Abstract
Introduction
Alcohol use
Recent studies have found that nearly 80% of college students consume alcohol to some degree during their college career. 1 NSDUH data found that within the 60% of full-time college students reporting drinking within the past month, 39% also reported binge drinking during that timeframe.1,2 Of particular concern is the increased number of drinks and higher frequencies of drinking male college students report compared to female students.3,4 Specifically, male college students tend to consume an average of nine drinks per week, compared to four for female students, and they tend to drink on an average of 9.9 out of 30 days, compared to women who reported drinking on an average of 6.8 out of 30 days. 5 A similar study found that 1-in-5 male college students reported consuming alcohol on campus, compared to 1-in-10 female students. 6 Increases in alcohol consumption are often established early in a student’s first year of college and maintained across their tenure. 7 Pressure to initiate heavy drinking patterns stems from various drinking motives (eg, emotional enhancement or avoidance, social facilitation, sensation seeking, perceived norms 8 ). Despite a host of research examining the consequences of frequent alcohol consumption including poorer life satisfaction, college students continue to drink at rates higher than the general population.9,10
Aggression
Also of significant concern is the high rate of aggression perpetration on college campuses. A large national survey found that in 2014, 27 000 criminal incidents occurred on college campuses within the U.S.; including, 17,441 forcible sexual offenses, 2,100 aggravated assaults, and 13,500 robberies. 11 With respect to physical violence within the context of a dating relationship, up to one fifth of college students have experienced at least one incident of physical violence perpetrated by their dating partner.12,13 Further, research has shown upwards of 75% of college students reported being both perpetrators and victims of some form of psychological violence within, and outside, the context of a dating relationship.14–16 Experiencing aggression, regardless of the context in which it occurs, increases the likelihood of suffering from additional psychological, social, and academic problems.17,18 For instance, being the victim of aggression is significantly related to poorer personal-emotional adjustment to college, 19 as well as poorer performance with regard to current academic achievement (GPA) and cumulative academic achievement (CGPA). 20 Victimization is also associated with increased alcohol use, 21 drug use, 21 mood disorders, 22 and somatic complaints. 23
The relationship between alcohol and aggression
The association between drinking and aggressive behavior has long been recognized as a significant problem among college students.24,25 Hingson and colleagues reported that at least 646,000 students are assaulted each year by another college student who has consumed alcohol and that approximately 97,000 students experience date-rape or sexual assault related to alcohol use. 26 In addition, upwards of 100,000 students ages 18-24 reported having been too intoxicated to know whether or not they consented to sexual activity. 26 In a national survey conducted by the Core Institute, researchers concluded that 43% of college students have experienced some threat of physical violence while consuming alcohol. 27 Further, substance use has been found to be common among both the victim and the perpetrator. 28 Physical victimization is associated with physical violence perpetration for males, however, for women, it is more strongly related to sexual victimization. 28 Higher frequencies of alcohol consumption and drinking larger amounts during each drinking episode are associated with violence for women only. 13 Because the prevalence of alcohol use, aggression, and the co-occurrence of both behaviors is a significant problem, it is vital to identify and more fully understand the variables that affect these relationships.
Theoretical models of the relationship between alcohol use and aggression suggest that acute alcohol intoxication interacts with other situational and dispositional variables to increase the risk of aggression. A theoretical model developed by Finkel (I3), posits that all risk factors of violence promote aggression through one of three processes: instigation, impellance, and inhibition. 29 According to this model, instigating factors are behaviors that trigger aggression (e.g., provocation), and impelling factors are dispositional or situational triggers that lead an individual to experience the need to aggress (e.g., dispositional aggressiveness). These instigating and impelling factors collectively constitute an individual’s readiness to aggress and interact such that when both are strong, an individual is most likely to be aggressive. 29 Contrarily, inhibiting factors are dispositional or situational factors that increase a person’s ability to overcome their urge to be violent (e.g., executive control). As such, violence only occurs when the strength of impelling and instigating factors is stronger than the inhibiting factors that prevent violence, making it vital to understand the relevant impelling, instigating, and inhibiting factors associated with violence. 29 This model is consistent with older and more researched biopsychosocial models which posit that alcohol use contributes to aggressive perpetration through its association with both distal and proximal variables, subsequently increasing the risk of violence when in the context of a conflict-oriented interaction.30,31 Distal factors related to violence are generally enduring and include childhood aggression, history of aggressive behavior, family history of substance use, individual substance use patterns, gender role expectations, norms regarding aggression, relationship dissatisfaction, and psychopathology. Proximal factors are more immediate conditions within a conflict that may increase risk for violence including intoxication, information processing deficits, and location/setting of the conflict.
An impelling variable that has received some attention in the literature as impacting the relationship between alcohol use and violence is an individual’s history of prior aggression, often studied via a person’s dispositional aggression, anger, or hostility. 32 For instance, in a series of studies, Giancola consistently found that violence perpetration in men high in aggressiveness increased with alcohol consumption compared to those low in dispositional aggression.33,34 Using a lab paradigm, Bailey and Taylor found that men high in dispositional aggression were more likely to become aggressive when provoked and while intoxicated compared to intoxicated men low in dispositional aggression. 35 Research showing a definitive link between past violent behavior, alcohol use patterns, and continued aggression is lacking, especially for college students. Thus, it may be that a more complete explanation requires examination of both intraindividual and interpersonal factors to better understand and prevent interpersonal aggression.36–38
The frequency and amount of alcohol consumption may be important factors to investigate in order to more fully understand the relationship between past aggression and continued violence perpetration. In a population study, Room and Rossow found that as alcohol use increases in a given country, so do rates of violence perpetration. 37 Furthermore, longitudinal studies show that heavy drinkers are more likely to be aggressive than light to moderate drinkers,39,40 and that higher frequency of intoxication is associated with increased rates of violence for those receiving substance use treatment. 41 In a study of the temporal association between alcohol use and aggression, Parks and colleagues found that the odds of experiencing aggression (perpetration and victimization combined) were 2% to 20% higher on heavy drinking days. 42 Results from these studies support a growing body of research suggesting that frequency and quantity of alcohol use may be causally linked to aggressive behavior. Results also show that this relationship is not always direct, and that impelling factors such as past aggressive behavior may also play an important role in continued violence perpetration. However, extant research has not examined how varying degrees of past aggression interact with varying degrees of past alcohol use concomitantly in maintaining or increasing rates of future aggression, particularly in a college student population. Therefore, the present study addresses this gap by using a highly controlled, lab-based paradigm to examine the association between alcohol use patterns, past violence, and continued aggressive behavior in male college students. To test our hypotheses, we dichotomized past aggression and past alcohol use into “low” and “high” levels. We hypothesize that aggression in the lab will be highest for participants who reported a history of “high” past aggression as well as “high” levels of past alcohol use. In contrast, we hypothesize that for those with a history of “low” past aggression, “high” past alcohol use will not have an effect on aggression on the Point Subtraction Aggression Paradigm (PSAP). We also hypothesize that for participants who have a history of “high” aggression but “low” past alcohol use, provocation in the lab will result in higher rates of aggressive responding than for participants with histories of “low” past alcohol use and “low” past aggression. We also examine whether these hypotheses are true for individuals who respond aggressively on the PSAP before (proactive aggression) or after (reactive aggression) being provoked by the fictitious competitor. We hypothesize that those with a history of “high” aggression and “high” past alcohol use will be more likely to aggress proactively than those with a history of “high” aggression but “low” alcohol use; and that for those who reported “low” levels of past aggression, alcohol will not have an effect on aggression in the lab. We also hypothesize that there will not be any significant effects for those who only aggressed reactively on the PSAP. In demonstrating the interactive effect of past aggression and alcohol use using a laboratory design, this study provides a valuable heuristic for future research to explore other combinations of impelling and instigating factors and their role in the perpetration of violence.
Methods
Participants
Participants were 160 male undergraduate college students aged 18-24 (
Demographic information for the overall sample.
Measures
Alcohol Use Disorders identification Test (AUDIT)
The AUDIT is a 10-item self-report measure that was used to assess alcohol use and related problems in the past 6 months. 43 A total score is computed by summing the scores ranging from 0 to 4 for each item. Scores of 8 or more are indicative of hazardous alcohol consumption and were used to define “high” alcohol use; scores of 7 or below defined “low” alcohol use. The AUDIT has been found to be valid and reliable.44,45 The present study yielded good internal consistency (α = .87).
Buss Perry Aggression Questionnaire (BPAQ)
The BPAQ is a 29-item measure of aggression in which participants rated various statements on a 5-point Likert-scale ranging from 1 (
Point Subtraction Aggression Paradigm (PSAP)
The PSAP is a laboratory based behavioral measure of aggression using a 25-minute computer-based task in which the participant ostensibly competes against another participant to earn money. 49 This operant task uses a 3-button response panel; button A earns money (reinforcement), button B subtracts money from the opponent (i.e., index of aggression), and button C protects from monetary subtractions by the opponent (escape). Every 100 presses of the A button earns the participant 10 cents (Fixed Reinforcement (FR)-100 schedule); for 10 presses of the B-button, the participant subtracts 10 cents from the opponent (FR-10 schedule); and for every 10 times the C button is pressed, the participant briefly protects their own earnings (FR-10 schedule). B-button responding is the dependent measure of aggression, an operationalized definition that is congruent with the definition of aggression proposed by Baron and Richardson, 50 in that the behavior is directed toward hurting the other person by taking money away from them. Subtraction of money from the participant’s total at random intervals by the computer served as the provocation. The computer provocation time intervals were randomized at the beginning of the study and were held consistent for each participant.
Using the PSAP as a measure of lab-based aggression has several advantages. First, the PSAP examines aggression in the context of a current situation, rather than employing the use of retrospective self-reports of aggression, which research has shown to be unreliable since many individuals tend to underreport their own aggression. 51 In addition, the PSAP offers participants two ways to respond in a non-aggressive manner (i.e., by earning money, and by protecting their earnings). As a criticism of the PSAP, Tedeschi and Quigley stated that it is difficult to assess whether the operationalized definition of aggression in the PSAP paradigm is in fact measuring aggression as opposed to competitive strategy because there is no measure of participants’ intentions or motivations. 52 Furthermore, prior research demonstrates the PSAP’s external validity by highlighting its ability to differentiate violent from nonviolent men, with those who have a history of violence exhibiting increased aggressive responding. 49 It has also been shown to have strong convergent validity with self-report measures assessing recent aggression and is associated with both past aggression and self-reported alcohol use.53,54 Taken together, this research lends support to the study of aggression under highly controlled conditions, and indicates that the PSAP is a valid measure of lab-based aggression.
Procedure
Institutional Review Board approval was obtained through the University prior to initiation of data collection. Eligible participants completed an initial consent, the AUDIT, and BPAQ online using surveymonkey.com, a secure online survey website. The first page of the survey displayed the consent form and presented participants an opportunity to click a button providing initial consent or declining consent and ending the survey. Individuals that consented then completed the AUDIT and BPAQ. Following the completion of online questionnaires, participants were invited via email to attend “Phase 2” of the experiment in the laboratory where they provided written consent. Upon arrival, each participant was introduced to a male or female confederate, whom they were led to believe was their opponent, and told that they would be competing in a button-pressing task in order to earn money; in actuality, money is regulated by the computer program which acts in the same previously-randomized manner regardless of “opponent” behavior. The task lasted for approximately 25 minutes. Participants were debriefed at the conclusion of Phase 2 and compensated for their time with $5 and extra course credit in their general Psychology course. No actual PSAP monies were earned or dispensed in Phase 2.
Results
Self-reported alcohol use and past aggression main effects
Effects of past alcohol use on individuals with histories of “High” aggression
Regression analyses were used to examine the main effects of self-reported alcohol use patterns (AUDIT) and self-reported past aggressive behaviors (BPAQ) on aggressive (B-button) responding on the PSAP. Consistent with our hypotheses, results revealed that individuals with a history of “high” past aggression were more likely to push the ‘B-button’ during the procedure, reflecting an aggressive response (regardless of whether this action was reactive or proactive to provocation), if they also reported a history of “high” alcohol use. This finding was primarily driven by a significant interaction between alcohol use patterns and the

The effects of history of aggression and current alcohol use on continued aggression for the overall sample.
Effects of past alcohol use on individuals with histories of “Low” aggression
Results of the main effect for individuals reporting “low” levels of past physical aggression showed that alcohol use habits did not significantly impact aggressive responding on the PSAP (
Effects of past aggression on “Low” versus “High” past alcohol use
Teasing apart the main effects revealed that the association between “low” versus “high” past alcohol use, aggression, and aggressive responding on the PSAP was significant for those with “low” past alcohol use (
Reactive versus proactive aggression
We also analyzed how the levels of past alcohol use and past aggression resulted in proactive aggression (i.e., pressing the B-button any number of times prior to being provoked), or reactive aggression (i.e., pressing the B-button any number of times after being provoked). Individuals who did not press the B-button were excluded from analyses examining proactive and reactive aggression. The analysis revealed that the above findings held true only for individuals who aggressed proactively on the PSAP (

The effects of history of aggression and current alcohol use on continued aggression for individuals who are proactively aggressive.
Discussion
This study furthers the body of literature regarding the impact of past aggression and alcohol use patterns on aggressive behavior by using a highly controlled, lab-based paradigm to examine the real-time association between alcohol use history, past aggression, and continued aggressive behavior in male college students. Based on Finkle’s I3 model of instigating, impelling, and inhibiting processes, 29 we hypothesized that 1.) aggression in the lab will be highest for participants who reported a history of “high” past aggression as well as “high” levels of past alcohol use; 2.) that for those with a history of “low” past aggression, “high” past alcohol use will not have an effect on aggression on the PSAP; 3.) that for participants who have a history of “high” aggression but “low” past alcohol use, provocation in the lab will result in higher rates of aggressive responding than for participants with histories of “low” past alcohol use and “low” past aggression; and 4.) that the findings will hold true for individual who aggressed proactively but not reactively on the PSAP.
Analyzing the entire college sample revealed that the interaction of alcohol use habits and past
A surprising finding that warrants additional study is that those with a history of “low” past alcohol use and “low” past aggression responded
Analyzing only those individuals who aggressed prior to being provoked revealed that individuals reporting both a history of “high” aggressive behavior and a history of “high” alcohol use were most likely to behave aggressively on the PSAP. However, these findings were not significant for individuals who aggressed reactively. According to Finkel’s I3 model of aggression, instigating factors are situational or contextual experiences leading to aggressive responding, whereas impelling factors are dispositional factors that predispose individuals to behave aggressively in response to instigation.57,29 Given past research linking provocation with subsequent aggressive behavior, 58 provocation thus instigates aggressive responding regardless of preexisting impelling forces and in spite of inhibiting forces. However, it may be that those individuals who responded aggressively prior to provocation differ in some way compared to those who aggressed in response to provocation. Researchers have identified a myriad of risk factors (ie, antisocial traits, temperament, aggressive norms) that serve as impelling forces, linked to both aggression and alcohol use independently and predictively.54,59,60 Though other potential correlates of alcohol and aggression were not assessed in the current study, these findings suggest that past aggression and alcohol use habits appear to be uniquely associated with unprovoked aggression.
Several strengths of the present study suggest that these findings make a substantial contribution to the empirical understanding of the relationship between alcohol use patterns and aggression in college students. Much of the extant research in this area has examined the association between alcohol use and intimate partner violence, as well as male-to-female perpetrated aggression.61,62 However, it is unclear if the results of such studies generalize to aggression more broadly. Importantly, our findings speak to the impact of alcohol use and previous aggressive behaviors on general aggression within a college sample. Additionally, research indicates that individuals underreport perpetration of aggression via self-report measures and that statistical adjustments utilizing participants’ self-reported social desirability cannot attenuate the impact of underreporting on violence related outcomes.61–63 As such, the use of laboratory based aggression paradigms, such as the PSAP, are especially important to help elucidate patterns of aggressive behavior that may otherwise be obscured by social desirability effects.
Limitations and recommendations
Prior research has demonstrated the construct validity of the PSAP as a measure of aggression, yet its ecological validity and predictive power remain unclear. Although individuals with a history of violence and alcohol use behave more aggressively in the lab, it is unclear whether these individuals will aggress in the future and against whom. Furthermore, the present research used a single self-report measure of alcohol use in the past 6 months. Prior research indicates that individuals may underreport their own alcohol use behaviors particularly when they are asked to report about use over long periods of time.64,65 Thus replications of these findings would benefit from the use of ecological, collateral, and interviewer facilitated assessments of alcohol use (eg, Timeline Followback66,67). Additionally, the relatively small sample size used in the study remains a limitation. Future studies would benefit from using a larger and more diverse sample size in order to increase the generalizability of the findings. Finally, the present study focused on alcohol use habits and aggression history as impelling factors for aggressive behavior in college populations. Finkel’s I3 model could be further validated by the simultaneous evaluation of the relationship between impelling, instigating, and inhibitory factors on aggression. For instance, future research utilizing the PSAP paradigm could examine instigating factors such as in-vivo alcohol administration, or negative mood induction, and inhibitory factors such as inducing negative alcohol use expectancies. Additionally, prospective data could elucidate the degree to which the interaction of these factors predicts risk for violence in naturalistic settings.
Conclusion
These findings have important clinical implications. Collegiate alcohol use and violence intervention and prevention programs may prove more effective when they incorporate information about an individual’s histories of violence and alcohol use. Research also suggests that the impact and cost-effectiveness of reducing college student drinking can be maximized when individuals are matched to interventions based on the severity of their alcohol use and related problems. 68 The current findings suggest that a person’s violence history interacts with alcohol use habits to produce risk for future violence. Thus, college student drinking interventions, particularly those aimed at harm reduction, may be able to reduce the incidence of drinking-related violence by assigning students to treatment according to both their alcohol use and violence histories.
