FS16_Introduction to meta

advertisement
Zurich-Basel Plant Science Center
PhD Program in Plant Sciences: Introduction to meta-analysis
Lecturer: Julia Koricheva
Location: tba
Dates: April 4 to 8, 2016
Credit Points: 2 ECTS
Course Objectives
This course aims to promote and facilitate the thoughtful and critical use of meta-analysis for
research synthesis in ecology by:
1. Explaining the principles and advantages of meta-analysis for research synthesis
2. Demonstrating the range of applications of meta-analysis in ecology
3. Promoting understanding of the assumptions and limitations of meta-analysis
4. Providing first-hand experience in question formulation, data extraction, database
design, use of software for meta-analysis and report preparation
Course Program
Day 1 (Monday, April 4th)
 Morning: lectures on history of meta-analysis, types of quantitative research
synthesis, conversion of ecological data to effect sizes, and question formulation
 Afternoon: data extraction exercise & Work in groups: students decide on inclusion
criteria and metrics of effect size for their own meta-analysis based on 8-10 studies
and begin data extraction
Day 2 (Tuesday, April 5th)
 Morning: lectures on combining effect sizes across studies and testing for moderators
in meta-analysis (meta-regression), practical on conducting meta-analysis using
OpenMEE software
 Afternoon: Work in groups: completing data extraction for own meta-analysis and
running meta-analysis and testing for moderators
Day 3 (Wednesday, April 6th)
 Morning: lectures on publication bias, dealing with varying research quality and nonindependence of observations
 Afternoon: Work in groups: testing for publication bias in own dataset, considering
sources of non-independence, finalizing analyses
Day 4 (Thursday, April 7th)
 Morning: lectures on format of meta-analysis report, review of case studies of metaanalysis in ecology, and critique of meta-analysis
 Afternoon: Work in groups: preparing Power Point presentation with results of the
analyses
Day 5 (Friday, April 8th)
 Morning: student presentations on the results of their analyses
 Afternoon: Individual consultations for students planning to do meta-analysis in
their own research projects
Prior Knowledge: No previous experience with programming languages is required. General
understanding of standard statistics (mean, standard deviation, analysis of variance,
correlation, regression) is required.
Number of Participants: 25
Individual Performance and Assessment:
Throughout the course students will work in small groups (3-4). Each group will be given a
topic for the analysis and a collection of 8-10 research papers on the topic. The students will
have to:
 Formulate questions for the analysis, particularly with respect to the potential sources
of variation in the effect
 Formulate inclusion criteria for the studies
 Decide on the metric of effect size to be used
 Design the data file and extract data from the research papers
 Conduct meta-analysis of the data extracted using MetaWin software
 Prepare a Power Point presentation on the results of their analysis
Download