When to use what test Flashcards
When to use logistic regression?
used for predicting binary outcomes of the dependent variable
Estimates odds probability of the DV occurring as values of IVs change
When to use linear regression?
predict outcome of a continuous variable
when to use one-sample t-test
wanting to see if a population mean differs from a particular value
Ex: is mean BP of patients 140 mmHg? - one particular value you now in advance
when to use paired samples t-test
before and after study (measuring something on a parient twice before and after treatment)
used to see if there is a difference in the means
matched pairs (if looking at tumor samples do study where we get a tumor sample from patient and nontumour sample and look at difference in expression in 2 samples
IV should have two groups
independent samples t-test
used to see if difference in means
RCT, case/control, cross-sectional
IV should have no more than 2 groups
when to use correlation?
to test if linear relationship between variables
When to use chi-square?
compares observed frequencies to expected
voting behavior in males and females. is there a difference by gender?
When to use ANOVA?
tests the significance of group differences between two or more groups
the IV has two or more categories
only determines that there is a difference between groups, but doesn’t tell which is
different
When to use multiple regression?
• used with several independent variables and one dependent variable
• used for prediction
• it identifies the best set of predictor variables
• you can enter many IVs and it tells you which are best predictors by looking at all of them
at the same time
• in sequential regression the computer adds the variables one at a time based on the
amount of variance they account for
When to use Mann-Whitney U Test?
When want to test difference between independent samples but independent t-test assumptions aren’t met
When to use Kruskal-Wallis?
When looking at difference between many groups and ANOVA assumptions not met.
Pops don’t have to be normal
When to use Wilcoxon Signed Rank?
When looking at differences between paired groups and paired t-test assumptions don’t work
When to use McNemar’s chi-square test
If have paired binary data and t-test assumptions aren’t met (Yes/No questions on 2 questionnaires administered before and after therapy)
When to use Spearman’s correaltion
When Pearson correlation assumptions aren’t met
What are non-parametric tests to look at regression?
LOWESS, Splines, Kernel estimation, Wavelets, GAMs (Generalized additive models)
