Optimising User Trust in Decision Aids in the Healthcare Domain: A national decision aid for medication prescription
Student thesis: Master Thesis and HD Thesis
- Mikkel Bay Revsbech Jensen
- Morten Adelsen Jakobsen
4. term, Engineering Psychology, Master (Master Programme)
To explore healthcare professionals' trust in decision aids and other healthcare professionals as well as which factors affect these perceptions, an experimental pairwise comparison study with 28 healthcare professionals and medicine students is conducted. The study is based on 10 perceptual statements about two adviser characters (a consultant and a decision aid) that were found, developed, and described based on semi-structured and contextual interviews with healthcare professionals and professors specialising in the Danish healthcare sector.
The experimental study shows that healthcare professionals perceive the statements to be true for the decision aid to a larger extent than the consultant for 3 statements. Healthcare professionals have likewise perceived the statements to be true for the consultant to a larger extent than the decision aid in 5 statements. The statements are not likely to be significantly different in 2 statements.
Based on the combined results of the exploratory and experimental studies it is concluded that healthcare professionals are not averse to advice from decision aids, according to the three factors perceived as more true for the decision aid. Furthermore, is is found that the decision aid can be improved on the remaining 8 factors to perform as well as or better than the consultant. Changes to the design of the decision aid are suggested.
The mixed method approach to explore people's perception of algorithms and algorithmic advice in the form of decision aids is found to provide useful results. The results provide insights into factors that affect healthcare professionals' perception of advice from different advisers. The results can be turned into useful improvements and considerations for the design process of decision aids and algorithmic advice.
The experimental study shows that healthcare professionals perceive the statements to be true for the decision aid to a larger extent than the consultant for 3 statements. Healthcare professionals have likewise perceived the statements to be true for the consultant to a larger extent than the decision aid in 5 statements. The statements are not likely to be significantly different in 2 statements.
Based on the combined results of the exploratory and experimental studies it is concluded that healthcare professionals are not averse to advice from decision aids, according to the three factors perceived as more true for the decision aid. Furthermore, is is found that the decision aid can be improved on the remaining 8 factors to perform as well as or better than the consultant. Changes to the design of the decision aid are suggested.
The mixed method approach to explore people's perception of algorithms and algorithmic advice in the form of decision aids is found to provide useful results. The results provide insights into factors that affect healthcare professionals' perception of advice from different advisers. The results can be turned into useful improvements and considerations for the design process of decision aids and algorithmic advice.
Language | English |
---|---|
Publication date | 6 Jun 2019 |
Number of pages | 52 |