Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Stephen T. Higgins


The aim of this study was to examine associations between biases in decision-making (delay discounting [DD], opportunity cost neglect [OCN], status quo bias [SQB]), educational attainment, and use of cigarettes and other tobacco and nicotine delivery products among women of reproductive age. Women of reproductive age are of special interest because of the additional risks that cigarette smoking or use of these other products represents should they become pregnant. Data were collected anonymously online in survey format using Amazon Mechanical Turk [AMT]. Participants were 800 women of reproductive age (24-44 years) from across the US. Half (n = 400) were smokers who reported current, daily smoking and half (n = 400) were never smokers who reported smoking less than 100 cigarettes in their lifetime. Participants reported smoking characteristics, plans to quit smoking, use of nicotine replacement therapies, use of other tobacco and nicotine delivery products, alcohol and drug use histories, and the Fagerström Test for Nicotine Dependence. Participants completed two measures for each of the three biases in decision-making, the Barratt Impulsiveness Scale [BIS-11], and two scales measuring short- and long-term propensity to plan for money expenditures [PPMS and PPML]. Educational attainment analyses compared three education groups: high school or less vs. some college (e.g. some college/A.A.) vs. B.A. or higher. DD was steeper among current vs. never smokers and for women with lower vs. higher levels of education, with no significant interaction between smoking and education. Modifying the instructions of the DD measure to make the zero option explicit reduced DD similarly across levels of smoking status and education. OCN was worse at lower vs. higher educational attainment on one OCN measure, with no significant effect of smoking status or interaction between opportunity cost neglect and educational attainment on either measure. No evidence was found for stronger SQB by smoking status or education. Smoking status was related to BIS Total, BIS Motor and Nonplanning subscales and to PPML in initial models but remained significant after adjusting for baseline differences in participant characteristics only for BIS Motor subscale and educational attainment was related only to BIS Nonplanning subscale. Preliminary comparisons of e-cigarette users to non-users suggest smokers using e-cigarettes only differ from smokers not using e-cigarettes on measures related to quitting smoking whereas within never smokers e-cigarette users demonstrated a pattern of riskier decision-making compared to non-users. Results confirm that DD and education are important to understanding the use of tobacco and nicotine products in women of reproductive age, and suggest that smoking and educational attainment are independently related to discounting rates. The observed explicit-zero framing effect suggests making alternatives more explicit when presenting choices may help reduce DD and lead to better decision-making, which has possible treatment implications. Results identify OCN as an additional decision-making bias to consider in understanding how low educational attainment might relate to smoking vulnerabilities. The preliminary examination of e-cigarette use suggests for women of reproductive age above age 24 years, e-cigarette use among current smokers may reflect desire or attempts to quit or cut back on smoking whereas e-cigarette use among non-smokers may be a marker of a more impulsive, riskier repertoire, although additional study of this question is needed.



Number of Pages

100 p.

Available for download on Friday, April 13, 2018