Prevalence and Characteristics of Manipulative Design in Mobile Applications Used by Children

Abstract

Key Points Question What types of manipulative design features exist in childrens’ mobile apps, and do inequities exist in young childrens exposure to manipulative design? Findings In this cross-sectional study of apps used by 160 children aged 3 to 5 years, the majority of apps were associated with manipulative design features that included parasocial relationship pressure, fabricated time pressure, navigation constraints, and use of attractive lures to encourage longer gameplay or more purchases, in addition to advertisement-based pressure; only 20% of apps had no manipulative design features. Children from lower socioeconomic strata played apps with more manipulative design. Meaning These results suggest that interactive designs that serve the interests of technology companies over the interests of children are common and deserve further study and regulation.

Dark commercial patterns

“…Furthermore, many stakeholders have recognised the heightened vulnerability to dark patterns of certain subsets of consumers, and some suggest that this is what warrants the urgency to regulate them (Chugh and Jain, 2021 [135]). For instance, in relation to dark patterns that unnecessarily impose transaction costs such as making it difficult to opt out, the Stigler Committee (2019 [49]) note that "users who are less tech savvy or do not have the extra time to devote to navigating byzantine opt out procedures will be less likely to persist so that they can express their authentic preferences in the transaction.…”

“…Further, these groups may preferentially include those who are already at some social disadvantages, such as elderly people with less developed technology skills or less educated people." Also, Radesky (2021 [135]) considers that five differences from adults make children more susceptible to dark patterns: having immature executive function; forming imaginative relationships with characters; being susceptible to rewards; being indifferent or unfamiliar with data privacy; and lack of understanding of virtual currencies.…”

“…But another possible explanation is that such practices may not be clearly unlawful under existing laws in some jurisdictions, such as general misconduct prohibitions, and/or present a high evidentiary burden under such laws. Various commentators have suggested that in particular dark patterns that are not clearly deceptive -in the sense that they do not necessarily cause a consumer to believe something that is not true -are unlikely to be captured by prohibitions on misleading or deceptive conduct as formulated in certain jurisdictions (CPRC, 2020 [148]; King and Stephan, 2021 [149]; ACCC, 2021 [88]; Chugh and Jain, 2021 [135]). Indeed, in their review of US case law, Luguri and Strahilevitz (2021 [25]) did not identify cases indicating whether practices that do not clearly constitute deceptive acts, such as nagging, price comparison prevention, intermediate currency, toying with emotion or confirmshaming, were unlawful.…”

“…This evolving nature favours a mix of a principle-and a rule-based approach to consumer law. Specifically, well-designed rule-based regulation, or bright-line rules, banning specific practices may help address certain dark patterns that are already in existence and considered harmful, as well as provide clarity to consumer authorities on when to intervene and businesses certainty in relation to specific designs (OECD, 2021 [9]; Chugh and Jain, 2021 [135]). But new dark patterns may not fit neatly into the scope of specific bans (OECD, 2021 [9]; Corones et al, 2016[184]).…”

“… To identify the prevalence of dark patterns (including possible third-party entities that might create them), mystery shopping or sweeping (similar to e.g. the 2022 EC study (EC, 2022 [29]) or SERNAC (2021 [72])) including with manual or automated screenshot tools, selective audits of user interfaces (Chugh and Jain, 2021 [135]; Luguri and Strahilevitz, 2021 [25]) and machine learning-based web crawling tools (similar to e.g. by Mathur et al (2019[24])) can be employed.…”