Examkin

Question

Difficulty: MediumInferential Statistical Tests

Cardiology researchers are assessing the impact of three distinct dietary interventions (Mediterranean diet, DASH diet, and a standard low-fat diet) on serum high-sensitivity C-reactive protein (hs-CRP) levels in patients with stable coronary artery disease. A total of 120 patients are randomly assigned in a 1:1:1 ratio to one of the three diet groups for 6 months. At the end of the study period, absolute hs-CRP levels (measured in mg/L) are recorded for each participant. Upon initial data analysis, the investigators note that the hs-CRP values are highly right-skewed across all three groups and fail statistical tests for normality. Which of the following is the most appropriate statistical test to compare the final hs-CRP levels among these three groups?

  1. Kruskal-Wallis testAnswer
  2. B
    One-way analysis of variance
  3. C
    Mann-Whitney U test
  4. D
    Chi-square test
  5. E
    Wilcoxon signed-rank test

Answer

The most appropriate test is the Kruskal-Wallis test.
The study design requires comparing a continuous numerical variable (serum hs-CRP levels) across three independent groups (Mediterranean, DASH, and low-fat diets). Because the data is explicitly described as highly right-skewed and non-normally distributed, a non-parametric statistical test is required. The Kruskal-Wallis test is the non-parametric equivalent of the one-way analysis of variance (ANOVA) and is the most appropriate choice for comparing continuous, non-normally distributed data across three or more independent groups.

Step-by-Step Solution

1
Identify the type of outcome variable.
The outcome is serum hs-CRP measured in mg/L, which is a continuous numerical variable.
The type of data determines the initial category of statistical tests (e.g., continuous vs. categorical).
2
Determine the number of comparison groups and their relationship.
There are three independent groups (Mediterranean, DASH, and low-fat diets) being compared.
Statistical tests differ based on how many groups are being compared (2 vs. 3+) and whether they are paired or independent.
3
Assess the data distribution to decide between parametric and non-parametric tests.
The data is highly right-skewed and not normally distributed, requiring a non-parametric test.
Parametric tests (like ANOVA) require normal distribution, while non-parametric tests (like Kruskal-Wallis) do not.
4
Select the specific non-parametric test for three or more independent groups.
The Kruskal-Wallis test is selected.
It is the standard non-parametric equivalent for comparing continuous outcomes across three or more independent groups.

Key Concept

Selecting the appropriate non-parametric statistical test (Kruskal-Wallis) for comparing continuous, non-normally distributed data across three or more independent groups.
Estimated Time:1m 0s
Rate this question

Topics

Inferential Statistical Tests

More questions from this topic