Fashion Trends in a Middle School

1
Fashion Trends in a Middle School

• Supercomputing Challenge
• Summer InstituteFinal ReportJuly 27,
  2006Capshaw Middle School - Team 01
• Theresa Anaya Burney Peter Colehour
• Susan Gibbs Thansewi J. Martinez
• Makoena Monese

2
Introduction

• Middle school students arrive at school wanting
  both to conform to fashion norms and to be
  recognized as individuals. Each student also has
  a relative degree of shyness or sociability.
  These factors affect each students
  susceptibility to a fashion trend. The research
  supports the relative shyness and resistance
  levels of students, and tipping points in the
  spread of a trend.

3
Executive Summary

• Our computational model
• Allows multiple trials of how the initial
  percentage of resistant students affects speed of
  a trends spread
• Imitates the real-world school environment
• Shows relative shyness and susceptibility of
  students
• Recognizes that the relative percentage of
  resistant students determines the speed of the
  trends spread.
• Can be used to study relevant middle school
  topics
• bullying
• spread of disease
• the susceptibility of individuals to drug,
  alcohol, tobacco use, and sexual activities.

4
The Project Hypotheses

• The StarLogo model will graph the spread of a
  trend, comparable to an infection model.
• The lower the initial percentage of followers,
  the slower the trend will spread.
• A tipping point may exist in the initial
  percentage of followers, above which the trend
  spreads much more quickly.

5
Mathematica Model

• Our infection model set two contact rates (30 of
  agents have fewer contacts) and used an infection
  variable for the percentage of the population
  that was susceptible (infection after 2
  exposures) or immune (after 4 exposures).
• dPI /dt rcPI (1-PI)
• Expected number of susceptible individuals who
  will become infected in the time ?t, assuming rc
  1.

6
Starlogo Model

• Our STARLOGO model allows students to manipulate
  the original percent of students characterized as
  followers (susceptible) and freethinkers
  (relatively immune).

7
Results

• All 3 Hypotheses proved
• The StarLogo model is comparable to an an
  infection model.
• The lower the initial percentage of followers,
  the slower the trend will spread.
• Two tipping points exist in the initial
  percentage of followers, above which the trend
  spreads much more quickly

8
Recommendations

• Students
• Expand model to apply to new issues
• Expand and refine model to show affect of
  advertising, new conflicting trends, and
• Conduct real-world experiments on the spread of
  trends
• Teachers
• Create a StarLogo TNG model
• Use model to entice students into Challenge

9
Acknowledgements

• We appreciate
• STI Team - for all their support and patience.
• Dylan Allegreti - for supporting the process and
  working with us.