Skip to main content

Table 2 Mediation of pictorial health warnings’ effect on parents’ likelihood of purchasing any sugar-sweetened beverage for their children

From: How pictorial warnings change parents’ purchases of sugar-sweetened beverage for their children: mechanisms of impact

Mediator

 

Warnings on mediator

Mediator on purchasing

Total effect

Mediated effect

n

a

b

c

a x b (95% CI)

Message reactions

     

Attention to the labels

322

1.48 (1.21, 1.76)

0.13 (-0.06, 0.31)

− 0.93 (-1.48, -0.38)

0.19 (-0.08, 0.49)

Negative emotional reactions

320

1.87 (1.66, 2.09)

-0.10 (-0.34, 0.14)

-0.53 (-1.17, 0.11)

-0.19 (-0.67, 0.28)

Thinking about harms of drinking SSBs

322

2.31 (2.07, 2.55)

-0.14 (-0.34, 0.07)

-0.42 (-1.09, 0.24)

-0.32 (-0.85, 0.18)

Attitudes and beliefs

     

Perceived amount of added sugar in SSBs

321

0.10 (-0.06, 0.26)

-0.23 (-0.55, 0.09)

-0.73 (-1.20, -0.26)

-0.02 (-0.09, 0.02)

Perceived healthfulness of SSBs for child

321

-0.26 (-0.44, -0.08)

0.64 (0.34, 0.94)

-0.60 (-1.08, -0.12)

-0.17 (-0.33, -0.05)

Appeal of SSBs for child

322

-0.08 (-0.30, 0.13)

-0.00 (-0.24, 0.23)

-0.74 (-1.20, -0.27)

0.00 (-0.03, 0.04)

Perceived tastiness of SSBs for child

321

-0.17 (-0.37, 0.03)

-0.01 (-0.26, 0.24)

-0.72 (01.19, -0.26)

0.00 (-0.05, 0.06)

Perceived likelihood of health problems

309

0.12 (-0.10, 0.33)

-0.40 (-0.65, -0.15)

-0.83 (-1.31, -0.35)

-0.05 (-0.14, 0.05)

Injunctive norms to limit SSBs for child

316

0.39 (0.11, 0.67)

0.10 (-0.09, 0.28)

-0.81 (-1.29, -0.34)

0.04 (-0.04, 0.13)

Intentions

     

Anticipated social interactions

322

1.73 (1.45, 2.01)

-0.02 (-0.20, 0.16)

-0.71 (-1.27, -0.15)

-0.03 (-0.34, 0.29)

Intentions to serve SSBs to child

320

-0.35 (-0.64, -0.05)

0.20 (0.03, 0.38)

-0.65 (-1.12, -0.18)

-0.07 (-0.21, -0.003)

  1. Note. SSB = sugar-sweetened beverages. The second column reports coefficients for the a pathway, regressing the mediator on treatment arm. The third column shows coefficients for the b pathway, regressing purchasing SSBs (yes/no) on the mediator, controlling for treatment arm. Coefficients for both pathways are presented in log odds metric. The final column shows mediated effects (a × b). The c pathway represents the coefficient from regressing the outcome on treatment arm; this coefficient may vary due to differing sample sizes. Bold indicates statistical significance, p < .05. Confidence intervals for mediated effects were bootstrapped