Modeling adoption, security and privacy of COVID-19 apps: findings and recommendations from an empirical study using UTAUT

Investigating Citizens' Acceptance of Contact Tracing Apps: Quantitative Study of the Role of Trust and Privacy

BACKGROUND

The COVID-19 pandemic accelerated the need to understand citizen acceptance of health surveillance technologies such as contact tracing (CT) apps. Indeed, the success of these apps required widespread public acceptance and the alleviation of concerns about privacy, surveillance, and trust.

OBJECTIVE

This study aims to examine the factors that foster a sense of trust and a perception of privacy in CT apps. Our study also investigates how trust and perceived privacy influence citizens' willingness to adopt, disclose personal data, and continue to use these apps.

METHODS

Drawing on privacy calculus and procedural fairness theories, we developed a model of the antecedents and behavioral intentions related to trust and privacy perceptions. We used structural equation modeling to test our hypotheses on a data set collected at 2 time points (before and after the launch of a national CT app). The sample consisted of 405 Irish residents.

RESULTS

Trust in CT apps was positively influenced by propensity to trust technology (β=.074; P=.006), perceived need for surveillance (β=.119; P<.001), and perceptions of government motives (β=.671; P<.001) and negatively influenced by perceived invasion (β=-.224; P<.001). Perceived privacy was positively influenced by trust (β=.466; P<.001) and perceived control (β=.451; P<.001) and negatively influenced by perceived invasion (β=-.165; P<.001). Prelaunch intentions toward adoption were influenced by trust (β=.590; P<.001) and perceived privacy (β=.247; P<.001). Prelaunch intentions to disclose personal data to the app were also influenced by trust (β=.215; P<.001) and perceived privacy (β=.208; P<.001) as well as adoption intentions before the launch (β=.550; P<.001). However, postlaunch intentions to use the app were directly influenced by prelaunch intentions (β=.530; P<.001), but trust and perceived privacy only had an indirect influence. Finally, with regard to intentions to disclose after the launch, use intentions after the launch (β=.665; P<.001) and trust (β=.215; P<.001) had a direct influence, but perceived privacy only had an indirect influence. The proposed model explained 74.4% of variance in trust, 91% of variance in perceived privacy, 66.6% of variance in prelaunch adoption intentions, 45.9% of variance in postlaunch use intentions, and 83.9% and 79.4% of variance in willingness to disclose before the launch and after the launch, respectively.

CONCLUSIONS

Positive perceptions of trust and privacy can be fostered through clear communication regarding the need and motives for CT apps, the level of control citizens maintain, and measures to limit invasive data practice. By engendering these positive beliefs before launch and reinforcing them after launch, citizens may be more likely to accept and use CT apps. These insights are important for the launch of future apps and technologies that require mass acceptance and information disclosure.

Antecedents predicting digital contact tracing acceptance: a systematic review and meta-analysis

An awareness of antecedents of acceptance of digital contact tracing (DCT) can enable healthcare authorities to design appropriate strategies for fighting COVID-19 or other infectious diseases that may emerge in the future. However, mixed results about these antecedents are frequently reported. Most prior DCT acceptance review studies lack statistical synthesis of their results. This study aims to undertake a systematic review and meta-analysis of antecedents of DCT acceptance and investigate potential moderators of these antecedents. By searching multiple databases and filtering studies by using both inclusion and exclusion criteria, 76 and 25 studies were included for systematic review and meta-analysis, respectively. Random-effects models were chosen to estimate meta-analysis results since Q, I 2, and H index signified some degree of heterogeneity. Fail-safe N was used to assess publication bias. Most DCT acceptance studies have focused on DCT related factors. Included antecedents are all significant predictors of DCT acceptance except for privacy concerns and fear of COVID-19. Subgroup analysis showed that individualism/collectivism moderate the relationships between norms/privacy concerns and intention to use DCT. Based on the results, the mean effect size of antecedents of DCT acceptance and the potential moderators may be more clearly identified. Appropriate strategies for boosting the DCT acceptance rate can be proposed accordingly.

The role of uncertainty and affect in decision-making on the adoption of AI-based contact-tracing technology during the COVID-19 pandemic

Objective

This study explores how negative affect, perceived net equity, and uncertainty influence the public's privacy decision-making regarding the adoption of contact-tracing technology based on artificial intelligence (AI) during the COVID-19 pandemic.

Methods

Four hundred and eighteen adults in the US participated in the study via Amazon Mechanical Turk in August 2020. Statistical analyses were performed using the PROCESS macro. Indirect effects and their significance were estimated using bias-corrected bootstrap confidence intervals (CIs) with resampling set to n = 5000.

Results

Perceived net equity was positively associated with low levels of perceived uncertainty regarding a COVID-19 contact-tracing application and intention to adopt it. Low levels of perceived uncertainty were positively associated with intentions to adopt such an application, thereby suggesting that a perceived level of uncertainty mediates the association between perceived net equity and adoption intentions. Anxieties regarding AI technology and COVID-19 risks both moderate the associations among perceived net equity, perceived level of uncertainty, and intentions to adopt the contact-tracing technology.

Conclusions

Our findings highlight how the differing sources of emotion influence the associations among rational judgment, perceptions, and decision-making about new contact-tracing technology. Overall, the results suggest that both rational judgments and affective reactions to risks are important influencers of individuals' perceptions and privacy-related decision-making regarding a new health technology during the pandemic.

The German COVID-19 Digital Contact Tracing App: A Socioeconomic Evaluation

The COVID-19 pandemic posed challenges to governments in terms of contact tracing. Like many other countries, Germany introduced a mobile-phone-based digital contact tracing solution ("Corona Warn App"; CWA) in June 2020. At the time of its release, however, it was hard to assess how effective such a solution would be, and a political and societal debate arose regarding its efficiency, also in light of its high costs. This study aimed to analyze the effectiveness of the CWA, considering prevented infections, hospitalizations, intensive care treatments, and deaths. In addition, its efficiency was to be assessed from a monetary point of view, and factors with a significant influence on the effectiveness and efficiency of the CWA were to be determined. Mathematical and statistical modeling was used to calculate infection cases prevented by the CWA, along with the numbers of prevented complications (hospitalizations, intensive care treatments, deaths) using publicly available CWA download numbers and incidences over time. The monetized benefits of these prevented cases were quantified and offset against the costs incurred. Sensitivity analysis was used to identify factors critically influencing these parameters. Between June 2020 and April 2022, the CWA prevented 1.41 million infections, 17,200 hospitalizations, 4600 intensive care treatments, and 7200 deaths. After offsetting costs and benefits, the CWA had a net present value of EUR 765 m in April 2022. Both the effectiveness and efficiency of the CWA are decisively and disproportionately positively influenced by the highest possible adoption rate among the population and a high rate of positive infection test results shared via the CWA.