社会化媒体调查的测量效果评估  

Evaluating Measurement Effects in Social Media Surveys

作　　者：朱文龙[1] 邵培基[2] 方佳明[2] 曹云忠[3] 

机构地区：[1]青岛理工大学商学院,山东青岛266520 [2]电子科技大学经济与管理学院,成都610054 [3]四川农业大学建筑与城乡规划学院,成都611830

出　　处：《系统管理学报》2015年第6期806-813,共8页Journal of Systems & Management

基　　金：国家自然科学基金资助项目(71101018);青岛市哲学社会科学规划资助项目(QDSKL150480);青岛理工大学社科基金资助项目(C-2014-011)

摘　　要：混合调查已成为非常重要的数据收集方式,社会化媒体为混合设计提供了新的平台。调查者可基于该平台产生的测量效果来实施混合调查,以获取同质性更高的调查数据,以便在统计分析前可对其采取相同的预处理方式。以称许性平衡量表(BIDR)为例,运用多组结构方程模型,研究了社会化媒体平台下匿名调查与非匿名调查间的测量效果异同。结果表明:两类调查具有相同的被试作答方式效果,但该效果不同于任何传统效果;两类调查具有相同的被试认知效果;两类调查能够产生相同的主效果,且它们与被试性别、年龄间的交互效果均不存在显著差异,但教育水平会对两类调查的被试作答方式产生显著差异的交互效果。Mixed surveys have been a very important method of data collection.Social mediaprovides a new platform for mixed design.Nowadays investigators can implement mixed surveys to obtain higher homogeneity data according to the measurement effects in order to take the same pretreatment method for them before statistical analysis.By using the MG-SEM,we study the similarities and differences between the measurement effects of the anonymous survey and non-anonymous survey based on social media,taking the Balanced Inventory of Desirable Responding（BIDR）as example.The results show that both surveys have the same response style effect,but this effect is different from any traditional effects.The two surveys have the same cognitive effect and they can lead to the same main effect.The interaction effects between survey method and gender,age,respectively,are not significantly different across the modes.However,the interaction effect of education on response style is significantly different across the methods.

关 键 词：网络调查 混合调查 社会化媒体 测量效果 交互效果 多组结构方程模型 

分 类 号：C31[社会学]