Topic 16: internal/external study validity Flashcards
(12 cards)
1
Q
Define internal validity
A
- Do the results represent truth of population = is there association or results due to something else
2
Q
Define external validity
A
- Can results be generalized to general population
3
Q
Give the factors to determine internal validity
A
1) Chance = random/sampling error
2) Bias = systematic error
3) Confounding
4
Q
How to determine validity of results affected by chance?
A
- Check if finding statistically significant - p-values/95% CI
- If association significant = conclude finding is not by chance
5
Q
How to minimize influence of chance in findings?
A
- Large sample size = reduce SE
6
Q
How to determine validity of results affected by bias?
A
- Based on researcher’s relection
- Convenince sampling + self-report = expected bias
7
Q
How to minimize influence of bias?
A
- Choose sample via randon sampling methods
- Use assessment tools = high accuracy
- Thorough investigation of data collected during + after in analysis
8
Q
How to determine validity affected by confounding?
A
- In non-experimental conditions = expect to some degree
-If unexpected results + not biological sense = confounding
9
Q
How can we minimize the influence of confounding in our findings?
A
- Identify all potential confounders pre-study
- Adjust for confounders during analysis
10
Q
Give factors to determine external validity
A
1) If findings can be generalized to general population
2) To other similar populations
11
Q
How to determine whether our results have low external validity?
A
- Going through literature + evidence of association = make judgement
- Cannot assume validity without investigation
12
Q
How to improve the external validity of our results?
A
- Avoid highly selected samples = to be able to generalize
- Random sampling
- If we intentionally chose specific characteristics = make sure sample represnts populations subgroup
