Beneficial or Harmful? How (Mis)Fit of Targeted Political Advertising on Social Media Shapes Voter Perceptions

Figures & data

Table 1. Results of the structural equation model controlling for autoregressive relationships.

Predictors	Dependent Variables
	Perceived Manipulative Intent (W2)	Perceived Benefits for Democracy (W2)	Perceived Harms for Democracy (W2)
	b (SE)	b (SE)	b (SE)
Perceived fit	−0.15 (0.06)*	0.14 (0.06)*	0.01 (0.06)
Perceived misfit	0.11 (0.07)	−0.04 (0.06)	0.10 (0.07)
Age	0.01 (0.01)	−0.01 (0.01)*	0.00 (0.01)
Gender (female = 1)	0.09 (0.16)	0.23 (0.12)	0.20 (0.16)
High education	−0.11 (0.22)	0.09 (0.17)	0.37 (0.22)
Medium education	−0.29 (0.19)	0.03 (0.15)	0.06 (0.18)
Political ideology (W2)	0.01 (0.03)	0.06 (0.03)	0.02 (0.03)
OBA knowledge (W2)	0.06 (0.05)	−0.14 (0.04)**	−0.06 (0.05)
Political interest	0.08 (0.08)	0.09 (0.05)	−0.08 (0.08)
Political distrust	0.11 (0.07)	−0.04 (0.06)	0.26 (0.07)***
Privacy concerns	0.16 (0.07)*	−0.04 (0.05)	0.14 (0.07)
Perceived manipulative intent (W1)	0.38 (0.07)***	—	—
Perceived benefits for democracy (W1)	—	0.39 (0.07)***	—
Perceived harms for democracy (W1)	—	—	0.32 (0.08)***
R²	0.34	0.36	0.31

Note. N = 428. OBA = online behavioral advertising. Robust maximum likelihood estimation was applied. Robust fit indices are reported χ² (617) = 857.83, p < .001, CFI = .97, TLI = .96, RMSEA = .03, SRMR = .05. The control variables age, gender, education, political interest, political distrust, and privacy concerns were measured in Wave 1 (W1); OBA knowledge and political ideology in Wave 2 (W2).

*p < .05; **p < .01; ***p < .001.

Figure 1. Model showing the relationships between the perceived fit and misfit of targeted political advertising (TPA) on perceived manipulative intent and perceived benefits and harms for democracy. N = 428. The paths were estimated applying robust maximum likelihood estimation. Robust fit indices are reported: χ² (617) = 857.83, p < .001, CFI = .97, TLI = .96, RMSEA = .03, SRMR = .05. Political interest, political distrust, political ideology, age, gender, education, online behavioral advertising (OBA) knowledge, and privacy concerns served as control variables. Dotted lines show autoregressive relationships. W1 = Wave 1; W2 = Wave 2. *p < .05; **p < .01; ***p < .001.