Term
|
Definition
| Used to compare two groups that are assumed to be independent (different populations). |
|
|
Term
|
Definition
| non-parametric version of t-test |
|
|
Term
|
Definition
| Used to compare one subject or population, usually a before and after. |
|
|
Term
| Wilcoxon signed-rank test |
|
Definition
| Non-parametric equivalent of PAIRED t-test. Use with ranked data. |
|
|
Term
|
Definition
| Tests if two or more PROPORTIONS are different. |
|
|
Term
|
Definition
| Used in place of Chi-square when there are fewer than 5 cases in any subgroup. |
|
|
Term
| Chi-square test for trend |
|
Definition
| To assess association between a nominal variable and an ordinal variable. |
|
|
Term
|
Definition
| Analysis of variance- looks for differences in means of at least 3 subgroups. |
|
|
Term
|
Definition
| Used to compare magnitude of variances in various subgroups. |
|
|
Term
|
Definition
| Non-parametric equivalent of the f-test. Used to compare continuous, non-normal distributions of variables. |
|
|
Term
|
Definition
| Used to assess the change in 2 or more continuous measurements made on the same person. Same analysis as ANOVA, but different data type. |
|
|
Term
|
Definition
|
|
Term
| Cox Proportional Hazards Regression |
|
Definition
| Used to assess relationships of factors (age, treatment group) with time-to-event curves (i.e., survival curves). |
|
|
Term
|
Definition
| A certain result is "x" times more likely to be observed in a patient with a certain condition than a patient without that condition. |
|
|
Term
|
Definition
| How well the study was conducted, i.e. were the individuals blinded? were there any fatal flaws? |
|
|
Term
|
Definition
| Is the study applicable to real-world patient populations? |
|
|
Term
|
Definition
Number with disease
Number unaffected |
|
|
Term
|
Definition
Odds of disease in exposed persons
Odds of disease in unexposed persons |
|
|
Term
|
Definition
Number with disease
Number at risk for disease |
|
|
Term
|
Definition
Number of new cases in a given time period
Number of persons at risk for a disease |
|
|
Term
|
Definition
Incidence (exposed)
Incidence (unexposed) |
|
|
Term
| Pearson's correlation coefficient |
|
Definition
| Used to measure the correlation within a continuous data set. r varies between 1 and -1. |
|
|
Term
|
Definition
| Correlation coefficient used for RANKED data |
|
|
Term
|
Definition
| Unordered, categorical data (dead or alive, hair color, etc.) |
|
|
Term
|
Definition
| Ordered, categorical data. (Agree, neutral, disagree) |
|
|
