Confounding, validity, reliability

Question 1

define: confounding
- verbal
- symbolic

Answer 1

there is an apparent assocation between X and Y which is also influenced by another variable

X –> Y
but also
X C
C –> Y

Question 2

bias vs confounding

Answer 2

bias: creates untrue assocation
confounding: creates true but misleading assocation

Question 3

how to determine confounding?

Answer 3

two by two table, with confounder labelled

disproportionate –> confounding

Question 4

3 ways to correct confounding at the design stage?

Answer 4

restriction: people with CV not included
random allocation (spread out)
matched participants (on CV)

Question 5

2 ways to correct confounding at analysis stage?

Answer 5

stratification: analyse results separately for with vs without CV
statistically modelling: accounts for CV

Question 6

construct validity

Answer 6

does the test measure what its meant to measure

Question 7

internal validity

Answer 7

did X really cause Y or is there a CV?

Question 8

external validity?

Answer 8

can the relationship be generalised ?

Question 9

test retest reliability

Answer 9

same results when test is repeated?

Question 10

inter rater reliability

Answer 10

same results between judges?