The moderating role of attitude strength in selective exposure to information

Individuals prefer to receive information that is consistent with their attitudes. Three experiments examined whether attitude strength moderates this selective exposure effect. Experiments 1A and 1B found that participants preferred attitudinally consistent information but that this effect was more pronounced to the extent that the attitude was strongly held. Experiment 2 replicated these findings and ruled out an alternative interpretation that a general tendency to hold strong attitudes rather than issue-specific attitude strength moderates selective exposure. Discussion concerns the implications of these findings and the possibility that other variables moderate the selective exposure effect.     

Distortion of probability and outcome information in risky decisions

Substantial evidence indicates that information is distorted during decision making, but very few studies have assessed the distortion of probability and outcome information in risky decisions. In two studies involving six binary decisions (e.g., banning blood donations from people who have visited England, because of “mad cow disease”), student and nonstudent participants distorted their evaluations of probability and outcome information in the direction of their preferred decision alternative and used these biased evaluations to update their preferences. Participants also evaluated the utilities of possible outcomes more positively when the outcomes could follow only from the preferred alternative and more negatively when they could follow only from the competing alternative. Such circular reasoning is antithetical to the normative consequentialist principles underlying decision analysis. Presenting numerical information as precise values or as ranges of values did not significantly affect information distortion, apparently because the manipulation did not affect perceived ambiguity as intended.     

A randomised controlled trial of the effectiveness of writing as a self-help intervention for traumatic injury patients at risk of developing post-traumatic stress disorder

The study investigated the effects of writing and self-help information on severity of psychological symptoms in traumatic injury patients at risk for developing post-traumatic stress disorder (PTSD). Patients attending Accident and Emergency (A & E), were screened for Acute Stress Disorder and randomised to an information control group (n = 36) or a writing and information group (n = 31). Participants in both groups received an information booklet one-month post-injury. Participants in the writing group also wrote about emotional aspects of their trauma during three 20-min sessions, five to six weeks post-injury. Psychological assessments were completed within one month and at three and six months post-injury. There were significant improvements on measures of anxiety, depression and PTSD over time. Differences between groups on these measures were not statistically significant. However, subjective ratings of the usefulness of writing were high. In conclusion, the results do not currently support the use of writing as a targeted early intervention technique for traumatic injury patients at risk of developing PTSD.     

Counting distinctions: on the conceptual foundations of Shannon’s information theory

Categorical logic has shown that modern logic is essentially the logic of subsets (or “subobjects”). In “subset logic,” predicates are modeled as subsets of a universe and a predicate applies to an individual if the individual is in the subset. Partitions are dual to subsets so there is a dual logic of partitions where a “distinction” [an ordered pair of distinct elements (u, u′) from the universe U] is dual to an “element”. A predicate modeled by a partition π on U would apply to a distinction if the pair of elements was distinguished by the partition π, i.e., if u and u′ were in different blocks of π. Subset logic leads to finite probability theory by taking the (Laplacian) probability as the normalized size of each subset-event of a finite universe. The analogous step in the logic of partitions is to assign to a partition the number of distinctions made by a partition normalized by the total number of ordered |U|² pairs from the finite universe. That yields a notion of “logical entropy” for partitions and a “logical information theory.” The logical theory directly counts the (normalized) number of distinctions in a partition while Shannon’s theory gives the average number of binary partitions needed to make those same distinctions. Thus the logical theory is seen as providing a conceptual underpinning for Shannon’s theory based on the logical notion of “distinctions.” This paper is dedicated to the memory of Gian-Carlo Rota—mathematician, philosopher, mentor, and friend.     

The role of visual perspective in information processing

Social events can be described from the perspective of either a person in the situation in which the event occurs (e.g., “John came into…”) or that of an outside observer (“John went into…”). We find that when individuals are disposed to form visual images, they have difficulty comprehending both verbal statements and pictures when the perspective from which the event is described differs from the perspective from which they have encountered similar events in daily life. Furthermore, the disposition to form visual images increases the intensity of emotional reactions to an event when the event is described from the perspective of someone in the situation in which it occurs. These effects are not evident, however, among individuals who typically process information semantically without forming visual images.     

Voters, Emotions, and Memory

In this paper we argue that emotional response to information encountered in campaigns has implications for the likelihood of remembering that information at the time of the vote. Most voting models are built on what voters remember—whether it is the placement of candidates on issues, or opened ended responses to prompts. That recall of memory is biased is well-known, but the extent to which those biases are driven by affective response to candidate information has not been studied. Using dynamic process tracing, we examine voters' emotional responses, information search, and candidate evaluations during a simulated presidential primary campaign. By manipulating anxiety and the amount of incongruent information that voters encounter, we can detect the direct influence of affect in information processing. We find evidence that voters are more likely to remember information that generates any affective reaction as opposed to information for which the subjects report no emotional response. However, we find little evidence that anxiety has a special role, compared to enthusiasm or anger, in increasing the likelihood that an individual item is remembered. This challenges a primary contention of the theory of Affective Intelligence that anxiety leads to more memories for campaign information.     

The Limitations of Heuristics for Political Elites

Despite the extensive literature on citizens' use of cognitive heuristics in political settings, far less is known about how political elites use these shortcuts. Legislative elites benefit from the efficiency of the accessibility heuristic, but their judgments can also be flawed if accessible information is incomplete or unrepresentative. Using personal interviews and a quasi-experimental design, this paper examines the use of the accessibility heuristic by professional legislative staff when assessing the importance of natural resources issues to their constituents. Staff members recall only a small subset of the relevant constituents in the district, and this subset is biased in favor of active and resource-rich constituents over other, equally relevant constituents. This paper provides a new application of cognitive psychology to political elites and addresses important normative questions about the importance of information processing for political representation. By drawing on the psychology literature on heuristics, this paper identifies the cognitive mechanisms of congressional representation and provides new evidence of old biases.     

Depressive symptoms and parenting competence: An analysis of 13 regulatory processes

Mechanisms that lead depressive symptoms to undermine parenting are poorly understood. This review examines cognitive, affective, and motivational processes thought to be responsible for the impact of depressive symptoms on parenting. We present a five-step, action-control model and review 152 studies relevant to 13 regulatory processes. Evidence suggests that depressive symptoms undermine parenting because they reduce child-oriented goals, undermine attention to child input, increase negative appraisals of children and parenting competence, activate low-positive and high-negative emotion, and increase positive evaluations of coercive parenting. Yet, this review reveals significant limitations in knowledge of these processes. Evidence that they mediate depression-parenting relations is scare; important processes remain unstudied; conceptions of regulation are undifferentiated; children’s contributions are largely unexamined; moderating variables are largely unexplored; and methods fail to capture the dynamics of processing input from children. Rigorous testing of such process models holds promise for clarifying the basis of depression-related parenting problems.     

Answer comparison in automated question answering

In the context of Question Answering (QA) on free text, we assess the value of answer comparison and information fusion in handling multiple answers. We report improvements in answer re-ranking using fusion on a set of location questions and show the advantages of considering candidates as allies rather than competitors. We conclude with some observations about answer modeling and evaluation methodology, arising from a more recent experiment with a larger set of questions and a greater diversity of question types and candidates.     

文本检索模型综述

文本检索是信息检索一个重要的分支。随着互联网信息的迅速膨胀,如何检索到用户最需要的信息变得越来越关键。文本检索模型是文本检索中的核心技术,其性能直接影响到搜索引擎的检索质量。本文对当前的经典检索模型及其研究进展进行介绍,并分析各个模型之间的优缺点。