23 Flashcards
1
Q
Internal validity
A
The extent to which the study findings are free of chance, bias and confounding
2
Q
What are p-values?
A
Probability of getting study estimate (or one further from the null), when there is really no association, just because of sampling error (chance)
If probability really low, then unlikely that estimate is due to sampling error (chance)
3
Q
Interpretation of a p value
A
The probability of finding an OR of 2.3 (or further from the null) when the null hypothesis is true is 0.03 (or 3%).
4
Q
T ype-II errors
A
Incorrectly fail to reject H0 when should have
Typically due to having too few people in the study
Statisticians can calculate power to find out how many participants are needed to minimise chance of a Type-II error
5
Q
Why p-values are problematic
A
- Arbitrary threshold
- Only about H0
- Nothing about importance
