Principal speaker
Associate Professor Sama Low-Choy
Meta-analysis is often conducted to summarise quantitative results from several similar studies, perhaps conducted on slightly different populations, sometimes with different methodologies. These results may be focussed on straightforward quantities, such as: counts, averages or proportions. Many meta-analyses may consider the result of a regression, ANOVA or SEM: effect sizes (e.g. of an intervention or treatment on an outcome); or differences (e.g. efficacy of a drug compared to a control).
The greatest challenge when extracting these quantities is sometimes the associated uncertainty. We explore basic statistical techniques for translating uncertainties expressed in different ways, e.g. standard errors, confidence intervals, t-statistics or graph-only. A key challenge is to ensure that quantities and uncertainties, are comparable across studies, i.e. measured on similar scales and expressed in the same way.
Software: We illustrate the methodology using outputs from a freely available software package, RevMan. We encourage you to install RevMan and refer to online resources for a HowTo on using software such as RevMan. Since we focus on interpretation (rather than software use), the information is relevant to use of these statistical models for meta-analysis using any package.
Pre-requisites: For those already involved in a meta-analysis, this session will revisit underlying assumptions. For newcomers to meta-analysis, this seminar will be more easily understood if you have attended or reviewed the text on the first three episodes, which cover: (1) a visual intro to MA: inputs & outputs; (2) reviewing literature for MA inputs, with stats implications; (3) qual lit review techniques for MA.
Level: This workshop will be suitable for advanced beginners, who have background (workshops / practice) in any of: preparing inputs for meta-analysis; similar statistical analyses; or have read the text.
Recommended Reading: This workshop refers to a book chapter, available through the Griffith library as an e-book. Low-Choy, S., Almeida, F., & Rose, J. (2021). "Combining study findings by using multiple literature review techniques and meta-analysis",Chapter 15 in E. Manu and J. Akotia (eds) Secondary Research Methods in the Built Environment, pp207-220
Event categories
RSVP
RSVP on or before Monday 6 June 2022 09.36 am, by email RED@griffith.edu.au , or via https://events.griffith.edu.au/D2VYPG