Epidemiology: Causation & Validity Flashcards
The ideal sample population is a selection of individuals that: __________________________________________ .
reflects the population being studied.
When the risk of disease is greater than what a multiplicative effect between two factors would predict, it indicates between the exposures.
Interaction
When a variable is associated with the exposure it means that the variable is seen to occur significantly more frequently among the group.
Exposed group
A cross-sectional study (can/cannot) establish causality between a risk factor and disease.
Cannot
Interaction between exposures can occur when there is a (confounding/causal) relationship between exposure and disease.
Causal
When the incidence of disease is due to the direct effect of two different exposures added together, it is referred to as (additive/synergistic) .
Additive
Confounding generally occurs when the two groups being compared (are/aren’t) similar to one another
Aren’t
When the effect of multiple factors added together leads to a higher rate of disease than expected, it is termed .
Synergism
A factor is (necessary/sufficient) in a causal relationship if disease always develops in its presence.
Sufficient
A factor is (necessary/sufficient) in a causal relationship if it is needed to instigate the development of disease.
Necessary
An association is (sensitive/specific) when a particular factor only causes one disease.
Specific
Randomization (does/does not) always eliminate sampling bias.
Does not
In (direct/indirect) causation there are intervening factors that cause disease along with the original risk factor.
Indirect
When the effect of multiple disease factors added together is less than expected, it is termed .
Antagonism
Interaction between risk factors for a disease process (does/does not) depend on the biology of the disease.
Does
is one tool that can be used to control selection bias where individuals get selected to enter the study based on chance.
Randomisation
The problem with a study with selection bias is that the results might lack (internal/external/internal and external) validity.
Internal and external
In order to have internal validity, the two populations being compared must have with the exception of the variable being tested.
similar baseline characteristics to one another
If the risk of cancer with one exposure is seven and the risk of cancer with another exposure is three, but the risk of cancer with both exposures is sixty, the relationship is (multiplicative/synergistic) .
Synergistic
In order to make sure multiple study groups are similar, the 3 methods we can use are:
randomization, restriction of the study population, and matching
Confounding is the term used in a study in which an _________ is the reason for an assumed causal relationship between an independent and dependent variable.
An alternative explanation
A confounder is a variable in a study that: _________________.
distorts the true relationship between an exposure and an outcome.
When judging an association for , considering alternate explanations ensures that all confounding variables have been taken into account.
Causality
If a certain exposure doubles the risk for disease, it is referred to as (additive/multiplicative) .
Multiplicative
A necessary and sufficient causal relationship (rarely/frequently) occurs because it is a consequence of a one-to-one relationship of exposure to disease.
Rarely
One criteria for causality is , meaning that the exposure to a factor always occurs before the disease develops.
Temporality
_____________ refers to a situation in which multiple factors have more than an additive effect on the pathogenesis of disease.
Interaction
The strength of an association to determine causality between two variables is measured by the (absolute/relative) risk, or odds ratio.
Relative
When looking at the relationship between obesity and heart disease, cholesterol levels can be a third variable, since increase in obesity is typically associated with an increase in cholesterol, which is also associated with an increased risk of heart disease. In this scenario cholesterol levels (is/is not) a confounder.
Is not
