Chapter 16 Questions

Complete the following questions. The questions may include pictures or graphics to illustrate or aid in solving the problem. You can check your answer by clicking View Answer. If the question is unclear, confusing, or if you need further clarification, send me an email.

1. Imagine we are interested in which of three methods of inducing anger in subjects results in the most intense emotions. These methods are 1) recollection, where subjects recall an incident in which they were actually angry, 2) insults, where the subjects are insulted by the experimenter, and 3) pain, where the experimenter behaves clumsily and repeatedly steps on the subject's toes. Given the following data (heart rates measured in three samples employing the three different induction techniques), perform an analysis of variance to determine whether the methods differ from one another. Complete an ANOVA summary table.

Answer

2. Using the previous data, why would we use an ANOVA here instead of multiple t-tests? What is family-wise error?

Answer

Family-wise error rate: The total probability of making a Type I error (saying differences in the means exist when there are none) for any group or family of comparisons. We use ANOVA to maintain the family-wise error rate at a level we wish it to be (usually .05). Computing multiple t-tests will result in a family-wise error rate greater than the individual error rate for any of the single comparisons.

3. Using the data from question one, which treatments differ from another? Compute a Fisher's LSD post-hoc analysis. Did the treatment differences have a large effect, or was it merely statistically significant? Compute the eta-squared value.

Answer

4. Explain the general linear model for the one-way analysis of variance.

Answer