American Board of Surgery Qualifying Exam (ABS QE) Practice Test

What type of data is suitable for a t-test?

1. Skewed, continuous data

2. Normal, continuous data

3. Ordinal data

4. Nominal data

The correct answer is: Normal, continuous data

The t-test is specifically designed to compare the means of two groups and is most suitable for normal, continuous data. This type of data must follow a normal distribution pattern, which allows for the underlying assumptions of the t-test to be satisfied. When the data is normally distributed, the t-test can accurately determine if there is a significant difference between the means of the groups being compared. Normal distribution implies that the data will cluster around a mean, with symmetrical tails on either side. Continuous data, which can take any numeric value within a range, allows for precise calculations of means and standard deviations, both of which are fundamental in performing a t-test. Hence, the capacity to apply statistical methods accurately hinges on the data meeting these criteria. In contrast, skewed data may violate the assumptions required for a t-test, leading to unreliable results. Ordinal data is categorical and ranks the order of observations without establishing a consistent distance between them, making it unsuitable for mean comparisons. Nominal data is categorical as well, representing distinct groups without any inherent order, which further disqualifies it from application in a t-test. Therefore, normal, continuous data is essential for the correct application of a t-test.