What should be included in the sign test write‑up?

What should be included in the sign test write‑up?

• A. The relationship between the calculated and critical value, how many participants there were, the significance level, whether it was one-tailed or two-tailed, and whether it was significant or not.
• B. The raw scores and the individual differences.
• C. The confidence interval and effect size.
• D. The exact p-values for all participants.

Answer: (A)

The key idea here is reporting how the observed sign-test result leads to a decision. In a write‑up you should present the comparison between the calculated statistic (the number of positive differences you observed) and the critical value that comes from the sign-test distribution, given your sample size and chosen alpha. You also need to state the sample size, the significance level, whether the test was one‑ or two‑tailed, and the final decision about significance. This set of details lets someone see exactly how you reached whether to reject the null. Raw scores and individual differences are data details, not the decision criteria you use to conclude the test. Confidence intervals and effect sizes aren’t required for a basic sign-test write‑up, though they might appear in more extended reports. Providing exact p-values for every participant isn’t meaningful in this context, since the test evaluates the aggregate direction of differences, not per‑participant statistics.

The key idea here is reporting how the observed sign-test result leads to a decision. In a write‑up you should present the comparison between the calculated statistic (the number of positive differences you observed) and the critical value that comes from the sign-test distribution, given your sample size and chosen alpha. You also need to state the sample size, the significance level, whether the test was one‑ or two‑tailed, and the final decision about significance. This set of details lets someone see exactly how you reached whether to reject the null.

Raw scores and individual differences are data details, not the decision criteria you use to conclude the test. Confidence intervals and effect sizes aren’t required for a basic sign-test write‑up, though they might appear in more extended reports. Providing exact p-values for every participant isn’t meaningful in this context, since the test evaluates the aggregate direction of differences, not per‑participant statistics.