Trust in search engines - developing a measurement tool and applying it in an experiment
The purpose of the PhD project is to investigate to what extent users trust the search engines they use and its impact on search behavior. Additionally, it is examined how trustworthiness of the search engine operator and reliance of the search engine algorithm contribute to trust.
The expected results determine misplaced and legitimate trust in search engines. Consequently, this furthers the discussion among civil society, researchers, and policymakers on the societal consequences of trust, the role of search engines, and the necessary user skills.
The project builds on interdisciplinary research on trust and selection behaviour with commercial search engines.
A variety of qualitative and quantitative methods is applied, such as a laboratory study, a survey and an experiment.
The developed trust measure for search engines may be adapted and re-used for novel AI applications.
A holistic conception of trust is applied to search engines. Trust is surveyed with several indirect items.
The research questions are applied to scenarios from the domains of health and finance. On the one hand, these topics form good preconditions for the observation of trust; on the other hand, the effects of trust can possibly be far-reaching. Trust is tested for Google and Ecosia, since the search engines are both contrary in their implemented technology and image as a brand.
A laboratory study with stimulated-recall interviews will first be used to identify indicators of trust in web searches and to determine specific trust-related behavior. From this, items for measuring trust in search engines will be generated and validated with the help of a survey. Finally, the validated trust measure will be used in an experiment to test the relations with trustworthiness/reliance and trust-related behavior. The combination of self-reporting and actual behavioral measurements has rarely been used in relation to trust in search engines before.
As part of the project, a survey to test search engine literacy has been developed and is open for re-use.
Funding period 12/2020-11/2023
Häußler, H.; Schultheiß, S.; Lewandowski, D. (2023). Is googling risky? A study on risk perception and experiences of adverse consequences in web search. Journal of the Association for Information Science and Technology, 1-14. https://doi.org/10.1002/asi.24802
Häußler, H. (2023). Do users trust search engines? And if so, why?: Developing a trust measure and applying it in an experiment. In: ACM SIGIR Conference on Human Information Interaction and Retrieval (CHIIR ’23), March 19–23, 2023, Austin, TX, USA. ACM, New York, NY, USA. https://doi.org/10.1145/3576840.3578280