Drinking Habits of Freshmen 17.871 Group Project

1
Drinking Habits of Freshmen17.871 Group Project
2
Question

• When freshmen live off campus, do they drink more
  than when they live on campus?

3
Why is this important?

• Influence University policy/government laws
  regulations
• Impacts student life on college campuses
• At MIT all freshmen must live on campus!

4
Alternate Explanations

• Past experience High School Parents Drinking
  Habits
• Living Environment Substance free-housing,
  single sex housing, roommates

5
Background

• Henry Wechsler Principal Researcher, Harvard
  School of Public Health
• College Alcohol Studies (CAS) every 3-4 years
• 1993 study prompted examination of heavy-episodic
  binge drinking.
• 90s alcohol abuse classified 1 public health
  problem facing college students by Surgeon
  General and CDC

6
Statistics

• 1 in 3 college students use of alcohol qualified
  for formal diagnosis of alcohol abuse
• Continued attempts in 1990s?Little change
• Efforts aimed at educational aspects no
  environmental changes (thought to be major cause
  of alcohol issues)

7
Method Data

• Random sample
• Relevant variables?STATA datasets to analyze
• Recoded and dummy variables created for
  comparison across datasets

8
Method Answering the Question

• Variables on and off for freshmen place of
  residence
• Measure for total drinks consumed in a 30-day
  period (total drinks drinking occasions x
  drinks per occasion)
• Compared percentage of on-campus freshmen who
  drank to the percentage of total alcohol those
  on-campus freshmen drank

9
Survey Questions

• How many occasions in the last 30 days have you
  had a drink?
• How many drinks, on average, did you drink each
  time?
• Combining the questions (problematic)

10
The data we have total_drinks histogram
On Campus Off Campus Mean 43 Mean
34 SD 68 SD 72
11
Occasions
Mean (occasions/past 30 days) 7.29 6.41
On Campus
Off Campus
Mean (average drinks/occasion) 4.95 4.30
Number of drinks
12
Occasions
Mean (occasions/past 30 days) 7.09 7.07
Alcohol allowed in students housing
No alcohol allowed
Mean (average drinks/occasion) 4.75 4.94
Number of drinks
13
Occasions
Mean (occasions/past 30 days) 6.67 9.45
Non-Greek
Greek
Mean (average drinks/occasion) 4.68 5.47
Number of drinks
14
Method Rival Explanations

• For Greek membership alcohol-free housing, we
  repeated the same analysis procedure
• ?Generated binary dataset for (X, not X)
  Xpossible explanatory variable
• ?Created table of (X, not X) total drinks
  consumed in the past 30 days
• ?Compared percentage of respondents who were X
  with percentage of total alcohol drunk by
  respondents who were X

15
Who consumes the alcohol?
Percent respondents Percent total drinks Difference
On 77.40 80.83 - 3.43
Off 22.60 19.17 - 3.43
Alcohol-free 70.29 69.37 - 0.92
Alcohol allowed 29.71 30.63 - 0.92
Non-Greek 85.10 78.69 - 6.40
Greek 14.90 21.31 - 6.40
Total alcohol consumed by the students surveyed
in 30 days previous to survey 3951 respondents
drank 163,058 drinks (average 41
drinks/person) Note Among those who
drank/answered both questions
16
Results

• As far as we can tell, when freshmen live on
  campus, they are no less thirsty than those who
  live off campus.
• This is true when we look at number of times they
  drink
• And number of drinks they have on each occasion.

17
But

• There may be unavoidable shortcomings, either in
  our methodology, or in the data.

• Do we look at total drinks?

• Or do we look at measures of drunkenness?

• We chose to attempt to quantify the number.
• This assumes an even distribution of sex
• And a normal distribution of weights.

18
I said heywhats going on?

• Perhaps people who drank in high school drink
  more in college.
• Engineering students may drink more because of
  chronic depression.
• Students at easy schools drink more.

• Data for high school drinking was lost
• Data for majors was not taken across all surveys
• Schools were not identified.

19
So why should schools care?

• Measures to force freshmen to live on campus may
  not be as productive
• Providing substance free housing may be a waste
• Because the most significant finding we had was

20
Further thought and research

• We have identified several distinct categories of
  drinkers which may be of interest to college
  policymakers.
• Greek/Non-Greek histogram
• A measure of what constitutes harmful behavior
  needs to be determined.
• Is binge drinking frequently the problem?
• Is binge drinking infrequently the problem?
• Are Greek organizations really to blame?

21
Questions?
22
Occasions
Non-Greek
Greek
Number of drinks