Belief and the problem of Ulysses and the sirens

While self-contained, this paper is a sequel to my Belief and the Will (1984); see theBibliographical Note at the end of this paper for a survey of the relevant literature since then and supplementary references. I have meanwhile benefitted greatly from Brad Armendt's commentary and Richard Foley's How should future opinion affect current opinion?, both presented at a symposium at the APA Central Division, April 1993.     

Intention-behaviour consistency: Effects of consideration set size,involvement and need for cognition

This study focuses on the strength of the relationship between behavioural intentions and actual behaviour in a multi-alternative choice context. Two separate moderating processes of intention-behaviour consistency were hypothesized, i.e. the amount of reasoning during intention formation, and the degree of confidence in the intention. Involvement (as an issue-specific factor), and need for cognition (as an individual difference factor) were investigated as antecedents of amount of reasoning. confidence in the intention was predicted from the size of the consideration set (i.e. the number of alternatives that one considers for choice), and involvement. The study comprised a longitudinal two-wave survey conducted before and after national elections in The Netherlands, in which pre-election voting intentions were compared with actual voting behaviour. A high degree of intention-behaviour consistency was found, which was significantly related to both amount of reasoning and confidence. The expected relations were found. The results extend current process models of attitude–behaviour relations. Furthermore, the results indicate that processes related to the consideration set size and content account for variance in intention-behaviour consistency in choice contexts that cannot be accounted for by traditional attitude-behaviour perspectives.     

What is morally at stake when using algorithms to make medical diagnoses? Expanding the discussion beyond risks and harms

In this paper, we examine the qualitative moral impact of machine learning-based clinical decision support systems in the process of medical diagnosis. To date, discussions about machine learning in this context have focused on problems that can be measured and assessed quantitatively, such as by estimating the extent of potential harm or calculating incurred risks. We maintain that such discussions neglect the qualitative moral impact of these technologies. Drawing on the philosophical approaches of technomoral change and technological mediation theory, which explore the interplay between technologies and morality, we present an analysis of concerns related to the adoption of machine learning-aided medical diagnosis. We analyze anticipated moral issues that machine learning systems pose for different stakeholders, such as bias and opacity in the way that models are trained to produce diagnoses, changes to how health care providers, patients, and developers understand their roles and professions, and challenges to existing forms of medical legislation. Albeit preliminary in nature, the insights offered by the technomoral change and the technological mediation approaches expand and enrich the current discussion about machine learning in diagnostic practices, bringing distinct and currently underexplored areas of concern to the forefront. These insights can contribute to a more encompassing and better informed decision-making process when adapting machine learning techniques to medical diagnosis, while acknowledging the interests of multiple stakeholders and the active role that technologies play in generating, perpetuating, and modifying ethical concerns in health care.

     

The effectiveness of narrative versus informational smoking education on smoking beliefs,attitudes and intentions of low-educated adolescents

Objective: This study tests the effectiveness of narrative versus informational smoking education on smoking beliefs, attitudes and intentions of low-educated adolescents.

Design: A field experiment with three waves of data collection was conducted. Participants (N = 256) were students who attend lower secondary education. At the first and third waves, they completed a questionnaire. At the second wave, 50.8% of the participants read a smoking education booklet in narrative form and 49.2% read a booklet in informational form. After reading, all participants also completed a questionnaire at wave 2.

Main outcome measures: Beliefs about negative consequences of smoking, attitudes towards smoking and intentions to smoke were measured.

Results: Repeated measures analyses with time as a within-subjects factor and condition as a between-subjects factor showed that beliefs about smoking were more negative at Wave 2 compared to Wave 1, irrespective of condition. However, attitudes towards smoking were more positive at Wave 3 compared to Wave 1 when participants had read the narrative version.

Conclusion: These results show that narrative smoking education is not more effective than informational smoking education for low-educated adolescents and can even have an unintended effect for this target group by making attitudes towards smoking more positive.     

The Effect of Catastrophe Potential on the Interpretation of Numerical Probabilities of the Occurrence of Hazards1

Two studies demonstrated that identical numerical probabilities of the occurrence of hazards are judged as higher when these involve potential catastrophes compared to noncatastrophic hazards. Fifteen hazards were presented that involve a potential catastrophe and 15 noncatastrophic hazards. Each hazard was given a numerical probability, which was either 1:10, 1:1,000, or 1:100,000. Numerical probabilities were rated as larger when these concerned hazards that have catastrophe potential compared to the noncatastrophic hazards, also when this effect was controlled for perceived benefits. Similar results were obtained in a second study, which controlled for possible confounds (e.g., base rate). The results suggest that verbal interpretations of numerical probabilities of the occurrence of hazards include more than only probability, for instance one's attitude toward the hazardous activity. Implications for risk communication are discussed.     

A Defence of van Fraassen's Critique of Abductive Inference: Reply to Psillos

Psillos has recently argued that van Fraassen's arguments against abduction fail. Moreover, he claimed that, if successful, these arguments would equally undermine van Fraassen's own constructive empiricism, for, Psillos thinks, it is only by appeal to abduction that constructive empiricism can be saved from issuing in a bald scepticism. We show that Psillos' criticisms are misguided, and that they are mostly based on misinterpretations of van Fraassen's arguments. Furthermore, we argue that Psillos' arguments for his claim that constructive empiricism itself needs abduction point up to his failure to recognize the importance of van Fraassen's broader epistemology for constructive empiricism. Towards the end of our paper we discuss the suspected relationship between constructive empiricism and scepticism in the light of this broader epistemology, and from a somewhat more general perspective.