Fitzpatrick, BG, R Scribner, AS Ackleh, J Rasul, GM Jacquez, N Simonsen, R Rommel. 2012.
“Forecasting the Effect of the Amethyst Initiative on College Drinking”. Alcoholism Clinical
and Experimental Research (impact factor: 3.34). 03/2012; 36(9):1608-13. DOI:10.1111/j.1530-
Title: Forecasting the Effect of the Amethyst Initiative on College Drinking
Authors: Ben G. Fitzpatrick PhD1,2 , Richard Scribner MD, MPH3 , Azmy S. Ackleh PhD4,
Jawaid Rasul, PhD5, Geoffrey Jacquez PhD5,6 Neal Simonsen PhD3, Robert Rommel PhD5
1Loyola Marymount University, 1 LMU Drive, Los Angeles, CA 90045
2Tempest Technologies, 8939 S. Sepulveda Blvd, Suite 506, Los Angeles, CA 90045 15
3Louisiana State University School of Public Health, 1615 Poydras St., Room 1535, New
Orleans, LA 70112
4University of Louisiana at Lafayette, P.O. Box 41010, Lafayette, LA 70504-1010
5Biomedware Incorporated, 3526 W. Liberty, Suite 100, Ann Arbor, MI 48103
6The University of Michigan School of Public Health, 1415 Washington Heights, Ann Arbor, MI 20
Corresponding Author : Ben G. Fitzpatrick PhD, Loyola Marymount University, 1 LMU
Drive, Los Angeles, CA 90045, 310.338.7892 (voice), 310.338.3768 (fax), email@example.com
Support : This research was supported by National Institute on Alcohol Abuse and Alcoholism 25 grant RO1 AA015573.
Background: A number of college presidents have endorsed the Amethyst Initiative, a call to
consider lowering the minimum legal drinking age (MLDA). Our objective is to forecast the 30
effect of the Amethyst Initiative on college drinking.
Methods: A system model of college drinking simulates MLDA changes through (1) a decrease
in heavy episodic drinking (HED) due to the lower likelihood of students drinking in
unsupervised settings where they model irresponsible drinking (misperception), and (2) an
increase in overall drinking among currently underage students due to increased social 35
availability of alcohol (wetness).
Results: For the proportion of HEDs on campus, effects of large decreases in misperception of
responsible drinking behavior were more than offset by modest increases in wetness.
Conclusions: For the effect of lowering the MLDA, it appears that increases in social
availability of alcohol have a stronger impact on drinking behavior than decreases in 40
Keywords: Amethyst Initiative, Modeling, College, Heavy Episodic Drinking, Misperception
College drinking is one of the most significant and complex public health problems 45
today. Heavy drinking among college students remains a pervasive problem that places students
at considerable risk for a variety of negative outcomes, including date rape, academic problems,
traffic accidents, and health problems ( Hingson, Heeren, Winter, and Wechsler, 2005; Wechsler
and Nelson, 2008). Alcohol use is embedded in the college lifestyle, resulting in enormous
social, economic, and health consequences among some of the nation’s finest students (Task 50
Force on College Drinking, 2002). Heavy episodic drinking is generally conducted in private,
among peers, and college students engage in the behavior in much higher proportions than do
other young adults (Schulenberg, et al., 2001; Carey, Scott-Sheldon, Carey, and DeMartini,
2007; Substance Abuse and Mental Health Services Administration, 2006; Timberlake, et al.,
Interventions to reduce the negative outcomes associated with college drinking have
been mixed. For example, reeducation programs targeting misperception of drinking norms
remain a popular intervention. Social norms researchers have found that college students
routinely misperceive the level of alcohol use among their peers (Baer, Stacy, and Larimer,
1991; Perkins, Haines, and Rice, 1991) and liberal perceptions of social norms for peer 60
drinking are consistently shown to be strong predictors of alcohol use among college
students (Baer, Stacy, and Larimer, 1991; Perkins, Haines, and Rice, 1991; Reis and Riley,
2000; Baer and Carney, 1993, Perkins, et al., 1999; Babor, Aguirre-Molina, Marlatt, and
Clayton, 1999; Thombs, Wolcott, and Farkash, 1997). Rather than address the population level
factors that lead to misperception of social norms in the first place, Social Norms Marketing 65
(SNM) interventions attempt to reeducate students by correcting their misperceptions of their
peers’ behavior in hopes of changing individual behavior. The results of these interventions
have been equivocal (Thombs, Wolcott, and Farkash, 1997; Haines and Spear, 1996; Werch, et
al., 2000; Wechsler, et al., 2003; Toomey, Lenk, and Wagenaar, 2007; DeJong, et al., 2006).
The limited effectiveness of college drinking interventions has led to calls to 70
reexamine the MLDA. One of these, called the Amethyst Initiative, asks chancellors and
presidents of universities and colleges across the country to sign on to a call asking elected
officials to revisit the 21 year old drinking age. So far over 100 chancellors and presidents
have signed on. The Amethyst Initiative statement argues that the 21 year old drinking
age is not working. Underage students are legally prohibited from purchasing and 75
possessing alcohol and the majority continue to drink. In addition, the statement argues the
rampant flaunting of the drinking laws by students has led to a “culture of dangerous binge
drinking” on many campuses. John McCardell, the original author of the initiative, has
described the mechanism by which this culture has evolved. He argues that because of the
21 MLDA underage students are precluded from drinking in supervised settings (e.g., bars, 80
school sponsored parties). As a result, they are more likely to drink in unsupervised settings
(e.g., off campus parties) where there are fewer constraints on excessive drinking. He
concludes that in these settings underage students who are new to drinking develop
misperceptions of the normal drinking behavior and binge drinking tends to be viewed as
normal (McCardell, 2008). This argument is consistent with literature that demonstrates 85
misperception of drinking norms predicts individual drinking (Baer, Stacy, and Larimar,
1991; Perkins, Haines, and Rice, 2005). Social norms theory argues that the effect of
misperceptions is rooted in a psychological attribution process in which the individual tends
to perceive the drinking actions of others as reflective of their individual drinking
temperament and align their behavior accordingly (Perkins, 1997; Prentice and Miller, 1993), 90
proponents suggest that the MLDA of 21 years of age is the problem. Critics of the
Amethyst Initiative argue that lowering the MLDA will increase the availability of alcohol
for both social and home consumption, and therefore increase drinking among the entire
population with disastrous results (Babor, 2008).
Lowering the current MLDA represents an enormous social experiment with potentially 95
major consequences. While there is considerable evidence indicating the harms associated with
lowering the MLDA with regard to the general population (Wagenaar and Toomey, 2002), there
is little in the way of observational evidence (Kypri, et al., 2006) to either support or oppose the
specific hypotheses regarding student drinking behaviors embedded in the Amethyst initiative.
A systems approach is one method of providing a forecast for the effects of a policy change prior 100
to carrying out interventions based on that policy. Public health researchers are beginning to see
the opportunities of moving from a purely inferential approach of experimental design and data
analysis to a more mechanistic, systems approach (Homer and Hirsch, 2006). In particular, the
social and economic cost of suboptimal policy decisions can potentially be mitigated by an
increased understanding of the potential consequences that a systems model can provide. 105
Materials and Methods
We have developed a systems model, referred to as SimHED (Ackleh, et al., 2009;
Scribner, et al., 2009), to simulate a college campus student population structured by drinking 110
behavior and drinking age. The model, a continuous dynamical systems compartmental model
derived using epidemiological reasoning, is provided in detail in the Appendix. The
compartmental structure involves two levels of structure in age, namely underage (U) and legal
age (L), as well as four drinking styles associated with college drinking, namely abstainers (1)
social drinkers (2), problem drinkers (3), and heavy episodic drinkers or HEDs (4). Abstainers 115
are defined as individuals who do not drink. Social drinkers are individuals who drink more
frequently than abstainers but do not belong to the other two compartments. Problem drinkers
are those who report at least two out of four indicators of problem drinking based on the CAGE
instrument (Mayfield, McLeod, and Hall, 1974), regardless of the amount or frequency of their
drinking. Heavy episodic drinkers are individuals who consume more than five drinks in a single 120
sitting on at least one occasion in the past two weeks. In this manner, the state of the system at
any given time is defined by 8 numbers,
43214321 ,,,,,,, LLLLUUUU
, which are the number of
individuals in each of the four drinking style compartments for the underage drinkers
4321 ,,, UUUU
) followed by the corresponding numbers for the legal age drinkers
4321 ,,, LLLL
Individuals move from underage to legal age purely by aging. We assume that the aging
process itself does not directly change the drinking style of the individual, so that transition from
the underage to legal age compartments preserves the drinking style. The drinking style
transition model includes three types of parameters that control transfers between the four
drinking style compartments: individual risk (Wechsler, Dowdall, Davenport, and Castillo, 1995, 130
Presley, Meilman, and Leichliter, 2002), social interactions (Reifman, Watson, and McCourt,
2002; McCabe, et al., 2005), and social norm misperception (Perkins, et al., 1999; Borsari and
The individual risk model handles transitions that depend only on individual factors (e.g.,
mood, developmental transitions, individual traits). This component of the system involves a 135
fraction of individuals with a particular drinking style transitioning to a different drinking style
over a period of time. These transitions are modeled by terms of the form
(in which N
represents U or L, r models the fraction of those individuals transitioning, and the subscript ij
represents the transition out of drinking style i into drinking style j). Since the movement is from
, this term is a positive term in the
equation and a negative term in the
Social interaction transitions depend on individuals from two separate groups coming
into contact with one another, much like an epidemiological model of disease transmission. That
is, an individual with a certain type of drinking style convinces another individual with a
different drinking style to change behavior. We model these transitions by terms of the
, proportional to the number of pairings available among the two different drinking 145
styles. In this case, the movement may be in either direction, i.e., from
or vice versa. So,
represents a net movement between the two compartments i and j, and thus
Social norms/misperception transitions occur due to perception of the level of a particular
drinking behavior. These movements occur in two situations: when abstainers become light 150
drinkers because they perceive an exaggeratedly large number of drinkers on campus, and when
social drinkers become HEDs because they perceive an exaggeratedly large number of HEDs.
Situation a) is modeled by the term
and situation b) is modeled by the
can be either 2 or 3. Again, N represents either U or L
here, an underage (U) or legal age (L) drinking style, as the changes in drinking style are 155
assumed to occur within an age group. The function M is the misperception function, modeling
how badly the students overestimate the fraction of individuals undertaking the role model
behavior (either the fraction of drinkers or the fraction of HEDs). Were M the identity function,
student perception would be entirely accurate. Research in social norms suggests (Reis and
Riley, 2000) that misperception is greatest when the model behavior is least prevalent and that 160
misperception decreases as the model behavior increases. For example, in an analysis of the
National College Health Assessment, 59.9% of students overestimate the drinking norms of their
peers at parties by three or more drinks on campuses where abstinence is the norm. Where six
drinks is the actual norm, 31.5% of students overestimate by three or more drinks (Perkins,
Haines, and Rice, 2005). For these reasons we have chosen the functional form 165
whose graph is illustrated in Figure 1.
Figure 1 is to be inserted here.
Figure 1. Example misperception function.
controls the level of misperception: as
the amount of
overestimation goes to 0, and M becomes the identity function. We use the term 170
“hyperparameter” here to distinguish from the basic transition rate parameters in the model and
to emphasize that those rate parameters depend on
The campus alcohol environment (i.e., level of campus wetness) is an additional
hyperparameter, w, which modifies the transfer rates between compartments as a function of
campus wetness. Each rate parameter
ijijij nsr ,,
depends linearly on the wetness. For example, 175
so that a completely “dry” campus, w=0, has rate parameter
completely “wet” campus, w=1, has rate parameter
A version of the model without age structure has been successfully calibrated to survey
data obtained from 32 campuses across the United States in the Social Norms Marketing
Research Project, SNMRP (DeJong, et al., 2006; Ackleh, et al., 2009; Scribner, et al., 2009). 180
We have reasonable estimates of the “wet” and “dry” rate parameter values, the wetness
levels for each of the 32 campuses, and the misperception levels. Table 1 contains parameter
estimates for the wet and dry rate parameters.
Value at w=0
Value at w=1
Table 1. Bounds for the rate parameters as estimated from SNMRP data.
The wetness hyperparameters range from 0.05 to 0.55 in the 32 schools from the SNMRP
data. Wetness correlates (
) with alcohol outlet density, a common measure of
availability, but clearly other phenomena are involved, such as enforcement and campus social 190
environment. However, lowering the MLDA would dramatically affect the effective alcohol
outlet density for underage students. We estimated the misperception hyperparameters for the
SNMRP campuses, and these values range from nearly 0 to 0.25.
We are interested in the exploration of the interplay between misperception and
availability. In particular, the Amethyst Initiative’s hypothesis is that a reduction in 195
misperception and attendant improper role model choices will lead to a reduction in HED
behavior. Toward that end, we consider a hypothetical college population behavior averaged
over a ten year period, and we conduct a series of experiments. We assume that the wet and dry
rate parameter values are the same for the under age and legal age groups; however, we take the
legal age group to have higher wetness (greater availability) and lower misperception levels. We 200
conducted a series of computer simulation experiments that we report here.
We begin our exploration by considering a campus with medium wetness of 0.30,
applicable to the legal age students, who also have a small amount of misperception (
In order to examine the effect of misperception, we consider a number of simulated “treatment” 205
scenarios in which we assume a range of effects on the wetness and misperception parameters
for the underage population and observe the resulting drinking behavior.
Our first simulation involves the assumption that the wetness hyperparameter for the
underage population is the same as for the legal age (w=0.30), but that the misperception is at a
higher level (
=0.25). Our treatment is assumed to reduce the misperception from 0.25 down 210
to 0.05, the level of the legal age students, with 100% treatment implementation (Figure 2 upper
panel). The x-axis in the upper panel Figure 2 denotes this linear change of misperception,
where 0% implementation corresponds to
=0.25 and 100% to
=0.05. We can see that the
reduction in Heavy Episodic Drinkers is relatively small. We should note that it is not clear that
the legal age population would actually have less misperception than the underage population; 215
however, such an assumption leads to a slightly conservative estimate on heavy episodic
A different approach might be to change the wetness parameter for the underage students.
Our second simulation, in the lower panel of Figure 2, shows the fraction of HEDs in the legal 220
and underage compartments as we reduce wetness in the underage group from 0.30 (0 percent
treatment on the x-axis ) down to 0.00 (100 percent treatment). Again, the legal age students
have a small misperception parameter (
=0.05) while the misperception of the underage
students is at a high level (
=0.25). One can see in the lower panel of Figure 2 that this
treatment has a much stronger impact on the HED fraction than does the reduction of 225
Figure 2 is to be inserted here.
Figure 2. HED fraction as a function of simple one-variable treatments. Upper panel
shows the effect of reducing misperception from 0.25 to 0.05. Lower panel shows the effect
of lowering wetness from 0.30 to 0.00. 230
Neither of these simulations captures the actual effect of the Amethyst Initiative’s MLDA
reduction. One might expect that wetness for the underage population is somewhat less than it is
for the legal age population (exactly how much is of course a difficult matter to resolve) while
the amount of misperception is greater for the underage population. We have conducted a
number of simulations in which wetness and misperception for the underage population are 235
changed simultaneously. As a first illustrative example, we continue with the hypothetical
campus having wetness of 0.30 and misperception of 0.05 for the legal age students. We have
also simulated a wet campus (w=0.55). We assume in this example that the underage population
has a wetness parameter of half that of the legal age population, and a misperception parameter
of 0.25, while the misperception parameter is at the high end of those we have inferred from 240
SNMRP data. The treatment is to increase the underage wetness to that of the legal age group,
while reducing the misperception to 0.05, so that at the end of the treatment, the legal age and
underage students have the same parameters.
In Figure 3, we show the effect of simultaneously increasing the wetness and decreasing 245
the misperception in the underage population. In each of the simulated experiments (both
moderate and wet), the underage population goes from half the wetness of the legal population to
fully as wet, while simultaneously going from high misperception (0.25) to the low level of the
legal age population (0.05). In each of the three panels, the qualitative trend is the same: HED
drinking among underage students is increased. 250
Figure 3 is to be inserted here.
Figure 3. HED fractions as a function of changing underage wetness and
misperception. The upper panel shows a campus with moderate wetness (w=0.30) among
the legal age students; the lower panel shows a wet campus.
With any computer simulation model of a real-world phenomenon, but most especially 255
with the highly challenging modeling problems of social systems, the prediction of events that
are out of the scope of observation must be viewed with some skepticism. We have endeavored
to calibrate and validate our model as accurately as possible (Ackleh, et al., 2009), and we have
also conducted a large number of simulation studies to examine the dependence on assumptions
about the legal age population, dropout and recruitment rates, and other parameters. Such 260
sensitivity analyses of the model to its parameters are a necessary step for developing confidence
in the resulting predictions. We have found across a wide spectrum of such analyses that the
structure of the basic findings presented here is remarkably consistent.
It is of course a risky exercise to attempt to predict possible behavior based on inferred
parameterizations. Rather than viewing these results as quantitative predictions of the actual
levels of drinking that will occur under an MLDA change, we prefer to interpret the model
output in terms of the trends and joint behavior as wetness index and misperception are changed
simultaneously. The inescapable conclusion is that the misperception must be very great among 270
the underage population and very significantly reduced by allowing the underage population to
drink in order to compensate for the increased availability of drinking venues to the underage
population. Indeed, further studies are required to quantify with accuracy how these parameters
might actually change in the presence of an MLDA reduction. We do, however, interpret our
simulations to date as pessimistic for the Amethyst Initiative’s proposal that an MLDA reduction 275
will have beneficial consequences for college drinking.
The preliminary insights provided by this model suggest that a reduction in the MLDA
may not produce the desired reduction in heavy episodic drinking that is the goal of the
Amethyst Initiative’s strategy, based on the Initiative’s reasoning about why “21 is not 280
working.”(20). The analysis we have conducted suggests that effects of a reduction in
misperception from the largest observed values to the lowest is overcome by a 25 percent
increase in campus wetness. One might expect a much larger increase in wetness from the
increased physical availability of alcohol associated with making the entire college population,
rather than approximately half, to be of legal drinking age. 285
The eight-compartment set of equations is given by
Ackleh, A.S., Fitzpatrick, B.G., Scribner, R., Simonsen, N., Thibodeaux, J., 2009. Ecosystem
Models of College Drinking: Parameter Estimation and Comparing Models to Data. 290
Mathematical and Computer Modelling , 50, pp. 481-97.
Babor, T., 2008. Amethyst initiative versus animal house: What college presidents can learn
from research on the Mediterranean drinking style. American Public Health Association Annual
Meeting; San Diego, CA, USA, October 25-29, 2008.
Babor, T.F., Aguirre-Molina, M., Marlatt, G.A., Clayton, R., 1999. Managing alcohol problems 295
and risky drinking. American Journal of Health Promotion, 14(2), pp. 98-103.
Baer, J.S., Carney, M.M., 1993. Biases in the perceptions of the consequences of alcohol use
among college students. Journal of Studies on Alcohol and Drugs, 54(1), pp. 54-60.
Baer, J.S., Stacy, A., Larimer, M., 1991. Biases in the perception of drinking norms among
college students. Journal of Studies on Alcohol and Drugs, 52(6), pp. 580-6. 300
Borsari, B., Carey, K.B., 2001. Peer influences on college drinking: a review of the research.
Journal of Substance Abuse,13(4), pp. 391-424.
Carey, K.B., Scott-Sheldon, L.A., Carey, M.P., DeMartini, K.S., 2007. Individual-level
interventions to reduce college student drinking: a meta-analytic review. Addictive Behaviors,
32(11), pp. 2469-94. 305
DeJong, W., Schneider, S.K., Towvim, L.G., Murphy, M.J., Doerr, E.E., Simonsen, N.R.,
Mason, K., Scribner, R. A., 2006. A multisite randomized trial of social norms marketing
campaigns to reduce college student drinking. Journal of Studies on Alcohol and Drugs, 67(6),
Haines, M., Spear, S.F., 1996. Changing the perception of the norm: a strategy to decrease binge 310
drinking among college students. Journal of American College Health, 45(3), pp. 134-40.
Hingson, R., Heeren, T., Winte, M., Wechsler, H., 2005. Magnitude of alcohol-related mortality
and morbidity among U.S. college students ages 18-24: changes from 1998 to 2001. Annual
Review of Public Health, 26, pp. 259-79.
Homer, J.B., Hirsch, G.B., 2006. System dynamics modeling for public health: background and 315
opportunities. American Journal of Public Health, 96(3), pp. 452-8.
Kypri, K., Voas, R.B., Langley, J.D., Stephenson, S.C.R., Begg, D.J., Tippetts, A.S., & Davie,
G.S., 2006. Minimum purchasing age for alcohol and traffic crash injuries among 15- to 19-
year-olds in New Zealand. American Journal of Public Health, 96(1), 126-131.
Mayfield, D., McLeod, G., Hall, P., 1974. The CAGE questionnaire: validation of a new 320
alcoholism screening instrument. American Journal of Psychiatry, 131(10), pp. 1121-3.
McCabe, S.E., Schulenberg, J.E., Johnston, L.D., O'Malley, P.M., Bachman, J.G., Kloska, D.D.,
2005. Selection and socialization effects of fraternities and sororities on US college student
substance use: a multi-cohort national longitudinal study. Addiction,100(4), pp. 512-24.
McCardell, J., 2008. The drinking age. The status quo has bombed. US News and World Report, 325
145(6) p. 11.
Perkins, H.W., 1997. College student misperceptions of alcohol and other drug norms among
peers; explaining causes, consequences and implications for prevention programs. In: The
Higher Education Center for Alcohol and Other Drug Prevention, Designing Alcohol and Other
Drug Prevention Programs in Higher Education, Newton: US Department of Education. 330
Perkins, H.W., Haines, M.P., Rice, R., 2005.Misperceiving the college drinking norm and related
problems: a nationwide study of exposure to prevention information, perceived norms and
student alcohol misuse. Journal of Studies on Alcohol and Drugs, 66(4), pp. 470-8.
Perkins, H.W., Meilman, P.W., Leichliter, J.S., Cashin, J.R., Presley, C.A., 1999. Misperceptions
of the norms for the frequency of alcohol and other drug use on college campuses. Journal of 335
American College Health, 47(6), pp. 253-8.
Prentice, D.A., Miller, D.T., 1993. Pluralistic ignorance and alcohol use on campus: some
consequences of misperceiving the social norm. Journal of Personality and Social Psychology,
64(2), pp. 243-56.
Presley, C.A., Meilman, P.W., Leichliter, J.S., 2002. College factors that influence drinking. 340
Journal of Studies on Alcohol Supplement (14), pp. 82-90.
Reifman, A., Watson, W.K., McCourt, A., 2006. Social networks and college drinking: probing
processes of social influence and selection. Personality and Social Psychology Bulletin, 32(6),
Reis, J., Riley, W.L., 2000. Predictors of college students' alcohol consumption: implications for 345
student education. The Journal of Genetic Psychology, 161(3), pp. 282-91.
Schulenberg, J., Maggs, J.L., Long, S.W., Sher, K.J., Gotham, H.J., Baer, J. S., Kivlahan, D. R.,
Marlatt, G. A., Zucker, R. A., 2001. The problem of college drinking: insights from a
developmental perspective, Alcoholism: Clinical and Experimental Research, 25(3), pp. 473-7.
Scribner, R., Ackleh, A., Fitzpatrick. B., Rommel. R., Jacquez, G., Simonsen, N., 2009. A 350
Systems Approach to College Drinking: Development of a Deterministic Model for Testing
Alcohol Control Policies. Journal of Studies on Alcohol and Drugs, 70(5), pp. 805-21.
Substance Abuse and Mental Health Services Administration, 2006. Results from the 2005
national Survey on Drug Use and Health: National Findings. (Office of Applied Studies,
NSDUH Series H-30, DHHS Publication No. SMA 06-4194) Rockville: DHHS. 355
Task Force on College Drinking, 2002. Call to Action: Changing the Culture of College
Drinking. Bethesda: National Institute of Alcohol Abuse and Alcoholism.
Timberlake, D.S., Hopfer, C.J., Rhee, S.H., Friedman, N.P., Haberstick, B.C., Lessem J.M.,
Hewitt, J.K., 2007. College attendance and its effect on drinking behaviors in a longitudinal
study of adolescents. Alcoholism: Clinical and Experimental Research, 31(6), pp. 1020-30. 360
Thombs, D.L., Wolcott, B.J., Farkash, L.G., 1997. Social context, perceived norms and drinking
behavior in young people. Journal of Substance Abuse, 9, pp. 257-67.
Toomey, T.L., Lenk, K.M., Wagenaar, A.C., 2007. Environmental policies to reduce college
drinking: an update of research findings. Journal of Studies on Alcohol and Drugs, 68(2), pp.
Wagenaar, A. C., and Toomey, T.L., 2002. Effects of Minimum Drinking Age Laws: Review
and Analyses of the Literature from 1960 to 2000, Journal of Studies on Alcohol and Drugs
Supplement (14), pp. 206-225.
Wechsler, H., Dowdall, G.W., Davenport, A., Castillo, S., 1995. Correlates of college student
binge drinking. American Journal of Public Health, 85(7), pp. 921-6. 370
Wechsler, H., Nelson, T.F., 2008. What we have learned from the Harvard School Of Public
Health College Alcohol Study: focusing attention on college student alcohol consumption and
the environmental conditions that promote it. Journal of Studies on Alcohol and Drugs, 69(4),
Wechsler, H., Nelson, T.E., Lee, J.E., Seibring, M., Lewis, C., Keeling, R.P., 2003. Perception 375
and reality: A national evaluation of social norms marketing interventions to reduce college
students' heavy alcohol use. Journal of Studies on Alcohol and Drugs, 64(4), pp. 484-94.
Werch, C.E., Pappas, D.M., Carlson, J.M., DiClemente, C.C., Chally, P.S., Sinder, J.A., 2000.
Results of a social norm intervention to prevent binge drinking among first-year residential
college students. Journal of American College Health, 49(2), pp. 85-92. 380