Modified t-Test Procedures

What is the primary statistical problem that arises when conducting multiple independent-samples t-tests to compare every pair of group means in an experiment?

A researcher conducts six separate t-tests to compare every pair of means among four experimental groups, using an alpha level of .05 for each test. Because each individual test maintains only a 5% chance of a Type I error, the overall probability of making at least one Type I error across all six tests also remains at 5%.

A researcher is comparing the mean scores of three different groups: a Control group, a New Treatment group, and a Traditional Treatment group. To find specific differences, they conduct three separate statistical comparisons (Control vs. New, Control vs. Traditional, and New vs. Traditional). Match each term to how it applies to this specific scenario of multiple comparisons.

A psychology researcher is analyzing data from an experiment with four distinct treatment conditions. Arrange the following testing scenarios in order of their cumulative probability of committing at least one Type I error (the familywise error rate), starting with the scenario that has the lowest probability (top) and ending with the highest probability (bottom).

You are designing the 'Data Analysis' section of a research proposal comparing four different types of memory training. Your goal is to identify all specific pairs of training types that differ from one another without exceeding an overall Type I error rate of $5\%$. Which of the following represents the most effective plan to construct this statistically valid framework?

Match each statistical concept related to the problem of multiple comparisons with its correct description.

Arrange the following steps in the logical order that explains how the 'multiple comparisons' problem develops when analyzing group differences in a research study.

A researcher argues that their study maintains a total risk of only $5\%$ for a false positive because every individual comparison was tested at an alpha ($\alpha$) of $0.05$. When evaluating this argument, we find it to be statistically invalid because the _____ error rate compounds with each additional test, making the overall risk of at least one false positive much higher than intended.

When analyzing an experiment with multiple group means, conducting several independent-samples $t$-tests simultaneously keeps the cumulative risk of mistakenly rejecting a true null hypothesis at exactly $5\%$.

When evaluating a study design that performs multiple independent-samples $t$-tests to compare all group means, a researcher must recognize that simultaneous tests compound the overall probability of committing a _____, requiring specialized statistical procedures to control the error rate.