non-parametric test Flashcards
non-parametric version of the t-test. comparing two samples
mann-whitney u-test
non-parametric version of ANOVA
3 or more groups
kruskal-wallis one-way analysis of variance
non-para version on a paired t-test
compares 2 paired groups
wilcoxon matched pairs
non-para version of a repeated measures ANOVA
compares 3 or more paired group
followed w post-test
friedman test
when do you use a non-parametric test?
-cannot assume normal population
- variance is unequal
- small samples
- nominal or ordinal data
Formula for % sensitivity
TP/TP+FN proportion of positive tests among
patients that have the disease
Formula for % specificity
TN/FP+TN proportion of negative tests among
patients that are healthy
PPV formula?
TP/TP+FP
Disease prevalence formula?
TP+FN/ TOTAL SUM
ability of
an analytical method to assess small variations of
the concentration of the analyte
Analytical sensitivity
ability of an
assay to determine specifically the concentration
or detection of the target analyte in the presence
of interfering substances
Analytical specificity
defined as the fraction of
patients with a disease that the assay correctly
identifies
Clinical sensitivity
defined as the fraction of
patients without a disease that the assay
correctly identifies
Clinical specificity
What would the null hypothesis be for a Chi-
square analysis?
H 0 = There is no relationship between variable
A(studying) and variable B (passing)
What would the research hypothesis be for a Chi-
square analysis?
H a= There is a relationship between variable A
(studying) and variable B (passing)
If the frequencies (counts) are <5, then it is best
to use an alternative to Chi-square called the
Fisher exact test
What happens if chi squared value is greater than critical value?
Significant, reject null
What happens if chi squared value is less than critical value?
Not significant, fail to reject null
If observed population is deemed as not significant, is it in hardy weinburg equilibrium?
Yes
Compares a sample mean to a given population
mean
One sample t-test
two-sample t-test • Compares two sample means • Is one sample mean significantly different from
the other?
Students t test
Requires a set of paired observations from a
normal population • compares one set of measurements
with a second set of measurements from the same sample • Can be used to compare “before” and “after”
scores
Paired t-test
one independent variable/factor WITH three or more levels. determines if observed differences among a
set of means are statistically significant from each other
One way ANOVA
Other independent variables/factors can be
added to the mix, looking for interactions between the independent variables Ex: Measure effects of 3 drugs
AND 3 diet regimens on blood pressure
Two way anova
Analogous to the paired t-test Aka “within-subjects” design. Advantage of using is that
individual differences are controlled. Includes multiple dependent variables, Scenario: Effects of three different medications on diastolic and systolic blood pressure (DBP, SDP)
MANOVA
