Page 158 - GHES-3-1

Global Health Economics and
Sustainability
Affect heuristics in substance use

Given the importance of social interactions in adolescent The study constructed indices to analyze adolescents’
decision-making regarding the consumption of alcohol, standard of living and degree of exposure to various risks
cigarettes, and illicit drugs, this study aims to analyze the based on multiple correspondence analysis (MCA) , the
2
relationship between adolescent consumption and the use standard of living index (SLI) , and the risk behavior index
2
of these substances by their peers from the perspective of (RBI) . A separate MCA was applied for constructing each
2
the affect heuristic. The multivariate probit model was used index. Table 1 presents the variables included in the MCA.
to achieve this objective because it allows for analysis of The indices were obtained from the lines of the first
the determinants of joint decisions to use substances. This dimension of each MCA for the results of the composition
study contributes to the literature by considering a possible of each of the constructed indices. Table A1 displays these
common affective impulse in substance consumption. results for the SLI and Table A2 presents the outcomes
It thus advances the literature by modeling joint use of obtained for the RBI. The indices are described by the
substances (i.e., alcohol, illicit drugs, and cigarettes) by following equations:
considering common unobservable factors in adolescents’
decision to consume these substances.   SL row − min (SL row  ) 
SLI = (1)
)
−
2. Materials and methods [max (SL row ) min (SL row 
2.1. Data   RB row − min (RB row  ) 
Microdata obtained from the NSHS and made available by RBI = [max ( RB ) min (RB row  (2)
1
)
−
the Brazilian Institute of Geography and Statistics (IBGE) row
were used to conduct this study. The NSHS contains a Both indices assumed values between 0 and 1; values
questionnaire administered to students from public and closer to 1 on the SLI indicated a worse standard of living,
private schools and aims to collect information on risks and values closer to 1 on the RBI indicated greater exposure
and protective factors that can impact adolescent health. to risk.
The NSHS encompasses elementary- and high-school Variables were defined to indicate the plans of
students enrolled in the 2019 school year. This study used adolescents after finishing school and were constituted
a sample of high school students who attended school as binary variables: The value 1 was assumed if the
regularly. The NSHS questionnaire contains information adolescent intended to continue studying and working
such as sociodemographic and economic circumstances, after high school and 0 if not, the value 1 was assigned if an
family contexts, health-related behaviors, and other risk adolescent intended only to study and 0 if not; the value 1
factors for adolescents. This study focuses on the relationship was allocated if the adolescent respondent intended only to
between the consumption of alcohol, cigarettes, and illicit work and 0 if not; the value 1 was ascribed if the adolescent
drugs among adolescents and their peers; therefore, the mentioned plans other than the previously mentioned
following variables were selected for the analysis. facets and 0 if not. These binary variables were compared
Three binary variables were defined as dependent against adolescent respondents who did not declare any
variables, which assumed a value of 1 if the adolescent had defined plans for their futures after completing high school.
consumed the respective substance (alcohol, illicit drug, or The mother’s education level was also included: A value
cigarette), and a value of 0 if otherwise. of 1 was ascribed for no education, 2 for incomplete

In this study, peers are the people reported by primary education, 3 for complete primary education,
the respondents consist of those who are part of the 4 for incomplete high school, 5 for complete secondary
adolescents’ coexistence friends, and parents. Thus, those education, 6 for incomplete higher education, and 7 for
representing the behavior of peers were considered as complete higher education.
explanatory variables of greatest interest and defined as Other individual characteristics such as the gender,
binary variables. A value of 1 was assigned to affirmative race, age, school type, and residential locations of the
answers for each of the following scenarios: A friend adolescents were also controlled: for instance, the value
consumes alcoholic beverages, a friend consumes any of 1 was ascribed for males, whites, adolescents attending
illicit drug, a parent smokes, and a parent consumes private school, urban residents, and respondents aged
alcoholic beverages. 11 – 19. Table 2 elucidates how the variables used in the

1. The NSHS microdata are public and were obtained directly from study were defined.
the official IBGE website at the following link: https://www.ibge.gov.
br/estatisticas/downloads-estatisticas.html?caminho=pense/2019/ 2 For a detailed explanation of multiple correspondence analysis see
microdados/ Hair et. al. (2009)

Volume 3 Issue 1 (2025) 150 https://doi.org/10.36922/ghes.3829