important definitions Flashcards
what does error mean?
the difference between an estimated/ measured value and a true value
what does bias mean?
systematic, non-random deviation of results or inferences from the truth, or processes leading to such a deviation
(leaning to one side)
why might error due to selection arise?
sample selection - participation effects mixed with exposure effects
when does diagnostic bias occur?
when diagnosis made on basis of exposure
when might self-selection bias occur?
when subjects make decisions e.g. whether or not to participate
‘healthy worker effect’ = workers self-selecting as more likely to be heathier as remain at work whereas unwell people retire
is participation bias a type of selection bias?
yes
when does misclassification bias occur?
when data is placed into categories
e.g. smokers more likely to be diagnosed with emphysema than non-smokers with emphysema
list some types of information bias
surrogate bias
recall bias
interviewer bias
sample?
small subset of population
population
every member of a group of interest
case
person with a disease of interest
control
person without a disease of interest
randomise
each unit has equal probability of being selected
controlled
compares the results obtained from an experimental sample against a controlled sample
ecological study
counting numbers of people affected in a group
register
an official list or record of names or items
counterfactual
‘counter to the facts’
things that don’t happen ‘counter to the fact’
would have expected them to be exposed
used to describe causation
analytical
relating to or using analysis or logical reasoning
sampling bias
selection of particular sampling units therefore differences between the sample and the population
confounding
influence of a third variable so as to distort the association being studied between two other variables
intention to treat
assess the effect of the treatment regimen by evaluating the basis of the intervention to treat a patient rather than the actual treatment given
placebo
inert substance given as medication for its suggestive effect
cause
a factor that a person experiences
causality
the relationship between cause and effect
descriptive
concerned with a classification or description
anecdote
a short account of an incident
cross-sectional
study what is happeneing in group at particular point in time
case-control
retrospective, compares control and cases for a expsoure
cohort
a group who share a characteristic at a specific point in time and a followed forward in time
experiemental
risk factors under control of the experimenter
observational
results obtained retrospectively or without a control group
retrospective
analysis of data collected in the past
prospective
expected to happen in the future
epdiemiology
incidence, distritbution, possible control of disease and other factors relating to health
