A Meta-Analysis of the Relationship between Cannabis,Opiates,Cocaine,Heroin,or other Illegal Drug Use and Student Academic and Behavioral Outcomes

Abstract

The meta-analysis, that included 75 studies, examined the relationship between illegal drug consumption, on the one hand, and student academic and behavioral outcomes, on the other, for the middle school to college grade levels. The meta-analysis first (research question #1) addressed whether there is a statistically significant relationship between illegal drug consumption and student academic and behavioral outcomes. A second question assessed whether there was a statistically significant relationship between the consumption of specific kinds of illegal drugs and student academic and behavioral outcomes (research question #2). The third analysis distinguished between the effects for educational- and behavioral-outcomes to see whether the consumption of illegal drugs was associated with one more than the other (research question #3). Fourth, there were analyses to determine whether the effects that emerged under the first two research questions differed by the age of the student (research question #4). The fifth analysis (research question #5) assessed whether the effects that emerged under the first two research questions differed by the race of the student. The results indicated widespread statistically significant effects for all the drugs under study. The extent of the effects were considerably greater for college students than they were for middle school students.

Keywords

urban education students school reform school improvement educational reform

Drug use has become one of the most hotly debated topics in education today, as well as in the world at large (Jeynes, 2002; National Center on Addiction & Substance Abuse, Columbia University, 2011). A myriad of scholars have focused on the salience of self-discipline regarding drug use, if children are to do well in school. Since the mid-1960s few topics have stirred human emotions as much as what society’s attitude should be toward the use of illicit drugs and whether there should be organized efforts to reduce the use of these drugs, particularly by youth and college-age individuals. Moreover, because there are such a wide variety of these drugs, one of the most debated questions regarding illicit drugs is which drugs should be discouraged the most and whether the effects of certain specific drugs are significant enough to be discouraged at all (Belenko, 2000; Levinthal, 2012).

To be sure, certain published reports regarding the decline in test scores, for example, have blamed increased usage of illegal drugs by high school students as partially responsible for the decline in SAT scores (Jeynes, 2007; Wirtz, 1977). The question remains, however, whether there are other factors that are more pronounced than the influence of these substances (Jeynes, 2011; Ritchie, 2016). Moreover, there may be subtle or complex interactions between various factors that, when considered, might substantially reduce the influence of illicit drugs on the scholastic and behavioral-outcomes of youth (Layton & Brown, 2012; Jeynes, 2010).

A growing number of studies have been done that examine the impact of substance abuse generally, and consuming illegal addictive drugs specifically, on the general population. A large number of these studies, however, have certain shortcomings such as focusing only on the general student population, addressing only certain substances, or having sample sizes that are too small from which people can generalize to specific population groups such as blacks and Latinos (Jeynes, 2002; Blavos et al., 2017).

There is a real need to determine what actually are the effects of consuming various drugs such as marijuana, cocaine, heroin, and opiates, in part, because this issue has become so emotionally charged (Harris, 2005; Husak, 2002; Pollan, 2001). There are several reasons that are the most salient in determining why substance abuse has become so emotionally charged. First, unfortunately most people’s views about a given substance generally have more to do with whether they have partaken a particular drug rather than they do with what objective evidence indicates (National Center on Addiction & Substance Abuse, Columbia University, 2011; Pollan, 2001). That is, people who consume a given substance usually want it legalized and those who abstain want the drug to be illegal (National Center on Addiction & Substance Abuse, Columbia University, 2011; Pollan, 2001). This reality is particularly of concern, because legalization of drugs then becomes a matter of how popular a drug is rather than the extent to which it does or does not have ill effects (Blavos et al., 2017; Pollan, 2001). Judging the appropriate legality of a drug based on its popularity becomes a very capricious undertaking, which can potentially do a great deal of harm. Once certain behavior is legalized, and people therefore become accustomed to its practice, making the same actions illegal becomes a herculean task (Brick & Erickson, 2013; Pollan, 2001). It would seem far wiser to base drug policy on enduring facts rather than the vagaries of public opinion (Jeynes, 2002; Brick & Erickson, 2013). Therefore, quantitative analysis is particularly helpful.

Second, those who have been victims of physically-, verbally-, or emotionally-abusive behavior resulting from a person engaged in substance abuse are more likely than most to oppose legalization of drugs, such as marijuana (Orford, 2013; Rogers, 2011). Whether one is a relative of a victim of-violence, abusive language, or behavior resulting from drug addiction or a relative of the abuser can also make a substantial difference in one’s perspective on whether various drugs should be legalized or not (Husak, 2002; Orford, 2013; Rogers, 2011). Once again, analyzing the collective results of studies can help one rise above the emotions that are often associated with this issue.

Third, a recent complicating factor is that some of these substances that are addictive can also serve as painkillers in particular and it is possible they might have other medicinal benefits (Husak, 2002). As one might imagine, this is the most controversial of the factors that have made this debate an emotional topic for two primary reasons. (a) There are a myriad of other painkillers to choose from that are much less addictive and much more mild in their side effects and undesirable characteristics (Blavos et al., 2017; Kleiman et al., 2017). (b) the research literature on the medical benefits of marijuana, opiates, and other drugs is inconclusive and subject to much debate. Drugs of this kind, almost by definition, may reduce pain (Blavos et al., 2017; Kleiman et al., 2017). However, the medical benefits beyond this are subject to much discussion and are inconclusive (Blavos et al., 2017; Kleiman et al., 2017). Nevertheless, the medical benefits might be real.

Fourth, the taking of certain drugs is regarded by a number physicians as using gateway drugs making the consumption of certain harder drugs, such as cocaine and heroin, more likely (Hall & Pacula, 2004; Orford, 2013; Rogers, 2011). There is some evidence to indicate that the partaking of substances such as marijuana and alcohol may make it more likely for individuals to take harder drugs at a later time (Hall & Pacula, 2004; Orford, 2013). Moreover, to the extent that less severe drugs such as alcohol and marijuana are legalized, simple logic dictates that drug cartels and other organized crime outlets will simply resort to pushing and distributing primarily hard drugs (Hesterman, 2013).

In the midst of a plethora of emotion-filled people arguing one way or another for the legalization of marijuana and loosening the restrictions on opiates and other drugs, there is a need to lay aside personal preferences that are largely based on one’s own consumption habits as well as those whom one loves. Instead, there needs to be an assessment of the effects of these drugs based on facts (Hall & Pacula, 2004; Rogers, 2011). As part of this overall assessment, a meta-analysis is needed to determine the relationship between the consumption of various drugs and student-academic and behavioral outcomes.

A meta-analysis statistically combines all the relevant existing studies on a given subject in order to determine the aggregated results of said research. Meta-analyses are probably the single most popular type of academic article, because they enable people to grasp what the overall body of research on a given topic indicates. A meta-analysis would enable social scientists to conclude which drug policies would be the most productive and the most efficacious.

Methods for the Meta-Analysis

Analytical Approach

This meta-analysis examined the relationship between the consumption of illicit drugs and student outcomes. The consumption of illicit drugs was defined as either taking a drug that is not legal to consume, as defined by U.S. federal law, or exceeds the amount prescribed by a qualified physician. The procedures employed to conduct the meta-analysis are outlined under this heading (Analytical Approach) and the following headings below: Data Collection Method, Statistical Methods and Effect Size Statistics, Study Quality Rating, and Defining of Variables. Each study included in this meta-analysis met the following criteria:

(1) It needed to examine illegal drug use and the specific independent variables in a way that could be conceptually and statistically distinguished from other primary variables under consideration. For example, if a researcher examined drug usage in conjunction with other independent variables that could not be statistically isolated from the other features; the study was not included in the analysis.

(2) It needed to include a sufficient amount of statistical information to determine effect sizes. That is, a study needed to contain enough information so that test statistics, such as those resulting from a t-test, analysis of variance, and so forth, were either provided in the study or could be determined from the means and measures of variance listed in the study.

(3) If the study used a control group, it had to qualify as a true control group and therefore be a fair and accurate means of comparison. Moreover, if the research utilized a control group at some times but not others, only the former comparisons were included in the meta-analysis.

(4) The study could be a published or unpublished study.

Due to the nature of the criteria listed above, qualitative studies were not included in the analysis. Although qualitative studies are definitely valuable, they are difficult to code for quantitative purposes and any attempt to do so might bias the results of the meta-analysis.

Data collection method (coding and rater reliability)

In order to obtain the studies used in the meta-analysis, a search was undertaken to locate the relevant studies on the consumption of the drugs under study. The first procedures to be used to locate these studies involved a computer search using 60 research databases (e.g., Psych Info., ERIC, Dissertation Abstracts International, Wilson Periodicals, Sociological Abstracts, and so forth) to find studies examining the consumption of the drugs under study. The search words included illegal drugs, illicit drugs, consumption, addiction, marijuana, cannabis, weed, pot, opioids, various specific types of opioids, cocaine, heroin, ecstasy, amphetamines, and many other similar terms. Reference sections from journal articles on drug usage were also examined to find additional research articles. The research team obtained a total of 106 studies that addressed the relationship under study, and found 75 studies that had a sufficient degree of quantitative data to include in this meta-analysis. Among the 75 studies that possessed a sufficient degree of quantitative data to include in this meta-analysis, the total number of subjects was approximately 525,394.

A number of different characteristics of each study were included for use in this study. These characteristics included: (a) report characteristics, (b) sample characteristics, (c) intervention type, (d) the research design, (e) the grade level or age of the students, (f) the outcome and predictor variables, (g) the attrition rate, and (h) the estimate of the relationship between student drug use and pupil academic- and behavioral-outcomes.

Report Characteristics—Each study entry began with the name of the author of the study. Then the year the study was recorded, followed by the type of research report. Research reports were defined either as a journal article, book, book chapter, dissertation, Master’s thesis, government, school or private report, conference paper, or other type of report.

Sample Characteristics included the number of students sampled, their locations, and how they were selected, for example, via random selection, stratified random selection, or via advertisement.

Intervention Type—The experimental or procedural manipulation used, if any, was recorded to determine the effects of the consumption of these largely illicit drugs.

Research Design—The studies in this meta-analysis were categorized into three basic types of designs. First, the studies that employed some type of manipulations to assess the effects of these drugs were noted. The second type of design included studies that took cross-sectional measures of these drugs without utilizing any type of manipulation. The third type of design involved the calculation of a correlational coefficient between the usage of these drugs and academic and behavioral outcomes.

For studies that employed a manipulation to assess the effects of the consumption of these largely illegal drugs, the following information was recorded (a) the length, frequency, duration, and total number of training sessions, (b) the method of training (workshop, individual meetings, phone calls, videotape, email communication, newsletter), (c) the type of behavioral or attitude-related outcome measure, (d) the unit of analysis (individual student or classroom) at which the effect size was calculated, and (e) the magnitude of the relationship between the consumption of these drugs and the measured outcomes.

For the cross-sectional studies and correlation studies, if it was available, the (a) socio-economic status of participants in the sample and (b) the types of behavioral and attitudinal measures that were used and recorded.

The grade level or age of the students was coded, including means and standard deviations when they were available.

The outcome and predictor variables from each study were coded to include the different drugs that were measured.

Attrition Rate—When available, the attrition rate of each study was coded.

The estimate of the relationship between the consumption of these largely illicit drugs with academic and behavioral variables—The process of the effect size estimation is described in the next section.

Statistical methods and the effect size statistic

Effect sizes were computed from data in such forms as t tests, F tests, p levels, frequencies, and r-values via conversion formulas provided by Glass et al. (1981). When results were not significant, studies sometimes reported only a significance level. In the unusual case that the direction of these not significant results was not available, the effect size were calculated to be zero.

For studies with manipulations the standardized mean difference was used to estimate the effects of illegal drug consumption. The d-index is a scale-free measure of the separation between two group means. Calculating the d-index for any comparison involves dividing the difference between the two group means by either their average standard deviation or by the standard deviation of the control group. In each meta-analysis, the experimental group mean was subtracted from the control group mean. Then the researchers divided the difference by their average standard deviation. As a supplement to these analyses, the Hedges’ “g” measure of effect size was used (Hedges & Vevea, 1998). Since it employs the pooled standard deviation in the denominator, it customarily provides a more conservative estimate of effect size. Hedges also provides a correction factor that helps to adjust for the impact of small samples.

For studies that involved cross-sectional measures but included no attempt to statistically equate students on third variables, the results from the t-tests, F-tests, and correlation studies provided by the researchers in the study were used. Probability values were used as a basis for computation only if the researchers did not supply any of information on the test statistics just mentioned.

Calculating average effect sizes

A weighting procedure was used to calculate average effect sizes across all the comparisons. First, each independent effect size was multiplied by the inverse of its variance. The sum of these products was then divided by the sum of the inverses. Then, 95% confidence intervals were calculated. As Hedges and Vevea (1998) recommend, all the analyses were conducted using fixed-error assumptions in one analysis and applied random-error assumptions in the other. The results presented here used analyses based on random-error assumptions. The rationale for presenting these results rather than those using fixed-error assumptions is to utilize analyses that yielded more conservative effect sizes.

Study quality rating

Two researchers coded the studies independently for quality, the presence of randomization, and whether the definitional criteria of a particular drug consumption are met. Study quality and the use of random samples were graded on a 0 (lowest) to 3 (highest) scale. Quality was determined using the following:

(1) Did it use randomization of assignment?

(2) Did it avoid mono-method bias?

(3) Did it avoid mono-operation bias?

(4) Did it avoid selection bias?

(5) Did it use a specific definition of a particular drug consumption?

It is important that if one is to conclude that relationships exist between independent and dependent variables (in this case drug consumption and student outcome variables) that the analyses of only high quality studies confirm the results.

Rater Reliability

The inter-rater reliability was calculated by computing percentage of agreement on: the definition of a particular drug consumption, issues of randomization, and quality of the study. A supplementary analysis was done to include only those studies with quality ratings with quality ratings of 2 to 3.

Illicit drug consumption and student academic achievement and behavior

This meta-analysis examined the relationship between illicit drug consumption and pupil academic achievement and behavior in students from middle school up until college. This meta-analysis’ first research question addressed whether there is a statistically significant relationship illegal drug consumption and student academic and behavioral outcomes. A second question assessed whether there was a statistically significant relationship between the consumption of specific kinds of illegal drugs (e.g., cocaine, marijuana, heroin, and opiates) and student academic and behavioral outcomes (research question #2). The third analysis distinguished between the effects for educational- and behavioral-outcomes to see whether the consumption of illegal drugs was associated with one more than the other (research question #3). Fourth, there were analyses to determine whether the effects that emerged under the first two research questions differed by the age of the student (research question #4). The fifth analysis (research question #5) assessed whether the effects that emerged under the first two research questions differed by the race of the student.

Defining of Variables

Independent variable

For the purposes of this study, illegal- or illicit-drug consumption was defined as partaking drugs that were either defined as against federal law or taking amounts of the drug that were beyond the legal limit defined by the patient’s doctor.

Dependent variables

Academic Achievement—How students performed in school defined by GPA, standardized tests, or less standardized measures such as teacher ratings.

School Behavior—How students behaved as measured by whether they were truant, suspended for school, got in fights with children, etc.

Tests of homogeneity were completed on the consumption of illicit drugs to gain a sense of the consistency of specific the illicit drugs consumption measures across studies.

Results

Overall, the results of the meta-analyses indicated that there is a relationship between the drugs under study, that is, cannabis, opiates, cocaine, or heroin for middle school through college age urban youth, as expressed in academic and behavioral outcomes combined.

Table 1 indicates that the effect sizes ranged from a high of 1.71 to a low of .00. It is interesting to note that although there were a number of individual studies in the meta-analysis that did not yield statistically significant results, none of the overall relationships were in the negative direction. The studies with the smallest samples produced the most extreme effect sizes on either end, consistent with the “funnel” pattern ideal in effect sizes (Greenhouse & Iyengar, 1994).

Table 1.

Studies Included in the Meta-Analysis Listed by Author, Year of Study, Sample Size, and a Variety of Other Characteristics.

Study and year (identified by lead authors)	Sample size	Distinctions of study	Grade or age of students	Effect size without sophisticated controls	Effect size with sophisticated controls
Lowry et al. (1994)	11,631	Oversampled for African Americans and Latinos	Grades 9–12	1.71	—
Brill (1982)	2,190	Examined High Schools in Jamaica and Los Angeles	College	1.40	—
Phillips et al. (2015)	57	Examined both general Marijuana use and being under the influence of Marijuana	College	1.17	—
Mintzer et al. (2005)	20	Examined former opioid users	College	1.10	—
Springer et al. (2007)	15,240	Examined 2003 youth at risk survey	Grades 9–12	1.00	—
Chabrol et al. (2010)	615	Focused on delinquent behaviors	Grades 9–12	0.99	—
Dorr et al. (2009)	364	Study done in Chile	Grades 9–12	0.80	—
Arafat and Yorburg (1973)	592	All the subjects were unmarried females	College	0.75	—
Miozzo et al. (2013)	453	Study done in Brazil	Grades 9–12	0.69	—
Martins and Alexandre (2009)	92,886	Youth at risk behavior survey	Grades 9–12	—	0.68
Ehrenreich et al. (2015)	619	Used a combination of teacher ratings and student reports	Grades 6–12	0.67	—
Myers et al. (2017)	65	Examined cognitive learning	College	0.65	—
Wong et al. (2013)	73,183	Examined a wide range of drugs	Grades 9–12	—	0.65
Kalet (2006)	734	Examined university in the Northeast	College	0.60	—
Finn (2012)	1,112	Examined both general Marijuana use and being under the influence of Marijuana	Grades 9–12	0.58	—
Enns et al. (2017)	1,599	Study done in Canada	High school and college	—	0.55
Kandel and Davies (1996)	7,611	Examined 54 schools in New York	Grades 7–12	0.51	—
Shewan and Dalgarno (2005)	126	Study done in Scotland	College	0.48	—
Author (2002)	18,726	Used NELS data set	High school	0.48	—
Vest et al. (2016)	143	Examined psychological disorders	College	0.47	—
Buchman et al. (2007)	50,000	Assessed students at two times during their childhood	High school	0.44	—
Tolomeo et al. (2018)	146	Examined males only	College	—	0.44
Ersche et al. (2006)	120	Study done in England	Post-college	0.42	—
Chabrol et al. (2008)	248	Examined delinquency	High school	0.42	—
Tarter et al. (1995)	180	Examined females only	Middle school and high school	0.41	—
Bolin et al. (2017)	8,471	Focused on Marijuana	College	0.41	—
Williams et al. (2007)	212	Focused on low-SES students	Grade 11	0.40	0.20
Anker et al. (1971)	7,032	Focused on various levels of student behavior	College	0.40	—
Guerra et al. (1987)	60	Focused on student psychological performance	College	0.40	—
Guinn (1977)	937	Examined Mexican Americans	College	0.39	—
Burge et al. (1995)	11,631	Examined both consumption and being under the influence	Grades 9–12	0.39	—
Ornstein et al. (2000)	111	Examined only male subjects	College	—	0.37
Ansary and Luthar (2009)	256	Study in Northeastern U.S.	Grades 10–12	—	0.35
Chabrol and Saint-Martin (2009)	312	Study done in France	Grades 9–12	0.35	—
Meier et al. (2015)	254	Study done in England	Grades 9–12	0.35	—
Felts et al. (1992)	3,064	Study done in North Carolina	Grades 9–12	0.34	—
Roy et al. (1990)	1,270	Study done in Australia	Middle school and high school	0.34	—
Scivoletto et al. (2002)	689	Study done in Brazil	Grades 9–12	0.32	—
Simon (1974)	414	Examined GPA at both high school and college levels	College	0.31	—
			Grades 9–12	0.32
Zimmerman and Schmeek-Cone (2003)	681	Examined African Americans	Grades 9–12	0.31	—
Schepis et al. (2018)	28,138	National survey of drug use and health	High school and college	—	0.30
Lin et al. (2009)	1,076	Data set from mid-west	Middle school and high school	—	0.30
Miron et al. (1997)	7,580	Focused on behavior	Grades 9–12	0.30	—
Author (2006)	18,726	Used NELS data set	High school	—	0.28
Arria et al. (2015)	1,117	Focused on college GPA	College	0.28	—
Castro et al. (1989)	1,081	Examined violent acts	Grades 9–12	0.25	—
Pilowsky et al. (2011)	343	Examined primarily minority subjects from all over the country	College	0.28	0.20
Melchior et al. (2017)	1,103	French Study examining whether students dropped out of High School	College	0.24	—
Egbochuku and Akerlele (2007)	800	Study in Nigeria	Grades 9–12	0.24	—
Hittner and Kennington (2008)	1,103	French Study examining whether students dropped out of high school	College	0.24	—
Brook et al. (2008)	749	Longitudinal study	Grades 9–12	—	0.22
Finnell and Jones (1975)	335	Focused on Marijuana	College	0.22	—
Ndete et al. (2010)	343	Study in Kenya	Middle school and high school	0.22	—
Yan (2005)	1,476	Used national educational longitudinal study	Grade 12	—	0.21
Strote et al. (2002)	14,000	Study in Spain	College	—	0.19
Silins et al. (2015)	2,179	Australian study	Grades 6–12 and college	0.23	0.18
Register et al. (2001)	3,525	Examined males only. 1/3 of the sample African American or Latino	Grades 7–12	—	0.22
Hernandis et al. (2002)	564	Study in Mexico	Middle school and high school	0.22	—
Affia et al. (2013)	1,133	Study undertaken in mid-Atlantic States	College	0.20	—
Horwood et al. (2010)	6,000	Australian study	Middle school and high school	—	0.18
Wang et al. (2014)	21,672	Study in China	Grades 7–12	—	0.17
Cox et al. (2007)	1,488	Sample nearly half African American	Grades 9–12	0.17	—
Suerken et al. (2016)	3,146	From 11 colleges in North Carolina and Virginia	College	0.16	—
Pardini et al. (2015)	2,775	Longitudinal study in Pittsburgh	Ages 11–16	0.15	—
Maloney et al. (2007)	1,125	Australian study	College	—	0.15
Zule et al. (2016)	1,985	Examined sexual behavior	College	0.14	—
D’Amico et al. (2016)	6,509	Southern California study	Middle school	—	0.12
Maggs et al. (2015)	4,925	Examined degree attainment	Grade 12	0.11	—
Pattek et al. (2017)	26,475	Canadian nationwide data set	Grades 9–12	—	0.10
Patrick et al. (2016)	10,020	Longitudinal study	Grade 12	—	0.09
Marie and Zolita (2017)	4,419	Study in the Netherlands	College	—	0.09
Knight (2015)	12,341	60% Minority Sample	Grades 7–8	0.09	—
Verweij et al. (2013)	3,337	Examined twins	College and beyond	—	0.08
Hill and Mrug (2015)	23,615	Focused especially on low-SES African Americans	Grades 6–12	—	0.06
Walsh and Frantz, (1972)	456	Study in New England	College	0.00	—
Goode, 1971	560	Study in New York state	College	0.00	—

Table 2 gives an overview of some of the key summary statistics of the studies included in the meta-analyses. The studies included in the meta-analysis tended to be relatively recent, indicating that most of the studies undertaken on the effects of illegal drugs on student scholastic achievement and behavior have been done relatively recently. The average year that a study done was 2004.7. The average sample size was 7,005.3. The sample sizes were often petty large. A high, 41.3%, percent of the studies possessed sample sizes of 2,000 or more. It is also important to note that the average quality of the studies undertaken was 2.00. This is pretty high. Most (about 76.0%) of the studies were rated either 3 or 2 on a 0 to 3 scale.

Table 2.

Means for Measures Assessing the Quality of Study, Year of Study, and Sample Size for the 75 Studies Included in the Meta-Analysis.

	Mean	Distribution of studies	Range
Year of study	2004.7	2012–2019 = 29	1971–2018
		2002–2011 = 24
		Before 2002 = 22
Sample size	7005.3	2,000+ = 31	20–92,886
		500–1,999 = 21
		1–499 = 23
Quality of study	2.00	3 = 21	0–3
		2 = 36
		1 = 15
		0 = 3

Table 3 presents the results of the correlation analyses that examined if there was any relationship between the quality of the study, the year of the study, the effect size, and the age of the students that were examined. All of the combinations of the relationships examined were near zero and not statistically significant.

Table 3.

Correlations between Measures Assessing the Quality of Study, the Year of Study, Average Age of the Students, and Sample Size for the 75 Studies Included in the Meta-Analysis.

	Correlation with year of the study	Correlation with effect size of the study	Correlation with quality of the study	Age of students
Year of study	____	−0.08	−0.09	0.16
Effect size from study	−0.08	____	0.08	−0.02
Quality of study	−0.09	0.08	____	−0.14
Age of students	0.16	−0.02	−0.14	_____

Effect Sizes for The Drugs Under Study Overall (Research Question #1)

Table 4 lists the effects for taking the drugs under study. For the entire sample (U.S. plus foreign) the effect size was .62 (.20, 1.04), p < .001, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was almost the same at, .61 (.19, 1.03), p < .001, standard deviation units. When including only the U.S. samples the effect sizes were similar at .63 (.21, 1.05), p < .001, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was very similar at .62 (.20, 1.04), p < .001, standard deviation units. In terms of the specific outcomes for academic achievement and behavior, the respective effects sizes were .45 (.09, .81), p < .01 and .89 (.31, 1.57), p < .001, without the use of sophisticated controls. With the use of sophisticated the results were .59 (.19, .99), p < .001 and .64 (.22, .96), p < .001, standard deviation units.

Table 4.

Effect Sizes for Consuming Illegal Drugs with 95% Confidence Intervals in Parentheses.

Type of outcome variable	Effect size without sophisticated controls	Effect size with sophisticated controls	Overall effect size
Overall (U.S. plus foreign)	0.62*** (0.20, 1.04)	0.61*** (0.19, 1.03)	62
Overall (U.S.)	0.63*** (0.21, 1.05)	0.62*** (0.20, 1.04)	63
Academic achievement	0.45** (0.09, 0.81)	0.59*** (0.19, 0.99)	0.51
Behavior	0.89*** (0.31, 1.57)	0.64*** (0.22, 0.96)	0.78
High quality studies (rating of 3)	87*** (0.33, 1.41)	0.66*** (0.28, 1.04)	0.78

Note. NA = not available.

p < .05. **p < .01. ***p < .001.

When only studies that were rated highest in quality (rated 3 out of 3) were included in the analysis, the results were .87 (.33, 1.04), p < .001, when sophisticated controls were not in use and .66 (.28, .74), p < .001, when they were.

Effect Sizes for The Specific Drugs Under Study (Research Question #2)

Table 5 indicates that the effects for taking marijuana was .41 (.14, .68), p < .01, of a standard deviation, for the entire sample included in the meta-analysis without the use of sophisticated controls. With sophisticated controls, the effect size was smaller, but was still statistically significant at .36 (.09, .63), p < .01, standard deviation units. When including only the U.S. samples the effect sizes were similar .40 (.14, .66), p < .01, of a standard deviation, for the entire sample included in the meta-analysis without the use of sophisticated controls. With sophisticated controls, the effect size was smaller, but was still statistically significant at .35 (.04, .66), p < .05, standard deviation units. In terms of the specific outcomes for academic achievement and behavior, the respective effects sizes were .27 (.04, .66), p < .05 and .52 (.18, .86), p < .01, without the use of sophisticated controls. With the use of sophisticated the results were .28 (.03, .53), p < .05 and .14 (.01, .27), p < .05, standard deviation units.

Table 5.

Effect Sizes for Consuming Marijuana and Opioids with 95% Confidence Intervals in Parentheses.

Type of outcome variable	Effect size without sophisticated controls	Effect size with sophisticated controls	Overall effect size
Marijuana consumption
Overall (U.S. plus foreign)	0.41** (0.14, 0.68)	0.36** (0.09, 0.63)	0.39
Overall (U.S.)	0.40** (0.14, 0.66)	0.35* (0.04, 0.66)	0.38
Academic achievement	0.27* (0.03, 0.51)	0.28* (0.03, 0.53)	0.27
Behavior	0.52** (0.18, 0.86)	0.14* (0.01, 0.27)	0.32
High quality studies (rating of 3)	58** (0.21, 0.95)	0.43** (0.12, 0.74)	0.51
Youth of color	98** (0.35, 1.61)	NA	98** (0.35, 1.61)
Opioid consumption
Overall (U.S. plus foreign)	0.33* (0.05, 0.61)	0.24* (0.04, 0.44)	0.27
Overall (U.S.)	0.33* (0.05, 0.61)	0.24* (0.04, 0.44)	0.27
Academic achievement	0.70** (0.20, 1.20)	0.25* (0.05, 0.45)	0.31
Behavior	0.32* (0.06, 0.58)	0.21* (0.03, 0.39)	0.22
High quality studies (rating of 3)	33* (0.06, 0.60)	NA	33* (0.06, 0.60)

Note. NA = not available.

p < .05. **p < .01. ***p < .001.

When only studies that were rated highest in quality (rated 3 out of 3) were included in the analysis, the results were .58 (.21, .95), p < .01, when sophisticated controls were not in use and .43 (.12, .74), p < .01, when they were.

Table 5 also lists the effects for taking prescription opioids. For the entire sample the effect size (U.S. plus foreign) was .33 (.05, .61), p < .05, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was, .24 (.04, .44), p < .05, standard deviation units. When including only the U.S. samples the effect sizes were similar at .33 (.05, .61), p < .05, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was .24 (.04, .44), p < .05, standard deviation units. In terms of the specific outcomes for academic achievement and behavior, the respective effects sizes were .70 (.20, 1.20), p < .01 and .32 (.06, .58), p < .05, without the use of sophisticated controls. With the use of sophisticated the results were .25 (.05, .45), p < .05 and .21 (.03, .39), p < .05, standard deviation units.

When only studies that were rated highest in quality (rated 3 out of 3) were included in the analysis, the results were .33 (.06, .60), p < .05, when sophisticated controls were not in use and .33 (.06, .60), p < .05, overall.

Table 6 lists the effects for taking cocaine. For the entire sample (U.S. plus foreign) the effect size was .60 (.22, .98), p < .01, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was a little higher at, .63 (.19, 1.03), p < .01, standard deviation units. When including only the U.S. samples the effect sizes were similar at .61 (.23, .99), p < .01, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was somewhat higher at .63 (.24, 1.04), p < .01, standard deviation units. In terms of the specific outcomes for academic achievement and behavior, the respective effects sizes were .61 (.24, .98), p < .01 and .53 (.18, .88), p < .01, without the use of sophisticated controls. With the use of sophisticated the results were .39 (.19, .59), p < .01 and .69 (.29, 1.09), p < .01, standard deviation units.

Table 6.

Effect Sizes for Consuming Cocaine, Heroin, and Hallucinogens with 95% Confidence Intervals in Parentheses.

Type of outcome variable	Effect size without sophisticated controls	Effect size with sophisticated controls	Overall effect size
Cocaine consumption
Overall (U.S. plus foreign)	0.60** (0.22, 0.98)	0.63** (0.19, 1.03)	0.62
Overall (U.S.)	0.61** (0.23, 0.99)	0.63** (0.24, 1.02)	0.62
Academic achievement	0.61** (0.24, 0.98)	0.39** (0.19, 0.99)	0.50
Behavior	0.53** (0.18, 0.88)	0.69** (0.29, 1.09)	0.65
High quality studies (rating of 3)	0.53** (0.33, 1.41)	NA	0.53** (0.33, 1.41)
Heroin consumption
Overall (U.S. plus foreign)	0.86*** (0.35, 1.37)	0.80*** (0.32, 1.28)	82
Hallucinogens consumption
Overall (U.S. plus foreign)	0.32* (0.05, 0.59)	0.64** (0.21, 1.07)	60
Behavior	0.32* (0.09, 0.81)	0.32* (0.19, 0.99)	0.32

Note. NA = not available.

p < .05. **p < .01. ***p < .001.

When only studies that were rated highest in quality (rated 3 out of 3) were included in the analysis, the results were .53 (.33, .73), p < .01, when sophisticated controls were not in use and .53 (.33, .73), p < .01, overall.

Table 6 also lists the effects for taking heroin. For the entire sample (U.S. plus foreign) the effect size was .86 (.35, 1.37), p < .001, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was a little smaller at, .80 (.32, 1.28), p < .001, standard deviation units.

The effects for taking hallucinogens appear at toward the bottom of Table 6. For the entire sample (U.S. plus foreign) the effect size was .32 (.05, .59), p < .05, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was a good deal higher at, .64 (.21, 1.07), p < .01, standard deviation units. In terms of the specific outcomes for behavior, the effects sizes were .32 (.09, .55), p < .05, without the use of sophisticated controls. With the use of sophisticated the effect size was .32 (.11, .53), p < .05, standard deviation units.

In Table 7 appears the effects for taking ecstasy. For the entire sample (U.S. plus foreign), the effect size was .83 (.24, 1.42), p < .01, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was at, .64 (.20, 1.08), p < .01, standard deviation units. In terms of the specific outcomes for academic achievement and behavior, the effects sizes were .67 (.22, 1.12), p < .01 without the use of sophisticated controls and .54 (.21, .87), p < .01 with them.

Table 7.

Effect Sizes for Consuming Ecstasy, Amphetamines, and Other Illegal Drugs with 95% Confidence Intervals in Parentheses.

Type of outcome variable	Effect size without sophisticated controls	Effect size with sophisticated controls	Overall effect size
Ecstasy consumption
Overall (U.S. plus foreign)	0.83** (0.24, 1.42)	0.64** (0.20, 1.08)	0.66
Academic achievement	NA	0.67** (0.22, 1.12)	0.67** (0.22, 1.12)
Behavior	NA	0.54** (0.21, 0.87)	0.54** (0.21, 0.87)
Amphetamine consumption
Overall (U.S. plus foreign)	0.87*** (0.35, 1.39)	0.77** (0.26, 1.28)	0.80
Overall (U.S.)	0.87*** (0.35, 1.39)	0.77** (0.26, 1.28)	0.80
Academic achievement	0.47* (0.08, 0.86)	NA	0.47* (0.08, 0.86)
Behavior	0.88*** (0.36, 1.40)	NA	0.88*** (0.36, 1.40)
Other drug consumption
Overall (U.S. plus foreign)	0.37* (0.05, 0.69)	0.14	26

Note. NA = not available.

p < .05. **p < .01. ***p < .001.

Table 7 also lists the effects for taking amphetamine. For the entire sample (U.S. plus foreign) the effect size was .87 (.33, 1.31), p < .001, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was a lower higher at, .77 (.24, 1.30), p < .01, standard deviation units. When including only the U.S. samples the effect sizes were similar at .87 (.33, 1.31), p < .001, of a standard deviation, without the use of sophisticated controls. standard deviation units. With sophisticated controls, the effect size was somewhat smaller at .77 (.24, 1.30), p < .01, standard deviation units. In terms of the specific outcomes for academic achievement and behavior, the respective effects sizes were .47 (.18, .73), p < .01 and .88 (.34, 1.42), p < .01, without the use of sophisticated controls and overall.

Table 7 includes the effects for taking other drugs. For the entire sample (U.S. plus foreign) the outcome was .37 (.05, .69), p < .05, of a standard deviation, without the use of sophisticated controls. With sophisticated controls, the effect size was no longer statistically significant at, .14, p = n.s.

The Effects for Educational- and Behavioral-Outcomes (Research Question #3)

In terms of the specific outcomes for academic achievement and behavior, the respective effects sizes were .45 (.09, .81), p < .01 and .89 (.31, 1.57), p < .001, without the use of sophisticated controls. With the use of sophisticated the results were .59 (.19, .99), p < .001 and .64 (.22, .96), p < .001, standard deviation units. The fact that when sophisticated controls were not in place the effects for taking illicit drugs was at a much higher level for student behavior than for student achievement, whereas the numbers were very similar and even somewhat greater for academic achievement when sophisticated controls were in place is an interesting finding and will be examined in the Discussion Section.

Differences by the Age of the Students (Research Question #4)

The analyses (see Table 8) indicate that generally as youth continue to partake of a given addictive drug, they become even more erratic in their behavior and academic performance. When comparing effects for different age groups, students in: (1) middle school, (2) high school, and (3) college, it is important that one compare by using the same drug, otherwise differences can emerge simply because: (a) as students become older they tend to become more likely to use harder drugs, which on average exert a greater downward pressure on scholastic- and behavioral-outcomes than do less hard drugs. (b) they reflect the fact researchers tend to examine the consumption of certain drugs more with particular age groups.

Table 8.

Effect Sizes for Consuming Marijuana by Age Group with 95% Confidence Intervals in Parentheses.

Type of outcome variable	College age	High school age	Middle school age
Marijuana consumption	0.55** (0.18, 0.92)	0.41** (0.13, 0.69)	0.18* (0.02, 0.34)

Note. NA = not available.

p < .05. **p < .01. ***p < .001.

Among the age groups represented, studies addressing the substance use of middle school pupils overwhelmingly focused on marijuana, because generally most students of this age have not partaken of hard drugs or opioids. Studies that sought to focus on opioid use were most likely to include college students. Marijuana was, by far, the substance most examined in these studies. Therefore, when considering the effects of drug consumption by age, marijuana was the substance chosen to give the most accurate effect sizes.

The results indicate that the effects sizes were substantially larger for older students than for younger students. For college students the effects were .55 (.18, .92), p < .01, which was the largest of the three groups examined. For high school students, the results were not quite as large as for the college students at .41 (.13, .69), p < .01 standard deviation units. For pupils attending middle school the effects were smallest of the three age groups and the level of statistical significance was at .05 rather than the .01 that emerged for the other two age groups. The effects were .18 (.02, .34), p < .05.

Although the trends by age are not reported here for other illegal drugs, in part because of the small number of students taking hard drugs at the middle school level made comparisons by age difficult, the trends in effect sizes becoming larger as the age of the students increased were apparent for other illegal drugs as well.

Differences by the Race of the Student (Research Question #5)

The meta-analytic results (see Table 9) yielded roughly the same pattern in effect sizes for children of color, as it did for the overall sample, when no sophisticated controls were used. When no sophisticated controls were utilized, the effect size for marijuana consumption was .40 (.12, .68), p < .01). When sophisticated controls were used the effect size was .25 (.09, .41), p < .01). The effect size for illegal drug usage overall was, just as in the case for marijuana, almost identical as that which emerged for the entire sample, that is, .61 (.22, 1.00), p < .001). The effect size was smaller when sophisticated controls were used at .35 (.06, .64), p < .05).

Table 9.

Effect Sizes for Consuming Marijuana Drugs Overall for Students of Color with 95% Confidence Intervals in Parentheses.

Type of outcome variable	Effect size without sophisticated controls	Effect size with sophisticated controls	Overall effect size
Marijuana	0.40** (0.12, 0.68)	0.25* (0.09, 0.41)	0.38
Overall drug use	0.61*** (0.22, 1.00)	0.35* (0.06, 0.64)	0.55

Note. NA = not available.

p < .05. **p < .01. ***p < .001.

In a related supplementary analysis, that is not one of the research questions and does not appear in a table, the effects of consuming drugs for females, .66 (.23, 1.09), p < .001, was numerically higher than for males .58 (.16, 1.00), p < .001, but the difference was not statistically significant.

Discussion

The results of this meta-analysis indicate that partaking of any of the illicit drugs examined was associated with lower academic and poorer behavioral-outcomes, in almost every case to considerable degrees. The consistency of these findings, across the broad spectrum of drugs that were examined, is noteworthy. This is especially important to note, because there are some people who will argue that hallucinogens are not as destructive to adolescent lives as other types of drugs (Bennett & White, 2015; Rosenthal, 2019). The consistency of these findings is also important, because there is clearly a disconnect between the data and the perceptions of pre-adolescents, adolescents, and young adults (Bennett & White, 2015; Binks et al., 2005).

Moreover, two other considerations are notable, in part because of the student ages and large sizes of some of the findings. First, given that struggles with illegal drugs and prescription opioids often begin in a person’s youth and are associated with depressed academic outcomes, the academic, parent, and teacher must consider not only the immediate impact of taking these drugs, but also the likely influences of these actions upon their futures (Brown & Lo, 2000; Hitchens, 2012). The reality is that the grades that one receives, particularly in high school and college, often play a major role in terms of affecting one’s future. Their roles are especially puissant when applying for a job on the next level of the educational ladder (Bernstein et al., 2019; Wai et al., 2005).

There is evidence that grades and SAT scores together are often very good predictors of one’s educational and occupational success (Bernstein et al., 2019; Wai et al., 2005). Grade point averages (GPAs) are often the best indicators of GPAs achieved at higher levels of education (Beard & Marini, 2018; Kobrin & Patterson, 2011). SAT (including PSAT) scores are frequently some of the best predictors of occupational success (Bernstein et al., 2019; Wai et al., 2005). Hence, the salience of this issue is even greater, when one recognizes that preventing illegal drug use not only influences a young person’s current life, but is likely to extend for decades to come (Anisman, 2015; Higgins, 1998). The reality is that certain decisions that pre-teens and teenagers make at this key point of their lives can weigh on the number of positive options available to them in the future (Anisman, 2015; Higgins, 1998).

Second, although the outcome variables were primarily divided up into academic and behavioral categories, some of the behavioral categories included or reflected acts of violence either to oneself or to others, for example, suicide and aggression. As it is, there is what can only be called a drug overdose epidemic across with United States in which, depending on the year that a person examines, there are about 70,000 deaths per year (U.S. Centers for Disease Control, 2018). This number is greater than the peak year of auto accident deaths {over 50,000 in 1972} and in the peak in H.I.V.-A.I. D.S. deaths {over 45,000 in 1995} (Katz & Sanger-Katz, 2018).

According to CBS News, drug overdoses are now the leading cause of death among Americans under 50 and the annual levels exceed the number of Americans who died in the entire Vietnam War (Reynolds, 2017). The overdose epidemic is so severe that it has caused the life expectancy of Americans to drop versus several years ago, for the first time since World War II (Katz & Sanger-Katz, 2018). McCarthy (2015) in the British Medical Journal notes that drug overdose has become the leading cause of death due to injury in the United States. Drug overdoses have become so common that police regularly administer naloxone/narcan in attempt to revive people, who are unconscious or semi-unconscious in attempt to save their lives (U.S. Government, 2017). What is clear is that whether it is accidental overdoses, acts of aggression or suicidal actions or thoughts, death and bodily harm either to others or oneself is so strongly associated with illegal drugs, that it is clear that the policies need to be propounded to address this problem.

Any full discussion of the effects of the use of illicit drugs must consider the initial causes of its use (Becker et al., 2004; Lyman, 2014; Scheier & Hanson, 2014). Without such a broad discussion, programs designed to halt the rise of illicit drug consumption will be limited in their success (Bennett & White, 2015; Binks et al., 2005). Nevertheless, if one only focuses on the causes without seriously examining the data on the widespread usage of illegal drugs overall (e.g., the extent to which youth have consumed some type of illegal drug whether it is marijuana, illegally obtained opioids), attempts at curbing the use of these substances will also fall short.

The gap between what the data show and what teens understand about the potential or even likely-consequences of the use of these drugs is of great concern (Bennett & White, 2015; Binks et al., 2005). A reassertion of the importance of drug education is necessary (Loconte, 2004; Ramsey & Morrison, 2010). In the final analysis, the results of the meta-analysis point to the need of a balanced approach to the discouraging drug addiction. On the one hand, the differences that often exist in the results between the effects of illegal drug usage when sophisticated controls are in place versus when they are not suggests that there are a number of causal agents at work that make substance addiction more likely. Unfortunately, with life many times comes various kinds of trauma that young people can experience such as parental marital dissolution, a break-up with a significant other, abuse from a family member or a student at school, gang violence, the death of a loved one, etc. These and other factors should be considered if one is to understand the causes of drug addiction and have greater compassion on those who ultimately partake (Jeynes, 2005; Ramsey & Morrison, 2010).

Effect Sizes for The Drugs Under Study Overall (Research Question #1)

The utilization of a meta-analysis on the effects of illegal drug use is very important in propelling the academic- and practical-examination of the use of illicit drugs to the next level. There are a few reasons for this. First, the results of meta-analysis are very difficult to explain away, especially when it includes such a large total sample of individuals (Kulinskaya et al., 2008). In this meta-analysis the samples totaled over 500,000 subjects. Moreover, many of the studies consisted of a large sample of randomly selected students over a nation or large geographic area.

By definition, a meta-analysis is a quantitative summary of the existing body of research. Localized studies with small sample sizes can possess results that may be subject to being explained away, because they are made up of a localized sample that is not representative of the general population. However, a meta-analysis of the type just described that statistically summarizes the existing body of research is one of the most highly reliable indicators of actual effect sizes (Greenhouse & Iyengar, 1994; Hedges & Vevea, 1998). Hence, given that this meta-analysis demonstrates nearly all of the 75 studies in the meta-analysis indicate a substantial negative relationship between consuming illegal drugs and student academic and behavioral-outcomes, it is more than clear that such a relationship exists (Greenhouse & Iyengar, 1994; Hedges & Vevea, 1998). Both the results and the ramifications need to be fully acknowledged.

Second, it is key that one understands that without one recognizing the primary results, society cannot proceed to the next step, which is developing a comprehensive plan to reduce the pre-adolescent, adolescent, and young adult-consumption of these drugs. Just what the best plan is can be touched on to some degree by a study such as this meta-analysis, but it cannot be fully elaborated on simply because what constitutes the best plan is not the focus of this meta-analysis. It is the case that historically speaking that when the U.S. government did a great deal to discourage illegal drug use, that the partaking of these drugs stabilized and even declined, Yes, that is true. However, making that observation does not point to all the specific solutions about what drug policy is best, but only suggests that widespread drug abuse education is likely part of the solution.

Effect Sizes for The Specific Drugs Under Study (Research Question #2)

It is also vital that scholars and educators appreciate the effects of individual drugs, including both the similarities and the differences between them. There is a general trend in the data that the effect sizes tend to be somewhat larger for student behavior than for their achievement. This is true for the effects of overall drug use, as well as for amphetamines, in which the overall effect sizes for student behavior approached .9 of a standard deviation. This is exceptionally large. If one were to ask most people their main concerns about the effects of illicit drugs on people, most would emphasize their tendency to influence behavior more than they would achievement. It is also likely that part of the reduced levels of achievement finds its roots in the altered behaviors and attitudes of the students (Jeynes, 2007; Schreiner et al., 2012).

Having said that most teachers, parents, friends, and acquaintances of a drug user are most concerned with changes in the user’s behavior, the long-term effects of reduced academic achievement can be quite substantial. To be sure, there is many a woman, especially, who may not partner up with a man who is an illegal drug user for fear of potential physical abuse (Brady et al., 2009; Finnegan & Kandall, 2015). That concern is real and should not be underestimated (Brady et al., 2009; Finnegan & Kandall, 2015). Nevertheless, it also true that the consequences of unnecessarily low academic achievement can reduce a student’s future prospects for success, advancement, and fulfillment, even if that individual overcomes addiction later in life. The scholastic achievement of students opens and closes doors for many years to come. Graduating from Harvard or Princeton will often open up opportunities decades after one’s graduation. Similarly, failing to graduate from high school can close doors of opportunity for long lengths of time. Moreover, to whatever extent taking illegal drugs increases the chances for getting in trouble with the police, consuming illegal drugs can create behavioral consequences that also have some long-term consequences.

The Effects for Educational- and Behavioral-Outcomes (Research Question #3)

There is some evidence that the effect sizes for consuming the drugs examined in this study were greater for student behavior than they were for academic outcomes. However, one should note that this pattern was only in place when the analyses were undertaken without sophisticated controls. When this relationship was examined for all the drugs combined, the effects for behavior were nearly twice the size as those that emerged for academic (.89 vs. 45).

However, when sophisticated controls were included in the analyses, the effects for than for behavioral measures were only somewhat greater than for academic measures (.64 vs. 59).

There are a number of different possibilities as to why the above findings turned out the way that they did. Some of the foremost explanations are included here. First, scholastic tests take place at certain specified times and major assignments are due on particular days. Students who engage in the use of illicit drugs may avoid taking large amounts of these substances, or perhaps may even abstain, when there are key test- or due-dates arising. In contrast, anti-social events and other misbehaviors can occur at any time. Second, the difference between behavioral and academic outcomes was especially large for marijuana, which is not as hard a drug as most of the drugs examined in this study. Therefore, the fact that pupils can plan for academic events may be especially salient in the case of marijuana, because its effects are usually not as long-lasting. The fact that there were more studies that examined marijuana than any other drug may have influenced the overall drug consumption results to some extent.

Third, the principle of Occam’s Razor encourages one to look for the simplest explanation may apply here, because it is often the most accurate. It may be that illegal drugs simply affect one’s emotional behavior, on average, more than a person’s intellect. Fourth, it may be because taking illegal drugs is an act focused on oneself rather than helping others that it only follows that the drugs would affect behavior and make one even more self-oriented even to the place of increasing the likelihood of anti-social and even violent behavior.

There are likely other contributing factors as well.

Differences by the Age of the Students (Research Question #4)

The results by age were likely some of the most important of the study. There is no question that the effects sizes for marijuana were substantially larger for older students than for younger students. The findings from this study support the notion that the effect sizes for illegal drugs are larger for older students, who likely have taken these substances for a larger period of time. The trend toward larger effect sizes for older students was noteworthy and undeniable. For college students the effects were .55 (.18, .92), p < .01, which was the largest of the three groups by a sizeable degree. Moreover, the effect sizes were three times the size of those for middle school students, .18 (.02, .34), p < .05.

Differences by the Race of the Student (Research Question #5)

The effects or taking illegal drugs for children of color were very similar to those that emerged for the overall sample, especially when models with no sophisticated controls were utilized. This builds confidence that the results of this study can be generalized to different racial groups and probably to other easily identifiable subgroups as well. When sophisticated controls were utilized, the effects of illegal drug use for children of color, especially African American and Latino students, was somewhat smaller than for the overall sample. This likely indicates that there were certain variables, such as level of parental education or other variables at work that helped account for the effects of these drugs on African American and Latino youth to a greater degree than in the case of their white (and in some cases Asian American) counterparts.

In a related supplementary analysis the effects for females was very similar to that for males, with the results for females yielding numerically higher effect sizes than was the case for males. However, the magnitude of these differences was not enough to be statistically significant.

Limitations of Study

The primary limitation of this meta-analysis, or any meta-analysis, is that it is restricted to analyzing the existing body of literature. A second limitation of a meta-analysis is that the social scientist is limited to addressing the same research questions addressed in the aggregated studies.

Further Conclusions

The Western world values freedom as one of its foremost values. Nevertheless, freedom has different manifestations. Many people think that freedom primarily means freedom of choice. However, total freedom of choice can produce bondage both in the life of the person doing the choosing and in the lives of others. If one thinks of freedom as freedom from bondage then that person will look upon freedom and the issue of addictive, often illegal, drugs very differently. The results of this study suggest that the issue of the consumption of cannabis, cocaine, heroin, and alcohol should be viewed in a broad context of considering its relationship to student academic achievement and behavior.

Implications for Public Policy

It should be noted that although this study does not directly test different public policies, it does have substantial implications for public policies. What is clear is that nations needs a drug policy that will decrease overall illicit drug use, because the evidence is so strong that there are severe consequences that are sadly associated with the use of these drugs.

Within the context mentioned above, the message that this study conveys is that wise public policy will seek to decrease the consumption of illegal drugs, rather than encourage its use. Just what policies are adopted to reduce the use of these drugs should fall within the parameters of what the results of this study indicate, but even within that purview the findings of this meta-analysis do not necessarily dictate one approach. Yes, these results suggest that given that the effect sizes were larger than the author and readers likely expected, drug education would certainly be important, as well as increased efforts at drug rehabilitation via programs such as Teen Challenge (Loconte, 2004). It is also clear that faith-based efforts have a much higher rate of success than others. To the extent that this is the case, wisdom would dictate that society should take these efforts seriously and they should be pursued (Becker et al., 2004; Loconte, 2004).

Final Thoughts

The results suggest that there needs to be a debate that, for the most part, is not taking place in America and in some other countries as well, who find the idea of legalizing some or all of the drugs examined in this study very attractive. That is, the United States has a large number of its inhabitants addicted to drugs and they need help in order to overcome their problem. Giving these individuals the help that they need will not only benefit them, but also those affected by their behavior both in their immediate family and beyond. Just how society decides to offer to this help is not the subject of this meta-analysis and is worthy of debate. However, in order to engage in this type of debate society, at large, needs to acknowledge drug addiction as the problem that it is.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

William H. Jeynes

Author Biography

William Jeynes, PhD, is Professor of Education at California State University, Long Beach. He graduated first in his class from Harvard University. He is also a Senior Fellow at the Witherspoon Institute in Princeton, New Jersey. He has served as a speaker and advisor for three presidential administrations, as well as for several G20 foreign governments, the UN, and the EU. He specializes in conducting statistical meta-analyses on a variety of public policy topics. Bill has more than 185 academic publications, including 17 books.

References

*Affia

A. M.

Garinger-Dykstra

L. M.

Caldeira

K. M.

Vincent

K. B.

Winick

E. R.

O’Grady

K. E.

(2013). Drug use patterns and continuous enrollment and college: Results from a longitudinal study. Journal of Studies on Alcohol and Drugs, 74(1), 71–83.

Anisman

(2015). Stress and your health: From vulnerability to resilience. Wiley-Blackwell.

*Anker

Milman

D. H.

Kahan

S. A.

Valenti

(1971). Drug usage and relate patterns of behavior in university students. I. General survey and marijuana use. Journal of the American College Health Association, 19(3), 178–186.

*Ansary

N. S.

Luthar

S. S.

(2009). Distress and academic achievement among adolescents of affluence: A study of externalizing problem behavior in school performance. Development and Psycholpathology, 21(1), 319–341.

*Arafat

Yorburg

(1973). Rug use and the sexual behavior of college Women. The Journal of Sex Research, 9(1), 21–29.

*Arria

A. M.

Caldeira

K. M.

Bugbee

B. A.

Vincent

K. B.

O’Grady

K. E.

(2015). The academic consequences of marijuana use during college. Psychology of Addictive Behaviors, 29(3), 564–575.

Beard

Marini

(2018). Validity of the SAT for predicting first year grades: 2013 SAT validity sample. College Board.

Becker

G. S.

Murphy

K. M.

Grossman

(2004). The economic theory of illegal goods: The case of drugs. National Bureau of Economic Research.

Belenko

S. R.

(2000). Drugs and drug policy in America: A documentary history. Greenwood Press.

10.

Bennett

W. J.

White

R. A.

(2015). Going to pot. . .why the rush to legalize marijuana puts America at risk. Center Street.

11.

Bernstein

B. L.

Lubinski

Benbow

(2019). Psychological constellations at age 13 predict distinct forms of eminence 35 years later. Psychological Science, 30(3), 444–454.

12.

Binks

Hoskins

Salmon

Benger

(2005). Prevalence and healthcare burden of illegal drug use among emergency department patients. Emergency Medicine Journal, 22(12), 82–83.

13.

Blavos

A. A.

Glassman

T. J.

Sheu

(2017). Marijuana and college students: A critical review of the literature. American Journal of Health Education, 48(3), 167–184.

14.

*Bolin

R. M.

Pate

McClintoch

(2017). The impact of marijuana use on academic achievement among college students. Social Science Journal, 54(4), 430–437.

15.

Brady

Back

S. E.

Greenfield

S. F.

(2009). Women and addiction: A comprehensive handbook. Guilford Press.

16.

Brick

Erickson

C. K.

(2013). Drugs, the brain, and behavior: The pharmacology of drug use disorders. Routledge.

17.

*Brill

N. Q.

(1982). Psychological effects of marijuana use. American Journal of Social Psychology, 3(1), 41–44.

18.

*Brook

J. S.

Stommel

M. A.

Chenshu

Brook

D. W.

(2008). The association between earlier marijuana use and subsequent health problems: A longitudinal study. American Journal of Addiction, 17(2), 155–160.

19.

Brown

(2000). The physical and psychosocial consequences of opioid Addiction: An overview of changes in opioid treatment. Australian and New Zealand Journal of Mental Health Nursing, 9(2), 65–74.

20.

Buchman

J. G.

O’Malley

Schulenberg

J. E.

(2007). The education-drug use connection: How the successes and failures in schools relate to adolescent smoking, drinking, drug use, and delinquency. New York: Taylor & Francis.

21.

*Buchanan

J. G.

Schulenberg

J. E.

(2007). The education-drug use connection: How successes and failures in school relate to adolescent, smoking, drinking, drug use, delinquency. Taylor & Francis.

22.

*Burge

Felts

Chenier

Parrillo

A. V.

(1995). Drug use, sexual activity, and suicidal behavior in U.S. high school students. Journal of School Health, 65(6), 222–227.

23.

*Castro

M. E.

Perez

M. A.

de la Serna

, et al (1989). Conducta antilsocially consume de divogas en adolescents espanoles. Revista Mexicana de Psciologia, 6(1), 27–34.

24.

*Chabrol

Chauchard

Girabet

(2008). Cannabis use and suicidal behaviors in High school students. Addictive Behaviors, 33(1), 152–155.

25.

*Chabrol

Rodgers

R. F.

Sobolewski

Van Leeuwen

(2010). Cannabis use and delinquent behaviors in a non-clinical sample of adolescents. Addictive Behaviors, 35(3), 263–265.

26.

*Chabrol

Saint-Martin

(2009). Cannabis use an delinquent behaviors in high school students. Addictive Behaviors, 34(2), 187–189.

27.

*Cox

R. G.

Zhang

Johnson

W. D.

Bender

D. R.

(2007). Academic performance and substance use. Findings from a survey of public school students. Journal of School Hath, 77(3), 109–115.

28.

*D’Amico

E. J.

Tucker

J. S.

Miles

J. N.

Ewing

B. A.

(2016). Alcohol and marijuana use trajectories in a diverse longitudinal sample of adolescents: Examining use patterns from 11 years. Addiction, 111(10), 1825–1835.

29.

*Dorr

Gorostegui

Viani

Dorr

M. P.

(2009). Adolescentes consumidores due marihuana: Implicaciones para la famillia y las escuela. Salud Mental, 32(4), 269–278.

30.

*Egbochuku

E. O.

Akerlele

J. O.

(2007). Stimulant use as correlate of abusive behavior among Nigerian undergraduates. College Student Journal, 41(1), 50–58.

31.

*Ehrenreich

Nahapetyan

Orpinas

Song

(2015). Marijuana use from middle to high school: Concurring problem behaviors, teacher-rated skills, and sixth-grade predictors. Journal of Youth & Adolescence, 44(10), 1929–1940.

32.

*Enns

Krebs

DeBeck

(2017). The costs of crime associated with stimulant use in a Canadian setting. Drug and Alcohol Dependence, 180(11), 304–310.

33.

*Ersche

Clark

London

Robbins

T. W.

DSahakian

B. J.

(2006). Profile of executive and memory function associated with amphetamine and opiate dependence. Neuropsychopharmacology, 31(5), 1036–1047.

34.

*Felts

W. M.

Chemier

Barnes

(1992). Drug use and suicide ideation among North Carolina public school students. American Journal of Public Health, 82(6), 870–872.

35.

*Finn

K. V.

(2012). Marijuana use at school and achievement-linked behaviors. High School Journal, 95(3), 3–13.

36.

Finnegan

L. P.

Kandall

S. R.

(2015). Women, children, and addiction. Routledge.

37.

*Finnell

W. S.

Jones

J. D.

(1975). Marijuana, alcohol, and academic performance. Journal of Drug Education, 5(1), 13–21.

38.

Glass

G. V.

McGaw

Smith

M. L.

(1981). Meta-analysis in social research. Sage.

39.

* Goode

(1971). Drug use and grades in college. Nature, 234(5326), 225–227.

40.

Greenhouse

J. B.

Iyengar

(1994). Sensitivity analysis and diagnosis. In Cooper

Iyengar

(Ed.), Handbook of research synthesis (pp. 383–398). Russell Sage Foundation.

41.

*Guerra

Sole

Cami

Tobena

(1987). Neuropsychological performance in opiate addicts after rapid detoxification. Drug and Alcohol Depdendence, 20(3), 261–270.

42.

*Guinn

(1977). The phenomenology of marijuana use among Mexican American youth. Journal of Psychodelic Drugs, 9(4), 3410–3343.

43.

Hall

Pacula

R. L.

(2004). Cannabis use and dependence: Public health and public policy. New York: Cambridge University Press.

44.

Harris

(2005). Opiates. Greenhaven Press.

45.

Hedges

L. V.

Vevea

J. L.

(1998). Fixed and random effects models in meta-analysis. Psychological Method, 3(4), 486–504.

46.

*Hernandis

S. P.

Diez

J. P.

Rouma

A. C.

(2002). El consume de inhallables y cannabis en la preadolescencia: Analis multivaridoe factores predisponentes. Anales de psicolgia, 18(1), 77–93.

47.

Hesterman

J. L.

(2013). The terrorist-criminal nexus: An alliance of international drug cartels, organized crime, and terror groups. Taylor & Francis.

48.

Higgins

S. T.

(1998). Applying behavioral economics to the challenge of reducing cocaine abuse. National Bureau of Economic Research.

49.

*Hill

Mrug

(2015). School-level correlates tobacco, alcohol, and marijuana use. Substance Use and Misuse, 50(12), 1518–1528.

50.

Hitchens

(2012). The war we never fought: The British establishment’s surrender to drugs. Continuum.

51.

*Hittner

J. B.

Kennington

L. E.

(2008). Normative perceptions, substance use, age of substance use, limitations, an gender as predictors of HIV-risky sexual behavior in a college student sample. Journal of Applied Biobehavioral Research, 13(2), 86–101.

52.

*Horwood

L. J.

Ferguson

D. M.

Hayatbakhsh

M. R.

Najam

J. M.

Coffey

Patton

G. C.

Silins

Hutchinson

D. M.

(2010). Cannabis use and educational achievement: Findings from three Australasian cohort studies. Drugs and Alcohol Dependence, 110(3), 247–253.

53.

Husak

D. N.

(2002). Legalize this!: The case for decriminalizing drugs. Verso.

54.

*Jeynes

(2002). The relationship between the consumption of various drugs by adolescents and their academic achievement. The American Journal of Drug and Alcohol Abuse, 28(1), 1–21.

55.

Jeynes

(2005). Effects of parental involvement and family structure on the academic achievement of adolescents. Marriage & Family Review, 37(3), 99–116.

56.

Jeynes

(2006). Adolescent religious commitment and their consumption of marijuana, cocaine, and alcohol. Journal of Health and Social Policy, 21(4), 1–20.

57.

Jeynes

(2007). American educational history: School, society and the common good. Thousand Oaks, CA: Sage.

58.

Jeynes

(2010). The salience of the subtle aspects of parental involvement and encouraging that involvement: Implications for school-based programs. Teachers College Record, 112(3), 747–774.

59.

Jeynes

(2011). Parental Involvement and Academic Success. New York: Taylor & Francis/Routledge.

60.

*Kalet

(2006). Recreational prescription drug use among college students. NADSPA Journal, 43(1), 19–39.

61.

* Kandel

D. B.

Davies

(1996). High school students who use crack and other drugs. Archives of General Psychiatry, 53(1), 71–80.

62.

Katz

Sanger-Katz

(2018, November 29). “The numbers are so staggering.” Overdose deaths set a record last year. New York Times. https://www.nytimes.com/interactive/2018/11/29/upshot/fentanyl-drug-overdose-deaths.html

63.

Kleiman

M. A.

Claukins

J. P.

Hawken

(2017). Drugs and drug policy. Oxford University Press.

64.

*Knight

K. N.

(2015). An analysis of the relationship between substance use selected demographics and academic achievement among middle school age youth in an urban midwestern county 76 (5-A) E, Kent State University in Kent, Ohio.

65.

Kobrin

J. L.

Patterson

B. F.

(2011). Contextual factors associated with the validity of SAT scores and high school GPA for predicting first-year college grades. Educational Assessment, 16(4), 207–226.

66.

Kulinskaya

Morgenthaler

Staudte

R. G.

(2008). Meta-analysis: A guide to calibrating an combining statistical analysis. John Wiley & Sons.

67.

Layton

Brown

(2012, September 24). SAT scores hit a four-decade low. Washington Post.

68.

Levinthal

C. F.

(2012). Drugs, behavior, and modern society. Allyn & Bacon.

69.

*Lin

L. A.

Walton

M. A.

Boner

E. E.

Blow

F. C.

(2016) Trajectories of nonmedical use of prescription opioids among adolescents in primary care. Addiction Research & Theory, 24(6), 514–520.

70.

*Lin

L. A.

Walton

M. A.

Boner

E. E.

Blow

F. C.

(2016) Trajectories of nonmedical use of prescription opioids among adolescents in primary.

71.

Loconte

(2004). Faith healing. Public Interest, 155, 89–104.

72.

*Lowry

Holtzman

Truman

B. T.

Kann

Collins

J. L.

Kolbe

L. J.

(1994). Substance use and HIV-related sexual behaviors among U.S. high school students: Are they related? American Journal of Public Health, 84(7), 1116–1120.

73.

Lyman

M. D.

(2014). Drugs in society: Causes, concepts, and control. Routledge.

74.

*Maggs

J. L.

Staff

Kloska

D. D.

Patrick

M. E.

O’Malley

P. M.

Schulenberg

(2015). Predicting young adult degree attainment by late adolescent Marijuana use. Journal of Adolescent Health, 57(2), 205–211.

75.

*Maloney

Degenhardt

Darke

Mattick

R. P.

Nelson

E. A.

(2007). Suicidal behavior and associated risk factors among opioid-dependent individuals: A case control study. Addiction, 102(12), 1933–1941.

76.

*Marie

Zolita

(2017). High achievers? Cannabis access and academic performance. Review of Economic Studies, 84(3), 1210–1237.

77.

*Martins

S. S.

Alexandre

P. K.

(2009). The association of ecstasy use and academic achievement among adolescents in two U.S. national surveys. Addictive Behaviors, 34(1), 9–16.

78.

McCarthy

(2015). Drug overdose has become the leading cause of death from injury in the US. British Medical Journal, 350, 3328.

79.

*Meier

M. H.

Hill

M. L.

Small

P. J.

Luther

S. S.

(2015). Associations of adolescent cannabis use with academic performance and mental health: A longitudinal study of upper middle class youth. Drug and Alcohol Dependence, 156(1), 207–212.

80.

*Melchior

Bolze

Fombonne

Surkan

P. J.

Pryor

Jauffret-Roustide

(2017). Early cannabis initiation and educational attainment: Is the association causal? International Journal of Epidemiology, 46(5), 1641–1650.

81.

*Mintzer

M. Z.

Copersino

M. L.

Stitzer

M. L.

(2005). Opioid abuse and cognitive performance. Drug and Alcohol Dependence, 78(1), 225–230.

82.

*Miozzo

Dalberto

E. R.

da Silver

D. X.

(2013). Consumo de substancias pscioativas emu ma amostra de adolescents e sua relacoio com o comportamento sexual. Journal Brasileiro de Psiquatria, 62(2), 93.

83.

*Miron

Serrano

Godas

, et al (1997). Ecstasy and “academic achievement. Analiss y modificacion de conducta, 23(88), 255–282.

84.

*Myers

C. E.

Rego

Haber

Morley

Beck

K. D.

Hogarth

Moustafa

A. A.

(2017). Learning and generalization from reward and punishment in opioid addiction. Behavioral Brain Research, 317(1), 122–131.

85.

National Center on Addiction & Substance Abuse, Columbia University. (2011). Adolescent substance use: America’s #1 public health problem. Columbia University.

86.

*Ndete

D. M.

Khasakhala

Mutiso

Ongecha-Owuor

F. A.

Kokonya

D. A.

(2010). Drug use in a rural secondary School in Kenya. Substance Abuse, 31(3), 170–173.

87.

Orford

(2013). Power, powerlessness, and addiction. Cambridge University Press.

88.

*Ornstein

T. J.

Iddon

J. L.

Baldacchino

A. M.

Sahakian

B. J.

London

Everitt

B. J.

Robbins

T. W.

(2000). Profiles of cognitive dysfunction in chronic amphetamine and heroin abuses. Neuropsychopharmacology, 23(2), 113–126.

89.

*Pardini

White

Xiong

Bechtold

Chung

Loeber

Hipwell

(2015). Unfazed or dazed and confused: Does early adolescent marijuana use cause sustained impairments in attention and Academic functioning. Journal of Abnormal Psychology, 43(7), 1203–1217.

90.

*Patrick

M. E.

Schulenberg

J. E.

O’Malley

P. M.

(2016). High school substance use as a predictor of college attendance completion and drop out: A national multicohort longitudinal study. Youth and Society, 48(3), 425–447.

91.

*Patte

Quan

Leatherdale

S. T.

(2017). Marijuana as predictors of academic achievement: A longitudinal analysis among youth in the COMPASS study. Journal of School Health, 87(5), 310–318.

92.

*Phillips

K. T.

Phillips

M. M.

Lalonde

T. L.

Tomochien

K. N.

(2015). Marijuana use, craving, and academic motivation, and performance among College students: An in-the-moment study. Addictive Behaviors, 47, 42–47.

93.

*Pilowsky

D. W.

Burchett

Blazer

Ling

(2011). Depressive symptoms, substance use, and HIV-related high risk behaviors among opioid-dependent individuals: Results from the clinical trials network. Substance Use & Misuse, 46(14), 1716–1725.

94.

Pollan

(2001). The botany of desire: A plant’s eye view of the world. Random House.

95.

Ramsey

N. A.

Morrison

C. R.

(2010). Youth violence and juvenile justice: Causes, intervention, an treatment programs. Nova Science Publishers.

96.

*Register

C. A.

Williams

D. K.

Grimes

P. W.

(2001). Adolescent drug use and educational attainment. Education Economics, 9(1), 1–18.

97.

Reynolds

(2017, June 6). Overdoses now leading cause of death for Americans under 50. CBS News. https://www.cbsnews.com/news/overdoses-are-leading-cause-of-death-americans-under-50/

98.

Ritchie

(2016). All that matters. Hodder & Stoughton.

99.

Rogers

(2011). Substance use and abuse. Rosen Educational Services.

100.

Rosenthal

M. S.

(2019). Drugs: Mind, body, and society. Oxford University Press.

101.

*Roy

Reynolds

Finlayson

P. F.

(1990). Adolescent marijuana use: Risk factors and implications. Australian and New Zealand Journal of Psychiatry, 24(1), 47–56.

102.

Scheier

L. M.

Hanson

W. B.

(2014). Parenting and teen drug use: The most recent findings from research, prevention, and treatment. Oxford University Press.

103.

*Schepis

T. S.

Teter

C. J.

McCabes

(2018). Prescribing drug use, misuse, and related substance use disorder symptoms vary by educational status and attainment in U.S. adolescents and young adults. Drug and Alcohol Dependence, 189, 172–177.

104.

Schreiner

L. A.

Louis

M. C.

Nelson

D. D.

(2012). Thriving in transitions: A research-based approach to college student success. University of South Carolina.

105.

*Scivoletto

Tsuji

R. K.

Najjar

Carmito , (2002). Use of psychoactive substances and sexual risk behavior in adolescents. Substance Use and Misuse, 37(3), 381–398.

106.

*Shewan

Dalgarno

(2005). Evidence for controlled heroin use? Low levels of negative health and social outcomes among non-treatment heroin users in Glasgow (Scotland). British Journal of Health Psychology, 10(1), 33–48.

107.

*Silins

Ferguson

D. M.

Patton

G. C.

Harwood

L. J.

(2015). Adolescent Substance use and educational attainment: An interactive data analysis comparing cannabis and alcohol from three Australian cohorts. Drug and Alcohol Dependence, 156, 90–96.

108.

*Simon

W. E.

(1974). Psychological needs, academic achievement, and marijuana Consumption. Journal of Clinical Psychology, 30(4), 496–498.

109.

*Springer

Peters

Shegog

White

D. L.

Kelder

S. H.

(2007). Methamphetamine use and sexual risk behaviors in U.S. high school: Findings from a National Risk Behavior Survey. Prevention Science, 8(2), 103–113.

110.

*Strote

Lee

J. E.

Wechsler

(2002). Increasing MDMA use among college students. Journal of Adolescent Health, 30(1), 64–72.

111.

*Suerken

C. K.

Reboussin

B. A.

Egan

K. L.

Sutfin

E. L.

Wagoner

K. G.

Spangler

Wolfson

(2016). Marijuana use trajectories and academic outcomes among college students. Drug and Alcohol Dependence, 162, 137–145.

112.

*Tarter

R. E.

Mizzich

A. C.

Hsieh

Parks

S. M.

(1995). Cognitive capacity in female adolescent substance abusers. Drug and Alcohol Dependence, 39(1), 15–21.

113.

Tolomeo

Matthews

Steele

Baldacchino

(2018). Compulsivity in opioid dependence. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 81(2), 333–339.

114.

U.S. Centers for Disease Control. (2018). Drug overdose deaths. U.S. Department of Health & Human Services. https://www.cdc.gov/drugoverdose/data/statedeaths.html

115.

U.S. Government. (2017). Opioid epidemic. U.S. Government Printing Office.

116.

*Verweij

Huizinmk

A. C.

Agnawal

Martin

N. G.

Lynskey

M. T.

(2013). Is the relationship between early-onset cannabis use and educational attainment causal or due to common liability. Drug and Alcohol Dependence, 133(2), 580–586.

117.

*Vest

Reynolds

C. J.

Tragresser

S. L.

(2016). Impulsity and risk for prescription opioid misuse in a chronic patient sample. Addictive Behaviors, 60, 184–190.

118.

Wai

Lubinski

Benbow

(2005). Creativity and occupational accomplishments among intellectually precocious youths: An age 13 to 33 longitudinal study. Journal of Educational Psychology, 97(30), 484–492.

119.

*Walsh

E. P.

Frantz

T. T.

(1972). The relationship of marijuana smoking to LSD, sex, age, and grades: Setting the rationale for use. Journal of College Student Personnel, 13(6), 518–521.

120.

*Wang

Deng

Zhou

Huang

Gao

(2014). A nonmedical use of prescription medicines among high school students: A cross-sectional study in southern China. Drug and Alcohol Dependence, 141(1), 9–15.

121.

*Williams

J. H.

Davis

L. E.

Johnson

S. D.

Williams

T. R.

Sanders

J. A.

Nebbitt

V. E.

(2007). Substance use and academic performance among African American high school Students. Social Work Research, 31(3), 151–161.

122.

Wirtz

(1977). On further examination. College Entrance Examination Board.

123.

*Wong

Zhou

Goebert

Hishinuma

E. S.

(2013). The risk of adolescent suicide across patterns of drug use: A nationally representative study of High school students in the United States from 1999-2009. Social Psychiatry and Psychiatric Epidemiology, 48(10), 1611–1620.

124.

*Zimmerman

M. A.

Schmeek-Cone

K. H.

(2003). A longitudinal analysis of adolescent substance use and school motivation among African American youth. Journal of Research on Adolescence, 13(2), 185–210.

125.

*Zule

W.A.

Oramasionwu

Evon

Hino

Doherty

I. A.

Bobashev

G. V.

Wechsberg

W. M.

(2016). Event-level analyses of sex-risk and injection-risk behaviors among nonmedical prescription opioid users. American Journal of Drug and Alcohol Abuse, 42(6), 689–697.