Statistics Flashcards
Gold standard for determining causality
RCTs
Prospective study grouped according to exposure status for a risk factor
Cohort study
Prospective study subjects randomly assigned to groups and followed
RCT
Retrospective study measurements taken after occurence of outcomes grouped by outcome
Case-control
Measurements of possible exposure and outcomes made at single point or period in time
Cross-sectional
Incorrectly rejecting the null hypothesis
Type I error
Incorrectly accepting the null hypothesis
Type II error
Number of new cases in a specific population over a specific period of time
Incidence
Total cases in a population at a specific point in time
Prevalence
Probability of testing positive if you have the disease
Sensitivity
Probability of testing negative if you do not have the disease
Specificity
Test to determine difference between means of two groups
t-test
Test for analyzing to known differences between two groups
z-test
Test used to determine whether there is a significant difference between expected and observed frequencies in one or more categories
chi-square
Test used to compare means among three or more normally distributed groups
ANOVA
Studies reporting positive results are more likely than studies reporting negative results, type of bias
Publication bias
Study staff can influence the results of a study; experimenters can interpret results incorrectly because of a tendency to identify information that supports their hypothesis. Blinding helps to ameliorate this type of bias
Observer bias
Study participants often have incomplete or skewed recollections of past events. This is particularly a problem in retrospective studies (eg, case-control studies)
Recall bias
Study subjects changing behavior because they are aware they are being observed, ie, they are on their best behavior
Hawthorne effect
Compares nonparametric data from the same group
Wilcoxon signed-rank test
Kaplan-Meier method for an event occurring at a certain point in time
Log-rank test
