This study does not involve any smoking cessation interventions. This study does involve experimental manipulation of nicotine/tobacco product price and availability to understand consumers' behavior. Participants will be provided with a commercially available e-cigarette to use during the study, if they wish.
This study uses a within subject design. Menthol cigarette smokers will complete questionnaires on a computer, sample a range of tobacco products, and purchase tobacco products in an online store under four different scenarios: a) cigarette flavor restricted and e-cigarette flavor restricted, b) cigarette flavor unrestricted and e-cigarette flavor restricted, c) cigarette flavor restricted and e-cigarette flavor unrestricted, d) cigarette flavor unrestricted and e-cigarette flavor unrestricted.
Participants will complete a 1) consent and assessment session, and a product sampling phase, 2) an ETM session and a 3) follow-up session:
In the consent and initial assessment session, participants will go through standard consent procedures and then provide a breath sample to confirm recent levels of smoking. They may be asked to provide a urine sample for an analysis of cotinine content. Participants will complete a timeline follow back to assess previous month's recent smoking, e-cigarette use, and consumption of nicotine products, and to determine the ETM budget. A Qualtrics survey will administer demographics questions, smoking assessments (Questionnaire on Smoking Urges-Brief, Minnesota Nicotine Withdrawal Scale, Fagerstrom Test of Nicotine Dependence, Perceived Health Risk for cigarettes and e-cigarettes), and delay discounting tasks. At the end of the session, participants will experience a trial of the ETM that will be used in the next session.
For the 7-day sampling phase, they will be provided a list of available flavors and identify 5 that they would like to try. They will leave with a sample of each of the 5 flavors to try during the next few days. Participants will also be provided with a sample of any other tobacco product they wish to examine and try. They may sample any products in the smoking lab before they leave.
In the ETM session, participants will buy tobacco products to use throughout the next 7 days. Participants will complete a total of 40 purchasing trials each for 7 days' worth of products. They will be exposed to 4 conditions with all cigarettes increasing in price and repeat the same conditions with all e-liquids increasing in price across trials. A balanced Latin square design will be used to present the following conditions:
In the Cigarette Flavor Restricted, only conventional tobacco cigarettes will be available. In the Cigarette Flavor Unrestricted, tobacco and menthol cigarettes will be available. In the E-Cigarette Flavor Restricted, only tobacco flavor e-liquids will be available. In the E- Cigarette Flavor Unrestricted, tobacco flavor e-liquid and a variety of other flavors will be available from five broad categories: tobacco, fruit, dessert, menthol, and coffee. A range of other tobacco products will be available across conditions, such as snus, lozenge, gums, dip, and nicotine pouch.
3\) In the follow-up session, participants will be able to return the e-cigarette if they wish. They will answer Perceived Health Risk assessments, complete hypothetical purchase tasks for cigarettes and e-cigarettes, rank order the products they sampled, and complete a liking scale for every tobacco product.
We will summarize demographic characteristics (e.g., age, race, income) and smoking-related measures (e.g., FTCD, QSU, PHR) using means, standard deviations, and percentages. For each product category in the ETM, we will test whether policy restrictions differentially influence purchasing of cigarettes, e-cigarettes, oral tobacco/nicotine products, and NRT products.
We will use hierarchical linear regression to evaluate differences in purchasing behavior, including a three-way interaction among the cigarette policy condition, the e-cigarette policy condition, and the log-transformed manipulated commodity price, with random intercepts for individuals. Model selection will be performed to identify the best-supported combination of predictors and interaction terms. We will conduct an exhaustive model search and select the optimal model based on the lowest Bayesian Information Criterion (BIC). Model posterior probabilities will be calculated using BIC following Barbieri and Berger \[29\].
Model summary statistics will be reported using Type III sums of squares and Kenward-Roger degrees of freedom \[30\]. All analyses will be performed in R (Version 4.3.0) \[31\], and statistical significance will be defined as p \< 0.05.
Additional analyses may be conducted.
