录音机辅助调查与面访调查收集性行为数据的可靠性比较

Comparative study on data collection of sexual behavior:audiotape self-administered questionnaires vs.face-to-face interviews

目的 比较调查员面访调查 (FTFI)与录音机辅助自填问卷调查 (A SAQ)收集艾滋病敏感行为数据的可靠性差异。方法 用统一的问卷制成FTFI和A SAQ两种版面。招募建筑工人、服装个体户、服务小姐三类调查对象随机分为两组 ,一组先采用听录音调查 ,3～ 5天后采用面访调查 ;另一组先采用面访调查 ,3～ 5天后采用听录音调查 ,两次调查完成后 ,抽取约 2 0 %调查对象参加查后访谈。结果 三类调查对象共 6 0 4人完成了 2次调查。不涉及个人隐私的非敏感问题 ,两种方法差异无显著的统计学意义 (P >0 0 5 ) ;发生过性关系的自我报告率 ,FTFI高于A SAQ(χ2 值分别为 11 373、9 70 1、2 5 0 5 3;P均 <0 0 1) ;多性伙伴和商业性性伙伴的自我报告率 ,A SAQ略高于FT FI(P>0 0 5 ) ;问卷应答的逻辑一致性程度 ,FTFI明显高于A SAQ(χ2 =76 10 7,P<0 0 1)。三类调查对象喜欢听录音的比例 (6 0 1% )高于喜欢面访的比例 (10 1% )。结论 与FTFI比较 ,A SAQ未显著提高收集艾滋病敏感行为数据的可靠性 ,对于本研究所选择的三类人群 ,应考虑与FTFI结合使用。

objective To probe the effect of audiotape self-administered questionnaires(A-SAQ)and face-to-face interviews(FTFI)on the reliability for measuring HIV-related sex behaviors.Method Two versions of the questionnaires,audiotape self-administered questionnaires(A-SAQ)and face-to-face interview questionnaries(FTFI),were designed with the same items.Three subgroups of subjects were recruited among construction workers,clothes salesmen and female sex workers. The samples of subjects were randomly assigned into two retest groups. Subjects of one group were interviewed by A-SAQ firstly and interviewed by FTFI secondly after three to five days. On the contrary,subjects of the other group were interviewed by FTFI firstly and interviewed by A-SAQ secondly after three to five days. Finally,20% qualified subjects,who completed the two different types of interviews,filled in the brief exit-interview questionnaires. Result Totally 604 participants completed the two retest interniews.There were no significant differences(P>0.05) between responses to the non-sensitive questions in A-SAQ and those to the same questions in FTFI.With regard to the sensitive behavioral questions,“ever had sex”were more positively reported in FTFI than in A-SAQ.(χ2=11.373, 9.701, and 25.053; P<0.01). “Multiple sex” and “commercial sex” were reported higher in A-SAQ than in FTFI (P>0.05).On the average,the responses to the questions in A-SAQ were less logically consisitent than those to the questions in FTFI(χ2=76.107, P<0.01). But results of the exit-interview data indicated that A-SAQ is more preferable to FTFI(60.1%>10.1%). Conclusion Comparicon with the FTFI,A-SAQ did not increase reliability of self-reports in HIV-related sensitive questions.For these three target populations,combination of these two interview methods,A-SAQ and FTFI,should be considered.