To examine the effect that a tough midsemester test has on...

To examine the effect that a tough midsemester test has on student evaluations of lecturers, a statistics lecturer had her class evaluate her teaching effectiveness before the mid-semester test. The questionnaire asked for opinions on a number of dimensions, but the last question is considered the most important: "How would you rate the overall performance of the lecturer?' The possible responses are: 1 = poor, 2 = fair, 3 = good, 4 = excellent. After a difficult test, the evaluation was redone. The evaluation scores before and after the test for each of the 40 students in the class are recorded. Do the data allow the lecturer to conclude that the results of the mid-semester test negatively influence student opinion? Sample statistics: $n$ (positive differences) $=19$; $n$ (negative differences) $=5$; $n$ (zero differences) $=16$.

To examine the effect that a tough midsemester test has on student evaluations of lecturers, a statistics lecturer had her class evaluate her teaching effectiveness before the mid-semester test. The questionnaire asked for opinions on a number of dimensions, but the last question is considered the most important: "How would you rate the overall performance of the lecturer?' The possible responses are: 1 = poor, 2 = fair, 3 = good, 4 = excellent. After a difficult test, the evaluation was redone. The evaluation scores before and after the test for each of the 40 students in the class are recorded. Do the data allow the lecturer to conclude that the results of the mid-semester test negatively influence student opinion?
Sample statistics: $n$ (positive differences) $=19$; $n$ (negative differences) $=5$; $n$ (zero differences) $=16$.

Labs

Want to see this concept in action?

NEW

Explore this concept interactively to see how it behaves as you change inputs.

View Labs

Key Concepts

Paired Sample Design

This concept involves collecting two related observations from each subject—one before and one after an intervention—to assess the effect of that intervention. In statistics, paired designs help eliminate the variability between subjects by allowing each subject to serve as their own control, making the study more sensitive to detecting changes or effects due to the treatment or event under investigation.

Nonparametric Sign Test

The nonparametric sign test is used when the assumptions required for parametric tests (like normality) are not met or when dealing with ordinal data. The test evaluates whether there is a significant difference in the median of paired differences by simply considering the direction (positive or negative) of the differences, while disregarding the magnitude of those differences.

Hypothesis Testing in Paired Comparisons

In the context of paired comparisons, hypothesis testing involves setting up a null hypothesis, typically that there is no effect or difference between the two conditions, against an alternative hypothesis that indicates a directional change (such as a negative effect). The statistical procedure determines whether the observed differences in the paired samples provide enough evidence to reject the null hypothesis in favor of the alternative.

Transcript

Please give Ace some feedback