Privacy policies of many COVID-19 contact tracing apps too difficult to read
Contact tracing apps for the current novel coronavirus disease (COVID-19) pandemic tend to have unreadable terms of agreement or privacy policies, according to a recent Singapore analysis. This may lead to insecurities about using these apps and may hamper contact tracing efforts.
“A readability analysis of privacy policies is timely and pertinent, given the avalanche of contact tracing applications, and the enforcement by governmental agencies for individuals to download and use,” the researchers said. “Individuals are more likely now to examine the accompanying privacy policies to understand what data is being shared and how their personal information is being protected.”
“Difficulties in comprehending the information contained within the privacy policies could result in reluctance to download and use such applications,” they added.
The analysis focused on the privacy policies of seven contact tracing apps used in different countries: COVIDSafe in Australia, BeAware in Bahrain, CoronApp in Colombia, GH COVID-19 Tracker App in Ghana, Rakning C-19 in Iceland, NZ COVID Tracer in New Zealand, and TraceTogether in Singapore.
To assess the policies, the researchers used the web-based calculator Readability Test Tool. Among the features it assesses are word count, Flesch reading ease, Flesch-Kincaid Reading Grade Level, Simplified Measure of Gobbledygook index, number of complex words, and Gunning-Fog score. All are well-validated indices that quantify readability in terms of sentence construction or the level of education needed to understand a text.
The Gunning-Fox score measures readability as an estimate of the number of years of formal education required to understand the document on first reading. The privacy policies of the contact tracing apps included saw scores ranging from 8.9 to 14.5, suggesting that these texts required high school- to college-level literacy to understand.
In addition, the maximum Flesch reading ease score was 58.8, out of a maximum of 100 points, while the lowest was 45.1. This tool quantifies readability in terms of sentence length and the use of polysyllabic words, and higher scores mean better readability.
Moreover, while the total number of words varied greatly (from 645 to 4,119 words), the percentage of complex words was relatively stable and ranged from 16.3 percent to 20.7 percent.
“With more countries now exiting their lockdowns, the use of contact tracing applications will become more commonplace,” the researchers said. “[I]t is essential that the public have access to readable terms of agreement or privacy policies, so that they are aware how their data is being collected, stored, and used.”
“By increasing the levels of trust that users have in how their data are used by the applications, more users will be confident in utilizing these applications. This will bode well as healthcare research drives into the age of big data to improve healthcare services for everyone,” they added.