Term
|
Definition
True Positive Rate
=TP/(everyone that is positive, TP + FN) |
|
|
Term
|
Definition
True Negative Rate
TN/(All Negatives) FP + TN
used to rule in disease.
If there is a very high TN rate that means a positive result is more likely to be positive |
|
|
Term
|
Definition
Snapshot in current time of disease and risk factors
Prevalence Sick/Population
|
|
|
Term
| Case Control Study compares two groups. Sick and not Sick. |
|
Definition
Watch people over time. Measure with odds ratio
Percent of sick people with a risk factor.
(a/c)/(/b/d)
a = exposed and diseased
c= not exposed and diseased
b= exposed and not diseased
d=not exposed and not diseased. |
|
|
Term
Cohort-
Compares group with exposure to groups without.
measure with RR |
|
Definition
Risk in exposed group (a/a+b)
divided by risk in unexposed group ) (c/c+d) |
|
|
Term
Drug Trials
phase 1,2,3,4 |
|
Definition
1- small # of healthy peeps
2- small number of sick
3- Random double blind
4- post marketing surveillance |
|
|
Term
|
Definition
Rate of TP versus all positives (FP)
Chance your positive test means your sick.
High Prevalence = High PPV |
|
|
Term
|
Definition
Chance that a negative test means your not sick.
Low prevalence means high NPV |
|
|
Term
|
Definition
Does the risk reduction mean lower sick rates
RRR= 1- RR |
|
|
Term
|
Definition
Risk diff between exposed and unexposed groups
AR
% Sick with Risk - % Sick without risk. |
|
|
Term
|
Definition
Difference in risk expressed as a ratio.
% Sick without Vaccine - % Sick with vaccine |
|
|
Term
| Number Needed to Treat or Harm |
|
Definition
How many people to treat to save 1, or how many people to expose to save 1
1/ARR
or
1/AR
ARR measures risk reduction and AR measures disease because of exposure. |
|
|
Term
|
Definition
| Picking wrong people for a study. |
|
|
Term
|
Definition
| Patients remember an exposure after learning they have disease |
|
|
Term
|
Definition
| Groups who know they are being studied behave differently |
|
|
Term
|
Definition
| Groups can tell which group they belong in. |
|
|
Term
|
Definition
| Doctor's belief system influences how they see the results |
|
|
Term
|
Definition
| When outside factors not being studied influence the outcome. Coal miners and fire fighters tend to smoke a lot. |
|
|
Term
|
Definition
| Early detection will increase rates of survival. So you have to compare severity of the disease at time of diagnosis. |
|
|
Term
|
Definition
|
|
Term
| SD and Standard Error of the Mean |
|
Definition
SD = measures variablity from the mean
SEM - sample mean from population mean
SD/ SQRT(sample size) |
|
|
Term
Distribution
1 SD
2- SD
3 - SD
% |
|
Definition
68% within +- 1
then 95%
then 99.7% |
|
|
Term
Error Types
type alpha
Type Beta |
|
Definition
Type 1 alpha - rejecting null in error
alpha is set as p value .05
Type 2 - accepting null in error
Power is 1-B (or prob of rejecting null when it is false)
a + b = 1 |
|
|
Term
SEM = Z
used to calculate CI |
|
Definition
|
|
Term
Relative Risk for Odds Ratio or Relative Risk includes 1
accept or reject Ho? |
|
Definition
|
|
Term
Mean difference between two variables includes 0.
|
|
Definition
|
|
Term
| Chi - squared versus ANOVA |
|
Definition
Checks for differences in differentiated groups.
ANOVA looks at the mean for 3+ groups.
Chi Square looks at percentage of each group. i.e 3/10 blacks of HTN. 5/10 whites. 7/10 Hispanics. |
|
|
Term
|
Definition
| difference between averages of two groups. Slugging percentage of two teams. |
|
|
Term
|
Definition
| Mean blood pressure of three groups. |
|
|
Term
|
Definition
R usually R^2
The rate at which something effects the outcome of something. does weather correlate to your utility bill? |
|
|
Term
| Consent of Parents for Minors |
|
Definition
| Everything but Sex Drugs and Rock and Roll (trauma) |
|
|
