Dining Services: who, where, and why

1
Dining Services who, where, and why

• Jason Bash
• Terri-Anne Segovia
• Linda Lomelino
• Patrick Doherty

2
Mission Statement

• Everyday, Dartmouth students travel to their
  favorite dining halls for their meals. There are
  various factors that could affect their
  preferences, in this survey we address three
  factors that could affect dining hall preference.
• Purpose to test and conclude whether or not
  there exists a dinning hall preference amongst
  different sexes, housing locations on campus, and
  classes.
• To test if dinning hall preference is independent
  of sex, location, and preference.

3
Dining Locations

• Food Court
• Homeplate
• Collis
• Lone Pine Tavern
• Hop

4
Dining Map
5
Dining Halls Collis, Thayer, Hop
6
Seating Capacity

• Collis 168 seats
• Food Court 300 seats
• Homeplate 242 seats
• Lone Pine Tavern 72 seats

7
Housing

• According to Dartmouth Statistics, the college
  houses a little over 2,700 students at a time,
  organized into 8 areas
• 85 of the student body lives on campus while 15
  live off campus

8
How housing on campus is clustered and organized

• Butterfield/Russell Sage/Gold Coast 174 beds
• Choates and North Hall 294 beds
• East Wheelock (Andres McCulloch Morton Zimmerman)
  314 beds
• Fayerweathers/Ripley/Woodward/Smith 363 beds
• Hitchcock and Massachusetts Row 358 beds
• The Lodge and Topliff/New Hampshire 367 beds
• The River and The Treehouses 506 beds
• Wheeler/Richardson and Affinity Programs (Asian
  Studies Center/Shabazz Center/Foley
  Cooperative/International/La Casa/Latin
  American/Native American) 279 beds

9
Sample Population

• Our goal was to survey a variety of students both
  male and female from different classes and
  different housing locations on campus
• Our method used the DND finder to randomly
  sample classes, and then supplemented that with
  class lists.
• Disregarded students living off campus and in
  fraternities/sororities

10
Sample Survey

• Class
• Sex
• Dorm (where do you live)
• Or do you live off campus

11
Responses

• 264 responses after eliminating those who lived
  off campus and in fraternities and sororities
• 146 male 118 female
• 164 freshman 100 non-freshman

12
Results overall student preference
13
Results overall student preference
14
Results dorms
15
Results dorms
16
Results male vs. female
17
Results male vs. female
18
2 Proportion Z-tests male vs. female

• One Tailed Tests
• Food Court
• Null Hypothesis Males and Females prefer Food
  Court in the same proportion.
• Alternative Hypothesis Males prefer Food Court
  in a higher proportion than Females.
• Z-Score -2.90
• p-value .00186
• Using a .05 significance level, we would reject
  the null hypothesis.

19
2 Proportion Z-tests male vs. female

• Home Plate
• Null Hypothesis Females and Males prefer Home
  Plate in the same proportion.
• Alternative Hypothesis Females prefer Home Plate
  in a higher proportion than Males.
• Z-Score 1.407
• p-value .0796
• Using a .05 significance level, we would not
  reject the null hypothesis.

20
Two Tailed Tests female vs. male

• Collis
• Null Hypothesis Males and Females prefer Collis
  in the same proportion.
• Alternative Hypothesis Males and Females prefer
  Collis in different proportions.
• Z-Score 3.285
• p-value .00101
• Using a .05 significance level, we would reject
  the null hypothesis.

21
Two Tailed Tests female vs. male

• The Hop
• Null Hypothesis Males and Females prefer The Hop
  in the same proportion.
• Alternate Hypothesis Males and Females prefer
  The Hop in different proportions.
• Z-Score -1.695
• p-value .090
• Using a .05 significance level, we would not
  reject the null hypothesis.

22
Two Tailed Tests female vs. male

• Lone Pine Tavern
• Null Hypothesis Males and Females prefer Lone
  Pine in the same proportion.
• Alternative Hypothesis Males and Females prefer
  Lone Pine in different proportions.
• Z-Score -1.86
• p-value .0628
• Using a .05 significance level, we would not
  reject the null hypothesis.

23
2 Proportion Z-Tests Freshman vs. Upper
Classmen (2 tailed)

• Food Court Null Hypothesis Freshman and Upper
  Classmen prefer Food Court in the same proportion
  - NOT REJECTED
• Home Plate Null Hypothesis Freshman and Upper
  Classmen prefer Home Plate in the same
  proportion. - REJECTED
• Collis Null Hypothesis Freshman and Upper
  Classmen prefer Collis in the same proportion. -
  REJECTED
• Hop Null Hypothesis Freshman and Upper Classmen
  prefer The Hop in the same proportion. - NOT
  REJECTED
• Lone Pine Null Hypothesis Freshman and Upper
  Classmen prefer Lone Pine in the same proportion.
  - NOT REJECTED

24
Chi Square Test

• Null Each Dining Hall is proportionately equally
  popular
• Alternative Some Dining Halls are more popular
  than others
• X2 46.70
• Significance Level .05
• Critical Value 9.488
• Using a .05 significance level, we would reject
  the null hypothesis.

25
Chi Square Test

• Null Each Dining Hall is equally popular to
  students living in the cluster with Butterfield.
• Alternate Some Dining Halls are more popular
  than others with students living in the cluster
  with Butterfield.
• X2 86.77
• Significance Level .05
• Critical Value 9.488
• Using a .05 significance level, we would reject
  the null hypothesis.

26
Chi Square Test

• Null Each Dining Hall is equally popular to
  students living in the Choates.
• Alternate Some Dining Halls are more popular
  than others with students living in the Choates
• X2 103.78
• Significance Level .05
• Critical Value 9.488
• Using a .05 significance level, we would reject
  the null hypothesis.

27
Chi Squared Test for Independence

• Is Dining Hall preference independent of a
• students Dorm?
• Null Hypothesis Dining Hall preference is
  independent of a students Dorm
• Alternate Hypothesis Dining Hall preference is
  not independent of a students Dorm

28
Observed vs. Expected
29
Chi Squared Test for Independence

• X2 30.93
• p-value .32
• Degrees of Freedom (Rows-1) x (Columns 1) 28
• Significance Level .05
• Using a .05 significance level, we would not
  reject the null hypothesis

30
Conclusions

• Collis/ Food Court significant difference in
  preference amongst males and females
• Hop/Lone Pine/Homeplate no significant
  difference
• Collis/Homeplate significant diff. In preference
  amongst freshman and non-freshman
• no significant difference at Food Court, Hop or
  Lone Pine.
• not all dining halls were equally preferred
  amongst the entire student body sampled

31
Conclusions continued

• From the results of our Chi Squared test for
  independence between dinning hall preference and
  dorm location we were unable to reject the null
  hypothesis that states that they were independent
  from each other.

32
What can DDS do with this information?

• Even though we only scratched the surface, we can
  conclude that preference is not related to size
• DDS could make a better use of its resources by
  focusing on the dining halls that are preferred
  by students
• They could also find out why some dining halls
  are preferred and add those factors to those that
  are unpopular

33
Sources

• Dartmouth students
• DDS Online
• http//www.dartmouth.edu/dds/index.html
• Housing Stats
• http//www.dartmouth.edu/tour/walking/06
  ass_row.html