random Flashcards
what is the fundamental problem of causal inference
you can never know what would have happened if you did things differently
- we cannot observe the counterfactual
what is random error
fluctuations in the profile of he sample
what is confounding bias
distortion of the casual association between two variables due to a common shared cause (a confounder)
how do we reduce confounding bias
conditioning
estimate the unconfounded effect of obesity on cancer we would need to condition on exercise levels
what are the three ways we condition to reduce confounding bias for example effect on obesity in cancer
restriction - restrict sample to single value of the confounder ie association in people who do zero exercise
stratification - calculate category specific effects for different levels of the cofounder - stratify across exercise levels
covariate adjustment - adjust for exercise covariate in a regression of obesity in caver
why does conditioning not completely get rid of confounding factors
other confounding variables that we did not measure ie unobserved confounding
error in our measure of exercise - imperfect conditioning - residual confounding
what is the effect on a larger sample size on error
reduces random error but has not effect on systematic error
how do precision and accuracy affect error
precision increases random error decreases = increasing sample size
accuracy increases as systematic error decreases - sample size does not affect this
what is the difference between measurement error and measurement bias
e - error in your measurement due to random factors eg weighing scales varying on climate
b - error in your measurement due to non-random factors ie weighing scales are broken
what is misclassification error
occurs when measurement error and measurement bias result in misclassification
what are the 6 types of bias in any study
inferential selection analytical information confounding experimenter
what are the 4 types of selection bias
selection - due to systematic difference between those sleeted into a study samples and those not selected
sampling - due to faulty sampling by investigator
participation - due to behaviour of participants
attrition - due to loss of participations from study
what is information bias
occurs due to systematic error in reporting, measurement or recording of information
what are the subtypes of information bias
response bias - people responding in inaccurate ways (acquiescence bias - people prefer yes, true or agree / social desirability bias - people downplay undesirable traits and exaggerate desirable ones)
recall bias - people have different abilities to remember past information
observer effect - people behaviour and response differently when they know they are being observed
what is experimenter bias and the subtypes
bias due to the behaviours and actions of the experimenter, whether conscious or unconscious
confirmation bias - more likely to accept findings that we expect and refute findings that we don’t
systemic bias - more likely to chase positive associations, seek results which support our carer
