Biostatistics & Epidemiology Reviewer: Non-Parametric Tests

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/18

flashcard set

Earn XP

Description and Tags

Vocabulary practice flashcards covering non-parametric statistical tests, including the Chi-Square Test, Kruskal-Wallis H-Test, and Spearman Rank-Order Correlation, based on Biostatistics & Epidemiology Modules 20–21.

19 Terms

1
New cards

Chi-Square Test (χ2\chi^2)

A non-parametric test used mainly for categorical data presented as frequencies or counts to determine whether there is a significant association or difference between categorical variables.

2
New cards

Chi-Square Test of Independence

A type of chi-square test used to determine whether two categorical variables are significantly associated with one another.

3
New cards

Chi-Square Goodness-of-Fit Test

A type of chi-square test used to determine whether observed frequency counts match expected frequency distributions or proportions.

4
New cards

Observed Frequency (OO)

The actual count or number of occurrences directly observed in the collected dataset.

5
New cards

Expected Frequency (EE)

The theoretical frequency expected in a cell if no association or difference exists, calculated in a contingency table as E=(Row Total)(Column Total)Grand TotalE = \frac{(\text{Row Total})(\text{Column Total})}{\text{Grand Total}}.

6
New cards

Chi-Square Formula

The mathematical formula χ2=∑(O−E)2E\chi^2 = \sum \frac{(O - E)^2}{E}, where OO is the observed frequency, EE is the expected frequency, and ∑\sum denotes the sum across all cells.

7
New cards

Contingency Table Degrees of Freedom (dfdf)

The number of independent values that can vary in an analysis, calculated for a contingency table as df=(r−1)(c−1)df = (r - 1)(c - 1), where rr is the number of rows and cc is the number of columns.

8
New cards

Fisher's Exact Test

A statistical test appropriate as an alternative to the chi-square test for a small 2×22 \times 2 contingency table when expected frequencies are too small (less than 5 per cell).

9
New cards

Kruskal-Wallis H-Test

A non-parametric test used to compare 3 or more independent groups using ranked data; it serves as the non-parametric alternative to the one-way ANOVA.

10
New cards

Mann-Whitney U Test

A non-parametric statistical test used specifically to compare 2 independent groups, as opposed to the Kruskal-Wallis test which is used for 3 or more independent groups.

11
New cards

Kruskal-Wallis Test Statistic Formula (HH)

The formula H=[12N(N+1)]∑(Ri2ni)−3(N+1)H = \left[ \frac{12}{N(N+1)} \right] \sum \left( \frac{R_i^2}{n_i} \right) - 3(N+1), where NN is the total number of observations, RiR_i is the rank sum for group ii, and nin_i is the number of observations in group ii.

12
New cards

Kruskal-Wallis Degrees of Freedom (dfdf)

The degrees of freedom for the Kruskal-Wallis H-test, calculated as df=k−1df = k - 1, where kk represents the total number of groups.

13
New cards

Dunn's Test

A post-hoc statistical test performed after obtaining a statistically significant Kruskal-Wallis H-test result to identify which specific pairwise groups differ.

14
New cards

Spearman Rank-Order Correlation (rsr_s or ρ\rho)

A non-parametric test that measures the strength and direction of the monotonic relationship between two ranked or ordered variables.

15
New cards

Monotonic Relationship

A relationship where two variables change together in a consistent order or direction (both increasing or one increasing while the other decreases), evaluated by Spearman correlation without requiring linearity.

16
New cards

Pearson Correlation

A parametric test that uses raw, actual values to assess linear relationships between continuous variables and requires the assumption of a normal distribution.

17
New cards

Spearman Correlation Formula

The formula rs=1−[6∑d2n(n2−1)]r_s = 1 - \left[ \frac{6 \sum d^2}{n(n^2 - 1)} \right] (used when there are no tied ranks), where dd represents the difference between paired ranks and nn is the number of paired observations.

18
New cards

Decision Rule for Significance (pp-value)

If p<0.05p < 0.05, reject the null hypothesis (H0H_0) indicating statistical significance; if p>0.05p > 0.05, fail to reject H0H_0 indicating no statistically significant difference or association.

19
New cards

Correlation vs. Causation

A fundamental principle in biostatistics stating that a statistical association or connection between two variables does not automatically establish that one variable causes changes in the other.