Which statistical test is most appropriate to analyze median weight loss (in kg) between semaglutide and tirzepatide?

• Independent t-test
• Chi-square test
• Mann-Whitney U test

---

What is the correct answer and why?

In this case, the most appropriate statistical test is dependent on the type of data that is being analyzed (continuous vs. categorical) and, if the data is continuous, whether it is parametric (normally distributed – i.e. bell-shaped distribution) or non-parametric (not normally distributed).

The “amount of weight loss” is a continuous variable. It can be measured on a near infinite scale — some patients may lose 2 kg and others may lose 20 kg or 40 kg. There are no “categories” (e.g. yes/no) for “amount of weight loss”, therefore, it is not a categorical variable.

Because the question stem is asking about a median amount of weight loss, the most appropriate statistical test should analyze the “rank” of weight loss between semaglutide and tirzepatide and not just the average (mean) weight loss between groups.

The most appropriate statistical test to compare the median value of a continuous variable between two independent groups is a Mann-Whitney U test.

An independent t-test would be appropriate to compare the average (mean) values between groups, assuming that the distribution of weight loss was normally distributed (parametric). A Chi-square test would be appropriate for comparing a categorical outcome (e.g. the percent of patients who lost more than 10% of body weight between the two groups).

---

---

---