Becky Liddle at Auburn University published a study in 1997 on disclosing sexual orientation in class. She taught four sections of the same class, and at the week of the final lecture she disclosed her lesbian identity to two of the sections, and withheld it from the two others.

She was concerned with the issue of whether disclosure would influence student evaluations of the course. The means and average variance for the two conditions, further broken down by gender of the students, are presented below. There were 15 students in each cell. Perform a two-way analysis of variance and draw the appropriate conclusions.

(The means are the same that Liddle found, but because I could not control for difference in midterm evaluations, as she did, the effect of gender is different from the effect she found.

The other effects lead to similar conclusions.)

Sexual Identification:

Disclose Not Disclose Mean Female 37.15 36.56 36.86 Male 33.00 33.00 33.00 Means 35.08 34.78 34.93 Average within-cell variance 5 20.74