Visual & Interactive Issues in the Design of Web Surveys

描述（申请人提供）：万维网作为调查数据收集工具的迅速接受提出了关于新方法如何运作的重要问题。网络调查的主要特点包括使用丰富的视觉呈现问题和与答卷人互动的能力。Web的快速发展使得对这些问题的仔细检查变得更加紧迫。这两套功能都没有被彻底探索，即使是早期的模式，网络提供了广泛增加的资源，用于视觉显示（网络调查问卷可以很容易地包含静态图片或视频剪辑）和交互（如浮动屏幕和滚动帮助定义）。我们的申请概述了一系列旨在解决有关这些问题的关键问题的研究。这些研究集中在网络调查，但我们相信，结果将推广到其他模式的数据收集，依赖于视觉呈现或纳入互动设计功能。 实验1-5考察了受访者如何解释网络问卷中的视觉线索。这些研究测试了一个普遍的命题，即问题呈现的偶然特征（例如，回答选项的间距，分配给不同回答选项的颜色）可以引起对其含义的意外推断。这些研究测试的预测来自一个理论框架，假设受访者使用简单的解释性推理来分配意义的视觉特征的问题。接下来的两个实验考察了将图像作为问题文本的补充的效果。图像必然是具体的，实验6测试的假设，这种具体性可能会导致受访者更狭隘地解释问题时，他们伴随着图像。实验7验证了图像中所描绘的项目可以作为被试判断的比较标准的想法。同样，这些研究的结果将导致关于使用图像作为口头问题的辅助手段所涉及的危险的实用指南。最后一系列的研究审查了答卷人何时可能利用调查表的互动功能。这些实验测试了三个一般性假设：受访者更有可能在以下情况下互动地利用他们所获得的信息：1）信息很容易获得，2）它显然是有帮助的，3）受访者有很高的动机寻求帮助。这六个实验将使人们更好地理解让受访者使用可以产生更好的调查数据的功能的方法。

DESCRIPTION (provided by applicant): The rapid acceptance of the Worldwide Web as a vehicle for survey data collection raises important questions about how the new method works. Key features of Web surveys include the use of rich visual presentation of questions and the capability of interaction with the respondent. The rapid growth of the Web makes a close examination of these issues even more urgent. Neither set of features has been explored thoroughly even with earlier modes and the Web offers widely increased resources for both visual display (Web questionnaires can readily incorporate still pictures or video clips) and interaction (such as, floating screens and scrolling for help with definitions). Our application outlines a set of studies designed to address key questions about these issues. The studies focus on Web surveys, but we believe that the results would generalize to other modes of data collection that rely on visual presentation or incorporate interactive design features. Experiments 1-5 examine how respondents interpret the visual cues in Web questionnaires. These studies test the general proposition that incidental features of the presentation of the questions (for example, the spacing of the response options, the color assigned to different response options) can give rise to unintended inferences about their meaning. These studies test predictions derived from a theoretical framework that assumes respondents use simple interpretive heuristics to assign meaning to visual features of the questions. The next two experiments examine the effects of including images as a supplement to the text of the question. Images are necessarily concrete, and Experiment 6 tests the hypothesis that this concreteness may lead respondents to interpret the questions more narrowly when they are accompanied by images. Experiment 7 tests the idea that the item depicted in an image may serve as a standard of comparison for respondents' judgments. Again, the results of these studies will lead to practical guidelines about the dangers involved in using images as an adjunct to verbal questions. The final series of studies examines when respondents are likely to take advantage of interactive features of a questionnaire. These experiments test three general hypotheses; respondents are more likely to utilize the information available to them interactively when 1) the information is easy to obtain, 2) it is clearly helpful, and 3) respondents are highly motivated to seek help. These six experiments would yield a better understanding of methods for getting respondents to use features that could yield better survey data.