Associate Professor Sama Low-Choy
Meta-analysis takes research to the next level, by considering the weight of evidence provided by a group of studies replicating a similar research method. Often a group of studies will all fit a similar model such as regression (or SEM) to estimate the effect of some intervention on an outcome. The aim is to obtain an overall picture of the consensus view of this effect size, across studies. This is a pair of seminars introducing the concepts of meta-analysis to a broad audience of researchers. We illustrate ideas using an example on how big data analytics impacts business performance. These seminars refer to recently published research (with innovations in methodology) on topics in Computer Science and Health. Part 4 - How meta-data underpins meta-analysis - statistical ideas Whilst many packages now make it easy to implement the statistical model for meta-analysis, most of the effort is actually in collating the data that goes into a meta-analysis. This seminar refers to international standards that help guide this process to be repeatable, methodical and relevant to the research question of interest. We show how decisions made in collating the meta-data going into a meta-analysis fundamentally underpin the statistical model and hence the findings. We use an example of a meta-analysis of how big data analytics impacts on business performance, described in a recent book chapter (Low-Choy, Almeida & Rose, 2021). Format - 1.5 hour seminar Related RED workshops - This is one of a series of sessions on meta-analysis, from literature review through to statistical analysis. Participants may find it easier to follow if they have done a first course in statistics, covering basic inference for a mean (e.g. t-tests, ANOVA and/or linear models a.k.a. regression). Reference - Low-Choy, S, F Almeida, and J Rose. "Combining study findings by using multiple literature review techniques and meta-analysis." Chapter 15 in eds (E Manu & J Akotia) Secondary Research Methods in the Built Environment (2021), p207.