One frequent assumption in statistical analyses is that observations are independent of one another (knowing one response tells you nothing about the magnitude of another response). How would you characterize the reaction-time data in Table 3.1, just based on what you know about how it was collected? (A lack of independence would not invalidate anything we have done with these data in this chapter, though it might have an effect on more complex analyses.)

5.10 High affiliation Low affiliation 5.05 5.00 a 4.95 4.90 4.85 4.80 Stable Unstable Stability From Cohen, Kaplan, Cunnick, Manuck, and Rabin (1992). u Figure 3.12 Immunity as a function of afifliation and stability.