At the time of writing, the UK Government is testing a contact tracing app. Contact tracing apps are seen as key in enabling an easing of lockdown measures, but are effective only if a majority of citizens use it. And that, in turn, could vary with the app’s features.
A team of researchers at Cass Business School – Caroline Wiertz, Aneesh Banerjee, Oguz A. Acar and Adi Ghosh – conducted a survey among the UK population, to investigate their preference for ten potential attributes of the app.
Then, based on the expressed preferences for different features, the researchers modelled the likely adoption rates for apps with specific combinations of features. They found that, to maximise adoption among UK residents, the contact tracing app would need to be presented as managed by the NHS, allow priority access to testing, and not lead to enforced isolation:
The study by Wiertz and her colleagues is both good news and bad news for the UK government. Good news, because the government is indeed branding its contact tracing app as an NHS offering. Bad news, because the government has limited capacity to offer widespread testing; and because the app would only help contain the virus if those flagged as potential carriers isolate themselves.
This study is also a really interesting example of the gap between perceptions and reality, and, thus, of the importance of understanding and responding to perceptions.
The study found that responsibility and oversight for the app was the most influential attribute, with a coefficient of 23.61%. Users would look favourably on an app managed by the NHS or, possibly, an independent oversight body. Conversely, they would feel uneasy about an app ran by the government, and they would be very much against an app managed by a large tech company.
This reaction may not be that surprising, given the current high level of positive sentiment towards the NHS (and despite the fact that, in the past, the NHS has shared patient data with big tech companies – for instance, the DeepMind data deal with Google); the growing unease with the aggressive data collection pursued by Facebook, Google, and other companies; and the fall-out from the Cambridge Analytica and other scandals.
However, the focus on who offers the app masks the fact that the type of contact tracing app currently being tested in the Isle of Wight (and which has been branded the NHSX app) is fundamentally different from the type of app proposed by the Apple-Google consortium. I unpack the differences in this blog post.
In the NHSX app, data are held centrally, whereas in the Google-Apple app data are kept locally (plus device-IDs are change regularly). This means that, bizarrely, the Apple-Google app is much better than the NHSX app, as far as users’ privacy and data security are concerned.
In turn, the low perceived importance of features such as the type of data collected, is at odds with the fact that collecting location data is not necessary for the functioning of the app (and is a privacy threat). The type of alert generated has an impact on the number of alerts received, which, in turn, could result in users getting tired of the app. And while users don’t seem to care whether the app can be used internationally is very much at odds with the fact that this disease can’t really be contained within national borders.
Despite its many limitations, this is a potentially useful tool to ease lockdown measures. What features would encourage you to download and use a contact tracing app?
Ana Canhoto 