Invisible transparency: How different types of ad disclaimers on Facebook affect whether and how digital political advertising is perceived

Abstract

One central measure set out in the regulation of digital political advertising (DPA) concentrates on transparency disclaimers to make users aware that the respective content was bought and targeted at them with a specific intention by an advertiser. However, we lack scientific evidence about if and how users perceive transparency disclaimers of DPA on social media. This article aims to provide first empirical answers to these questions by drawing on a two‐part eye‐tracking study with 177 participants that compares the effect of different prominent ad disclaimers (i.e., versions previously [V1] and currently [V2] used by Facebook as well as a self‐designed [V3] disclaimer version) on the perception of DPA. We show that most users do not fixate on the ad disclaimers regardless of their prominence, nor does the prominence of the disclaimers affect the perceived intrusiveness and acceptance of the DPA. However, the recall and ad recognition were significantly lower for the less prominent ad disclaimers used by Facebook compared to our self‐designed more prominent version, pointing to the shortcomings of the platform's current transparency rules. Altogether, our study allows a more substantiated discussion about how DPA is recognized and evaluated by users, which contributes to the debate about incorporating regulations for DPA.

For your eyes only? An eye-tracking experiment investigating microtargeting transparency, visual attention and critical processing

“…Within existing disclosure research, results regarding the correct recallment of the disclosure are very diverse (Binder et al, 2022;Evans et al, 2017;Jost et al, 2022;Kruikemeier et al, 2016). The current study found the same results with regard to recallment, in our control condition with the sponsored disclosure, 60% of our participants recalled the correct disclosure, while in the experimental condition, 37% of our participants recalled the correct disclosure, and the same number of participants incorrectly recalled a sponsored disclosure.…”

“…However, research shows that these disclosures seem insufficient in informing users and increasing transparency (Binder et al, 2022;Binford et al, 2021;Jansen & Krämer, 2023;Jost et al, 2022).…”

“…In the field of PMT, research shows that even though they are promising, targeting disclosures do not always get recognized or recalled by users (Binder et al, 2022;Dobber et al, 2023;Jost et al, 2022;Kruikemeier et al, 2016). However, in one of the first eyetracking studies regarding disclosures on microtargeted ads, Jost and colleagues (2022) found that while a prominent disclosure did not lead to more visual attention to the disclosure, participants were still better able to recall sponsored ads in a timeline.…”

Time for transparent targeting: an investigation of targeting disclosures, coping mechanisms, credibility, and political attitude

“…However, more prominent disclosures seem to be effective in informing users of the persuasive or targeted nature of a message. Prominent disclosures are PREPRINT: DISCLOSURES AND COPING WITH MICROTARGETING 6 those that are more easily seen by users and therefore lead to higher levels of recognition that an advertisement is an advertisement (Amazeen and Wojdynski, 2020) For instance, research shows higher levels of advertising recognition with a highlighting square around the ad and the disclosure (Jost et al, 2022). Other work suggests that disclosures that include information that aligns with the regulations in the DSA are noticed more easily (compared to disclosures investigated in prior advertising research) (Dobber et al, 2023).…”

“…Moreover, the hover feature can show participants information in a more fluent manner than Facebook's 'Why am I seeing this ad' because the popup will (dis)appear based on the mouse movements and is therefore less interrupting and more readily available, while still being applicable in the platforms corporate design, compared to large boxes drawn around an advertisement (Jost et al, 2022). While existing work underlines the importance of investigating new types of disclosures that adhere to the regulations of the DSA (Dobber et al, 2023;Jost et al, 2022), we examine a newly developed targeting disclosure with red and underlined text, as well as the option to see the actual datapoints that were used regarding the targeting parameters. To the best of our knowledge, targeting disclosures such as these have not been studied with regard to PMT.…”