PHCY 500 Biostats Exam Review

mean (SD)

A hint for determining if data is continuous is if it’s reported as:

median (range)

A hint for determining if data is categorical or nonparametric is if it’s reported as:

false alarm

Type 1 error can be thought of as:

oops missed it

Type 2 error can be thought of as:

selection bias

bias introduced by how participants are chosen

interview bias

bias introduced by how interviewer behavior influences repsonses

recall bias

bias associated with remembering

publication bias

bias introduced by selectively omitting data from publication

external validity

validity of if the data can be generalizable beyond the sample

internal validity

validity of if the study was designed well and if it controlled confounders

construct validity

validity of if the test measured a specific construct like it intended

interrater reliability

reliability of extent of agreement between two or more raters

test/retest reliability

reliability of degree to which test or instrument scores are consistent from one point in time to another

prevalence

all individuals affected at a particular time (think past tense)

incidence

new individuals that will become affected over a period of time (think future tense)

relative risk

likelihood of disease development in exposed vs not exposed in future time (incidence)

odds ratio

current odds of disease in exposed vs. not exposed in current/past time (prevalence)

statistically significant

If the confidence interval contains 1, the data cannot be considered _________

Chi Square

a test to compared two categorical variables

Fisher exact

this form of Chi square should be used if n<5

McNemar

this test should be used when comparing two categorical variables in a paired manner

Students T test

a test to compare a continuous variable with a categorical variable that contains 2 groups

Mann Whitney U

this form of student t test should be used if the data is nonparametric

ANOVA

a test to compare a continuous variable with a categorical variable that contains 3 or more groups

Kruskal Wallis

this form of ANOVA should be used if the data is nonparametric

Bonferroni correction

if an ANOVA test is performed and a significant difference is found, a __________ should be applied

paired t test

a test to compare a continuous variable with a categorical variable that contains 2 groups that are not independent of each other

Wilcoxon

this form of paired t test should be used if the data is nonparametric

strength and direction

correlation is a measurement of what 2 things?

Pearson

this correlation coefficient is used is used when the data is parametric

Spearman

this correlation coefficient is used when the data is nonparametric

Linear regression

this test estimates the relationship between a continuous DV and one or more continuous IVs

unstandardized (B)

if a question uses the terminology “every one unit change”, which regression coefficient do you use?

standardized (beta)

if a question uses the terminology “largest magnitude of effect”, which regression coefficient do you use?

logistic regression

this test measures relationship between a categorical DV and one or more IV, categorical or continuous; the choices are dichotomous, and the results are reported as an odds ratio

survival analysis

an analysis that models time to event data

cox proportional hazards

model that interprets survival analysis data of parametric populations

Kaplan Meier

model that interprets survival analysis data of nonparametric populations

reference group

when you dummy code and assign dichotomous variables as 0 or 1, what is the 0 group called

prediction

the regression tests are used to make a ________