Inside the ‘black box’: a thought-listing study examining the effects of political (mis)targeting

Political parties increasingly invest their efforts in strategies like political microtargeting, in which voters receive targeted campaign messages that match their interests and predispositions (Kruikemeier et al., 2022; Zuiderveen Borgesius et al., 2018). Previous studies have examined the effects of this technique on voter turnout (Aggarwal et al., 2023), trust in democracy (Matthes et al., 2022), and voters’ evaluation of political candidates (Broockman & Green, 2014; Krotzek, 2019). Less scholarly attention has been devoted to the way voters process politically targeted messages, causing the processing of these ads to remain a so-called “black-box”. Yet, based on the Elaboration Likelihood Model (ELM), information processing plays an important role in the way individuals form their attitudes and behaviour (Petty & Cacioppo, 1986). We argue that studying information processing provides us with the knowledge necessary to understand how people process politically targeted ads and the consequences thereof. To move forward, we conduct an online thought-listing experiment to examine the effect of targeted political ads on participants’ thoughts, the subject of their thoughts (issue vs. political party), and the sentiment of their thoughts. Previous research has shown that the way people respond to new information strongly depends on whether this information is congruent with their predispositions (Erisen et al., 2018). People tend to defend their prior beliefs by counter-arguing incongruent information presented to them, increasing their cognitive load (Kunda, 1990; Nir, 2011). Scholars have demonstrated that motivated reasoning has implications for party preferences (Slothuus & de Vreese, 2010), candidate evaluations (Redlawsk, 2002), and issue preferences (Meffert et al., 2006), amongst others. However, it remains unclear if voters process politically targeted ads in a different way than non-political ads, and whether this has an effect on their cognitive load. Additionally, we don’t know whether respondents experience more cognitive load about the issue or the sender of a targeted message. Therefore, we formulated the following research questions: RQ1: Is there a difference in the number of thoughts reported by participants that were correctly targeted versus those that were mistargeted? RQ2: Do participants that were (mis)targeted report more thoughts about the issue or political party of the ad? As participants defend their predispositions against incongruent information, they become more negative toward this information (Erisen et al., 2018). Furthermore, we expect that governmental parties will receive more negativity compared to oppositional parties, which results in the following hypotheses: H1: Participants who were (partly) mistargeted will report more negative thoughts than participants who were targeted correctly. H2: The sentiment of reported thoughts differs between governmental parties and oppositional parties. The sentiment towards parties in government will be more negative than the sentiment towards oppositional parties. We expect the effects mentioned in RQ1, RQ2, and H1 will be stronger for people with less political knowledge, as their predispositions about politics are less strong. This leads to the following hypothesis: H3: The aforementioned effects (number of thoughts, subject of thoughts, and sentiment of thoughts) are stronger for people with less political knowledge than for people with high political knowledge.