Theoretical and empirical models describing how voters form perceptions of political candidates assume that such perceptions are independent of each other, even though decades of evidence in cognitive science have shown that context influences the perceptions of various stimuli. In this research note, we argue that such perceptions depend on the full range of available ideological platforms. Data from three survey experiments in Israel provide strong support for the hypothesis that voters consistently view candidates as more centrist when a more extreme candidate appears next to them on the ideological spectrum. Our results imply that voters consider the full spectrum of political actors when they form opinions about the ideological stance of any candidate, and the same pattern holds for the perception of the ideological position of parties. A long tradition of research has analyzed how citizens form opinions about the ideological position of political parties and political candidates (Koch 2001; Adams 2012; Fortunato and Stevenson 2013). The significance of this literature is that it links empirical models of spatial voting (Jessee 2009) to research seeking to explain the strategic repositioning of political platforms (Adams et al. 2004; Adams and Somer-Topcu 2009; Ezrow et al. 2011). Interestingly, while much of the formal modeling literature on spatial voting assumes that voters’ perceptions of each party’s policy positions “correspond exactly with the policy promises the party elites issue to the public” and that “all voters Israel Waismel-Manor is a tenured senior lecturer in the School of Political Science at Haifa University, Haifa, Israel. Gabor Simonovits is a doctoral candidate in the Department of Politics, New York University, New York, NY, USA. The authors are indebted to Mik Laver, Walter Mebane, Jim Adams, and the reviewers for all their insightful comments and suggestions along the way. *Address correspondence to Israel Waismel-Manor, Amadrega 4018, School of Political Science, University of Haifa, 3498838, Israel; e-mail: wisrael@poli.haifa.ac.il. Public Opinion Quarterly, Vol. 81, No. 3, Fall 2017, pp. 759–768 doi:10.1093/poq/nfx017 Advance Access publication June 14, 2017 Downloaded from https://academic.oup.com/poq/article-abstract/81/3/759/3868270/The-Interdependence-of-Perceived-Ideological by New York University user on 14 September 2017 have identical perceptions of each party’s policy positions” (Adams 2012, 403), recent empirical studies on voter perceptions in multi-party democracies fail to support these assumptions (Adams et al. 2004; Adams 2012). We show that one reason that perceptions depart from what parties and candidates say is that perceptions are interdependent, and vary systematically based on the range of salient parties and candidates. Most explanations for this apparent disconnect rely on the assumed lack of ability and/or motivation of voters to follow political campaigns (e.g., Zaller 1992). On the other hand, a growing body of empirical research focuses on the perceptual biases that individuals exhibit when they form their perceptions of political platforms (Conover and Feldman 1989; Lauermann 2013). This paper suggests and tests a new mechanism affecting the perceived ideological position of political parties and candidates. Namely, because voters evaluate platforms vis-à-vis each other, their perceptions about them are interdependent (Callander and Wilson 2006, 2008). Our argument is based on the findings of cognitive psychology that when individuals evaluate stimuli, their perception of those stimuli depends on the context in which they are evaluated (Parducci 1965). Such contextual effects have been found in the fields of visual perception (Laberge and Brown 1986), pain research (Watkinson et al. 2013), and marketing (Hutchinson 1983; Simonson 1989). These studies show that when evaluating a stimulus, people identify what they perceive to be the extreme end values and accordingly work out a range that serves as the context for evaluating the stimulus (Yeung and Soman 2005). In other words, the perceived intensity of various stimuli depends on the distribution of the stimuli to which people are exposed, with individuals exposed to more extreme stimuli evaluating others as less intense. Analyzing the effect of question wording and response options on various attitude measures, public opinion scholars have come to very similar conclusions. A large body of evidence (e.g., Schwarz et al. 1985; Zaller 1992; Rockwood, Sangster, and Dillman 1997) shows that when answering closed questions, survey respondents are influenced by the range of response options. In fact, recent research (Broockman 2016; Levendusky and Malhotra 2016) finds that providing individuals with a wider range of policy alternatives to choose from can lead to radically different conclusions about the distribution of mass opinion about political issues. Applying these notions to perceptions about the ideology of political actors, including both parties and individual candidates, implies that changes in the ideological distribution of platforms running for office systematically alter voters’ perceptions about the positions taken by these actors. Given that even in multi-party systems voters are generally capable of perceiving the ideological direction of parties accurately (Blais et al. 2009), such contextual effects are expected to influence the perceived ideological distance of each party from the center. Thus, we hypothesize that after the entry of an additional Waismel-Manor and Simonovits 760 Downloaded from https://academic.oup.com/poq/article-abstract/81/3/759/3868270/The-Interdependence-of-Perceived-Ideological by New York University user on 14 September 2017 candidate or party at one end of the spectrum, the position of the candidates or parties that occupy the same side will be perceived as more moderate. Data and Measurement To test this hypothesis, we use data from three experiments embedded in public opinion surveys fielded in Israel before the general elections of 2009 and 2015. Each experiment measured the perceived ideological position of several candidates and parties. To induce variation in the perceived distribution of candidates and parties on the political spectrum, each experiment manipulated the set of candidates or parties that respondents were asked to consider.

中文翻译：

---

意识形态立场的相互依存

描述选民如何形成对政治候选人的看法的理论和经验模型假设这种看法是相互独立的，尽管认知科学中数十年的证据表明，背景会影响对各种刺激的看法。在本研究报告中，我们认为这种看法取决于所有可用的意识形态平台。来自以色列三项调查实验的数据为以下假设提供了强有力的支持，即当在意识形态范围内出现更极端的候选人时，选民始终认为候选人更中间派。我们的结果意味着选民在对任何候选人的意识形态立场形成意见时会考虑所有政治行为者，同样的模式也适用于对政党意识形态立场的看法。研究的悠久传统分析了公民如何形成对政党和政治候选人意识形态立场的看法（Koch 2001；Adams 2012；Fortunato 和 Stevenson 2013）。这些文献的重要性在于它将空间投票的经验模型（Jessee 2009）与试图解释政治平台战略重新定位的研究联系起来（Adams 等人，2004 年；Adams 和 Somer-Topcu 2009 年；Ezrow 等人，2011 年）。有趣的是，虽然许多关于空间投票的正式建模文献都假设选民对每一党政策立场的看法“与政党精英向公众发布的政策承诺完全一致”，并且“所有选民以色列韦斯梅尔庄园都是终身高级以色列海法海法大学政治学院讲师。Gabor Simonovits 是美国纽约州纽约市纽约大学政治系的博士研究生。在此过程中，作者感谢 Mik Laver、Walter Mebane、Jim Adams 和审稿人的所有有见地的评论和建议。*地址通信至 Israel Waismel-Manor, Amadrega 4018, 海法大学政治学院, 3498838, Israel; 电子邮件：wisrael@poli.haifa.ac.il。民意季刊，卷。81, No. 3, Fall 2017, pp. 759–768 doi:10.1093/poq/nfx017 Advance Access 出版物 2017 年 6 月 14 日 下载自 https://academic.oup.com/poq/article-abstract/81/3/ 759/3868270/The-Interdependence-of-Perceived-Ideological by 纽约大学用户在 2017 年 9 月 14 日对各方的政策立场有相同的看法”（Adams 2012, 403），最近关于多党民主国家选民看法的实证研究未能支持这些假设（Adams 等人，2004 年；Adams 2012 年）。我们表明，看法与政党和候选人所说的不同的一个原因是，看法是相互依存的，并且根据突出的政党和候选人的范围而系统地变化。对这种明显脱节的大多数解释都依赖于假定的选民缺乏能力和/或动机来追随政治运动（例如，Zaller 1992）。另一方面，越来越多的实证研究侧重于个人在形成对政治平台的看法时表现出的知觉偏见（Conover 和 Feldman 1989；Lauermann 2013）。本文提出并检验了一种影响政党和候选人意识形态立场的新机制。即，由于选民相互评估平台，他们对平台的看法是相互依赖的（Callander and Wilson 2006, 2008）。我们的论点基于认知心理学的发现，即当个人评估刺激时，他们对这些刺激的感知取决于他们被评估的背景（Parducci 1965）。在视觉感知（Laberge 和 Brown 1986）、疼痛研究（Watkinson 等人，2013 年）和营销（Hutchinson 1983 年；Simonson 1989 年）等领域都发现了这种背景效应。这些研究表明，在评估刺激时，人们会确定他们认为的极端值，并相应地制定出一个范围作为评估刺激的背景（Yeung 和 Soman 2005）。换句话说，对各种刺激的感知强度取决于人们所接触的刺激的分布，暴露于更极端刺激的个人认为其他人的强度较低。分析问题措辞和回应选项对各种态度措施的影响，舆论学者得出了非常相似的结论。大量证据（例如，Schwarz 等人 1985；Zaller 1992；Rockwood、Sangster 和 Dillman 1997）表明，在回答封闭式问题时，受访者会受到回答选项范围的影响。事实上，最近的研究（Broockman 2016；Levendusky 和 ​​Malhotra 2016）发现，为个人提供更广泛的政策选择可能会导致关于大众对政治问题的意见分布的完全不同的结论。将这些概念应用于对包括政党和个人候选人在内的政治行动者意识形态的看法，意味着竞选公职平台意识形态分布的变化系统地改变了选民对这些行动者所采取立场的看法。鉴于即使在多党制中，选民通常也能够准确感知政党的意识形态方向（Blais 等人，2009 年），这种情境效应预计会影响各方感知的意识形态距离中心。因此，我们假设在从 https://academic.oup 下载的附加 Waismel-Manor 和 Simonovits 760 进入后。com/poq/article-abstract/81/3/759/3868270/The-Interdependence-of-Perceived-Ideological 由纽约大学用户于 2017 年 9 月 14 日在光谱的一端候选人或政党，候选人的立场或站在同一边的政党将被视为更温和。数据和测量 为了验证这一假设，我们使用了 2009 年和 2015 年大选前在以色列进行的民意调查中嵌入的三个实验的数据。每个实验都测量了几个候选人和政党的感知意识形态立场。为了引起候选人和政党在政治光谱上的感知分布的变化，每个实验都操纵了受访者被要求考虑的候选人或政党集。站在同一方的候选人或政党的立场将被视为更温和。数据和测量 为了验证这一假设，我们使用了 2009 年和 2015 年大选前在以色列进行的民意调查中嵌入的三个实验的数据。每个实验都测量了几个候选人和政党的感知意识形态立场。为了引起候选人和政党在政治光谱上的感知分布的变化，每个实验都操纵了受访者被要求考虑的候选人或政党集。站在同一方的候选人或政党的立场将被视为更温和。数据和测量 为了验证这一假设，我们使用了 2009 年和 2015 年大选前在以色列进行的民意调查中嵌入的三个实验的数据。每个实验都测量了几个候选人和政党的感知意识形态立场。为了引起候选人和政党在政治光谱上的感知分布的变化，每个实验都操纵了受访者被要求考虑的候选人或政党集。每个实验都测量了几个候选人和政党的意识形态立场。为了引起候选人和政党在政治光谱上的感知分布的变化，每个实验都操纵了受访者被要求考虑的候选人或政党集。每个实验都测量了几个候选人和政党的意识形态立场。为了引起候选人和政党在政治光谱上的感知分布的变化，每个实验都操纵了受访者被要求考虑的候选人或政党集。