Participants who did not already have a strong opinion about legalizing recreational marijuana looked at a website that argued for legalization in Tennessee. They wore an eye tracker that recorded how many minutes they spent looking at the high-quality arguments instead of the celebrity endorsements and images. A month later, they were asked whether they favored legalization. If the central route produces a lasting attitude, then spending more time looking at the actual arguments should be associated with a greater likelihood of still favoring legalization a month later.
Minutes spent looking at high-quality persuasive information are associated with the likelihood of favoring legalization of recreational marijuana in Tennessee.
The data for this exercise were made up and included 200 participants. Minutes is the amount of time each participant spent looking at the high-quality arguments, so it is a continuous variable. Favor_1 has two categories. Participants who favored legalization a month later were coded 1. Those who were opposed, undecided, or did not answer were coded 0. I used logistic regression because Favor_1 only has two possible categories, while Minutes is continuous.
| Logistic Regression Results | ||||
| Odds Ratios with 95% Confidence Intervals | ||||
| term | Odds_Ratio | CI_Lower | CI_Upper | P_Value |
|---|---|---|---|---|
| (Intercept) | 0.099 | 0.044 | 0.205 | 0.0000 |
| IV | 1.167 | 1.116 | 1.227 | 0.0000 |
| Linearity of the Logit Test (Box-Tidwell) | |||
| Interaction term indicates violation if significant | |||
| term | Estimate | Std_Error | P_Value |
|---|---|---|---|
| (Intercept) | −2.566 | 1.100 | 0.0196 |
| IV | 0.262 | 0.290 | 0.3657 |
| IV_log | −0.032 | 0.078 | 0.6869 |
| Inflection Point of Logistic Curve | |
| Value of IV where predicted probability = 0.50 | |
| Probability | Inflection_Point |
|---|---|
| 0.5 | 14.965 |
Looking at the curve, the chance of favoring legalization goes up as the number of minutes goes up. The odds ratio for Minutes is 1.167 and the p-value is below .001. For each additional minute, the odds of favoring legalization increase by about 1.17 times. The confidence interval is 1.116 to 1.227, which stays above 1. Since the p-value is below .001, the result is statistically significant and supports my hypothesis. The IV_log p-value on the Box-Tidwell table is about .69. Since that is not significant, the linearity assumption holds. The inflection point is about 14.97 minutes. This is the point where the chance of favoring legalization reaches 50 percent. After about 15 minutes looking at the arguments, participants were more likely than not to favor legalization a month later. The eye tracker recorded how many minutes the participants spent looking at the arguments, but it does not tell what they were thinking while they looked at them. I can’t say the amount of time caused them to favor legalization a month later. Time spent looking at the celebrity endorsements or images was not measured in the data, so this test is only looking at the central route, not the peripheral route.