The study of trust has advanced tremendously in recent years, to the extent that the goal of a more unified formalisation of the concept is becoming feasible. To that end, we have begun to examine the closely related concepts of regret and forgiveness and their relationship to trust and its siblings. The resultant formalisation allows computational tractability in, for instance artificial agents. Moreover, regret and forgiveness, when allied to trust, are very powerful tools in the Ambient Intelligence (AmI) security area, especially where Human Computer Interaction and concrete human understanding are key. This paper introduces the concepts of regret and forgiveness, exploring them from social psychological as well as a computational viewpoint, and presents an extension to Marsh's original trust formalisation that takes them into account. It discusses and explores work in the AmI environment, and further potential applications.

Many people are now influenced by the information and advice they find on the Internet, much of it of dubious quality. This article describes two studies concerned with those factors capable of influencing people's response to online advice. The first study is a qualitative account of a group of house-hunters attempting to find worthwhile information online. The second study describes a survey of more than 2,500 people who had actively sought advice over the Internet. A framework for understanding trust in online advice is proposed in which first impressions are distinguished from more detailed evaluations. Good web design can influence the first process, but three key factors-source credibility, personalization, and predictability-are shown to predict whether people actually follow the advice given.

To observe the course of the development of automatic processes in word recognition, two experiments were performed in which a distracting word interacted with a target stimulus. In Expt 1 the subjects named a simple line drawing, and in Expt 2 they named a word appearing in a rectangle. The distracting word (i) was a semantic associate of the target stimulus, (ii) was graphemically similar to the name of the target, (iii) was phonemically similar to the name of the target, (iv) was unrelated in meaning, orthography and phonology, or (v) was replaced with an unpronounceable letter string. Three groups of subjects were tested. Skilled adult readers were compared with two groups of children. Matched groups of children were used, with similar ages (9-12 years old) and non-verbal spatial abilities, but dissimilar reading ages. All groups of subjects showed semantic interference and graphemic priming from the distractor word in Expt 1, and the children showed phonemic priming. In Expt 2, only the poor readers showed semantic interference, and no groups showed graphemic or phonemic effects. The sensitivity of the poor readers to the distractors indicates that the semantic, graphemic and phonemic codes of words are generated by these subjects, and that the difference in reading age between them and the good readers can be attributed to post-lexical processes such as text integration and articulatory planning and execution.

No abstract

No abstract