Previous Research and Theories

• B.F.Skinner: Law of Effect where behavior which is reinforced tends to be repeated

  • Where both negative and positive reinforcements will impact the way behaviors are repeated.

  • Where behavior that is not reinforced will be extinguish

• Murphy 06: Relative Reinforcing Efficacy of Alcohol Among College Student Drinkers

  • relative reinforcing efficacy is a theoretical model of substance use

  • Where substance use has high reinforcement-- and behaviors are altered to match resources to ensure reinforcement value

  • Where easy access (low price) has higher reinforcement

• Hallgreen 16: Psychometric Properties of the Adolescent Reinforcement Survey Schedule – Alcohol Use Version with College Student Drinkers

  • Where the ARSS-AUV where a reinforcement value to alcohol-- meant to predict use of alcohol based on daily activities

• Brown-Rice 15: Differences in College Greek Members’ Binge Drinking Behaviors: A Dry/Wet House Comparison

  • Where greek students report higher use and problems with alcohol use

  • Where problematic drinking behaviors were scored lower for greek members that do not drink than those who do

• Labrie 07:Reasons for Drinking in the College Student Context: The Differential Role and Risk of the Social Motivator Reasons for Drinking in the College Student Context: The Differential Role and Risk of the Social Motivator

  • Previous research suggested coping mechanisms influences drinking

  • Social interactions will be a positive reinforcement

  • Found that social camaraderie is more likely to influence higher levels of drinking than that of coping mechanisms

  • Higher prediction in females

Interaction Rules

• Non-drinkers conformity level to drink is influenced by number of drinkers it is linked to.

• A non- drinkers dependency to drink is based on its influence from others .

  • Where:

    • peer-avg mean ([drinking] of link-neighbors)

let my-drinking [drinking] of self

set drinking ((my-drinking * (1 - trr)) + (peer-avg * trr))

• Where trr is a random number between 0 to 1 representing a percentage, and drinking is a random number from 0 to 10

Simulation Plans

• After hitting “go once,” individuals with a drinking level lower than 2 will “die” or otherwise leave the group, because they will likely not be influenced by drinkers.

• During the simulation, individuals with a drinking level greater than 8 will form new networks, because those who are drinkers tend to be more social.

• As the simulation runs, individuals will converge toward the middle of the drinking spectrum, putting most individuals around a 5 or 6 drinking level.

Implications & Future Directions

These findings can help show the spread of drinking culture based on the people one is surrounded with. This simulation and model results would help to explain why people in college are more likely to drink, even if they did not before college. If you have people around you that are drinking regularly, you are more likely to find reinforcement similar to your peers drinking. An intervention that could be done is a seminar discussing these results which could influence who incoming college students become friends with. This model is limited due to programer usage, by being able to code larger scale and more communities it would allow for a higher probabilistic outcome in favor of the hypothesis.

Future directions to this model include adding turtles who trr does not change if they have a high initial trr, adding reinforcement subscale levels that show what activities are more reinforcing than others, adding specific turtle to have a higher influencing rate than others, and finally by making some greek life alcohol-free communities and alcohol-use to show how drinking spreads in these communities.