W1 & W2 Flashcards
What is a systematic review?
Review of a clearly formulated question
Uses systematic + explicit methods to identify, select + critically appraise relevant research, to collect + analyses data from the studies that are included in the review.
List the 5 steps involved in a systematic review
Formulate the q
Search for studies
Asses quality of studies
Summarise evidence
Interpret the new findings
List some characteristics of a systematic review
Clear question that needs answering
Exhaustive search strategy
Clear inclusion + exclusion criteria
Explicit + transparent methods that are accountable, replicable + updatable
Clear logical rationale that is reported to the reader
A systematic review is a review of a clearly formulated question using what kinds of methods?
Systematic + explicit methods
What comes under the search process for systematic reviews?
Searching multiple bibliographic databases
Utilise advanced search techniques
Why search multiple bibliographic databases?
To ensure you have been systematic + comprehensive in your approach
Why Utilise advanced search techniques
To ensure all relevant studies are identified.
File drawer problem
Publication bias
When the outcome of an experiment affects the decision to publish it
Reasons for the file drawer problem
Journals want to publish clear findings
Non-sig (null) findings are often ambiguous
Harking
Hypothesising after the results are known
Why is harking really problematic?
Due to impacting on how the p-values are interpreted.
P-hacking
Doing whatever to the data to get a sig. result
Data fishing
Digging through large datasets to hunt for associations/differences will most likely give you a few.
What does a researcher decide?
Which observations are recorded
Which factors to control for
How terms are defined
When were p values developed
1930s
What are p values calculated from?
Mean scores
Variance
Sample size
p value
<0.05
Significant
Reject null hypothesis (no difference between the means)
Type 1 error
Finding a sig difference when there actually isn’t one.
Incorrect rejection of a true null hypothesis
What does a type 1 error usually lead researchers to do?
Conclude that a supposed effect or relationship exists when actually it doesn’t;t
How should p-values be interpreted?
In context of their effect sizes + prior likelihood.
What guidelines are there for helping what science is real and what isn’t?
PRISMA guidelines
What can a small sample mean?
Can often mean the study is underpowered
Problematic w/ imprecise measurement.
Chance for a type 1 error
1 in 20
Preregistration
Outlining methods + analysis strategy before conducting your research,
