meta analyses evidence combining Flashcards
how meta analysis work in research
1
Q
what is meta analysis
A
- combination of evidence from multiple studies
- systematic and quantitative approach to synthesising evidence
- Cochrane database
- compare a bunch of papers statistically
- problem with file drawer and biased reporting
2
Q
what is a narrative review
A
- selects articles (subjective)
- risk of citation bias
- systematic reviews would be slightly better
3
Q
what are the guidelines for a meta analysis
A
- Participants
- Interventions
- Comparisons
- Outcomes
- Study design
4
Q
what is ego depletion
A
- self control or will power draws upon limited pool of mental resources
- depleted self-control increases suboptimal ways
- publication bias
- real effect is closer to zero
5
Q
what are the competitive reaction times
A
- in competition with someone more likely to become aggressive
- everyone wants a method of scoring but don’t want to use someone else’s
6
Q
what are the counteracting biases
A
- identify publication bias
- identify P-Hacking
- correct effect size estimates
- improve estimates resulting from meta-analysis
7
Q
what are incomplete reporting
A
- study is published but doesn’t have all the data in
- Lakens et al., 2017 found 7% used reporting guidelines & 67% list all individual effects
8
Q
what are the recommendations for solutions
A
- future proof meta analysis
- quality control
- PRISMA
- pre-register
- share data and code
- consult statistician or librarian
9
Q
what are the statistics used in meta analyses
A
- types of effect sizes
- Cohens d
- Eta squared
- odds ration
- Z test
- measure of heterogeneity
