ABSTRACT
A crucial ques tion for the design of an invest ig a tion is how to maxim ize the gener al iz ab il ity of the results. The ques tion is more import ant for factor analysis than for the usual stat ist ical proced ure because signi fi c ance tests are rare and because many possib il it ies exist for capit al iz ing on chance. This chapter is concerned with noting the condi tions under which the results from a factor analysis can be gener al ized across samples of indi vidu als or vari ables so that recom mend a tions for the design of factor-analytic studies can be made.
