An experiment was conducted at a large state university to determine whether two different instructional methods for teaching a beginning statistics course would yield different levels of cognitive achievement. One instructional method involved using a self-instructional format, including a sequence of slide-tape presentations; the other method utilized the standard lecture format. The 100 students who registered for the course were randomly assigned to one of four sections, 25 per section, corresponding to the combinations of one of the two methods with one of two instructors. The results obtained from identical final exams given to each section are summarized in the following table.

a. What do the results suggest about the comparative effects of the two instructional methods?
b. Classify each factor as either fixed or random, and explain your classification.

c. Using the ANOVA table, perform the appropriate F tests for each of the four types of factor classification schemes. Compare the conclusions reached under each scheme.
d. What factors should be controlled for in this experiment?
e. Given a continuous variable C to be controlled for, write an appropriate regression model for this data set that takes C into account. What general method of analysis is characterized by such a model?

Method Instructor Lecture Self-instruction 71.2. S= 13.8 Y 80.2. S 12.1 77.5, s 14.1 A. F-73.8s-1 1.7 Source d.f MS INSTRUC METHOD NSTRUC METHOD Error 6.2500E-02 6.0081E + 03 1.7556E +02 16141E +04 6.2500E-02 10081E +03 17556E 02 16814E+02 96