Session 7 - Comparison of variables between two groups using distribution free methods Flashcards

1
Q

What are other names for Non-paraametric tests?

A

-Distribution free
-Rank based

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What does the sign test check?

What is the main assumption?

What is its parametric equivalent?

A
  1. Median
  2. Continuous outcome
  3. one-sample t-test
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Wilcoxon signed-rank test

Checks?

Main assumptions?

Equivalent?

A

Median

Symmetric variable

one-sample t-test

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Wilcoxon signed-rank test

-Hypothesis checks?

-Assumptions?

-Parametric equivalent?

A
  • Mean rank difference
  • The difference has a symmetric distribution
  • two-sample paired t-test
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Kruskal-Wallis

Hypothesis?

Assumptions?

Parametric equivalent?

A

H0: Identical populations (in terms of ranks)
H1: At least one is different

  • As a rank-based method, the null and alternate hypotheses are in terms of ranks (not means, can be in terms of medians if additionally assume equal shape)
  • One-way analysis of variance (ANOVA)
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Cuzick non-parametric test for trend

-Hypothesis?

-Assumptions?

-Parametric equivalent?

A

-Hypothesis
H0: θ1 =θ2 =···=θk
H1: θ1 <θ2 <···<θk

-Assumptions
Assumes same shape and scale (to test for trend)

-Parametric equivalent
Chi-squared test for trend

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q
A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly