Final Activity Report Summary - CSEE (Cognition in structured electronic environments)
Research carried out in this project aimed at studying strategies underlying judgments of trust, authority and reputation in information search tasks. Web users massively rely on search engine results to find information they are looking for and they regularly face the problem of deciding "which link to follow" or "whom to trust". This raises a number of technological, ethical, and cognitive issues on the use of search engines as tools to select trustworthy sources of information.
From a psychological point of view, studying online information search behaviour proved to be an ideal testbed to understand how people make trust-based decisions in real-world tasks.
We started by analysing the psychological literature on credibility, persuasion and trust in the WWW and identified a number of key methodological and conceptual issues deserving particular attention. Research on Web credibility has focused, in particular, on the study of evaluative judgments of sources of information, in ideal, decontextualised conditions. It has not taken into account the fact that decisions made in ecologically valid conditions are typically based on fast and shallow judgments. Information search typically involves decisions based on simple cues to predict the expected trustworthiness of a source.
The focus of our research has been, hence, on simple heuristics driving what links Web users follow or what search result they select in trust-based decision conditions. We presented and published a number of analyses with the rationale for the study of heuristics involved in trust-based information search, and proposed a taxonomy of cues available on the Web involved in this class of tasks.We designed and implemented a platform based on the Google API to retrieve and manipulate search engine results and use them for Web-based experimental research. This platform was used to carry out a large-scale study on search engine use investigating how user behaviour varies as a function of available cues and different types of task constraint. We identified in particular the effects of a number of cues (such as length, depth and nature of the links) on judgments of trust and the resulting biases affecting user behaviour. Further empirical protocols were prepared to conduct research on the effect of social cues (such as those afforded by Web 2.0 services) on trust perception, to be carried out in a forthcoming series of Web experiments.
The overall outcome of this project is a methodological framework and a number of empirical results towards the study of heuristic processes driving information search and judgment of trust in the WWW. Further empirical research will need to assess the global biases resulting from these heuristic processes on how users access and select online information and estimate its credibility. Our results also suggest a number of directions relevant to the design of Web services to harness the potential of authority and reputation indicators available in the WWW, with a specific attention on social software.
From a psychological point of view, studying online information search behaviour proved to be an ideal testbed to understand how people make trust-based decisions in real-world tasks.
We started by analysing the psychological literature on credibility, persuasion and trust in the WWW and identified a number of key methodological and conceptual issues deserving particular attention. Research on Web credibility has focused, in particular, on the study of evaluative judgments of sources of information, in ideal, decontextualised conditions. It has not taken into account the fact that decisions made in ecologically valid conditions are typically based on fast and shallow judgments. Information search typically involves decisions based on simple cues to predict the expected trustworthiness of a source.
The focus of our research has been, hence, on simple heuristics driving what links Web users follow or what search result they select in trust-based decision conditions. We presented and published a number of analyses with the rationale for the study of heuristics involved in trust-based information search, and proposed a taxonomy of cues available on the Web involved in this class of tasks.We designed and implemented a platform based on the Google API to retrieve and manipulate search engine results and use them for Web-based experimental research. This platform was used to carry out a large-scale study on search engine use investigating how user behaviour varies as a function of available cues and different types of task constraint. We identified in particular the effects of a number of cues (such as length, depth and nature of the links) on judgments of trust and the resulting biases affecting user behaviour. Further empirical protocols were prepared to conduct research on the effect of social cues (such as those afforded by Web 2.0 services) on trust perception, to be carried out in a forthcoming series of Web experiments.
The overall outcome of this project is a methodological framework and a number of empirical results towards the study of heuristic processes driving information search and judgment of trust in the WWW. Further empirical research will need to assess the global biases resulting from these heuristic processes on how users access and select online information and estimate its credibility. Our results also suggest a number of directions relevant to the design of Web services to harness the potential of authority and reputation indicators available in the WWW, with a specific attention on social software.