大学生艾滋病传播潜在风险网络测试问卷重复测试一致性分析

Analysis on the consistency of repeated response from online survey on the potential risk of HIV/AIDS transmission among College Students

目的 通过重复应答的网络问卷,分析大学生艾滋病相关网络问卷调查的应答一致性,探讨网络调查问卷是否存在随意或者未真实回答等情况,为评价网络问卷调查质量提供依据。方法 数据来源于2018-2020年开展的大学生与艾滋病相关性行为和性向自我认知情况的网络调查,选择同一名大学生重复填写2次的问卷进行分析,使用Excel 2007和SPSS19.0进行数据统计分析。结果 本次网络问卷2次应答平均间隔时间(45.32±19.76)天,所有题项完全一致被调查者占64.32%,有1个或以上题项应答不一致的占35.68%,不同题项不一致比例为3.00%~18.44%,性行为特征、自我性向认知判断不一致比例要高于年龄、性别、年级等基本情况。从一致性指标来看,年龄、年级的一致性较好,性行为特征、情感特征、自我性向认知和判断等问题应答一致性一般,Kappa指数在0.52~0.78之间。结论 本次大学生艾滋病网络问卷调查存在重复应答不一致的情况,尤多见于情感隐私、敏感问题和容易产生歧义的题项,提示网络调查问卷存在一定数量的随意或者未真实回答的情况,数据分析或今后开展类似调查时均需综合考虑调查结果的可信度。

Objective To analyze the consistency and false answers of HIV/AIDS related network questionnaire among college students through repeated response. To discuss the facticity of network questionnaire and provide basis for evaluating the quality of online survey. Methods The data was from an online survey of College Students’ AIDS related behavior, sexual orientation and self-cognition conducted from 2018 to 2020. The questionnaires responded twice by the same college student were selected for analysis. The data were statistically analyzed by Microsoft Office Excel 2007 and SPSS19.0. Results In this online questionnaire, the average interval of the twice responds was 45.32±19.76 days.64.32% of the respondents answered all the 15 questions twice were completely consistently and 35.68% responded inconsistently at least one or more questions. The proportion of inconsistencies in different questions was 3.00%-18.44%.The proportion of inconsistencies in sexual behavior characteristics, sexual orientation cognitive questions was higher than that in questions of age, gender, grade and other basic situation. By consistency index, the consistency of age and grade questions is good, and sexual behavior characteristics, sexual orientation cognition questions is lower with Kappa index between 0.52-0.78. Conclusions There were inconsistent responses on repeated responding questionnaires during HIV/AIDS online survey among college students, especially the questions about emotional privacy, sensitive issues and ambiguous items. It is indicated that there are a certain number of false or unreal answers during online survey, which should be comprehensively considered in data analysis and surveys.