Continued Deficits of the Perceived Harms of Cigarette Smoking Among Young Adults: Insights from Participants on Amazon Mechanical Turk

Young adults are an age group that has been the target of multiple education campaigns regarding the harms of cigarette smoke exposure. Although smoking rates of conventional cigarettes have declined in the US, continued usage is noted among a select subset of the population aged 18 –30 years. This group either serves as current or soon-to-be future caretakers of young children who will likely spend a significant amount of time exposed to their cigarette byproducts. Understanding their current knowledge base of the dangers of secondhand smoke and the means to address these risks are of major importance for the welfare of young children. We had utilized a brief and limited sample size survey to explore the utility of using an online task marketplace on Amazon.com® to assess secondhand smoking beliefs and financial incentives among young adults. The pilot study had a particular focus on the perception and means to minimize smoke exposure among children residing with young adults serving as their primary caretaker or fellow home residents.

Introduction

A successful, multipronged public health marketing campaign may be attributed to the decline in the cigarette smoking in the US. Approximately 18% of adults had reported to be currently smoking cigarettes in 2013, as compared to 25% in the 1990 [1,2]. However, there remains a recalcitrant population of young adults who continue to smoke despite being raised in a society with media that had inundated them with messages regarding the harms of cigarette products. Approximately 19% young adults (between the ages of 18-24) reported to be current smokers in 2013, of which a 6-7% greater rate of smoking was noted among males [1]. When viewed in the context of potential secondhand smoke exposures (SHSe) to children, high rates of smoking among young adults are alarming. SHSe has been attributed to a myriad of clinical morbidities in children [3,4]. Preschool-aged children are in particular at a greater risk of the harms of SHSe secondary to the greater length of time they reside in indoor settings [5]. Since the mean age for first time mothers and fathers is approximately 26-27 years of age, it is of critical importance to better understand the effectiveness of educational messaging of the harms of SHSe to children that target young adults [6].

Beyond the importance of acquiring SHSe knowledge is also the practice of enforcing the reduction of smoke exposures in the home setting. Multiple strategies of varying complexities are available to reduce home smoke exposures among pre-school aged children, but one that is frequently studied and recommended by various public health organization is the use of a home smoking ban [7-9]. Though the number of smoke free home rules have increased over the past decade from 9.6% to 46.1% with a household with one smoker, it is quite evident that the SHSe educational marketing efforts are still inadequate [10]. This is particularly impactful among families with a lower socioeconomic status with higher rates of smoking and subsequent respiratory morbidities [11,12].

We have utilized the online task marketplace of Amazon® Mechanical Turk (MTurk) to determine its potential utility in capturing knowledge of basic pediatric SHSe concepts among young adult smokers and the means by which home smoking exposures are reduced in those identifying themselves as active smokers. We will also briefly explore the use of incentives as a means to encourage a reduction in home smoke exposures as select studies have shown utility in this modality and current healthcare policy initiatives have been directed to this arena [13]. This low-cost, pilot study is using MTurk given previous investigations that have largely sampled young adults using this modality [14-16].

Methods

Participants were recruited by crowdsourcing via the MTurk website that allows the cost-effective, yet reliable commission of participants for payable tasks. MTurk allows anonymous participants or “workers” to complete short tasks (HITs). The authors of the study served as “requesters” who post the HITs, with details on the study, to an Amazon.com-based MTurk website. The participants are paid upon satisfactory completion of all contents of the study’s survey.

All recruited participants were between the ages of 18-30 and whose servers indicated residence within the US. They had successfully completed over 95% of their previous assigned HITs. MTurk participants who agreed to the contents of the study description and instructions were directed to a 39-item online survey administered using Qualtrics online software (Qualtrics Labs Inc., Provo, UT). Each was paid $0.30 upon confirmation of completion. Despite the low reimbursement, review of recent comparative studies have not shown a significant difference in reliable responses among MTurk participants and sampling via non-online based modalities [17]. Though participants are anonymized on the MTurk website, pertinent details regarding the subjects were captured in the Qualtrics-based survey. Pre-validated survey questions exploring demographics, baseline smoking activity and SHSe beliefs were derived from the American Academy of Pediatric’s (AAP) Survey, Questionnaires and Assessment Tools [18]. Preschool-aged children were defined as being less than 5 years of age. Current smoking status was defined as the use of at least 1 cigarette in the past 30 days.

Financial incentive characteristics queries included the type and amount of incentives to achieve either a 6-month total abstinence from smoking or cessation within the home setting (e.g., home smoking ban). Incentive strategies were subdivided based on participant identification as caretakers of preschool-aged children. Given the small sample size of active smokers in our pilot study, we have used descriptive statistics to summarize our findings.

Results

The sampled population was overwhelmingly single white males with a college-level education

The sampled population (n= 210) consisted of predominantly male participants (n = 149) (Table 1). Slightly over 2/3rd of participants were White, but less than 10% identified themselves as being Black or Hispanic. Asians had the largest representation among all non-White ethnicities. 70% of participants were non-married and 27% had reported being in a domestic relationship – the latter characterized as being either married or living with a partner. Approximately ½ of all participants had a college degree, while 43% had a level of education training falling short of this status. The majority had an annual income less than $70000. About 30% had an income less than $30000, which may reflect their current student status or being under the continued dependency of their guardians.

Participants questioned the chronologic and minimal dose-response effects of SHSe

Among the recruited participants, 32% (n = 67) stated that they were current smokers. Within this subset, approximately 20% stated that they served as the primary caretakers of pre-school aged children. Queries of pediatric-specific SHSe beliefs among all participants noted that 37% were either unsure or did not believe that cigarette smoke exposure had lasting effects on a child’s health. Approximately 1/3 of participants felt fully certain that smoking cigarettes in a room in which a child resides has a lingering impact over many days on pediatric wellbeing. Participants felt either unsure or fully confident of the existence of safe SHSe levels in 15% and 12% of responses, respectively. Despite the varied responses of the chronologic and dose exposure effects of secondhand smoke, over 95% believed that cigarette smoke exposure is linked with respiratory issues in children.

Smokers identifying themselves as the primary caretakers of preschool-aged children have a higher motivation and involvement in the implementation of a home smoking ban

Among the limited individuals who reported to be current smokers and serving as caretakers of pre-school aged children (n=12), greater than 80% reported either a partial or no limitations in smoking within the home (Table 2). This is similar to the 55 participants who identified themselves as being current smokers and not serving in a childcare role, in whom greater than 90% reported limited to no home smoking restrictions. Primary caretakers of children had a greater participatory role as decision maker – most often with a spouse or partner. Regardless of the type of reported home smoking ban, smoke exposures continued to be reported in the homes of all current smokers. Greater levels of exposure were noted in those who did not have a childcare role. The motivation to cease individuals from smoking within the home was most prominent in those serving as caretakers of children.

Current smokers not serving in a childcare role are less likely to enforce a home smoking ban despite motivation with monthly cash incentives

Among current smokers caring for preschool-aged children, approximately 40% would require as little as $25/month to enforce a 6-month home smoking ban (Table 3). Incremental increases of monetary incentives yielded a less motivated population; yet when offered $100/month, only two individuals were either unsure or unwilling to alter their practices. Those smokers not caring for preschool children required higher level of incentives to be motivated to enforce an equivalent home smoking ban. A larger percentage of individuals appeared to be recalcitrant to lower level of incentives. Increasing incentives resulted in approximately 40% unwilling to accept $100/month; ultimately, 10% would not consider $1000/month sufficient to prohibit smoking within the home for 6-months.

Discussion

The study reflects both the utility and limitations of using an online database of participants. The objective of this work was to serve as a pilot study into the greater use of using MTurk, or related online crowdsourcing platforms, to capture young adult perceptions of the harms and techniques to remediate SHSe in the homes of young children. Young adulthood represents an important stage in life in which long-term cigarette smoking behaviors are often established [19]. The majority of our participants were white males with a college-level education. Though women serve largely in the role of primary caretakers of children in society, they only represented approximately 30% of our patient population. Black and Hispanic minorities were also underrepresented, when compared to 2013 US Census data [20]. Among the 32% of participants identifying themselves as current smokers, about 1/5th served in a childcare role. The number of young adult smokers captured in our study is consistent with previous US-based population studies [19]. Moreover, despite the extensive public health marketing campaigns educating the public of the harms of SHSe on children, a select segment of the young adult population continues to be unaware of key pediatric SHSe concepts. As to be expected, current smokers caring for children were more likely to describe involvement in developing the characteristics of a home smoking ban. Despite continued SHSe of varying levels in the home, caretakers currently smoking were far more likely to be motivated to enforce a home smoking ban. This was subsequently reflected in the low level of incentives that they reported was needed to enforce a complete home smoking ban. Without the motivation of caring for a child, young adult smokers were both less involved in the decision making and had diminished motivation to change their smoking behaviors.

MTurk has been utilized by various medical, marketing and political researchers [21-23]. To the best of the authors knowledge, our work is the first to utilize this online platform to explore concepts and strategies that relate to pediatric SHSe. We had specifically targeted young adults given their soon-to-be evolving role into parenthood. This study has the potential to be valuable for not only US-based research, but also for global tobacco investigators where cigarette usage is increasing among young adults [24,25]. Given this concerning trend, along with low quit rates in both genders appreciated in various global settings, our sampling strategy may be a cost-effective approach to understand and trial tobacco-related messaging among young adults [24]. MTurk usage for global researchers may be further supported with participants representing over 66 countries, with the majority of non-USA based users coming from India [14].

Beyond the basic SHSe conceptual issues we had explored, we had also attempted to use the online forum to understand the role of incentives to increase motivation for enforcing a home smoking ban. Those who did not serve in a childcare role were less motivated in general to enact a home smoking ban and required far larger amounts of financial incentives to consider instituting such a policy. Given that young adults in the US are often burdened with education debt and/or not earn higher levels of annual income, financial incentives may be a reasonable strategy to alter behavior patterns in those resistant to change [26]. Moreover, the differences in earned annual incomes become even more prominent in Black and Hispanic populations, who net significantly less than their White counterparts [27]. Aggressive marketing campaigns by cigarette companies have specifically targeted these lower income groups (including at a global level); however, financial incentives may provide a channel to persuade individuals to consider discontinuing smoking and thus “buying time” for them to consider adopting smoking cessation practices [28-30]. Though no incentive program has been developed to alter adult smoking behaviors for pediatric endpoints, this work may be informative for future research that is looking to use financial incentives beyond the typical focus of yielding abstinence among primary smokers [13,31].

This study had several notable limitations which are noteworthy for not only this pilot study but also for expansion of its objectives using an online platform. Our study population was not reflective of the US general population, nor of expected US current smokers. There was an underrepresentation of non-White ethnic groups, with the exception of Asians. Less than 10% of participants identified themselves as Black and Hispanic, which is notable given that they represent 13.2% and 17.1% of the US population, respectively [20,21]. Moreover, young adult US-born Hispanics and Blacks have higher levels of social and/or occasional smoking as compared to Whites [21]. Both ethnic groups are also likely to have their first child earlier with the average age at first birth for Blacks and Hispanics being 22.7 years and 23.1 years, as compared to 26.0 years for Whites [32]. Therefore, our recruited participants may have under-represented at-risk children with SHSe. We also did not explore in-depth the indoor smoking behaviors of the primary caregivers in the home setting, independent of visitors in the home who may be smoking. This limitation may have underestimated the significance of the primary caregiver’s smoking behaviors that may be exposing the preschool-aged child to the highest SHSe burden. Future work may as well benefit from exploring the SHSe beliefs among the population of individuals who do not accept the linkage between SHSe levels and the severity of disease symptoms. These individuals may have historical or ongoing observations of smoke exposures that may have not readily resulted in clinical manifestations of symptoms, which thus may be enlightening for future educational efforts. Our population was further limited by the disproportionate number of male participants, despite women usually serving in the role as the primary caretaker of children [33]. Our sampling approach did not reflect demographics of select social media sites whose users are comprised of more women and non-White minorities than evident in our sampled population [34]. We also used a more liberal definition of current smoker, as defined by the use of any cigarettes within 30 days. Though different than select national sampling strategies defining current smoking as both ≥ 100 cigarettes during one’s lifetime and smoking “every day or some days”, we believe we had captured more occasional and social young adult smokers that may be contributing to SHSe. We were further limited by exploring the responses of conventional cigarette users only and future work will incorporate the input of consumers of electronic cigarettes, given its increasing prevalence of usage and recognition of the harms of its gaseous byproducts [35-37]. Our comparisons for home smoking ban features could have been enhanced by capturing the characteristics of tobacco restrictions among non-smoking caregivers. We had chosen to not capture the home smoking restrictions of this particular group since caregivers are often the primary source for home SHSe or in the optimal position to minimize clinically-impactful smoke exposures on their children [38]. Lastly, our method of exploring the motivation to change behavior patterns using escalating incentives is limited by a survey-based methodology and may be enhanced using contingent valuation scenarios among caretakers whose children are clinically impacted by home exposures. The contingent valuation method may be able to better determine how much a caretaker would be willing to pay to reduce SHSe due to its associated morbidities.

Therefore, using an online platform like Mturk may serve beneficial to rapidly and cost-effectively recruit participants for pediatric-centric outcomes. We were able to capture a proportion of young adult smokers in whom select characteristics mirror national smoking rates. This is insightful given that young adulthood represents a time when long-term cigarette smoking behaviors, including perceptions of pediatric smoke exposures, becomes indoctrinated in this population of likely first-time parents [6,39]. Among those identifying themselves as current smokers, it appeared that our participants who served in a childcare role were motivated enough to immediately adopt or be receptive to incentives to yield a home smoking ban. The major limitation to our work appears to be the ability to recruit an adequate young adult population that accurately reflects all ethnic, gender, educational and socioeconomic differences seen among the heterogeneous smoking young adult population. Future studies will explore the role of other online platforms (e.g., social media) in targeting a more ethnically diverse, lower income population with higher levels of home SHSe. This demographic would be more relevant for researchers focused on addressing the growing global cigarette consumption among young adults. Furthermore, translating the incentive-compatible paradigm to real-world trials may involve the provision of monetary incentives and observation of participants’ behavior changes. Ultimately, this work is pursuing a new avenue of exploring perceptions and motivations to alter SHSe among the youngest and most vulnerable of children.

1. Centers for Disease Control and Prevention (2013) Trends in Current Cigarette Smoking Among High School Students and Adults, United States, 1965–2011 .
2. Centers for Disease Control and Prevention (2014) Current Cigarette Smoking Among Adults—United States, 2005–2013. Morbidity and Mortality Weekly Report 63: 1108–1112 .
3. Oberg M, Jaakkola MS, Woodward A, Peruga A, Prüss-Ustün A (2011) Worldwide burden of disease from exposure to second-hand smoke: a retrospective analysis of data from 192 countries. Lancet 377: 139-146.
4. Office on Smoking and Health (US) (2006) The Health Consequences of Involuntary Exposure to Tobacco Smoke: A Report of the Surgeon General, Atlanta (GA): Centers for Disease Control and Prevention (US) .
5. Breysse PN, Diette GB, Matsui EC, Butz AM, Hansel NN, et al. (2010) Indoor air pollution and asthma in children. Proc Am Thorac Soc 7: 102-106 .
6. Centers for Disease Control and Prevention (2015) Key Statistics from the National Survey of Family Growth .
7. Nicholson JS, McDermott MJ, Huang Q, Zhang H, Tyc VL (2014) Full and Home Smoking Ban Adoption After a Randomized Controlled Trial Targeting Secondhand Smoke Exposure Reduction, Nicotine Tob Res 17: 612-616.
8. Rees VW, Keske RR, Blaine K, Aronstein D, Gandelman E, et al. (2014) Factors influencing adoption of and adherence to indoor smoking bans among health disparity communities. Am J Public Health 104: 1928-1934 .
9. Eakin MN, Rand CS, Borrelli B, Bilderback A, Hovell M, et al. (2014) Effectiveness of motivational interviewing to reduce head start children's secondhand smoke exposure. a randomized clinical trial. Am J Respir Crit Care Med 189: 1530-1537 .
10. King BA, Patel R, Babb SD (2014) Prevalence of smoke free home rules--United States, 1992-1993 and 2010-2011. MMWR Morb Mortal Wkly Rep 63: 765-769 .
11. Spanier AJ, Beck AF, Huang B, McGrady ME, Drotar DD, et al. (2015) Family hardships and serum cotinine in children with asthma, Pediatrics 135: e416-423 .
12. Butz AM, Halterman JS, Bellin M, Tsoukleris M, Donithan M, et al. (2011) Factors associated with second-hand smoke exposure in young inner-city children with asthma. J Asthma 48: 449-457 .
13. Hamilton FL, Greaves F, Majeed A, Millett C (2013) Effectiveness of providing financial incentives to healthcare professionals for smoking cessation activities: systematic review. Tob Control 22: 3-8 .
14. Ipeirotis P (2010) Demographics of Mechanical Turk. CeDER Working Papers .
15. Ross J, Irani L, Silberman MS, Zaldivar A, Tomlinson B (2010) Who are the crowdworkers?: shifting demographics in mechanical turk. In Proceedings of the 28th of the international Conference Extended Abstracts on Human Factors in Computing Systems. ACM New York NY 2863‐2872 .
16. Boynton MH, Richman LS (2014) An online daily diary study of alcohol use using Amazon's Mechanical Turk. Drug Alcohol Rev 33: 456-461 .
17. Goodman JK, Cryder CE, Cheema A (2013) Data Collection in a Flat World: The Strengths and Weaknesses of Mechanical Turk Samples. J Behav Decis Making 26: 213–224 .
18. American Academy of Pediatrics (2013) Julius B Richmond Center of Excellence. Surveys, Questionnaires, and Assessment Tools .
19. Lariscy JT, Hummer RA, Rath JM, Villanti AC, Hayward MD, et al. (2013) Race/Ethnicity, nativity, and tobacco use among US young adults: results from a nationally representative survey. Nicotine Tob Res 15: 1417-1426 .
20. U.S. Census Bureau (2015) State and County Quick Facts .
21. Rand DG (2012) The promise of Mechanical Turk: how online labor markets can help theorists run behavioral experiments. J Theor Biol 299: 172-179 .
22. Berinsky, AJ, Huber, GA, Lenz, GS (2012) Evaluating online labor markets for experimental research: Amazon.com’s Mechanical Turk. Political Analysis 20: 351–368 .
23. Chandler J, Mueller P, Paolacci G (2014) Nonnaïveté among Amazon Mechanical Turk workers: consequences and solutions for behavioral researchers. Behav Res Methods 46: 112-130 .
24. Giovino GA, Mirza SA, Samet JM, Gupta PC, Jarvis MJ, et al. (2012) Tobacco use in 3 billion individuals from 16 countries: an analysis of nationally representative cross-sectional household surveys. Lancet 380: 668-679 .
25. Ng M, Freeman MK, Fleming TD, Robinson M, Dwyer-Lindgren L, et al. (2014) Smoking prevalence and cigarette consumption in 187 countries, 1980-2012. JAMA 311: 183-192 .
26. Fry R (2014) Young Adults, Student Debt and Economic Well-being.” Washington, D.C.: Pew Research Center’s Social and Demographic Trends project .
27. National Center for Education Statistics (2014) Annual Earnings of Young Adults .
28. Ayo-Yusuf OA, Olutola BG, Agaku IT (2014) Permissiveness toward tobacco sponsorship undermines tobacco control support in Africa. Health Promot Int 31: 414-422 .
29. Agaku IT, Omaduvie UT, Filippidis FT, Vardavas CI (2014) Cigarette design and marketing features are associated with increased smoking susceptibility and perception of reduced harm among smokers in 27 EU countries. Tob Control.
30. Hafez N, Ling PM (2005) How Philip Morris built Marlboro into a global brand for young adults: implications for international tobacco control. Tob Control 14: 262-271 .
31. Sigmon SC, Patrick ME (2012) The use of financial incentives in promoting smoking cessation. Prev Med 55: S24-32 .
32. Martinez G, Daniels K, Chandra A (2012) Fertility of men and women aged 15-44 years in the United States: National Survey of Family Growth, 2006-2010. Natl Health Stat Report 51: 1-28 .
33. Alecxih LM, Zeruld S, Olearczyk B Characteristics of Caregivers Based on the Survey of Income and Program Participation. NFCSP: Selected Issue Briefs .
34. Duggan M, Ellison NB, Lampe C, Lenhart A, Madden M (2015) Social Media Update 2014. Pew Research Center.
35. Zhu SH, Gamst A, Lee M, Cummins S, Yin L, et al. (2013) The use and perception of electronic cigarettes and snus among the U.S. population. PLoS One 8: e79332 .
36. McAuley TR, Hopke PK, Zhao J, Babaian S (2012) Comparison of the effects of e-cigarette vapor and cigarette smoke on indoor air quality. Inhal Toxicol 24: 850-857 .
37. Cheng T (2014) Chemical evaluation of electronic cigarettes. Tob Control 23: 2: ii11-7 .
38. Rosen LJ, Noach MB, Winickoff JP, Hovell MF (2012) Parental smoking cessation to protect young children: a systematic review and meta-analysis. Pediatrics 129: 141-152 .
39. Hammond D (2005) Smoking behaviour among young adults: beyond youth prevention. Tob Control 14: 181-185 .