Case Study

Based on the researcher's approach, explain how the dependent-samples design is being conceptualized as a one-sample tt-test. Specifically, explain what happens to the paired data structure and what the null hypothesis expectation of a population mean (μ0\mu_0) of 00 represents in the context of this meditation study.

Case context: A health psychologist is conducting a study to examine the effect of a new meditation app on participants' daily stress levels. She measures the daily stress levels (on a scale from 0 to 50) of 20 participants at baseline (before using the app) and again after two weeks of daily app usage. To analyze whether there is a change, she subtracts the post-test score from the pre-test score for each participant to obtain a set of difference scores, intending to run a one-sample tt-test on these difference scores.

Question: Based on the researcher's approach, explain how the dependent-samples design is being conceptualized as a one-sample tt-test. Specifically, explain what happens to the paired data structure and what the null hypothesis expectation of a population mean (μ0\mu_0) of 00 represents in the context of this meditation study.

Sample answer: By converting each participant's pre-test and post-test stress scores into a single difference score, the researcher mathematically reduces the paired data structure into a single distribution of difference scores. The analysis then proceeds exactly like a one-sample tt-test on these scores. Testing the sample mean difference against a hypothetical population mean (μ0\mu_0) of 00 represents the baseline expectation that using the meditation app has no overall effect or causes no change in stress levels between baseline and the two-week follow-up.

Key points:

  • Paired measurements are reduced to a single distribution of difference scores.
  • The analysis is conducted as a one-sample tt-test on the difference scores.
  • The population mean (μ0\mu_0) is set to 00.
  • The value of 00 represents the baseline expectation of no change or effect in stress levels.

Rubric: The response should be evaluated on the following criteria: 1. Explains that converting to difference scores collapses/reduces paired measurements into a single distribution. 2. States that the analysis is a one-sample t-test performed on these difference scores. 3. Explains that the population mean of 0 represents the null or baseline expectation that the app has no effect or causes no change in stress levels.

0

1

Updated 2026-05-26

Contributors are:

Who are from:

Tags

KPU

Research Methods in Psychology - 4th American Edition @ KPU

Related