Gerald's dependent measure is the order in which people...

Gerald's dependent measure is the order in which people turned in their exam (first, second, third, etc.). Can Gerald use a $t$ test on his data? Why or why not? What would you advise Gerald to do in future studies?

Key Concepts

Nonparametric Tests

Nonparametric tests provide alternatives to parametric tests when the data do not meet the necessary assumptions, such as normality or interval-level measurement. These tests work by ranking data rather than relying on the specific distances between data points, making them more suitable for ordinal data. For studies involving rankings or orders, nonparametric methods, like the Mann-Whitney U test or Kruskal-Wallis test, would be more appropriate.

Measurement Scales

Measurement scales define how variables are quantified and what statistical methods are appropriate for analysis. In particular, ordinal scales rank order data without equidistant spacing between values, which means that the differences between consecutive ranks are not necessarily equal. This property limits the applicability of techniques that assume a continuous scale, such as those requiring interval data.

Parametric Test Assumptions

Parametric tests, such as the t test, assume that the data are measured on an interval or ratio scale and follow a normal distribution with equal variances across groups. These tests also require that the distances between data points are meaningful. Using an ordinal measure, where the differences between positions (like first, second, third, etc.) are not uniformly interpreted as equal, violates these assumptions and renders the t test inappropriate.

Transcript

Please give Ace some feedback