high fives

27
High Fives Vinny Hageman Cory Sonderschafer

Upload: marvin

Post on 16-Feb-2016

58 views

Category:

Documents


0 download

DESCRIPTION

High Fives. Vinny Hageman Cory Sonderschafer. History . Originated in African-American urban culture as a greeting Gained popularity around 1980 in sports (baseball and basketball) for use as a congratulations Went mainstream in 80’s - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: High Fives

High FivesVinny Hageman

Cory Sonderschafer

Page 2: High Fives

History • Originated in African-

American urban culture as a greeting

• Gained popularity around 1980 in sports (baseball and basketball) for use as a congratulations

• Went mainstream in 80’s• Now a worldwide gesture

used for congratulations, celebration, greeting, etc.

Page 3: High Fives

Description

• Wanted to find out how strangers would respond to us offering them a high five

• Wanted to further explore their responses by comparing:– How males react to how

females react– How people react to

someone in casual attire to how they react to someone in Flyers attire

Page 4: High Fives

Procedure• Collected our data by offering high fives to every third person and recording

their response (high five or no high five) and their gender– Cory – casual attire, “hey high five,” went first– Vinny – Flyers attire, “let’s go Flyers,” went second

• Did this in Montgomery Mall, Valley Square Shopping Center, and Doylestown (Main Street, State Street, and Court Street)

• Tests of significance: – 1 Proportion Z Test on proportion of people who DID high five

• Ho: p0 =.2• Ha: p0 >.2

– 2 Proportion Z Test on proportion of males who DID high five vs. females who DID high five

• Ho: pmale = pfemale

• Ha: pmale > pfemale

– 2 Proportion Z Test on proportion of people who high fived Cory in a casual shirt vs. Vinny in a Flyers shirt

• Ho: pregular = pflyers

• Ha: pregular < pflyers

Page 5: High Fives

Montgomery Mall• Majority of females responded with no high five

– Same for casual and Flyers• Majority of males responded with a high five for Flyers

– Opposite for casual• Males responded with a high five more than females

– Same for casual and Flyers

Yes

No

Yes

NoCa

sual

Flye

rs

0 2 4 6 8 10 12

Montgomery Mall

FemaleMale

Montgomery MallCasual FlyersYes No Yes No

Male 5 7 11 3Female 4 9 4 6

Page 6: High Fives

Valley Square

• Majority of females responded with a high five– Same for casual and flyers

• Majority of males responded with a high five for Flyers– Opposite for casual

• Males responded with a high five more than females for Flyers– Opposite for casual

Yes

No

Yes

No

Casu

alFl

yers

0 2 4 6 8 10 12

Valley Square

FemaleMale

Valley SquareCasual FlyersYes No Yes No

Male 6 4 8 2Female 4 12 5 9

Page 7: High Fives

Doylestown • Majority of females responded with no high five

– Same for casual and Flyers• Majority of males responded with a high five

– Same for casual and Flyers• Males responded with a high five more than females

– Same for casual and Flyers

Yes

No

Yes

No

Casu

alFl

yers

0 2 4 6 8 10 12 14

Doylestown

FemaleMale

DoylestownCasual FlyersYes No Yes No

Male 5 6 11 1Female 7 5 8 6

Page 8: High Fives

Analysis

• Majority of females in Doylestown and in Montgomery Mall responded with no high five– Female patterns were always same for casual and

flyers• Majority of males everywhere responded with

a high five for Flyers• Majority of males in Montgomery Mall and in

Valley Square responded with no high five for casual

Page 9: High Fives

Conclusions

• If you are a white, average-looking young man in Greater Philadelphia Region (specifically Bucks County/Montgomery County)…– Majority of females will not high five you– Majority of males will not high five you…

• UNLESS you are obviously a Flyers fan!

• Males in Doylestown are most friendly to strangers– Females are least friendly

• Females in Valley Square are most friendly to strangers

Page 10: High Fives

Test #1: 1 Proportion Z Test

Assumptions:

State:1. SRS2. n0 p0 ≥ 10

n0(1 − p0) ≥ 103. pop ≥ 10n

Check:1. Assumed2. 148*.53 = 78.44 ≥ 10

148*.47 = 69.56 ≥ 103. Passed

Page 11: High Fives

Test #1: 1 Proportion Z Test

P (z ≥ 9.946) < 0.0001

= 9.946

We reject Ho because p-value is < α = .05. We have sufficient evidence that the proportion of people who responded with a high five is greater than .20.

Test of Total High Fives Test Proportion

Attribute (categorical): Total

Attribute: Total78 out of 148, or 0.527027, are 1 Alternative hypothesis: The population proportion for 1 is greater than 0.2 .

The test statistic, z, is 9.946.

If it w ere true that the population proportion of 1 w ere equal to 0.2 (the null hypothesis), and the sampling process w ere performed repeatedly, the probability of getting a z value this great or greater w ould be < 0.0001.

Note: This probability w as computed using the normal approximation.

Ho: p0 =.2Ha: p0 >.2

Page 12: High Fives

Test #1 Chart

53%

47%

TotalTotal Yes Total No

•Percentage of people who responded with a high five is higher than those who responded with no high five

Page 13: High Fives

Test #2: 2 Proportion Z Test

Assumptions:

State:1. 2 independent SRS2. n1 p1 ≥ 10

n1(1 − p1) ≥ 10n2 p2 ≥ 10 n2(1 − p2) ≥ 10

3. pop1 ≥ 10n pop2 ≥ 10n

Check:1. Assumed2. 69*.67 = 46.23 ≥ 10

69*.33 = 22.77 ≥ 1079*.41 = 32.39 ≥ 1079*.59 = 46.61 ≥ 10

3. Passed

Page 14: High Fives

Test #2: 2 Proportion Z Test

P (z ≥ 3.18) = 0.00074

= 3.18

We reject Ho because p-value is < α = .05. We have sufficient evidence that the proportion of males who responded with a high five is greater than the proportion of females who responded with a high five.

Test of Male vs. Female Yes PropCompare Proportions

Attribute (categorical): Male

Attribute (categorical or grouping): Female

In Male 46 out of 69, or 0.666667, are 1 In Female 32 out of 79, or 0.405063, are 1 Alternative hypothesis: The population proportion for 1 in Male is greater than that for 1 in Female

The test statistic, z, is 3.18.

If it w ere true that the tw o population proportions w ere equal (the null hypothesis), and the sampling process w ere performed repeatedly, the probabilityof getting a value of z this great or greater w ould be 0.00074. Note: This probability w as computed using the normal approximation.

Ho: pmale = pfemale

Ha: pmale > pfemale

Page 15: High Fives

Test #2 Graph

Male

Female

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Male vs. Female

Yes No

•Percentage of males who responded with a high five is higher than the percentage of females who responded with a high five

Page 16: High Fives

Test #3: 2 Proportion Z Test

Assumptions:

State:1. 2 independent SRS2. n1 p1 ≥ 10

n1(1 − p1) ≥ 10n2 p2 ≥ 10 n2(1 − p2) ≥ 10

3. pop1 ≥ 10n pop2 ≥ 10n

Check:1. Assumed2. 74*.42 = 31.08 ≥ 10

74*.58 = 42.92 ≥ 1074*.64 = 47.36 ≥ 1074*.36 = 26.64 ≥ 10

3. Passed

Page 17: High Fives

Test #3: 2 Proportion Z Test

P (z ≤ -2.634) = 0.0042

= -2.634

We reject Ho because p-value is < α = .05. We have sufficient evidence that the proportion of people who responded with a high five to someone in Flyers attire is greater than the proportion of people who responded with a high five to someone in casual attire.

Test of Casual vs. Flyers Yes PropCompare Proportions

Attribute (categorical): Casual

Attribute (categorical or grouping): Flyers

In Casual 31 out of 74, or 0.418919, are 1 In Flyers 47 out of 74, or 0.635135, are 1 Alternative hypothesis: The population proportion for 1 in Casual is less than that for 1 in Flyers

The test statistic, z, is -2.634.

If it w ere true that the tw o population proportions w ere equal (the null hypothesis), and the sampling process w ere performed repeatedly, the probability of getting a value of z this small or smaller w ould be 0.0042. Note: This probability w as computed using the normal approximation.

Ho: pregular = pflyers

Ha: pregular < pflyers

Page 18: High Fives

Test #3 Graph

Casual

Flyers

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Casual vs. Flyers

Yes No

•Percentage of people who to responded with a high five to someone in Flyers attire is higher than percentage of people who responded with a high five to someone in casual attire

Page 19: High Fives

Conclusions (Inf.)

• More than half of all people in Greater Philadelphia Area (Bucks County and Montgomery County specifically) will respond to the offer of a high five with a high five

• Males in this area are more likely to respond with a high five than females

• People in this area are more likely to respond to someone in a Flyers shirt with a high five than someone in casual attire

Page 20: High Fives

Application

• 53% of all people responded with a high five• Knowing this, we are now more likely to offer a

high five to a stranger– Greater than 50% chance that they will accept– Why would we do this?

• It’s fun• It’s an easy way to meet friendly people

Page 21: High Fives

Sources of Bias/Error

• Weather - more people would have been outside (Doylestown, Valley Square) if it was nicer out

• Odor- some areas smelled bad (Valley Square next to DSW) causing people not to be there

• Mood - someone would have been more inclined to respond with a high five if they were in a good mood

• Already saw us doing work – people would have been more inclined to respond with a high five if they knew what we wanted ahead of time (or less inclined if they’re jerks)

Page 22: High Fives

Sources of Bias/Error (Cont.)

• Age – children were reluctant to high five us (probably because of a “don’t talk to strangers” mentality)

• Group – people with SO’s/families seemed more hesitant than individuals

• Two different people collecting data – people could be more/less inclined to high five us for personal reasons (our appearances, etc.)

• Fans of a rival team/city – would have been less inclined to high five Vinny

Page 23: High Fives

Sources of Bias/Error (Cont.)• Flyers were in Stanley Cup Finals during data collection –

caused fans to be “flyered up” and therefore more likely to high five Vinny

• Flyers played later that night – fans would have been more likely to high five Vinny

• Outcome of Flyers game night before– fans could have been more/less likely to high five Vinny depending on if they won/lost night before (depends on the type of fan)

• Recording same person more than once – repeats were more likely to respond with a high five again

• Knowing someone – acquaintances were more likely to respond with a high five than a stranger

Page 24: High Fives

Personal Opinion

• Results could be more accurate– Remove sources of error

• Go places…– Big enough to not have any repeats– Far away enough from home to not know anyone

• Have one person only collect data (change into Flyers shirt) and other only record data

• Just wouldn’t be as much fun if we changed it– We actually had a good time with this project– Collecting data was enjoyable

Page 25: High Fives

Activity Introduction• Some people gave sloppy

high fives– Weak effort– Resulted in near misses

• Can be embarrassing for both parties

• People need instruction…

Page 26: High Fives

Activity

Perfect High Five Technique

1. Turn to your neighbor2. Stare at their elbow3. High five them repeatedly

Works every time!

Page 27: High Fives

FIN