ABSTRACT

Random models that contain only nested factors are generalizations of the one-fold nested design considered in Chapter 4. As will be illustrated in this chapter, confidence intervals for functions of variance components in nested designs can be constructed using the results of Chapters 3 and 4. Sokal and Rohlf (1969, p. 260) used a two-fold nested design to collect the data shown in Table 5. l.l. These data are from a biological experiment involving 12 female mosquito pupae that were randomly assigned into three rearing cages. The reported responses are independent measurements of the left wings of the mosquitoes.