Statistical Tests Flashcards
(31 cards)
What does the Chi-squared test assess?
It tests the association between 2 categorical variables, such as determining whether one treatment is better than another.
What type of data does the Chi-squared test use?
Categorical data (nominal/ordinal).
What type of test is the Chi-squared test?
It is a non-parametric test.
When is Fisher’s exact test used?
It is used for very small samples (<20) to calculate exact probability.
What does Fisher’s exact test determine?
It determines whether there is a non-random association between 2 categorical variables in a 2x2 categorical table.
What is the alternative to the Chi-squared test?
Fisher’s exact test, especially when sample sizes are small or expected cell counts are <5.
What does a parametric test assume?
It assumes normal distribution and homogeneity of variance.
What are examples of parametric tests?
t-test, ANOVA, Pearson correlation, and linear regression.
What type of data do non-parametric tests use?
Ordinal or non-normally distributed data.
What are examples of non-parametric tests?
Mann-Whitney U, Wilcoxon signed rank, Kruskal-Wallis, Spearman correlation.
What is the Wilcoxon Rank-Sum Test used for?
It is used for paired/matched data or small samples (n<25).
What is the Mann-Whitney U test used for?
It is used for unpaired/unmatched ordinal data.
What does the Friedman Two-Way ANOVA by Ranks test?
It is a non-parametric test for >2 matched/paired groups.
What does the Kruskal-Wallis One-Way ANOVA test?
It is a non-parametric test for >2 unmatched groups.
What does the independent (unpaired) t-test compare?
It compares means of two different groups.
What assumptions does the independent t-test make?
Data is normally distributed and has equal variances.
What does the paired t-test compare?
It compares means of the same group at two time points.
What does One-Way ANOVA test?
It compares means of >2 independent groups.
What is the null hypothesis?
It is the default assumption that there is no effect, no difference, or no association between variables.
What does a p-value < 0.05 indicate?
You reject the null hypothesis, suggesting an effect exists.
What does a p-value ≥ 0.05 indicate?
You fail to reject the null hypothesis, indicating insufficient evidence of an effect.
What does ANOVA tell you?
It tells you if there is a difference among groups but not where the difference lies.
Why are post-hoc tests needed after ANOVA?
To determine which specific group pairs differ and to control for type 1 error due to multiple comparisons.
What is the Bonferroni Correction?
It is a post-hoc test that is very strict.
