【作者】 方佳明；

【导师】 邵培基；

【作者基本信息】 电子科技大学 ， 管理科学与工程， 2009， 博士

【摘要】 截至2008年底,中国网民规模已达到3.0亿人,互联网普及率达到22.6%,网民规模跃居世界第一。随着互联网的发展和网民数量的大幅攀升,网络调查在中国正被应用得越来越普遍。相比于传统调查,网络调查具有成本低、调查范围广、调查周期短、易于调查敏感性问题等优点。但是,调查低参与率和样本覆盖率偏差问题是影响网络调查结果有效性的两个主要原因。由于样本覆盖率偏差可以采取混合调查模式加以缓解,调查低参与率成为网络调查者面临的最主要问题。目前国内外研究主要从调查者问卷设计视角考察网络调查低参与率的原因,很少有被调查者心理因素视角的研究。国内网络调查参与的相关实证性研究尤为缺乏,大部分研究都在介绍和论述网络调查的优缺点,以及对国外相关研究和论述的转述。本论文正是针对网络调查低参与率问题,从外在影响的激励因素和内在影响的心理因素视角对网络调查参与意愿的形成机理作了探索性和创新性研究,主要内容和相关结论包括：(1)研究了网络调查(基于BBS和基于网站)和传统纸笔调查的测量不变性,给出了在AMOS中如何实现两个潜变量构成的模型的测量不变性检验。研究结果表明虽然三种调查模式下得到的对潜变量的测量具有不同的均值和方差,但是,这三种调查形式具有强因子测量不变性。研究说明在涉及自我评估的实证研究中,恰当设计的网络调查具有和传统纸笔调查相同甚至更优的可靠性,因此可以采用混合调查模式缓解网络调查的覆盖率偏差问题。(2)使用荟萃分析较以往研究更加准确地考察了物质性激励,事后激励以及或有激励在影响网络调查参与中的作用。使用荟萃分析技术对以往的相关研究结果进行基于独立于样本量的效应值统计量计算。结果表明物质性激励、事后激励以及或有激励能够促使被调查者完成问卷；和激励在传统调查中对调查参与的促进作用相比,其在网络调查中的作用较小；或有激励是效果最稳定的一种激励策略；精神激励则对调查参与的提升具有最弱的有效性。(3)研究了网络调查中所有八种不同形式的激励策略对被调查者调查参与意愿的影响。结果表明不同激励策略在网络调查中对被调查者调查参与意愿提升具有不同的有效性。必然激励对参与意愿的提升最具有效性,且必然激励不同的激励实现手段(现金、实物、虚拟产品)之间不存在显著性的差异。(4)研究了调查者心理因素对网络调查激励有效性的调节作用。鉴于激励视角的研究难以解释在控制了激励的类型和激励物价值情况下,激励对调查参与的影响仍然展现出不同有效性,通过被试内设计以及混合设计方法在控制了具体的激励策略之后,研究了心理感知因素(主题卷入、身份认知)对网络调查参与意愿的影响。研究结果证实了心理感知因素负向调节了激励对网络调查参与的有效性,并且这种调节作用的强度甚至会完全抵消激励本身对参与意愿的影响。(5)通过设计三个不同的研究方案,采用被试内重复测量技术研究了不同身份认知的调查发起者和实施者对调查参与意愿的影响。结果表明在网络调查中调查发起者身份的重要性远远超过调查实施者,当调查者发起者是低身份认知时委托高身份认知的实施者不一定能对被调查者的调查参与意愿产生显著地提升作用。(6)基于理性行为理论(Theory of Reasoned Action, TRA)构建了一个潜在被调查者网络调查参与意愿形成过程的心理模型并在控制了年龄、性别、教育程度和网络使用年限等人口统计学变量之后实证检验了各因素对调查参与意愿的影响。同时,进一步考察了调查参与意愿是否受到人格因素影响。通过引入人格因素,检验了有些潜在被调查者是否天生厌恶参加调查的问题。更多还原

【Abstract】 With the development of Internet and the soaring number of cyber citizens in China (up to Dec.2008, Netizens and Internet penetration rate in China have reached 3 billion and 22.6%, respectively), Web-based or Web survey is more and more frequently used in such fields as commercial, government, and academic research. However, theoretical research related to web survey is far behind the practice in China. The focus of the thesis is on investigation of factors pertaining to the participation intention of web survey from incentive and psychological perspectives. Research and according findings involving in the thesis as follows:Firstly, author considered the measurement invariance between web survey and traditional surveys under the Chinese culture or the collectivistic culture. This thesis employed multi-group confirmatory factor analysis to examine measurement invariance of personal innovativeness scales under traditional paper and pencil survey, Web survey based on BBS, and Web survey based on websites, respectively. The results indicated that the three survey modes had measurement invariance in the domain of the self-evaluation. However, because of situation-dependence caused by culture surroundings, the three modes revealed different means and variances of latent variables. The results demonstrated that a well design web survey had equal psychometrics to, if not more than, paper and pencil survey.Secondly, the thesis investigated the effectiveness of incentive in Web survey using meta-analysis. Three meta-analyses examined the effect of incentives in web surveys. The first analysis securitized the impact of material incentives on retention. A significant effect was revealed, indicating that material incentives motivated people to complete a Web survey. The second analysis checked the impact of promised incentives on retention and the results indicated that promised incentives could urge the participants to complete the Web survey. The last analysis examined contingent incentives on retention. The results also demonstrated that offering contingent incentives could exert a positive impact on retention. Thirdly, eight types of incentives used in Web surveys were examined, and results indicated that different incentive tactics exerted different effects on web survey participation intention. Postpaid incentive, no matter of cash, real material, virtual material, exerted the most effect on participation intention, while psychological incentive exhibited the least effect on survey participation intention.Fourthly, in view of the fact that incentive perspective research cannot explain the phenomenon that even after controlling for the types and value of the incentive, web surveys show distinct participation rates in practice. This thesis considered whether potential respondent’s psychological factors (topic involvement and status cognition) could moderate the effect of incentive on participation intention. After controlling for the type of incentive, the results demonstrated that significant negative moderation effects appeared, and the strength of effects even could negate the role of incentive.Fifthly, designing three distinct experiments and employing within-subject measurement, the thesis inspected roles of status cognition of survey sponsor and investigator on participation intention. The results illustrated that the role of sponsor’s status cognition was far more important than that of investigator. When the survey sponsor possesses low status cognition and assigns a survey to a high status cognition investigator, participation intention cannot be necessarily increased.Sixthly, and last, the thesis proposed a research model from a psychological perspective based on theory of reasoned action. We empirically examined the role of attitude, social norm, moral obligation, trust in sponsor of survey, topic involvement, topic sensitivity, and reputation of the sponsor. The results indicated that attitudes, social norm, moral obligation, status cognition, and reputation of sponsor could positively influence the intention to participate in web survey after controlling for demographic factors such as age, gender, education level and net usage (year). Additionally, personality of potential respondents was investigated so as to verify the conjecture that some individuals were naturally unlike to participate in Web surveys. The results vindicated the conjecture.更多还原