ERIC: Common Measures & Stats Flashcards
Measures of Frequency (ERIC)
1 of 4 common epi measures
how often something occurs
(not a proportion like in bio)
3 common frequency measures (ERIC)
- risk/”incidence”
- rate/”incidence rate”
- prevalence
Risk/”incidence” forumula (ERIC)
1 of 3 common frequency measures
of new cases this year/total population this year = part:whole of cases this year
Another name for risk/”incidence”
attack rate…which isn’t actually a rate gah! (is also know as…)
Risk/”incidence” definition (ERIC)
1 of 3 common frequency measures
probability of an unaffected person developing a health outcome/dx, during a given time period
“Incidence rate” definition (ERIC)
1 of 3 common frequency measures
describes how quickly a health outcomes is moving through a population in terms of # of cases and person-time
“Incidence rate” formula (ERIC)
1 of 3 common frequency measures
of NEW cases/total amount of time that each participant in a study contributed to the study
Prevalence definition (ERIC)
1 of 3 common frequency measures
proportion of population that already has a dx/health outcome
Prevalence formula (ERIC)
1 of 3 common frequency measures
of affected individuals/total populations
Measures of Association (ERIC)
1 of 4 common epi measures
used to compare relationships/impacts between various specific exposures and health outcomes
Measures of Significance (ERIC)
1 of 4 common epi measures
p-value (usually 0.5) threshold used to determine whether an association between an exposure and health outcome valid or just d/t chance
Measures of Precision (ERIC)
1 of 4 common epi measures
shows how precise data measurements are
(by using confidence intervals (ranges of uncertainty) to express the variability in means, risks ratios, rate ratios, odds ratios, etc calculated for a study)
4 types of Measures of Association (ERIC)
- risk difference
- risk ratio
- rate ratio
- odds ratio
Risk difference formula
1 of 4 types of measures of association
(# of cases in exposed group/total # at risk in exposed group)/(# of cases in control group/total # at risk in control group)
Risk ratio formula
1 of 4 types of measures of association
RATIO of risk in exposed group : risk in unexposed/control group
Ratio > 1 = exposure has harmful effect
Ratio < 1 = exposure has protective effect
Rate ratio formula
1 of 4 types of measures of association
RATIO of the RATE in exposed group : Rate in unexposed/control group
Ratio > 1 = exposure has harmful effect
Ratio < 1 = exposure has protective effect
Odds ratio formula
1 of 4 types of measures of association
Risk difference definition
1 of 4 types of measures of association
aka “attributable risk”; compares risk differentials between groups to show how much risk is d/t an exposure under investigation
Risk ratio definition
1 of 4 types of measures of association
describe/measure the strength of an association between an exposure and a health outcome (in terms of risk)
Rate ratio definition
1 of 4 types of measures of association
describe/measure the strength of an association between an exposure and a health outcome (in terms of rate)
Odds ratio definition
1 of 4 types of measures of association
an approximation of risk ratio or rate ration for the purposes of use in cohort studies (allowing researchers to generalize results to broader populations?)
2 possible outcomes of risk difference assessment
- Positive difference- exposure in question confers higher risk of a specific health outcome
- Negative difference- exposure in question confers protective effect against a specific health outcome
Null value
A risk, rate, and/or odds ratio value to suggests the exposure in question has not effect on/association with health outcome at all.
Usually equals 1.0 but thresholds are set based on the nature of the study
Null hypothesis
proposes that there is no association between an exposure and health outcome under question
Confidence Interval
method for expressing uncertainty; large/wide intervals bad, small/tight intervals good
involves a range of plausible values for the measure of an association (i.e. between and exposure and a health outcome)
Point estimate
research/statistics jargon
a specific value used to estimate a mean, standard deviation, etc for a population characteristic under investigation
i.e. risk ratio, rate ratio, or odds ratio
Parameter
research/statistics jargon
a measurable characteristic of a population, such as a mean or a standard deviation
2 common types of epi measurement estimates (ERIC)
- crude values
2. adjusted values
Crude estimates
1 of 2 common types of epi measurement estimates
involves raw numbers/counts w/o nuance (such as differences between different sub-pops within the counted pop)
Adjusted estimates
1 of 2 common types of epi measurement estimates
involves making adjustments to raw numbers/counts so that the impact of differences sub-pops within the counted pop can be accounted for)