Table 3.
Results of ordinal logistic regression and ANOVA testing ad exposure frequency on risk and benefit recall and recognition
| Outcome | Overall | Experimental Condition | |||
|---|---|---|---|---|---|
| 1 Ad Exposure | 2 Ad Exposures | 4 Ad Exposures | X2-value, p-value | ||
| n (%) | n (%) | n (%) | n (%) | ||
| Recall of risk information | |||||
| 0 | 218 (36.3) | 92 (46.2) | 73 (36.1) | 53 (26.5) | 20.93, p < 0.001 |
| 1 | 202 (33.6) | 63 (31.7) | 71 (35.2) | 68 (34.0) | |
| 2+ | 181 (30.1) | 44 (22.1) | 58 (28.7) | 79 (39.5) | |
| Recall of benefit information | |||||
| 0 | 97 (16.1) | 37 (18.6) | 30 (14.9) | 30 (15.0) | 9.34, p = 0.009 |
| 1 | 256 (42.6) | 99 (49.8) | 77 (38.1) | 80 (40.0) | |
| 2+ | 248 (41.3) | 63 (32.2) | 95 (47.0) | 90 (46.0) | |
| Mean (SE) | Mean (SE) | Mean (SE) | Mean (SE) | F-value, p-value | |
| Recognition of risk information | 4.61 (0.09) | 4.30 (0.09) | 4.61 (0.09) | 4.91 (0.09) | 11.89, p < 0.001 |
| Recognition of benefit information | 5.07 (0.08) | 4.95 (0.08) | 5.01 (0.08) | 5.23 (0.08) | 3.17, p = 0.043 |
*
Significant differences for recall of risk information were between 4 vs. 1 and 4 vs. 2 exposures. Significant differences for recall of benefit information were between 2 vs. 1 and 4 vs. 1 exposures. Significant differences for recognition of risk information were between 2 vs. 1 and 4 vs. 1 exposures. For recognition of benefit information, significant differences were between 4 vs. 1 exposures. To be considered significant, p-values had to be below the Bonferroni adjusted p-value < 0.0167.