The workshop is structured in theoretical sessions encompassing: objectives of meta-analysis, step-by-step methodology, types of meta-analysis models, assessment of heterogeneity in the published outcomes, evaluation of publication bias, meta-regression and presentation of results. In the practical sessions, a series of meta-analyses addressing food safety/quality research questions will be performed using the R software. At the end of the workshop, if requested, the instructors will provide advice to the participants on their particular problems related to meta-analytical modelling.

The three-day sessions are scheduled as follows:

Day 1 (Morning):

09h00 – 11h00: Opening Keynote Lecture: The relevance of conducting meta-analysis in food safety research. Introduction to meta-analysis. Objectives and statistical significance (Dr. Anderson Sant’Ana)

11h00 – 12h00: Theory: Coding study characteristics. Effect size and types of variables. Computation of effect size and precision from primary studies (Dr. Ursula Gonzales-Barron).

Day 1 (Afternoon):

14h00 – 15h30: Theory: Effect sizes based on proportions and two-group binary data. Combining effect sizes. Fixed-effects versus random-effects model (Dr. Ursula Gonzales-Barron).

15h30 – 17h00: Practice: Searching the literature and extracting data and study characteristics from primary studies. Getting acquainted with the R environment. Meta-analytical data preparation, importing datasets (Dr. Vasco Cadavez).

Day 2 (Morning):

09h00 – 10h00: Theory: Effect sizes based on two-group means and correlations. Fixed-effects versus random-effects model. Assessing heterogeneity. Graphical displays (Dr. Ursula Gonzales-Barron).

10h00 – 12h00: Practice: Worked examples for proportion data and two-group binary data: (i) Meta-analysis for estimating overall pathogen prevalence in a foodstuff; (ii) Meta-analysis of relative risk for summarising the overall effect of an intervention. Fixed-effects versus random-effects. Evaluation and interpretation of results. Selecting between meta-analysis models (Dr. Vasco Cadavez).

Day 2 (Afternoon):

14h00 – 17h00: Social gathering

Day 3 (Morning):

09h00 – 10h00: Theory: Identifying and quantifying heterogeneity. Publication bias. Confidence versus prediction intervals (Dr. Ursula Gonzales-Barron).

10h00 – 12h00: Practice: A worked example for two-group means: Meta-analysis of mean difference for summarising the overall effect of an intervention. Explaining heterogeneity. Bubble plots. Assessing for publication bias (funnel plot) (Dr. Vasco Cadavez).

Day 3 (Afternoon):

14h00 – 15h30: Theory: Meta-regression. Introduction of study characteristics as moderating variables in multilevel meta-analysis models (Dr. Ursula Gonzales-Barron).

15h30 – 17h30: Practice: A worked example of meta-regression: Meta-analysis and predictive microbiology modelling. An application of a multilevel meta-analysis model based on a predictive microbiology equation (Dr. Vasco Cadavez).

The participants will get a (voluntary) homework assignment that they can send to the teachers no later than four weeks after the course. In case of a positive assessment of the homework assignment, the participants will get a certificate that they passed this MSc-level course and that it had a work load of 70 hours, which corresponds to 2.5 ECTS credit points.

PhD students can alternatively choose to write a short paper, where they analyse their own data set with the methods that they learned in the course. If they send the paper to the teacher no later than three months after the course and their paper is positively assessed, they will get a certificate that they passed this PhD-level course and that it had a work load of 170 hours, which corresponds to 6 ECTS credit points.