History
Steps of Synthesis Research
Hard Decisions of Syntheses
Meta-Analysis
100

In 1976, Gene Glass coined this term for the statistical analysis of a large collection of results from individual studies.

What is meta-analysis?

100

In a regular experiment, we use participants. This is what takes its place in a research synthesis.

What are studies?

100

This hard decision asks whether a search should include unpublished research.

What is Publish or Perish? 

100

Meta-analysis happens during this step of a research synthesis.

What is Step 4: Analysis and Interpretation?

200

In 1904, this statistician published what’s considered the first meta-analysis.

Who is Karl Pearson?

200

The step where we search for databases, journals, and unpublished work for studies relevant to our synthesis. 

What is data collection?

200

This is the saying for the argument to exclude flawed studies.

What is "garbage in, garbage out"?

200

This method sorts studies into three piles and lets the largest pile win.

What is vote counting?

300

These methods took off in this decade, when social science research and interest in social policy grew tremendously.

What are the 1960s?

300

This is the recommendation to ensure there are minimal coding mistakes. 

What is having two or more people code the studies, and then checking inter-rater reliability?

300

The alternative to including everything is to use only studies that meet this standard (from Slavin).

What is the best evidence approach? 

300

Computing separate average effect sizes for subsets of studies, like print vs. video ads, is called this analysis.

What is moderator analyses?

400

This collaboration, begun in 2000, produces syntheses on social policy topics like education, social welfare, and crime and justice.

What is the Campbell Collaboration?

400

A synthesis reports no search terms and no inclusion criteria. Name the step and the consequence.

What is Public Presentation (Step 5), and no one can replicate the synthesis?

400

You can’t tell whether studies share one true effect, so you run both models and compare. Name both models and the decision.

What are fixed effects and random effects (#5)?

400

Because the assumptions behind meta-analysis are generally known, some _______ ____ can be removed. (Be ready to explain!)

What is statistical bias?