she is interested in examining how identity is formed and revised over time and which individual, relational, and cultural factors influence this dynamic.
She is also strongly interested in methodological and statistical issues related to social research, such as testing for factorial validity of measurement instruments, research synthesis, systematic review, and meta-analysis. In the last years she has given several courses and lectures on meta-analysis.
She is affiliated with various associations and she is a member of the Member of the Founding Executive Council of the Society for the Study of Emerging Adulthood (SSEA).
Her complete CV and list of publications is available at www.elisabettacrocetti.com