Keystone Light: Dartmouths Beer of Choice

1
Keystone Light Dartmouths Beer of Choice

• Kathy Birchall, Anne Abbott, and Andrea Abeita
• Math 5
• Winter 2004

2
Say What?

• Benjamin Franklin once said, Beer is proof that
  God loves us and wants us to be happy.
• How does this apply to Dartmouth students?

3
FYI

• Dartmouths social scene is dominated by Greek
  houses, instead of bars and clubs
• Kegs have been forbidden by the College, so the
  Greek houses have been forced to find another
  cheap alternative to supply beer to the students
  their choice has been Keystone Light 30-packs
• All three members of our group are Greek-
  affiliated seniors and have noticed the abundance
  of this beer on campus

4
Goals

• To determine whether or not the constant presence
  of Keystone Light at the Greek houses makes it
  more recognizable to senior student taste testers
  that belong to a Greek organization
• How much beer are we drinking? If the majority
  of students can distinguish Keystone Light from
  Amstel Light (a similar beer), what does this say
  about the rate of student beer consumption?
• Are we drinking too much? Or just products of
  our environment? How much does being a part of
  the Greek system affect beer tasting ability?

5
Our Pre-Test Beliefs

• A majority of Greek affiliated seniors tested
  will be able to taste a difference between the
  beers.
• More precisely, we believe that the majority of
  those who can taste a difference will identify
  the Keystone Light.
• Even more precisely, we believe that at least 80
  of the testers will correctly identify the
  Keystone Light

6
Null Hypothesis

• Greek affiliated seniors that drink beer
  regularly and can taste a difference between the
  beers will not be able to recognize Keystone as
  opposed to the other light beer supplied.
• If p is the percent that can recognize it, then
  the null p 0.5

7
Alternate Hypothesis

• The majority of Greek affiliated seniors will be
  able to correctly identify Keystone
• More precisely, p gt 80 (use to calculate the
  power)

8
Keystone Light VS. Amstel Light
9
The Test

• Testers drank from cups A and B, both of equal
  color, foam, and temperature without knowing
  which beer was Keystone Light, or that the other
  beer was Amstel Light
• They were asked if they could tell a difference.
  If they could, then they were asked which one was
  Keystone

10
Bulls Eye!!

• Most of the students tested CAN IDENTIFY KEYSTONE
  LIGHT!!!

11
Results

• 94.1 (32/34) students drank beer frequently
• 96.9 (31/32) could detect a difference between
  Keystone and Amstel Light
• 87.1 (27/31) correctly identified the Keystone
  Light

12
Type One- Alpha Error

• Leaving a 5 significance level, or chance of a
  Type One (Alpha) Error, P 0.65 (the critical
  region)
• But we got 0.87!!!
• Therefore, we have less than a 5 chance of a
  type one error!

13
POWER!!

• 99.9 Power!!
• The chance of a Type 2 (Beta) error is therefore
  only 0.001

14
We Must Accept Our Alternate Hypothesis!!!

• P 87.1
• We hit our mark with room to spare!
• We believe this test is statistically significant
  and is an accurate reflection of what we tested.

15
Contributing Factors

• The requirements to be an ED (experienced
  drinker) limited our pool of test subjects to
  Greek- affiliated seniors
• Due to Winter Carnival (RAGE!) we were not able
  to find enough testers over the weekend so we had
  to move it to Monday night, which is when Secret
  Societies hold their meetings.
• Being the day after Carnival, some students were
  less than thrilled to drink any more alcohol-
  even a sip!!

16
Contributing factors continued

• Consequently, many usually enthusiastic
  drinkers were unable/unwilling to participate.
• We kept the beer samples properly chilled until
  the test was performed, unlike basement
  situations where the beer temperature tends to
  fluctuate (sometimes it is consumed when it is
  cold, other times it has been sitting on a pong
  table for hours and has grown warm).

17
Contributing factors continued

• We could only afford to buy one other type of
  beer. If we had been able to run the test using
  either a different type of beer or had more than
  one other alternative sample, our results may
  have been different
• Is Amstel Light too different than Keystone?? Is
  being light not enough??

18
Someone turned 21!

• We held the test in Andreas room in KDE
• Someone turned 21 on Monday night and people were
  celebrating!
• Many of our subjects were taking part in the
  birthday bash
• The fact that they may have already had tossed
  back a few could have inhibited their ability to
  detect Keystone...

19
No worries though!!

• The majority, 87.1, still correctly identified
  Keystone!
• Dartmouth students must really spend some quality
  time with this beer if they can spot it so
  easily!

20
Directions for Future Research

• Ideally another test would include the following
• Expand the number of sample beers to four or five
  to ensure a greater variety and would decrease
  the chance of lucky guesses
• More time to recruit more subjects
• Not be performed directly post-Winter Carnival
  weekend

21
Credits

• Thanks to all the participants and Professor
  Leibon and Professor Rudel for their online
  examples and support
• RAGE ON DARTMOUTH!!