Trust is crucial for human interaction with artificial intelligence (AI) and is frequently measured through questionnaires or rating scales. One commonly used questionnaire in AI research is the Trust between People and Automation scale (TPA). However, its psychometric quality has yet to be examined in the context of AI. More recently, a Trust Scale for Explainable AI (TXAI) was recommended but not empirically evaluated. In this study, we assessed the psychometric qualities of both scales, using confirmatory and exploratory factor analyses to test the scales’ validity and coefficients α and ω for reliability estimation. Our results suggested good psychometric quality for the TXAI after removing one item. Concerning the TPA, acceptable quality was only achieved when using a two-factor model (trust and distrust) and after removing two items. We provide recommendations for using the two scales and evidence to distinguish trust and distrust as separate psychological constructs.
User experience (UX) research relies heavily on survey scales to measure users' subjective experiences with technology. However, repeatedly raised concerns regarding the improper use of survey scales in UX research and adjacent fields call for a systematic review of current measurement practices. Therefore, we conducted a systematic literature review, screening 153 papers from four years of the ACM Conference on Human Factors in Computing Systems proceedings (ACM CHI 2019 to 2022), of which 60 were eligible empirical studies using survey scales to study users' experiences. We identified 85 different scales and 172 distinct constructs measured. Most scales were used once (70.59%), and most constructs were measured only once (66.28%). The System Usability Scale was the most popular scale, followed by the User Experience Questionnaire, and the NASA Task Load Index. Regarding constructs, usability was the most frequently measured, followed by attractiveness, effort, and presence. Furthermore, results show that papers rarely contained complete rationales for scale selection (20.00%) and seldom provided all scale items used (30.00%). More than a third of all scales were adapted (34.19%), while only one-third of papers reported any scale quality investigation (36.67%). On the basis of our results, we highlight questionable measurement practices in UX research and suggest opportunities to improve scale use for UX-related constructs. Additionally, we provide six recommended steps to promote enhanced rigor in following best practices for scale-based UX research.
Past research has demonstrated that aesthetics affect users' experiences in various ways. However, there is little research on the impact of interface aesthetics on user performance in a smartphone app context. The present paper addresses this research gap using an online experiment (N = 281). Two variants of the same web app were created and manipulated in their aesthetics. Participants were randomly assigned to either variant and asked to explore the app before answering questions concerning the app's content. Results showed a significant positive effect of aesthetics on perceived usability and aesthetics. Furthermore, results point toward a positive impact of interface aesthetics on performance (i.e., the number of questions answered correctly). Thus, results indicate that a visually appealing smartphone web app increases users' subjective experience and objective performance compared to an unaesthetic app. This suggests that user interface aesthetics impact users' experiences and provide stakeholders with quantifiable value and competitive advantage.
INTRODUCTION Code status discussions are useful for understanding patients’ preferences in the case of a cardiac/pulmonary arrest. These discussions can also provide patients with a basis for informed decision-making regarding life-sustaining treatment. We conducted a survey to understand current practices and perceptions of code status discussions in a tertiary-care Swiss hospital. METHODS We performed systematic interviews across different departments of the University Hospital of Basel. We interviewed 258 physicians and 145 patients who were hospitalised between May and July 2018 using a questionnaire designed to assess the use of code status discussions and to gauge patients’ individual experiences and opinions. RESULTS A total of 61.4% of patients did not recall having had a code status discussion during the hospital stay. However, a higher proportion of medical patients compared to surgical patients recalled having had a discussion (43.6 vs 22.4%, p = 0.03). For 9 out of 38 (23.7%) patients who did recall the discussion, there was a lack of agreement between the preference given in the interview regarding resuscitation measures and the documented code status in the medical electronic chart. Furthermore, a majority of physicians (72.4%) recalled defining a do-not-resuscitate (DNR) status for a patient without prior discussion with the patient. Physicians who recalled determining the DNR status without patient consultation reported conflicts with patients and relatives regarding code status at a higher rate compared to physicians who did not define DNR status without consultation (62.4 vs 39.4%, p <0.001). CONCLUSION A majority of patients do not report having discussed code status during their hospital stay and physicians frequently omit such discussions, thereby potentially failing to attend to patients’ preferences for care. Physician training regarding code status discussions may improve the quality of informed decision-making and patient-centred care.
Objectives: Prior research found the gut microbiota-dependent and pro-atherogenic molecule trimethylamine-N-oxide (TMAO) to be associated with cardiovascular events as well as all-cause mortality in different patient populations with cardiovascular disease. Our aim was to investigate the prognostic value of TMAO regarding clinical outcomes in patients after out-of-hospital cardiac arrest (OHCA). Methods: We included consecutive OHCA patients upon intensive care unit admission into this prospective observational study between October 2012 and May 2016. We studied associations of admission serum TMAO with in-hospital mortality (primary endpoint), 90-day mortality and neurological outcome defined by the Cerebral Performance Category (CPC) scale. Results: We included 258 OHCA patients of which 44.6% died during hospitalization. Hospital non-survivors showed significantly higher admission TMAO levels (μmol L−1) compared to hospital survivors (median interquartile range (IQR) 13.2 (6.6–34.9) vs. 6.4 (2.9–15.9), p<0.001). After multivariate adjustment for other prognostic factors, TMAO levels were significantly associated with in-hospital mortality (adjusted odds ratios (OR) 2.1, 95%CI 1.1–4.2, p=0.026). Results for secondary outcomes were similar with significant associations with 90-day mortality and neurological outcome in univariate analyses. Conclusions: In patients after OHCA, TMAO levels were independently associated with in-hospital mortality and other adverse clinical outcomes and may help to improve prognostication for these patients in the future. Whether TMAO levels can be influenced by nutritional interventions should be addressed in future studies.