Herbert [ 1987], p. 146.
Quadland et al. [ 1988].
Masia [ 1964].
8 Trend analysis is here defined as significant at an alpha level of < .10,
X2 = 2.83, df = 1,p <.10, that is, the increase in respondents practicing safer sex was
great enough to be distinguished at.10. See Table 16.1. 9
Consequently, they were not significant at p < .05.
Baseline: 8/26; follow-up, 14/26; percent increase: +23%, X2 (1) = 2.84, p <
Baseline: 28/85; follow-up, 34/79; percent increase: +10%, X2 = 3.908, df
1, p < .05.
That is, 100% use of condoms for all penoanal activity, whether insertive
Rosser [ 1989], Table 17.3, for a detailed presentation of these figures. 14
Given the findings in Chapter 9, in most cases it is more likely to be a
reintroduction than an introduction.
Monogamous at baseline: SAP: 10/28, ESS: 5/26; monogamous at followup: SAP: 12/28 (+7%), ESS: 3/26 (-7%); X2 = 7.98, df = 1, p < .10.
This hierarchy is worked out on the following principles: avoidance of
anal intercourse is safer than consistent condom usage because condoms may leak
or break, while condom usage in turn is safer than monogamy because partners
may be unfaithful.
At least as related to the behavior of homosexually active men with respect
to safer sex. Most health behavior change models to date have been based on
behaviors like smoking and obesity, where a reduction in the targeted behavior is
considered a success. Because one unsafe sexual encounter is sufficient to
transmit HIV, we are dealing with a qualitatively different situation.
A stepwise discriminant function analysis was performed using the
change in twenty-two sexual, prophylactic, and drug behavior variables between
baseline and follow-up as predictors of membership in the five groups. Four
discriminant functions were calculated, with a combined X2 = 70.08, df = 40, p < .005. After removal of the first function, the discriminating power of the
remaining functions was marginally significant, X2 = 39.60, df = 27, p < .055.
Following removal of the second function, the discriminating power of the
remaining functions was not significant, X2 = 10.40, df = 16, ns. The first
two functions accounted for 44% and 42% of the between group variability,