had sought out other AIDS education during the period of the study. Control subjects may best be conceptualized as being a waiting list group, as all respondents were encouraged to participate in some form of counseling or group intervention at the conclusion of the study.
Unless otherwise stated, for statistical purposes, an alpha level of p <.05 was selected as evidence of a significant result, with three other levels of significance (p <.01; p <.005; and p <.001) determined a priori, to be reported. To investigate the concomitants of safer sex, chi-square tests were performed on frequency data involving categorical data.28 Interval data were analyzed using t-tests.29 Other statistics were employed as specified in the text.
To investigate the interrelationship of various sexual behaviors (see Chapter 8), data for each sample were subject to orthogonal factor analysis (varimax rotation) yielding three factors, the matrices of which were then input into a factor replication program, assessing Cattell's S index. To investigate how various interventions affected change in sexual behavior, a discriminant function analysis of the data was performed (see Chapter 16).
Four interventions and a control condition were carried out as part of the investigation into the efficacy of AIDS education interventions in the NZ study (see Figure 2.1). As well, data were collected both prior to and subsequent to the Grim Reaper campaign, in order to investigate the effects of such public awareness campaigns on homosexual behavior. These interventions are described in detail in Chapters 15 to 17.