Use LEFT and RIGHT arrow keys to navigate between flashcards;
Use UP and DOWN arrow keys to flip the card;
H to show hint;
A reads text to speech;
40 Cards in this Set
- Front
- Back
Variable
|
Characteristic that takes on different values in different persons, places or things. Able to quantify information
|
|
Discrete variable
|
characterized by gaps or interruptions in values that it can assume
Example: # of siblings = must be 0, 1, 2, 3, 4…… it cannot be 5.3 siblings |
|
Continuous variable
|
variables that can take on any value
Ordinal variables are usually considered continuous |
|
Categorical Variables
|
Must assign a number to, but that number in itself means nothing
|
|
Independent Variables
|
These are “manipulated” by the experimenter
Ex: In epidemiology the exposure is the independent variable and the disease or outcome is the dependent variable |
|
Dependent variables
|
are the outcome
|
|
Regression is as simple as Y=mx + b
|
X=independent variable; y= dependent variable
|
|
Study Design Sequences
|
used to answer research questions, and create new knowledge
|
|
Case-series Study Design
|
-A series of cases
-No control/comparison group -Cannot calculate relative risk, attributable risk or odds ratios -Cannot makes statements about preventing disease -Causation cannot be shown |
|
Ecological Study Design
|
-Compare groups, but do not have data on individuals
-Used to generate rather than test hypotheses -Apt to ecological fallacies |
|
Cross-sectional study
|
-Observational, no intervention
-Effective at providing a picture of disease, health or psychological phenomenon -effective at assessing several variables at once and associations between those variables |
|
Cross-sectional study
|
-Disease and exposure data collected simultaneously
-Can determine distribution of disease and exposure -Problem: Temporal sequence not always known (ex: runners/thin) |
|
Case-control or retrospective study
|
-Observational
-Compare the past exposures of cases and controls -Collect data after onset of disease & Look backward in time to find the exposure |
|
Prospective studies or “cohort studies”
|
-All subjects are “well” at the onset
-Data collected before the disease, temporal sequence known -Provides incidence data -Divide on some exposure (ex: smoking) and look for outcome in future |
|
Experiments
|
-intervention occurs
-randomization and manipulation |
|
Quasi experimental design
|
-there is manipulation but not randomization
|
|
Cohort studies
|
-Cohorts are cohorts because of some common exposure
-Follow forward in time and look for an outcome -Gets incidence rates -Can add new exposures/outcomes -Temporal sequence know |
|
Relative Risk
|
Shows strength of association of exposure and disease
|
|
Relative Risk
|
-if RR > 1.0 then the factor is said to be causative
-if RR= 1.0 the factor is unrelated to the disease, "the factor is just as likely to occur among those with the factor as those without it" -if RR< 1.0 the factor is said to be "protective” |
|
Attributable Risk
|
-measure of how much disease could be eliminated if the exposure was not present
-estimates the proportion of disease among the exposed that is attributable to the exposure or the proportion of the disease that could be eliminated if people did not |
|
Population attributable risk
|
-indicates the proportion of all cases in the total defined population that can be ascribed to a factor
-the % of disease among the entire population that can be attributed to the exposure |
|
Bias in case control studies
|
*selective memory- those with the disease remember better than those without
*Selective survival-the result of differences between those who survive and those who live |
|
Odds ratio
|
-an estimate of the relative risk, will be similar unless the incidence is high in the exposed group
|
|
internal validity
|
can we ascribe the observed effect to the exposure / variable under investigation
|
|
External validity
|
do the results apply to others or other situations? Can we generalize the results beyond the observed findings to some universal statement?
|
|
7 main threats to validity
|
1. Testing: people improve over time
2. Instrumentation: same instruments should be used 3. Maturation: Over time people mature and maturation will change them whether or not they are in your study 4. Regression toward mean 5. Selection bias 6. History: Outside interference to study 7. Attrition: drops out of study |
|
bias
|
misclassify variables in a consistent way that tips the scale in a certain direction
|
|
Reliability
|
-measure of consistency & repeatability
- Does NOT mean accuracy |
|
Inter rater reliability
|
results might differ when 2 or more people rate something
|
|
Intra rater reliability
|
-same person rates observations
- same person might rate things differently |
|
Calculate Inter rater reliability & Intra rater reliability
|
# agreements / total # of possible combinations
|
|
Validity Components
|
sensitivity & specificity
|
|
sensitivity
|
if you have the disease/outcome how accurate is your measurement instrument in identifying that
|
|
specificity
|
if one truly doesn't have the disease, how well does the measurement tool accurately detect no disease.
|
|
Calculate sensitivity
|
(true positives) / (true positives + false negatives)
|
|
Calculate specificity
|
true negatives / all that are truly negative
|
|
Positive predictive value:
|
Proportion of true positives among those who test positive
a / (a + b) |
|
Negative predictive value
|
proportion of true negatives out of the total who test negative
d / (d + c) |
|
Confounding
|
occurs when the crude odds ratio or RR differs from the adjusted ones
occurs we say that an interaction has occurred |
|
interaction
|
occurred when a variable affects an outcome differently at different levels of another variable
|