Abstract
Previous research suggests truancy is related with increased risks for substance use and combustible nicotine and marijuana use. However, limited research has investigated the link between truancy and vaping nicotine and marijuana. Further, much past research on the relationship between truancy and nicotine and marijuana use employed data collected before the onset of the COVID-19 pandemic and recent shifts in patterns of adolescent truancy. This study utilized data from the 2024 Florida Substance Abuse Survey (FYSAS) to examine the link between truancy and nicotine and marijuana vaping. Findings indicate that as the number of school days skipped increases, the likelihood of vaping and combustible use of nicotine and marijuana in the past 30 days increases. Policy implications are discussed.
Introduction
Previous research documents links between truancy and substance use in adolescence and adulthood (Farrington, 1996; Flaherty et al., 2012; Henry & Huizinga, 2007; Henry & Thornberry, 2010; Rocque et al., 2017; Vaughn et al., 2013). Overall, this research indicates that students who engage in truancy (i.e., skipping school without a legitimate excuse) are more likely to use alcohol, tobacco, marijuana, and other illicit substances compared to students who attend school regularly (Henry & Huizinga, 2007; Henry & Thornberry, 2010; Maynard et al., 2017; Rocque et al., 2017; Vaughn et al., 2013). One potential explanation for this connection, may be that students who skip school have more opportunities to engage in unstructured and unsupervised socializing which increases the likelihood of engaging in deviant behavior and substance use (Augustyn & McGloin, 2013; Leimberg & Lehmann, 2022; Osgood et al., 1996).
However, the majority of this research was conducted before recent shifts in truancy patterns among American youth. Prior to the COVID-19 pandemic, studies estimated that approximately 11% of adolescents (between the ages of 12 and 17) engaged in truancy in the United States (Maynard et al., 2017; Vaughn et al., 2013). Recent data, however, suggests that there have been substantial upward shifts in the prevalence of truancy in the 2020 s (Vaughn et al., 2026). To illustrate, a study by Dee (2024) estimates that more than 28% of public school students missed 10% or more of the 2021–2022 academic school year. In addition, 20 states reported that 30% or more of their students missed three or more weeks of school in the 2022–2023 academic school year (Unites State Department of Education, 2025). For instance, the reported rate of chronic absenteeism (missing 10% or more of the academic year) in the state of Florida was 34% in the 2022–2023 academic school year (Unites State Department of Education, 2025).
A recent study by Vaughn et al. (2026), using data from a nationally representative sample of US adolescents aged 12–17, estimated that 23.4% of adolescents engaged in truancy in 2022 and 25.7% of adolescents engaged in truancy in 2023. Increased levels of truancy after the COVID-19 pandemic may be the result of an array of individual, familial, and contextual risk factors. For example, Lester and Michelson (2024, p. 1) hypothesized that the increase in school absences might be a result of increased “emotionally based school avoidance (EBSA).” EBSA is characterized by lower school attendance or avoidance of school and is proposed to manifest as the result of multiple risk factors and reinforced through stress and anxiety that is associated with school attendance (negative reinforcers) and engagement with others or valued activities outside of school (positive reinforcers) (Lester & Michelson, 2024). EBSA is subsequently related to negative educational outcomes, such as higher levels of truancy. The pandemic may have exacerbated both the risk factors associated with EBSA and the reinforcers to contribute to higher truancy post-pandemic. Regardless of its cause, recent increases in truancy rates raise substantial concerns about potential downstream consequences of skipping school – including potential increases in unstructured socializing, delinquency, and substance use (Farrington, 1980; Mazerolle et al., 2019; Rocque et al., 2017; Vaughn et al., 2013).
A wealth of previous research links truancy and with substance use (Gakh et al., 2020; Maynard et al., 2017; Rocque et al., 2017; Vaughn et al., 2013). For instance, a study by Maynard et al. (2017) revealed that truancy is associated with tobacco use, alcohol use, binge drinking, marijuana use, and the use of other illicit substances. Studies by Henry et al. (2009) and Henry and Huizinga (2007) also revealed that truancy is a significant predictor of initiating marijuana, tobacco, and alcohol use in adolescents. Moreover, a study by Henry and Thornberry (2010) indicates that once students begin substance use, truancy is associated with escalating patterns substance use. Similarly, a longitudinal study by Rocque et al. (2017) found that truancy in adolescence is associated with a greater likelihood problematic drinking at age 18 and at age 32. Taken together, the existing research suggests that truancy is associated increased risks of several forms of substance use, increased risks of initiating substance use in adolescence, and escalations in frequency of substance use among those adolescents who are already using illicit substances. Much of the existing research on the relationship between truancy and substance use suggests the relationship may be explained by increased unsupervised and unstructured time to socialize with other adolescents, and potentially reduced social control through reduced engagement in prosocial institutions such as school (Henry et al., 2009; Henry & Huizinga, 2007; Henry & Thornberry, 2010).
However, much of the research linking truancy with substance use has examined traditional forms of nicotine and marijuana use, with very limited research dedicated to exploring how truancy may be linked with vaping nicotine and marijuana. Vaping quickly became a popular mechanism for consuming nicotine and marijuana after e-cigarettes (vaporizers) were made available in the early 2000s. And in 2014, vaping became the most popular mode of nicotine use among adolescents (Centers for Disease Control and Prevention (CDC), 2022). Recent estimates from the 2024 National Youth Tobacco Survey suggest that 3.5% of middle school students and 7.8% of high school students report vaping nicotine in the past 30 days. Meanwhile, only 1.1% of middle school students and 1.7% of high school students report smoking cigarettes in the past 30 days (Centers for Disease Control and Prevention (CDC), 2024). Research examining risk factors for adolescent vaping and overall substance use initiation is of significant importance as previous literature links early initiation of substance use with greater risks of substance dependence in adulthood (Guo et al., 2010; Nelson et al., 2015; Van de Ven et al., 2010).
Previous research suggests that many adolescents who would not use nicotine or marijuana through traditional mechanisms, due to perceptions of risk, chose to vape due to perceptions of lower health-related risks with consuming substances through a vaporizer (Choi & Forster, 2014; Kaisar et al., 2016). As a result of perceived differential risk profiles, engaging in vaping nicotine and marijuana may be associated with different risk factors than consuming these substances through traditional mechanisms. To illustrate, several research studies have revealed differential relationships between smoking nicotine and marijuana and vaping these substances for delinquency (Jackson et al., 2019), health-related behaviors (Jackson et al., 2020), and the use of other illicit substances (Boccio & Jackson, 2021; Kristjansson et al., 2015). For instance, an article by Jackson et al. (2019) revealed that adolescents who vape marijuana are at greater risk of delinquent involvement compared to youths who ingest nicotine and marijuana through combustible means. Similarly, a study by Leal (2024) revealed that vaping marijuana is associated with a greater likelihood of criminal involvement than using marijuana through traditional means. Moreover, a study by Leal et al. (2026) revealed that vaping marijuana is more consistently associated with adverse mental health outcomes when compared to smoking marijuana and several other forms of marijuana use. One potential explanation for the consistent associations between marijuana vaping and adverse outcomes may be explained by higher THC concentrations found in marijuana vaping products compared to more traditional forms of use. Current estimates suggest the THC concentration in vaping products can be as high as 90% compared to often 20% or less in combustible marijuana (Northwestern Student Affairs, 2026; Washington State Liquor and Cannabis Board, 2026). Of note, many studies examining associations with modality of marijuana and nicotine use reveal the most unfavorable outcomes for respondents who report use of nicotine and marijuana through both combustible means and vaping (Boccio & Jackson, 2021; Kristjansson et al., 2015). Given evidence of differential relationships tied to mode of nicotine and marijuana use, it is possible that truancy may also hold differential relationships with different modes of nicotine and marijuana consumption.
The few studies that have explored the connection between truancy and vaping suggest that vaping nicotine is associated with school-risk behaviors including truancy and low grade point average in high school seniors (McCabe et al., 2017). In addition, there is also limited research employing data from the 2018 Monitoring the Future cohort linking marijuana vaping with truancy (Kritikos et al., 2021). As such, this study will extend the prior literature by exploring whether truancy holds similar relationships with vaping nicotine and marijuana along with traditional forms of use in a recent (2024) sample of adolescents. This study will expand the prior literature in two ways. First, this study will explicitly study the link between truancy and nicotine and marijuana vaping in a large sample of middle school and high school students. Second, this study will examine the relationship between truancy and nicotine and marijuana use in a contemporary sample (2024) since recent shifts in trends of truancy and chronic absenteeism. The findings of this study will shed light on connections between truancy and vaping activities and connections between truancy and traditional forms of substance use in a time period when rates of truancy have been rapidly shifting.
Methods
Data
This study uses data from the 2024 cohort of the Florida Youth Substance Abuse Survey (FYSAS). The FYSAS is an annual survey of students enrolled in public middle schools and high schools in the state of Florida. The primary purpose of the survey is to assess prevalence and trends in substance use in adolescents in Florida. Each year the survey is sponsored and administered by a partnership between the Florida Departments of Children and Families, Education, and Health. While the survey is primarily focused on adolescent substance use, each year there are a variety of questions concerning daily activities, school engagement, delinquency, and peer relationships.
Each year a representative sample of students are selected to participate in the survey through a two-stage cluster sampling strategy. First, groups of public middle schools and high schools (excluding adult education, correctional, and special education schools) are selected. Probability of school selection is tied to enrollment size, wherein, larger schools have a higher probability of selection. Then, classrooms are selected to participate within the selected schools. In 2024, 405 middle schools and 356 high schools were selected to participate. Three hundred and six of the high schools and 346 of the middle schools elected to participate with greater than 48,000 students completing the survey. A series of validation checks were carried out where surveys were removed where participants appeared to exaggerate substance use and deviant behavior, reported using a fictitious drug, recorded an inconsistent pattern of responses, completed less than a quarter of the survey items, or completed the survey for the wrong grade group leading to a final sample of 44,755 respondents (Florida Youth Substance Abuse Survey, 2024).
For the analyses of this study, respondents who did not have complete data on the truancy, nicotine use, marijuana use, or race measures were removed leading to a final analytic sample of 40,047 respondents. Missing data on the covariates was handled using multiple imputation with chained equations to produce and merge 20 datasets. Descriptive statistics are provided prior to imputation. All models were estimated employing data prior to and after imputation. Results from both sets of models yielded the same general pattern of findings.
Measures
Outcome Measures
Nicotine Vaping
Nicotine vaping was measured using a single item where respondents were asked to indicate “On how many occasions (if any) have you vaped nicotine (e-cigarettes, vape pens, JUUL). . .during the past 30 days?” Response options included “0 occasions” (0), “1–2 occasions” (1), “3–5 occasions” (2), “6–9 occasions” (3), “10–19 occasions” (4), “20–39 occasions” (5), and “40 or more occasions” (6). For the sake of the analyses of this study, this measure was employed in two ways. First, this measure was recoded as a dichotomous indicator of vaping nicotine in the past 30 days where 0 = has not vaped nicotine in the last 30 days and 1 = has vaped nicotine in the last 30 days. Second, this measure was retained in its original coding strategy as a measure of frequency of nicotine vaping in the past 30 days (0–6). Descriptive statistics for this variable and all other variables and scales in this study are displayed in Table 1.
Descriptive Statistics.
Cigarette Use
Cigarette use was measured using a single item where respondents were asked “How frequently have you smoked cigarettes during the past 30 days?” Response options included “not at all” (0), “less than one cigarette per day” (1), “one to five cigarettes per day” (2), “about one-half pack per day” (3), “about one pack per day” (4), “about one and one-half packs per day” (5), and “two packs or more per day” (6). This item was employed as a dichotomous indicator of cigarette use (0 = no; 1 = yes) and as a measure of frequency of cigarette use (0–6).
Dual Nicotine Use
A dual nicotine use measure was constructed using the previous two measures. This item is coded so that respondents who indicated that they both “vaped” nicotine and used cigarettes were coded as “1” and all other respondents were coded as “0.”
Marijuana Vaping
Marijuana vaping in the past 30 days was measured using a single item where respondents were asked “on how many occasions (if any) have you vaped marijuana (e-cigarettes, vape pens, JUUL). . .during the past 30 days?” This item had the same response categories as the nicotine vaping measure and was employed in the same two ways. First, this measure was recoded as a dichotomous indicator of marijuana vaping in the past 30 days (0 = no; 1 = yes). Second, it was employed as a measure of marijuana vaping frequency in the past 30 days (0–6).
Marijuana Use
General marijuana use was measured using a single item where respondents were asked to indicate “on how many occasions (if any) have you used marijuana or hashish. . .during the past 30 days?” This item included the same response categories as the previous measure and was employed as both a dichotomous indicator of marijuana use (0 = no; 1 = yes) and as a frequency measure of marijuana use (0–6).
Dual Marijuana Use
Using the previous two measures a dual marijuana use measure was constructed. Specifically, respondents who indicated that they “vaped” marijuana and “used marijuana or hashish” in the past 30 days were coded as ‘1’. Respondents who did not use marijuana through either mechanism or only reported “vaping” or “using” were coded as ‘0’.
Predictor Measures
Truancy
Truancy was measured in two ways using a single item asking respondents “during the LAST FOUR WEEKS, how many whole days have you missed school because you skipped or ‘cut’?” Response options included “none” (0), “1” (1), “2” (2), “3” (3), “4–5” (4), “6–10” (5), and “11 or more” (6). First, this item was recoded as a dichotomous indicator of truancy in the last four weeks where 0 = did not skip or cut any days and 1 = has skipped or cut days in the last 4 weeks. Then, this variable was also employed in it’s original coding strategy as a measure of number of days of truancy. This measure is coded so that higher numbers reflect skipping or cutting more days of school. These truancy measures are similar to previous truancy measures used with FYSAS data (Boccio et al., 2025)
Controls
All models in the study were estimated controlling for age, gender, racial and ethnic identity, living in a rural area, maternal education level, alcohol consumption, and exposure to substance using peers. Age was measured using a single item where respondents were asked to indicate their age (10–19 or older). Gender was measured using a single item where respondents were asked if they are female (0) or male (1). Racial and ethnic identity was measured using a series of five mutually exclusive dichotomous indicators (non-Hispanic Black, Hispanic, non-Hispanic White, non-Hispanic Other, Multiracial). Hispanic is coded so that 0 = non-Hispanic, Black, Other, or Multiracial, and 1 = Hispanic or Hispanic-White. Non-Hispanic Black was coded so that 0 = Hispanic, White, Other, or Multiracial, and 1 = non-Hispanic Black. Non-Hispanic White is coded so that 0 = Black, Hispanic, Other, or Multiracial, and 1 = non-Hispanic White. Non-Hispanic Other is coded so that 0 = Black, Hispanic, White, or Multiracial, and 1 = non-Hispanic Asian, non-Hispanic Native American or Alaskan Native, non-Hispanic Pacific Islander or Native Hawaiian, or non-Hispanic “Other” racial identity. Multiracial is coded so that 0 = only reported one racial identity (or Hispanic-White) and 1 = reported more than one racial/ethnic identity. For ease of interpretation, non-Hispanic White is used as the reference category for the analyses of this study.
Living in a rural area was measured using a single item where respondents were asked where they lived at the time. Response options included “in a city, town, or suburb,” “on a farm,” and “in the country, not on a farm.” Living in a rural area was coded so that 0 = lived “in a city, town, or suburb,” and 1 = lived “on a farm” or “in the country, not on a farm.” Maternal education was assessed using a measure tapping the respondents’ mothers’ highest level of education. Response options ranged from “completed grade school or less” to “graduate or professional school after college” (0–5). A measure of alcohol consumption was employed as a proxy for other forms of substance use and was measured using a single item where respondents were asked to indicate the number of times they have consumed alcohol in their lifetimes. This item has been recoded as a dichotomous indicator of alcohol use where 0 = no reported alcohol use and 1 = reported alcohol use.
Finally, substance using peers was measured using a scale created from four measures tapping peer use of nicotine and marijuana. Specifically, respondents were asked “Think of your four best friends (the friends you feel closest to). In the past year (12 months) how many of your best friends have:” “smoked cigarettes,” “used marijuana,” “vaped nicotine (e-cigarettes, vape pens, JUUL),” or “vaped marijuana (e-cigarettes, vape pens, JUUL)?” Response options for these items ranged from “none” (0), “1” (1), “2” (2), “3” (3), to “4” (4). Responses to these four items were summed together to create a scale of substance using peers (α = .88) where higher numbers reflect higher exposure to nicotine and marijuana using peers.
Analytic Strategy
The analytic strategy for the study proceeded in a number of steps. First, we examined the descriptive statistics for the nicotine use, marijuana use, and truancy variables in the sample. Second, we employed logistic regression to examine associations between the dichotomous measure of truancy and likelihood of nicotine and marijuana use. Then, we examined relationships between number of days of truancy and the likelihood of nicotine and marijuana use. Third, we utilized negative binomial regression to examine relationships between truancy and frequency of nicotine and marijuana use (as the frequency measures are overdispersed) among respondents who reported using nicotine or marijuana in the past 30 days. Finally, we employed negative binomial regression to examine relationships between number of days of truancy and frequencies of nicotine and marijuana use among respondents who reported nicotine or marijuana use in the past 30 days.
Results
First, we examined descriptive statistics for nicotine use, marijuana use, and truancy. As can be seen in Table 1, 7.15% of the sample reports vaping nicotine, 1.19% report smoking cigarettes, and 0.69% report dual nicotine use in the past 30 days. Meanwhile, 5.30% report vaping marijuana, 5.99% report using marijuana, and 4.27% report dual marijuana use in the past 30 days. Nearly half (45.52%) of the sample reports some level of truancy in the last four weeks with 54.48% of the sample reporting skipping zero days, 10.76% reporting skipping one day, 9.91% reporting skipping two days, 8.89% reporting skipping three days, 9.53% reporting skipping four to five days, 3.95% reporting skipping six to 10 days, and 2.47% of the sample reporting skipping 11 or more school days in the last four weeks.
In the next step of the analysis we employed logistic regression to examine associations between truancy and likelihood of nicotine and marijuana use. Examination of Table 2 reveals that truancy is positively and significantly associated with the likelihood of nicotine vaping (OR = 1.551, p <.01) and cigarette smoking (OR = 1.457, p <.01) in the past 30 days. These finding indicate that respondents who have skipped or cut school in the past four weeks are more likely to use nicotine in the same period. Examination of the lower half of the table reveals that truancy is also significantly and positively associated with the likelihood of marijuana vaping (OR = 1.576, p <.01), marijuana use (OR = 1.784, p<.01, and dual marijuana use (OR = 1.599, p <.01). 1 Similar to the nicotine findings, these results indicate that respondents who report skipping school in the past four week are more likely to use marijuana in the last 30 days.
Logistic Regression Models of the Association between Truancy and Nicotine and Marijuana Use (N = 40,047).
p <.05. **p < .01.
We then employed logistic regression to examine associations between the number of days respondents have been truant in the past 4 weeks and likelihood of nicotine and marijuana use in the past 30 days. As can be seen in Table 3, missing days of school is positively and significantly associated with nicotine vaping and cigarette use. Specifically, the Odds Ratios (OR) appear to indicate that as the number of days missed increases, the likelihood of vaping nicotine also increases. Of note, the ORs appear to increase in a stepwise fashion up to approximately six to 10 days missed for nicotine vaping (1 day OR = 1.259, p <.01; 2 days OR = 1.481, p <.01; 3 days OR = 1.547, p <.01; 4 to 5 days OR = 1.682, p <.01; 6 to 10 days OR = 2.008, p <.01). These findings appear to indicate that as respondents miss more days they are more likely to vape nicotine up to six to 10 days missed. After this point, however, the likelihood of vaping nicotine no longer appears to increase. Model 2 appears to indicate that the number of days missed tends to hold a similar relationship with cigarette smoking, as ORs generally increase as the number of days missed also increases. Examination of the lower half of the table reveals a similar pattern of findings where the more school days respondents skip (up to four to five days) the more likely they are to vape marijuana, use marijuana, or engage in dual marijuana use. To illustrate, for the model examining marijuana vaping the ORs for numbers of days skipped increases in a stepwise fashion from missing one day (OR = 1.273, p <.01), to two days (OR = 1.390, p <.01), to three days (OR = 1.762, p <.01), and then peaks at four to five days (OR = 1.860, p <.01), after this point, however, the ORs appear to decrease. This pattern of findings appears to suggest that missing additional days of school increases the likelihood of marijuana vaping, however, after a point, missing additional days is not associated with additional increases in the likelihood of marijuana vaping. Examination of the models predicting marijuana use and dual marijuana use reveals a similar pattern where the likelihood of marijuana use increases as number of days skipped increases with some leveling off after four to five days missed. Of note, for the marijuana use and dual use models the ORs appear to increase again for respondents who miss 11 or more days of school.
Logistic Regression Models of the Association between Truancy and Nicotine and Marijuana Use (N = 40,047).
Note. All models estimated controlling for age, gender, Black, Hispanic, Other, Multiracial, rural, maternal education, alcohol consumption, and substance using peers.
p < .05. **p < .01.
In the next step of the analysis we examined relationships between truancy and frequency of nicotine and marijuana use among respondents who report nicotine or marijuana use respectively in the past 30 days. Examination of Table 4 reveals that the dichotomous measure of truancy is positively and significantly associated with frequency of marijuana vaping (IRR = 1.071, p <.05) among respondents who report vaping marijuana at least one the past 30 days. 2 This finding indicates that among respondents who report vaping marijuana at least once in the past 30 days, those who skipped school in the past four weeks tended to report higher frequencies marijuana vaping in the same time period compared to those who did not skip school. In contrast, the dichotomous measure of truancy is not significantly associated with frequency of nicotine vaping, cigarette smoking, or marijuana use frequency among respondents who report engaging in the respective form of substance use at least one in the prior 30 days.
Regression Models of the Associations between Truancy and Frequency of Nicotine and Marijuana Use among Respondents who Report Nicotine or Marijuana Use.
Note. Models are restricted to respondents who reported using the substance of interest at least once in the past 30 days.
p <.05. **p < .01.
The lower half of Table 4 displays models examining associations between the number of days respondents skipped and frequencies of nicotine and marijuana use among respondents who reported nicotine or marijuana use respectively in the past 30 days. As can be seen, as number of days skipped increases, reported frequencies of nicotine vaping (IRR = 1.017, p <.01), marijuana vaping (IRR = 1.025, p <.01), and marijuana use (IRR = 1.015, p <.05) also increase. Overall, these findings indicate that respondents who skip more days of school tend to vape nicotine and use marijuana at higher frequencies. To garner a better understanding of this pattern we graphed predicted frequencies of nicotine and marijuana use as a function of number of days skipped in Figure 1.
Predicted frequency of nicotine and marijuana use in the last 30 days as a function of number of days truant with 95% confidence interval.
As can be seen in (Figure 1a–d), predicted frequencies of nicotine vaping (Figure 1a), marijuana vaping (Figure 1c), and marijuana use (Figure 1d) appear to increase as the number of days skipped increases. To illustrate, (Figure 1a) displays the predicted frequency of nicotine vaping among respondents who report vaping nicotine at least one over the past 30 days as a function of number of school days skipped. The overall pattern indicates that as number of school days skipped increases, nicotine vaping frequency also increases among those who report engaging in nicotine vaping at least once in the past 30 days. Similar patterns emerge for marijuana vaping (Figure 1c) and marijuana use ((Figure 1d)). In contrast, as can be seen in the (Figure 1b), skipping additional days of school does not appear to have the same relationship with cigarette smoking frequency among respondents who report smoking cigarettes in the past 30 days.
Discussion
The results of this study yielded two major findings. First, truancy was positively and significantly associated with the likelihood of nicotine and marijuana vaping in the past 30 days. The results also indicate that truancy is associated with cigarette smoking and general marijuana use in the same time period. These findings are consistent with a number of previous studies linking truancy with increased risks of substance use (Flaherty et al., 2012; Henry & Huizinga, 2007; Henry & Thornberry, 2010; Rocque et al., 2017; Vaughn et al., 2013). Further examination of the number of days skipped or cut reveals that as students miss more days of school, the likelihood of engaging in nicotine and marijuana vaping along with traditional forms of nicotine and marijuana use increases in a dose-response fashion. In essence, the more days students miss the more likely they are to use nicotine and marijuana. One possible explanation for this relationship may be that as students skip days of school they may have more opportunities to engage in unstructured socializing with their peers (who are also skipping school) outside of the supervision of their parents. This unsupervised and unstructured socializing may then create more opportunities to engage in delinquent behavior including substance use (Augustyn & McGloin, 2013; Leimberg & Lehmann, 2022; Osgood et al., 1996).
Of note, the ORs for this relationship appear to indicate that the likelihood of nicotine and marijuana use increases as the number of days skipped increases but appears to level off around four to five days or six to 10 days depending on the model. This pattern of results may suggest while skipping days of school increases risks for substance use, students who are skipping substantially more days may not be engaging in the same forms of activities outside of school as those missing four to five days (20% to 25% of the last four weeks) or six to 10 days (30% to 50% of the last four weeks). Keep in mind the truancy measure assesses the number of days students have skipped in the past four weeks (20 school days), as such respondents missing 11 or more days have missed more than 50% of the past four weeks. This particularly high level of truancy (seen in less than 2.5% of the sample) may be associated with different patterns of behavior and associations when out of school. While there is no universal definition as to what distinguishes “high” and “low” levels of truancy (e.g., Attwood & Croll, 2006; Vaughn et al., 2013), there is evidence that higher levels of truancy are associated with more substance use relative to lower and moderate levels of truancy (Vaughn et al., 2013). Additional research will be needed to explore connections between very high levels of truancy and substance use, especially among samples with data collected after the COVID-19 pandemic.
The second major finding to emerge is that magnitude of truancy is associated with increased frequencies of nicotine vaping, marijuana vaping, and marijuana use among respondents who report using these substances at least once in the past 30 days. These findings indicate that as the number of days students skip or cut school increases frequencies of nicotine vaping, marijuana vaping, and marijuana use tend to increase as well. In essence, as students miss more days their substance use patterns also tend to increase. These findings are consistent with previous research by Henry and Thornberry (2010) which suggests that truancy is associated with escalating patterns of substance use among adolescents who have already initiated use. Additional research will be needed to explore the mechanism of this relationship. Perhaps as students miss more days of school they have more time to engage in deviant behavior and therefore have more time to engage in vaping, and marijuana use. Again, additional research will be needed to explore the social contexts of substance use among adolescents skipping days of school.
The findings of this study need to interpreted taking into account five limitations. First, the data employed in the study are cross-sectional in nature limiting our ability to make causal inferences. Second, the wording of the general marijuana use measure does not exclude marijuana vaping. As such, some respondents who only vaped marijuana may have responded affirmatively to the marijuana use measure raising concerns about comparability between the marijuana vaping and marijuana use measures and the dual use measure. Ancillary analyses employing a restricted version of this measure excluding respondents who reported vaping marijuana led to a similar pattern of findings indicating that truancy is positively and significantly associated with marijuana use (excluding vaping). However, additional research will be needed to explore potential differences in the relationship between truancy and vaping marijuana and marijuana smoking (or other forms of marijuana use) specifically. Third, the FYSAS includes a representative sample of Florida middle school and high school students and, as a result, the findings may not be generalizable to adolescents in other states or regions of the country. Future research will be needed to explore these relationships in a nationally representative sample. Fourth, the FYSAS is a self-report survey and therefore, all measures of truancy and substance use were self-reported by the respondents. As a result, some respondents may have underreported their substance use or truancy due to social desirability bias. Finally, as this is a school-based survey administered in class, students who were truant on the day of survey administration are not represented in the data. As such, students with high levels of truancy may be underrepresented in this study.
The results of this study have several implications for policy. Previous research indicates that truancy interventions which target the underlying causes of truancy are the most effective (Dembo & Gulledge, 2009; Mazerolle et al., 2019). Truancy is a complex issue that often has multiple explanations that vary from one youth to the next. If substance use is a reason behind truanting behavior, interventions designed to target that risk factor would be worthwhile. However, there is currently an absence of research on interventions designed to reduce vaping among youths who engage in truancy. In a recent scoping review of adolescent vaping interventions, many interventions are implemented in schools, as part of educational campaigns, and through policy regulation (DiCasmirro et al., 2025). Findings from this review indicate that many of the school-based and educational campaign interventions were associated with increased knowledge of vaping, a lower likelihood of vaping, and lower intentions to vape. Such interventions could potentially be adapted to reduce vaping among youths who engage in truancy. These might be especially beneficial when implemented in schools and designed to target other underlying causes of truancy. For instance, the Ability School Engagement Program (ASEP), a truancy prevention intervention, has shown many ancillary benefits. A component of the intervention is designed to identify the specific causes of truanting among youths and then develop a tailored action plan designed to reengage the youths in school. Tested under randomized controlled field trial conditions, results indicate that those who participated in ASEP had lower self-reported antisocial behavior and official arrests (Bennett et al., 2018; Mazerolle et al., 2019). Interventions, such as ASEP, may therefore be beneficial in reducing not only the likelihood of truancy, but the likelihood of vaping if that plays a role in contributing to truancy.
Footnotes
Acknowledgements
The analyses, conclusions, views, and opinions presented here are those of authors alone and should not be attributed to any of the organizations that sponsor the Florida Youth Substance Abuse Survey.
Ethical Considerations
This study was determined “exempt” by the IRB of The University of Texas at San Antonio (FY20-21-130)
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Data is available on request through the Florida Departments of Children and Families