Using Social Media, Crowd-Sourcing, & Web-Based Mapping ...

CALIFORNIA STATE UNIVERSITY, NORTHRIDGE

The Pizza Map: Using Social Media, Crowd-Sourcing, & Web-Based Mapping to

Analyze Regional Pizza Preferences in the United States

A thesis submitted in partial fulfillment of the requirements

for the degree of Master of Arts in Geography,

GIS Program

By

Mark F. Turgeon

December 2013

ii

The thesis of Mark F. Turgeon is approved:

_________________________________ __________________

Dr. Regan Maas Date

__________________________________ __________________

Dr. Soheil Boroushaki Date

__________________________________ __________________

Dr. Steven Graves, Chair Date

California State University, Northridge

iii

Acknowledgements

This thesis would not be possible without the guidance and support of many

individuals. Thank you to my family, friends, classmates, CSUN Geography Faculty and

everyone else that contributed in making Geography of Pizza a reality.

First, I would like to express my deepest appreciation to my committee chair, Dr.

Steve Graves, who really understood my love and passion for pizza and encouraged me

to write this. I would also like to thank my other members, Dr. Regan Maas, and Dr.

Soheil Boroushaki for their guidance and support in learning new GIS applications and

tools such as Python scripting and Web-Based mapping design. These applications and

tools were very essential to have an online presence.

I would also like to thank Chad March, Geoffrey West, and many other GIS

classmates for their support along the way. Also, a big shout out to all the faculty and

geography students from departments across the United States, that participated and

contributed VGI data through the survey.

Above all, I would like to thank my amazing wife Sara for all of her love, support,

patience, encouragement, and pizza opinions. I am completely grateful.

iv

TABLE OF CONTENTS

Signature Page.....................................................................................................................ii

Dedication...........................................................................................................................iii

List of Tables......................................................................................................................vi

List of Figures....................................................................................................................vii

Abstract.............................................................................................................................viii

Chapter 1 - Introduction.......................................................................................................1

Chapter 2 - Background.......................................................................................................3

2.1. Origin of Pizza..................................................................................................3

2.2. Italian Immigration & Pizza in the United States.............................................3

2.3. World War II & Americanization of Pizza through Franchises.......................6

Chapter 3 - Review of Literature.......................................................................................11

3.1. Social Media & Crowd Sourcing....................................................................11

3.2. (VGI) Volunteered Geographic Data..............................................................13

Chapter 4 - Methodology...................................................................................................16

4.1. Regional Pizza-Styles.....................................................................................16

4.2. Survey Instrument & Questions......................................................................19

4.3. Geography of Pizza Website...........................................................................24

4.4. Beta Testing Survey Instrument......................................................................28

4.5. Promoting Thesis on Facebook.......................................................................28

4.6. Promoting Thesis on Twitter..........................................................................34

Chapter 5 - Results.............................................................................................................38

v

5.1. Survey Results from Social Media Promotion................................................38

5.2. Target Emailing..............................................................................................40

5.3. Organizing Total Survey Responses for Maps...............................................45

5.4. Identity & Place Map Results for Regional Pizza-Styles...............................65

Chapter 6 - Discussion.......................................................................................................75

References..........................................................................................................................78

vi

LIST OF TABLES

Tables 1, 2. Setting up properties and location values in a Google Fusion table..............20

Table 3. Regional Pizza-Styles Chart for Geography of Pizza Website............................26

Table 4. Results for the 1st Week of Promoting on Facebook & Twitter.........................40

Table 5. States that were targeted during the email promotion campaign.........................42

Table 6. Non-targeted states during the email promotion campaign.................................43

Table 7. Results for most preferred pizza-styles from targeted states...............................44

Table 8. Results of most preferred pizza-styles from non-targeted states.........................45

Table 9. Total pizza preference survey responses in the United States.............................48

Table 10. Most preferred pizza-styles in the U.S. from total survey responses................49

Table 11. Excel spreadsheet of pizza percentage maps format.........................................52

Table 12. XY to Line Tool Excel format for radial flow maps.........................................65

Tables 13-19. Radial flow maps and stats of survey responses........................................67

vii

LIST OF FIGURES

Figure 1. Most Popular Franchise Pizza in the United States............................................10

Figure 2. Pizza preference survey instrument questions in a Google Form......................22

Figure 3. Original Facebook posting about Geography of Pizza.......................................29

Figures 4-7. Facebook shares by friends for Geography of Pizza.....................................30

Figure 8. First Twitter tweet promoting Geography of Pizza............................................34

Figures 9-11. More Twitter tweets for pizza project.........................................................35

Figures 12, 13. First week of survey responses in Google Maps......................................38

Figure 14. Google map of the total U.S. pizza preference responses................................46

Figure 15. Main United States Pizza Preference Map.......................................................50

Figures 16-27. U.S. Pizza Preference Percentage Maps....................................................53

Figures 28-34. U.S. Pizza Preference Radial Flow Maps..................................................67

viii

ABSTRACT

The Pizza Map: Using Social Media, Crowd Sourcing, & Web-Based Mapping to

Analyze Regional Pizza Preferences in the United States

By

Mark F. Turgeon

Master of Arts in Geography

GIS Program

At its core this thesis explores the geography of place and identity through the

lens of pizza. Place not only dictates variations of regional pizza styles, but also pizza

taste preferences. The study aims to explore the history of pizza in the United States, its

spread in popularity, the development of regional pizza styles, and, through a survey, the

pizza preference of Americans today.

A secondary aspect of this study was the use of web-based tools, mapping, crowd-

sourced data, and social media websites to conduct research. Much of the research

conducted for this thesis involved reviewing articles of (VGI) Volunteered Geographic

Information and how effective and trustworthy this data is in today's society. This study

was conducted at a time when crowd sourced data was being utilized in a growing

number of forms, and explores the usages and limitations of crowd sourced data.

1

Chapter 1 - Introduction

I grew up in the middle of Connecticut, halfway between New York City and

Boston. These two cities were areas of major settlements for Italian immigrants and not

surprisingly the location for the first pizzerias on American soil. I originally come from

an Italian heritage and pizza is one our family's traditional foods. While living on the East

Coast, I've developed a taste preference for quality, thin crust pizza and have become

very opinionated about pizzerias across the U.S. After my I moved to California, I began

to notice that there were a number of regional American pizza styles, and that people

from different regions were just as passionate about their pizza style preference as I was

about mine.

Others have made similar observations. "The type of pizza that most people like

today is usually the one they grew up with" (Reinhart, 2003). In some communities there

are still great local independent pizzerias. In recent decades, in most parts of the country,

franchise pizzerias such as Pizza, Hut, Dominos, and Papa John's have come to dominate

the landscape. How people, especially those removed from the pizza hearth-places of

America acquire their taste preferences is an important aspect of this study. When it

comes to taste preference for pizza, place has a significant effect. Place not only dictates

variations in pizza style, but seems to establish general patterns of pizza taste preference.

A secondary goal of this thesis is the exploration of new possibilities for research

using social media, crowd sourced data and web-based mapping. It appears, at least for

topics that excite the imagination of social media users, that these new web-based tools

have proven to be effective in terms of generating discourse regarding the similarities and

differences in regional food preference and identity.

2

I decided to use pizza as a lens through which to view issues of place, identity and

new methods of research because I love pizza and believe it is a simple, accessible food

that is capable of eliciting strong opinions for a large number of volunteer survey

subjects. I am particularly interested in how place affects identity because I am an East

Coaster, living in California who is constantly curious about how geography affects

identity.

3

Chapter 2 - Background

2.1. Origin of Pizza

The word pizza is derived from the Latin word, "pinsere", which means to crush

or grind dough, preparing for use as a crust. The flatbread, which is the foundation of the

pizza, was invented by the Egyptians, who cooked the dough by using cone shaped oven,

invented by them 6,000 years ago (Buonassisi, 1982). Egyptian ingredients and flat bread

preparation techniques was embraced by the Romans and Greeks because they had

similar foods. The first traditional pizza that we know of today was probably constructed

in Naples, Italy in the early 1800’s. Don Raffaele Esposito created a signature pizza

named “alla Margherita," named for Queen Margherita. The Queen was very pleased

with this particular pizza because it was made with mozzarella, tomatoes, and fresh basil,

which comprised the colors of the Italian flag, white, red and green. This particular pie

became popular in Italy and the rest of the world and is considered the most traditional

style pizza. The formation of the “pizza margherita” remains a major influence on all

pizza makers but was not the impetus for its ultimate global popularity.

2.2. Italian Immigration & Pizza in the U.S.

The first flow of Italian immigration to the United States began as early as the

colonization of America (Lord, 1905). The first reliable accounting of the number of

Italian-born immigrants living in the United States is the 1900 census. This record shows

that Italians migrated all over the United States, but large populations settled in New

York, Boston, Philadelphia, and Chicago. This early census data of Italian immigration in

the United States showed where the Italian neighborhoods formed into clusters, which

4

were frequently called “Little Italy.” These neighborhoods remain today, and they help

preserve a number of Italian traditions including authentic recipes. Italian immigrants did

not identify themselves as “Italian” until they came to America (Benedetto, 2000). Italy

was very disjointed as a nation and people usually identified themselves regionally. Food

became a way for the Italian immigrants to connect to one another.

Indeed food has proven to be a major component of what it means to be Italian.

This identity, perhaps more than any other American immigrant group has tied their

group identity to a collective notion of cuisine. Italian immigrants in America began to

bond over their similarities, rather than their regional differences (Cinotto, 2004). Food

and drink is very important, perhaps even sacred to Italian families. The majority of the

money each Italian family brought home was spent on food for the family (Cinotto,

2004). Food is so important to Italian families because it is completely imbedded in the

culture as something used for much more than nutritional purposes; food is an expression

of love. Pizza became a way for Italian immigrants to connect to the “old country” and

also to connect with each other. In America, pizza was always a marker for identity

(Cinotto, 2004).

Human identities are shaped by the space in which people live and socialize.

Migrant identities are bound to the geographies of multiple places. In the Italian

neighborhoods of New York, food production, trade, and consumption had a significant

impact on social space (Cinotto 2004). Italy was a distant place to New York, so food for

the Italian immigrants became embedded as a sense of attachment. In “urban villages” the

identity of the community is made coherent through the widespread sharing of common

symbols in the form of food and food ways.

5

Food is intimately connected to identity and place. One objective of this research project

is to explore the way place has affected the look and taste of pizza. In this country of many

ethnicities, where many people live far from where they are born, particular foods such as pizza,

cheese, BBQ, gumbo and clam chowder assume the role of an icon for a place. People like to be

reminded of home and have a sense of community and region to provide an anchor of identity

through food (Shortridge, 1999). Today, deep-dish pizza is no longer considered an immigrant

tradition, but a Chicago icon and as central to the Windy City as Wrigley Field (BBC Travel). It

has been widely speculated that the Chicago Deep Dish partially grew out of the geography and

weather of the city -- it is cold, and windy, and the Deep Dish pizza is hearty, dense and warm. In

New York people are always on the go and in a hurry. New York-style pizza is designed and sold

in wide slices that can easily be folded and eaten while walking.

Italian immigrants from different regions of Italy coming together and living in

the same neighborhood cluster and sharing different styles of food such as pizza, pasta,

and other fine dishes, formed the foundation for the eventual spread and Americanization

of these dishes (Cinotto, et al. 2003).

Italians living in these clustered neighborhoods largely consumed food from

home. Gurney (2011) writes about the reclaiming of authentic, regional Italian cuisine.

She argues that many of Italian dishes belie the regional foods eaten in Italy – they

represent food the immigrants developed and enjoyed within their own households long

before Americans began eating them in restaurants.

Food products such as olive oil, cheese, meat, and macaroni were normally

expensive foods back in Italy. In the United States the Italian immigrants were able to

buy these goods at reasonable prices at Italian grocery stores. Food gives identity to

neighborhoods. Food is one of the most integral components of culture, and "pizza" and

6

"Italy" go hand in hand (Cinotto, et al. 2003). The Italian immigrants in East Harlem,

New York, settled one of the first authentic Italian neighborhoods in America. During the

interwar years, from 1900-1950 there was a focus on the role of food -as a cultural

feature, non-verbal form of communication, and metaphor of self among immigrants.

Food served as a “construction” of two places: one actual - the neighborhood - and the

other one remembered and imagined – the old country (Cinotto, et al. 2003). "Foods

strongly associated with an area can easily take on symbolic qualities, and their

consumption can evoke both personal nostalgia and community pride. By studying such

culinary traditions, we can improve understanding of how modern Americans strive to

create touchstones for group identity." (Shortridge, 2003).

Geography informs identity through food. The state of Maine has become a

“spatialization of food- ways and the interconnections between people, food, and places.”

An interest in Maine’s food ways has helped the state to become something of a culinary

destination (Lindenfeld and Silka, 2011). The Food Network also contributed to the

growth of tourism in Maine, featuring many popular destination spots throughout the

state serving authentic New England dishes such as lobster, New England clam chowder,

etc. The culture and identity of people are influenced by the geography of place.

2.3. World War II and the Americanization of Pizza through Franchises

Popularity of pizza soared after World War II, because tens of thousands of

Americans, who were not of Italian descent, were exposed to Italian food during the

campaign to wrest Italy from the Fascists. Many other American soldiers found

themselves visiting New York, Boston and Chicago during the war years, exposing them

7

to Italian-Americans and their cuisine. American soldiers in Italy had been exposed to

pizza and when they returned home they created a demand for pizza that was far greater

than it was before the war (Miller, 2006). The spread of pizza and Italian food seeped out

of Little Italy's and into America. In the 1930's, a guide to dining out in New York

defined pizza as “an inch thick, potato pan-cake, sprinkled with Parmesan cheese and

stewed tomatoes," and it has been vastly changed since then (Cinotto, 2004). President

Eisenhower, said he had eaten better pizza in New York than in Naples, Italy and that the

best pizza was American-made. This one quote alone from President Eisenhower, in

addition to returning troops may have sparked the interest for pizza in the United States.

While Italian immigration and the spread of pizza began in New York City, it

quickly extended into small towns throughout America (Brixius and Kupfer, 2008). A

small town named Old Forge located between Scranton and Wilkes-Barre, Pennsylvania,

became known as the “Pizza Capital of the World" (Brixius and Kupfer, 2008). Old

Forge is a village of about 10,000 people and a cluster of 10 pizzerias located on the main

street. The pizza is entirely its own style with its own unique history.

Many Italian immigrants are concentrated in the Philadelphia and Pittsburgh

areas, but a large number of immigrants scattered to small towns of Pennsylvania such as

Scranton and Old Forge to find work in the coal and steel industries (Brixius and Kupfer,

2008).

Old Forge-style pizza is different than other regional pizzas because of its

rectangular shape with a slightly thicker, pan seared crust and unique blend of cheeses.

Mama’s is a famous pizzeria in Old Forge – many locals grew up on this style of pizza.

This pizzeria has a special history, a memory that is woven together with the flavors,

8

textures, and atmosphere of the place (Brixius and Kupfer, 2008). Pizzerias, particularly

in a place such as Old Forge, are spaces meant to be shared, not only with each other, but

with our past selves as well (Brixius and Kupfer, 2008).

While the spread of pizza originated in Italian enclaves, pizza is now a fully

American food. Historic Italian neighborhoods in the United States relate to the original

immigrants and their authentic food styles and recipes. Many of these historic Italian

neighborhoods have disintegrated or became smaller as Italian-Americans followed the

flight of others to the suburbs and married outside their ethnicity, but the traditions and

food preferences still live among thousands of Italian households and Italian restaurants

scattered across the United States.

The creators of Pizza Uno in Chicago created a pizza for the American palette -

thick, doughy and with as much as a half-pound of toppings baked into a deep dish pan,

akin to what Americans would call a “cobbler” were it a fruit pie. The popularity of

Chicago style pizza spawned imitators in Chicago first, then in many other parts of the

United States. During the 1950s, numerous pizza shops began speckling the American

heartland and California. Several started franchising their name and recipes. As often as

not, these new pizza chains were operated by those without Italian heritage. Almost

always, the pizza was adapted to non-Italian (or Italian-American) palates. The

Americanization of pizza had begun. Pizza Hut was founded in Wichita, Kansas in June

15th 1958 (Singer, 2011). Dominos was founded in Ypsilanti, Michigan in June 10th

1960 (Singer, 2011). Most major pizza chains originated in middle America, far, far away

from hubs of Italian immigrants. It was this push-- the franchising and mass production

9

of pizza that popularized “American” pizza. From here, the idea of pizza spread and

seemingly took on different incarnations in each region of the country.

There are clear and distinct density hubs of fast food restaurants in various parts

of the country. This may be explained through land cost and lifestyle. In the areas of the

country where the cost of land is relatively lower, fast food pizza restaurants seem to

have grown. It can also be speculated that fast food pizza was unwelcome in cities where

there was already an abundance of authentic pizza, frequently served in “Italian

restaurants”, with a full menu of options and waitresses. Many of the biggest pizza chains

started out in Midwestern locations, and only later did they diffuse into higher cost areas

of the country (Gordon and Sterrett, 1999). (SHAKEYS – out here is an exception – LA

has high land costs and they are a pretty old chain).

The big pizza chains were all started by business people and not by individuals

who have a passion and love for making pizza. You can visit any website like Papa

John's, Pizza Hit or Little Caesars and you'll find heartwarming stories of young people

who overcame their modest circumstances to achieve great wealth by building big and

have a successful business selling pizza (Brixius, 2008).

10

Figure 1. Most Popular Franchise Pizza in the United States

11

Chapter 3 - Review of Literature

3.1 Social Media & Crowd Sourcing

Social media websites such as Twitter, Facebook, YouTube, Instagram, and Flickr

are valuable sources that store and constantly update a wealth of untapped data from real

people online. These websites are very popular and are only now being used as research

tools. I chose to use social media to gather data on pizza preferences. There is a need for

more information about regional cuisines (Shortridge, 2003). “We still lack basic

knowledge and analysis of regional cuisines. This gap exists despite renewed public

interest in all things related to food...the main problem is scarcity of appropriate place-

related data" (Shortridge, 2003). Because geographically detailed consumption

information is nearly nonexistent, one solution is to ask people about their preferences”

(Shortridge, 2003).

Various philosophical geographers have long recognized the role of media and

communication in shaping space and place using language, maps, and GIS (Sui and

Goodchild, 2011). There are new interests and topics in the GIScience field, such as

GeoWeb, Digital Earth, CyberGIS, and cloud computing. “Perhaps one productive way

for GIScience to proceed is to ride on the discipline of geography’s communication turn

and the spatial turn that is evident in media studies” (Sui and Goodchild, 2011). Today's

geographers are more interested in media and communication studies than ever, as social

media has really given maps a new platform. The Association of American Geographers

even just recently formed a Social Media and GIS specialty group (Adams, 2009).

One of the technological changes of GIS during the past 10 years is the growing

convergence of GIS and social media (Sui and Goodchild, 2011). Many people share

12

maps online and they easily go viral through social media websites. A map can

communicate an idea, argument, or concept quickly, almost instantly, and effectively.

This type of visual and almost instant communication is perfect for the internet age way

of communication.

Potential problems with social media and crowd sourced data might include false

information that is possibly not verifiable (Meier, 2011). Accurate information that

arrives too late is meaningless and useless. Crowd sourced information can provide a fast

situational awareness, especially when applied to a live map. Many people believe that

verifying crowd sourced data is either too challenging or impossible. Meir (2011)

demonstrates that concrete strategies do exist when reliable people scrape the internet to

dig for evidence that helps support the contributed crowd-sourced data is authentic.

Sometimes live crowd-sources maps do not have time to be verified. Social media

websites such as Twitter, Facebook, Flickr, Skype and YouTube all play significant roles

particularly in times of crisis. The live, crowd-sourced maps deployed in response to the

recent disasters in Haiti, Libya, and Egypt provided great contextual information on the

evolving situations in real time (Meier, 2011).

The GPS Waze app is an example of a tool that uses crowd sourced data in real

time. Users constantly contribute information to the maps, including accidents, speed,

and whether there is state trooper hiding around the next bend. Users are also constantly

asked to verify the information others have shared, so that in this way all information is

self-verified and constantly updated by the app.

13

3.2 (VGI) Volunteered Geographic Data

The potential challenges related to possible false information on the other hand

are not new issues. Survey methods have been employed by geographers for years. In the

present day, the method of survey distribution is the main differentiating factor.

Volunteered geographic information is the harnessing of tools to create, assemble, and

disseminate geographic data provided voluntarily by individuals (Goodchild, 2007). VGI

is important these days because there is already a community online who contribute data,

but more people are volunteering every day especially when emergency disasters occur

around the world such as Hurricane Sandy & Katrina, the tsunami in Japan, the Arab

Summer, and Occupy Wall Street.

Our earth’s population is six billion and growing and each and every one of us has

some form of geographic knowledge from books, TV, magazines, internet and our own

surroundings. Anyone with access to the internet can potentially be used as sensors to

contribute trustworthy geographic data on the internet that can be used in a GIS or other

geographic software’s for analysis and decisions. There are countless websites in which

contributors have shared geographic data that was used and made a difference in life

(Goodchild, 2007). Some websites include OpenStreetMap, Google Earth, WikiMapia,

P3DM Where?, and Panoramio all of which provide rich and abundant geographic data

and geo-referenced information from volunteered contributors.

Every human is able to act as an intelligent sensor, perhaps equipped with such

simple aids as a GPS, Smartphone or even the means of taking measurements of

environmental variables. The main goal Goodchild (2007) states is to ensure quality data,

14

being able to detect and remove errors, and build a level of trust and assurance from

contributors.

There remain many issues in the development of spatial data infrastructures

because of restrictions on spatial data sharing and legal issues (Goodchild, 2007). A new

option called VGI Volunteered Geographic Information might effectively supplement the

traditional efforts of mapping agencies and the power of remote sensing, filling in

obvious gap in the current processes of acquiring geographic information (Goodchild,

2007).

It is important to think about what motivates people to contribute VGI. One of the

reasons pizza preferences are an ideal subject for crowd-sourced, web-based mapping is

that people are in general opinionated about their pizza preferences and many people

relish in the opportunity to share their opinions. People on Facebook and Twitter are

always expressing their opinions about anything, especially if they see a picture of your

food that you cooked or have eaten at a restaurant. Yelp is another great website where

people can express their opinions about businesses and rate them. Many people rely on

Yelp’s opinions because they matter when you want to find the best pizza place, a

trustworthy dentist, or auto repair shop.

The question is, “What motivates people to voluntarily contribute information?”

(Coleman, et al. 2009). “Volunteered Geographic Information” or VGI is a termed coined

by (Goodchild, 2007) that defines the user generated geospatial content in various

websites and to satisfy the needs within the industries of government, and social

networking communities. Many large companies have already discovered the importance

of VGI, and that government mapping agencies have the potential to harness this new

15

media and VGI to improve their own detections and geospatial data updating processes

(Coleman, et al. 2009).

People from all over the world contribute spatial data voluntarily using Google

Maps, TeleAtlas, Navteq and TomTom. This increased usage of crowd sourced

geospatial information in today’s world makes it important to understand and identify the

quality of contributed data. There are challenges involved in describing and classifying

both the types of people who volunteer geospatial information and the nature of their

contributions (Coleman, et al. 2009). Apps like Waze, which Google bought for a billion

dollars, almost nullify the validity of the question. Crowd sourced data is self-verified by

its very nature. It’s almost an obsolete question, because it works, it exists and is a staple

of our modern world.

There have been major advances in positioning, Web mapping, cellular

communications and wiki technologies that have outpaced the original visions of

Geographic Spatial Data Infrastructures around the world (Coleman, et al. 2009).

Collaborative Web-Based Mapping efforts like Open Street Map, Tagzania,

Wayfaring.com, and the People’s Map now enable experts and amateur enthusiasts alike

to be able to create and share limited geospatial data. (Coleman, et al. 2009).

16

Chapter 4 - Methodology

4.1 Regional Pizza Styles

An early and important task was to create a typology of American pizza styles.

There were dozens of possibilities, but Singer’s American Pizza Styles and Components

(2011) proved exceptionally useful. Singer’s book consists of Wikipedia articles put

together in order to describe the many American pizza-styles. This book serves as a solid

and generally bias-free reference to help select a set list of pizza styles to use for survey

construction. I referenced information from Singer’s book, a pizza-styles chart from Jeff

Varasano's website www.varasanos.com, along with my personal experience to create a

working list of popular pizza styles in the United States. For the purposes of the survey, I

decided to narrow down the choices of the vast number of pizza styles available in the

United States. In the end, I settled on nine different and well-known pizza-styles and two

additional extremely popular styles I decided to call All-American Franchise Pizza &

Franchise-Style and Build-Your-Own Pizza. I wanted to create a manageable list with the

most popular options. There are countless styles of pizza that are very popular in a certain

town, for example, the Old Forge style in Pennsylvania, but this is a style of pizza

unknown to most people. I also decided to leave out size-based pizza styles such as the

DC Jumbo Slice, as I consider these styles to be novelties. Considering the small size and

scope of this project, I knew I had to stick with the most popular pizza styles, which I

then used to create the survey instrument questions pizza-styles chart on

www.geographyofpizza.com. The following is the final list of pizza styles I settled on for

this project:

17

All-American Pizza: The delivery staples such as Pizza Hut, Dominos, Papa John’s and

Little Caesars. These pizzas are commercial, fast-food grade and churned in large

numbers. I wondered if naming this style “All-American” might skew survey takers, and

I am curious what would happen if, in the future, I changed the name to “Fast Food

Pizza,” or “Corporate Pizza.”

California-Style: This pizza was popularized in 1980 by Ed LaDou (the “Prince of

Pizza”). This is a single serving pizza that is combined with New York and Italian thin

crust and topped with nontraditional toppings such as avocados, salmon, BBQ chicken

and fresh California produce. Wolfgang Puck and California Pizza Kitchen both further

popularized this style of pizza from a gourmet food trend to a mass consumer food

product.

Chicago-Style Deep/Stuffed: The Chicago-style “deep-dish” was invented at the original

Pizzeria Uno in Chicago in 1943. This style of pizza is cooked in a pan and is made with

a heavy thick buttery crust and then layered with large amounts of cheese, Italian sausage

or meat chunks, large vegetables, with a chunky tomato sauce on top. This is very hearty

and comforting pizza that takes a while to cook and is eaten with a knife and fork.

Chicago-Style Thin Crust: This pizza style is thin with a firm crust, a little crunchy and

a slightly doughy top. It features a liberal amount of tomato sauce with no visible

chunks. This pizza is highly spiced and herbal and is shaped and sliced with a tavern cut

with small rectangles, squares and wedges.

18

Detroit-Style: This style of pizza was developed in Detroit, Michigan and is also known

as “square pizza”, similar to Sicilian-style pizza. The crust is the noteworthy feature of

this style, because it is baked twice with a chewy, well done buttery flavor.

Franchise-Style and build-your-own: A franchise pizzeria such as Hungry Howie’s,

Donato’s, PizzaRev or Shakey’s. These pizzas are made for profit only with a business

aspect. The ingredients are usually not fresh and have high sodium content.

Neapolitan-Style: This style of pizza is a thin, single serving pie that is a little crispy and

fragrant. It is made with fresh mozzarella and basil and topped with a San Marzano

tomato sauce that is a little chunky and savory. This style of pizza is usually cooked in a

wood-fired or coal-fired oven with high heat.

New Haven-Style: This style of pizza originated in New Haven, CT and is a variation of

a Neapolitan-style but is larger and also known as “apizza.” The crust is thin, chewy, and

tender with a dark crisp and bitter flavor from a coal-fired oven. This pizza usually has an

oblong shape and is served as a whole pie on a wax paper atop of sheet pan.

New York-Style: A thin-crust, hand tossed dough with large round pizza pies. Usually

sold as a wide and foldable slice that can be eaten while walking. This pizza is typically

served with no more than two toppings and mainly served with condiments such as

oregano, garlic powder, Parmesan cheese, and dried chili pepper.

Sicilian/Grandma-Style: A Sicilian pizza features a thick crust that is chewy in the

center and has a slight crunch on the outside. The Grandma-style usually has a thinner

crust. Both of these styles of pizza use traditional ingredients and can be found with the

19

sauce layered either on top or below the cheese. These pizzas are cooked in a sheet pan

and are cut into squares slices.

St. Louis-Style: This style of pizza has a thin, cracker-like crust that uses no yeast. The

main characteristic of this pizza is the Provel cheese that was originally developed in St.

Louis in the 1950’s and is processed with a combination of cheddar, Swiss and provolone

cheeses. This pizza is cut into a “party cut,” with little squares, rectangles, and wedges,

similar to a Chicago-Style Thin Crust.

Other Pizza: This pizza can be any other style that is not from my selected list. Other

styles may include: Greek Pizza, Trenton Tomato Pies, Old Forge Pizza, Midwest Pizza,

D.C. Jumbo Slice, etc.

4.2. Survey Instrument and Questions

The purpose of the survey instrument was to gather data on American’s

pizza-style taste preferences and how their taste preferences correlate to the geographic

locations of the survey respondents. I used Google Fusion tables software to create an

online survey, which then fed into a Google Document spreadsheet that held all the data.

I titled the form The Pizza Map and then created my survey questions. Designing the

survey question was a process of trial and error. I had to think about what exactly it was I

was interested in learning from the data, and what was the best, most neutral way of

asking these questions. I ended up with nine total survey questions (see Table xx). After I

created my survey question through the Google Document, I linked a Google Fusion

table to the form. The Google Fusion table is a large spreadsheet where I can store all my

20

survey answers and also be able to collect geographic data. A vital step in the creation of

the Fusion Table was to figure out how to create a location-based column that would then

read the geographic data, such as a zip code, and sync this information onto the map.

After I set up the column and labeled it as a “location” field, it was necessary to go back

to the Google Form and set up a script and activate it. I went back to the spreadsheet of

the form and under Tools> Script editor, a project window appears on the form. The

next step was to set up the form’s properties by adding rows with “(name)” and “(value)”

placeholders.

Table 1: Setting up a Google Fusion Property Table

Since I’m collecting location data two more rows need to be added to allow the script to

correctly geo-code my data and place the coordinates into the “location” column that

were created when the fusion table was added.

Table 2: Adding Location Values to the Google Fusion Table Properties Setup

21

After my project properties were created, I saved the project and the next step was to

authorize the script. The code for the script can be found here:

http://kh-samples.googlecode.com/svn/trunk/code/appsscript.js

The code was placed into the Google script editor and to authorize the script it must be

saved and from the Select Function > onFormSubmit, click the Run button and this will

authorize the script to add an authorization token to script project properties. The final

step I had to do was add triggers to the script. These triggers run and sync the script

directly from my survey form to the fusion table.

Once the script was ready to go I performed a handful of personal tests to make

sure the survey worked and was able to collect data in the Google form and Fusion table.

The tests were a success because the data was being stored properly and the spreadsheet

synced to the Fusion table.

22

Figure 2: Pizza Preference Survey Questions Using a Google Form

Below is the actual Pizza survey questionnaire that was used to collect data on

www.geographyofpizza.com

23

24

4.3. Geography of Pizza Website

After a successful survey test, I created a website to have an online presence. I

had a few domain names in mind for the site, but was most excited about

www.geographyofpizza.com. I thought it was the perfect domain name because its

keyword enriched and relevant to my subject of pizza and geography. I looked at many

options for building a website, but ended up using Google Sites because it is free, easy to

embed Google products such Google Docs (Survey Form) and a Google fusion table with

a map.

I created my website logo to look like a pizza compass or a North arrow, with

wedge slices pointing North, South, East, and West. When you first log onto the website,

the home landing page features my selected pizza-styles description chart. I believe these

styles of pizza are the most popular in the U.S.. This pizza-styles descriptive chart (Table

xx) is very detailed and talks about all the components of each pizza-style such as their

ingredients, shape, size, how they are cooked, and authentic images of what they look

like. Also, in the pizza-styles chart, I added a column that features hyperlinks that

connect to sample pizzerias which sell particular pizza-styles. I created these sample

pizzerias in case survey takers were confused by which style to choose. Often people do

not know exactly what style of pizza they prefer, and they only know which restaurant

they frequent. I also included these links to pizzerias in hopes that the establishments

themselves would see their links on my website and be more likely to re-tweet my survey

on Twitter. On the landing page I have a quick summary write-up describing my thesis

project saying I have a dual hypothesis, and provide links to take the survey and

25

contribute VGI. Other links I have featured on this website include an About me page,

and recommended pizza related links.

I decided that anonymity was a very important part of the survey. Many people do

not like to readily share their personal information. We are all weary of websites where

we need to enter our addresses and email. I wanted a zip code to be the only identifying

factor, so that everyone would feel able to safely contribute their data. Often when I

spoke to people in person, the first question they asked was, “Do I have to put in my

email?” I know from personal experience that if I need to enter personal information, I

will most likely stay away from that particular website.

26

Table 3: Regional Pizza-Styles Descriptive Chart on Website

27

28

4.4. Beta Testing Survey Instrument

During beta testing for my survey instrument I sent my website link,

www.geographyofpizza.com, to fifty close friends, fellow students and the CSUN

Geography faculty via Facebook and email, and asked them to participate in beta testing

for the survey. The test went really well and I was excited to see the data of the pizza-

styles they preferred and where they lived. During beta testing, a few people pointed out

that a few of my questions in the survey were moderately biased, as a survey question

already had a default answer in drop down arrow selection. I agreed and immediately

fixed all survey questions that had a drop down menu in order to not display a default

answer. Initially, as part of my survey, I asked about the type of business people would

rather support, Mom and Pop, Franchise or No Preference. I noticed that every single

respondent chose Mom and Pop, which I found to be extremely unlikely. I then changed

the wording of the question to: Where do you usually buy your pizza? I felt this question

would yield a more appropriate unbiased answer. I immediately noticed the effects of the

change of wording. Many people chose establishments such as Dominos, which is a

franchise. I think this shows that people want to support Mom and Pop’s, or want to

appear that they do, but the reality of their choices is quite different. Also, this question

provides additional data indicating if the respondent shops at mom and pop

establishments or franchise pizzeria.

4.5. Facebook Promotion

Since a big part of my thesis is using the power of social media to find answers and

crowd source the data, I used Facebook and Twitter to promote

29

www.geographyofpizza.com. I find that Facebook is an effective tool when trying to

reach out to friends, family, and friends-of-friends, but the scope of Facebook can only

reach so far. Twitter is the more effective tool when trying to reach out or be accessible to

strangers.

Figure 3: Original Facebook Posting about Geography of Pizza

1/30/2013 at 1:24 pm

After this posting went out on Wednesday 1/30/2013 to my friends on Facebook,

throughout the day 12 of 443 of my friends shared my thesis project to other friends.

There were also a number of additional shares by friends of friends, creating a chain

reaction of real people contributing (VGI) Volunteered Geographic Information into the

Pizza-Style Preference Survey.

30

Figure 4: Friends Posting on Facebook about Geography of Pizza

31

Figure 5: More Friends Posting on Facebook about my Thesis

32

Figure 6: My Professor Dr. Graves Posting on Facebook

33

Figure 7: More Facebook Promotion for Geography of Pizza

People on Facebook each shared the link using their own language and generating

their own discussion. I noticed that my website, in general, makes people hungry.

34

4.6. Twitter Promotion

I joined Twitter for the express purpose of promoting my survey. I studied other

Twitter accounts to try to get a feel of what gets re-tweeted. I post lots of maps and

information about pizza-- links I think people will be interested in. Twitter is great for

reaching out to strangers.

Figure 8: First Twitter Tweet Promoting My Thesis Project

Day 1. January 30th 2013

35

Figure 9: Twitter Tweets Reaching Out to Other Geographers and Pizza Places

36

Figure 10: More Tweeting Promotion

My twitter profile is @thepizzamap. I wanted this account to have a theme of

Geography / GIS and pizza, so I searched for my favorite pizzerias on Twitter, and other

accounts related to pizza and/or geography to follow them, hoping they would return the

favor and add me to build a following. I started tweeting to famous pizzerias and asked

them to help promote my website and the pizza survey instrument. I was very pleased

that a few well known pizzerias in the U.S. shared my pizza thesis survey and began to

37

follow me on Twitter. It was very exciting to have places like Lou Malnati’s in Chicago,

Buddy’s Pizza in Detroit, Sac’s Pizza in New York, and Frank Pepe in New Haven all

following you and talking about your project.

My wife and a couple friends have also tweeted on this same day Jan 31st 2013,

just to get things moving. One of the coolest things about Twitter is you can reach out to

people directly and share with them. This is exactly what I did to talk about my project

and reach out to other Twitter handles around the U.S. and internationally. The more you

tweet the better and if someone else re-tweets your tweet it’s even better.

The second day of tweeting, I reached out to other geographers and geo-bloggers

such as Matt Rosenberg who is also a CSUN alumni and a writer for About.com. He and

other Geo-Bloggers helped me to get more data because they tweeted about my site.

Figure 11: Other Geographers and Pizza Places Tweet About Geography of Pizza

38

Chapter 5 - Results

5.1. Survey Results From Social Media Promotion

Figure 12: Day 1 Results in Google Map of Facebook Shares and Tweets

Day 1 after Facebook (12) sharing and minimal tweeting = (127 survey takers)

39

Figure 13: Day 2 Results in Google Map of Facebook and Twitter Shares

Day 2 Major Tweeters 3 (UK Twitter Follower GeoBlogs)

Below is a table representing the dates from the first original posting on Facebook

& Twitter on Wednesday 1/30/2013 through the next 3 days of survey responses. The

first day of promotion has the highest response rate of 26 VGI contributors, while

Saturday came in second with 14 responses during the first week of promoting

www.geographyofpizza.com.The lower Saturday numbers are probably due to it being a

weekend. I have found that the most effective times to post are during working hours.

40

Table 4: A Table of the Results for the First Week of Promoting on Facebook &

Twitter

Geography of Pizza Promotion through Reddit & MetaFilter WebBlog

Without my knowledge, a couple people posted a link to my website on Reddit &

Metafilter, which are both high traffic weblog sites. These postings definitely helped

contribute survey responses quickly. What is really fascinating about these postings are

the comments people made about pizza. You can read and see peoples true identity and

opinions about pizza from all over the world, and learn about their favorite toppings and

styles they prefer most.

5.2. Target Emailing (2/11/2013 - 2/25/2013 15 Days)

As I watched my map grow with VGI day to day, I was able to see which states

needed additional data. The majority of data contributed was coming in daily from

California and the Tri-state area of the East coast, because these are places either where I

live and go to school or where my family and friends are currently located. After twelve

41

days of promoting the pizza survey, the data available was insufficient. I decided to

personally ask friends who were from areas of the country that had an insufficient data to

help me get survey responses. For example I had very few responses from the Plains

states, so I emailed friends from North and South Dakota and Nebraska and asked them

to help me promote my survey. People were very receptive and willing to help. This

method did increase my data load and helped me get the word out even more for the

states that needed data, but it still wasn’t enough. I then decided to email fellow

geographers located in geography departments at universities and colleges across the U.S.

to help partake in my survey.

I constructed a professional letter describing myself and my thesis plan. I

explained that the survey is completely anonymous and I asked for help in sharing my

website with other faculty and students within their Geography Department. I came a

across a website that offered many links for departments in the U.S. The website list was

very detailed because it listed each state in alphabetical order and supplied links to the

universities and colleges that have Geography Departments.

I created an Excel sheet to help keep track and store all targeted states that

needed data and I emailed several geography departments in alphabetical order by state

name. My first day of emailing was on February 11, 2013, and the states included

Alabama, Alaska, Arizona, Arkansas, Colorado, Delaware, Washington D.C., Idaho,

Iowa, Kansas, and Kentucky. My second day of emailing continued on the 13th of

February, reaching out to Georgia, Hawaii, Louisiana, and Maine. Then again on the 14th

of February, emailing Nevada, and New Hampshire Geography Departments. On

February 21st, this was my final day of emailing departments in states where I needed

42

data for my pizza map. The states included North and South Dakota, Montana, Wyoming,

and Vermont.

I found that this method of targeting geography departments was extremely

successful. I was blown away by the support I got from these strangers in various

geography departments across the country. Professors loved the idea and passed the

survey along to their students. Of all the people I contacted, I only got one negative

response. A professor in Vermont said it was unheard of to contact departments in this

manner and that I was conducting research in a highly unprofessional way. She thought it

was some kind of spam. Again, this was an isolated response. All other professors were

incredibly helpful, eager and kind. Perhaps the overarching scope of this thesis is to

propose that the old ways of doing research are over, and it is time to embrace all the

tools available to us today. Facebook and Twitter by themselves were not enough to get

me the data that I needed, but reaching out through personal, targeted emails helped to fill

the gaps in data.

The following states contributed more than 10 responses during the Target Emailing

campaign:

Table 5: States Targeted During Email Campaign

43

The first email I sent out to target Geography Departments to take my Pizza

Preference survey was on Monday, February 11th, 2013. The next day, February 12th

2013, I received 157 total survey responses and out of this number, 121 of those

responses came from targeted Geography Departments. The schools with the highest

response rates were from Kentucky, with 70 responses and Kansas, with 34 responses.

Other states whose geography departments contributed data were Colorado, with 26

responses on February 13, 2013, Louisiana, between the days of February 14th and 15th

2013 with a combined total of 34 responses, and New Hampshire, with 21 responses on

2/22/2013.

Table 6: Non-Targeted States During Email Campaign

California was a major contributor of survey responses, especially during February 13-

15th, with a combined total of 101. Other states which contributed over 25 responses

through the survey instrument, while not being targeted are Illinois, Missouri, New York,

and Texas.

44

Table 7: Results for Most Preferred Pizza-Styles from Targeted States

2/11/13 - 2/25/2013

New York-style pizza was the most preferred with 79 votes followed by Neapolitan-style,

which was ranked 2nd with 67 votes, and All-American-style ranked 3rd with 60 votes.

The following chart, created with Excel, is similar to the chart above, but these

answers were from states that were not targeted during the Geography Departments email

campaign.

45

Table 8: Results of Most Preferred Pizza-Styles from Non Targeted States

New York-style is the most preferred again with 175 votes, followed by Neapolitan-style

with 106 votes, and in third is Chicago-style Deep dish.

5.3. Organizing Total Survey Responses for Maps

As of July, 7th 2013, 1,516 people have contributed VGI worldwide through the

survey instrument stating their favorite pizza-style and where they are from. One of my

goals for this project was to create maps that told a story of which pizza-styles people

preferred most in the United States. The majority of the data collected through the survey

instrument were people from the United States. For the purposes of this thesis, I will only

analyze the data from the United States.

I decided to have a cutoff of 1,377 responses of the overall data from people

living in the United States to use to make the representative maps. The maps I chose to

46

create are a main United States pizza-preference map, pizza-style percentage maps and

radial flow identity maps based on where a person is from and where they are living now.

Figure 14: Google Map Results of 1,377 Responses Through The Pizza Preference

Survey

U.S. Pizza Preference Survey Map 7/7/2013

As you can see from the map above, there is pizza data contributed from each

mainland U.S. state. Every time a person takes my survey, that data is stored into a

Google Doc and is also reflected into this map with a red pin.

47

Below, is part of a spreadsheet table I created to help store response rates for each

pizza style on a state level. The pizza preferences with the most responses or votes wins

that state.

48

Table 9: A Total of 1,377 U.S. Responses for Pizza Style Preferences by State

Total U.S. Survey Responses

49

Table 10: The Most Preferred Pizza-Style in the U.S. from the Total Responses

New York-style pizza is the most favored pizza in the United States with 367

votes. Neapolitan-style pizza came in 2nd with 229 preferred votes, followed by Chicago

Deep/Stuffed pizza with 195 preferred votes.

While looking at the complete U.S. survey response table above, you can see that

for the state of Connecticut there are responses for six of the pizza-styles that people

prefer. However, New Haven-style pizza received the highest rate (17), therefore that

particular pizza-style preference will “win” the state of Connecticut. Another example

would be the state of Pennsylvania. People have responded from Pennsylvania indicating

a preference for seven different pizza-styles, but New York-style got the most votes (10),

so that pizza-style wins the state of Pennsylvania.. Some of the states need more data

because the response rates for the pizza-styles are tied, therefore a winning pizza-style

could not be determined for that state. Again, only 1,377 survey responses in the U.S.

were used to create these maps. I do intend to incorporate all additional data to new maps

at a future time.

Using the Excel table (Table 9) above, I created the main Pizza Preference U.S.

Map and imported it into ArcGIS 10.0 and joined together with a U.S. boundary shape

file. After the join, I exported that shape file, saved it as a zip file, which was converted

into a GeoJSON format. I linked the GeoJSON file to a JavaScript, HTML, and a CSS

50

script that were part of the leaflet API map, which I then used to create the interactive

digital map (Fig. 14) below of the main Pizza Preference U.S. map.

Figure 15: Main Pizza Preference U.S. Map

1,377 U.S. responses

The results for the Main Pizza Preference U.S. Map (1,377 responses) show that

New York-style pizza is the most preferred in the majority of the states. What is also

interesting in this map are that almost all location specific pizza-styles are the preferred

style in their representative state. For example, in Illinois, Chicago Deep is the most

preferred. In Connecticut, New Haven-style pizza is the most preferred. In Michigan,

Detroit-style is most preferred, and in Missouri, St. Louis-style is most preferred. The

only state that did not win its pizza-style is California, as New York style won the state

with a higher response rate. Also, the states that prefer All-American pizza are located in

51

the middle of the country presumably because that is where the big franchise pizza places

were founded and they maintained their stronghold.

The method of organizing my data in an Excel spreadsheet helped me to make

percentage calculations for each pizza-style by state. The calculations I performed were

taking the total number of responses for each pizza-style and dividing that number to the

overall total responses for that particular state and multiplying that by 100. In order to

create my percentage maps representing each pizza-style, I exported the Google Doc into

a separate Excel spreadsheet and started to manipulate the data. I created a column that

stored all of the states, followed by a column that had a total number of survey responses

for that particular state, and the remaining columns were the pizza-styles All-American,

California, Chicago Deep, Neapolitan, etc.

52

Table 11: Example Excel Spreadsheets to Create Pizza Preference Percentage Maps

I used the percentage numbers from (Table 11) to create maps for the eleven pizza-styles

selected for this project.

53

Figure 16: All-American Pizza U.S. Preference Percentage Results Map

A total of 124 people have chosen All-American pizza. This style of pizza is most

popular in the middle of the United States. South Dakota had a total of eighteen responses

and seven of those are for All-American pizza which is equivalent to (38.9%). Oklahoma

had a total three responses and one response was for All-American for a (33.3%). Utah

had a total of nine responses with three choosing All-American giving that state a

(33.3%). This map also reminds me of the republican red state map, because most of the

All-American pizza lovers are located in the red states.

54

Figure 17: California-Style Pizza U.S. Preference Percentage Results Map

This California percentage map is very interesting because the people that prefer

this pizza-style the most are not in California. This pizza is not very popular in other

states. The highest percentage for California-style pizza comes from the state of

Delaware with a total number of responses one out of two choosing California pizza to

give this state (50%). Arkansas had a total of three responses and one of those was for

California pizza for (33.3%). New Mexico only had a total of three responses and one of

those responses was for California pizza for (33.3%). I have no real explanation for these

results, as the common philosophy of California style is that it is based on abundance of

55

fresh produce. A total of 90 people chose this style through the survey instrument across

the U.S..

Figure 18: Chicago-Deep/Stuffed Pizza U.S. Preference Percentage Results Map

Chicago Deep / Stuffed pizza is pretty popular all across the U.S. with a total

number of 195 people preferring this style of pizza. The states with the highest

percentage rates which prefer Chicago Deep pizza are Rhode Island at (50%), with one

response out of two total. Illinois at (39%), with twenty-eight responses out of seventy-

56

two. Georgia is at (33.3%) with a five responses out of fifteen total. Oklahoma, and

Montana are (33.3%) also.

Figure 19: Chicago-Thin-Crust Pizza U.S. Preference Percentage Results Map

Chicago-thin-crust pizza has a solid popularity across the U.S. with a total of 101

preferring this style. Minnesota has the highest percentage rating at (25%), with seven

responses out of twenty-eight chose Chicago-thin-crust pizza. Illinois is at (24%), with

seventeen responses out of seventy-two choosing Chicago-thin pizza.

57

Figure 20: Detroit-Style Pizza U.S. Preference Percentage Results Map

A total of 35 people from the survey results prefer Detroit-style pizza. Michigan

has the highest percentage preference for Detroit-style pizza at (42%), with ten responses

out of twenty-four choosing Detroit pizza. Iowa came in second at (40%), with two

responses out of five for Detroit pizza. Oklahoma is at (33.3%), with one out of three

choosing this pizza. Detroit-style pizza seems only popular in the middle of the country.

Only a couple states in New England have a taste preference towards this style, perhaps

because it’s very similar to a Sicilian square pizza, which is also popular in New

England. The north and west of the country have no preference for this style of pizza at

all, probably because they’ve never been exposed to it.

58

Figure 21: Franchise-Build-Your-Own Pizza U.S. Preference Percentage Results

Map

Franchise-build-your-own pizza seems moderately popular in many states across

the U.S. with 52 total people choosing this style. The highest percentage rate for this style

of pizza was in Mississippi at (50%) with one out of two responses choosing Franchise-

BYO pizza. Other states with a high percentage rate for this pizza include North Dakota

with three out of twenty-five responses choosing this pizza for (11%). South Dakota,

Utah, Idaho, Alabama, and Oregon, all with over 11% of a preference.

59

Figure 22: Neapolitan-Style Pizza U.S. Preference Percentage Results Map

Neapolitan pizza is the second most popular style of pizza behind New York with

229 people choosing this style of pizza. The highest percentage preference rate came out

of the state of New Mexico at a high (67%). Only two responses out of three choose

Neapolitan for NM. Delaware is at (50%) with one out of two responses choosing

Neapolitan pizza.

60

Figure 23: New Haven-Style Pizza U.S. Preference Percentage Results Map

New Haven-style pizza is very popular in the New England area of the United

States with Connecticut having the highest percentage of (52%). Seventeen responses out

of a total thirty-three chose New Haven pizza in CT. Rhode Island has the second highest

percentage at (50%), with one out of two responses choosing New Haven pizza. Iowa

came in third at (20%), with one response out of five choosing this pizza. This style of

pizza is only preferred in select states of the country. Most states show no preference or

knowledge of this type of pizza. A total of 49 people chose this style of pizza across the

U.S.

61

Figure 24: New York-Style Pizza U.S. Preference Percentage Results Map

New York-style pizza seems to be the most popular and preferred pizza across the

country with 367 people choosing this pizza style. New York has a (59.2%) preference

rate, with forty-five responses out of the total seventy-six choosing New York pizza.

West Virginia came in first with a 100% (although there were only two responses from

that state choosing NY pizza), followed by Nevada at (75%), with three responses out of

four choosing NY pizza and Arkansas at (67%), with two responses out of three. These

percentage maps are skewed, as some high percentage states had a small number of votes.

62

Figure 25: Other Pizza U.S. Preference Percentage Results Map

There were 45 people that chose the option of Other pizza. A high percentage rate

came from the state of Maine at (16.7%), with one response out of six choosing Other

pizza. Vermont came in second at (14.3%), with one response out of seven choosing

Other pizza. My guess would be that Greek pizza would be their option because that style

of pizza is very popular in the New England area.

63

Figure 26: Sicilian-Style Pizza U.S. Preference Percentage Results Map

There were 40 people that chose Sicilian / Grandma pizza across the United

States. Maine has the highest percentage for this particular style at (16.7%), with one

response out of six choosing Sicilian/Grandma pizza. Alabama came in second at

(11.1%), with one response out of nine choosing this pizza-style.

64

Figure 27: St. Louis-Style Pizza U.S. Preference Percentage Results Map

A total of 50 people chose St. Louis style pizza. The highest percentage for this

style of pizza is in Montana at (66.7%), with two responses out of three choosing St.

Louis-style pizza. Iowa came in second at (40%), with two responses out of five choosing

St. Louis pizza. Missouri had a total of seventeen responses out of forty-five choosing St.

Louis-style pizza giving the state a (38%) preference.

65

5.4. Results for Identity and Place through Pizza

One of the best ways to depict people’s identity through the lens of pizza was to

create radial flow maps with the location data collected through the survey instrument.

Radial flow maps show the movement of some phenomenon, normally goods or people,

from one place to another (Akella, 2011). One aspect of the survey involved the zip code

of where the survey taker was born and where he or she lives now. I was able to extract

these zip codes from both locations (if two zip codes were provided) and geo-code them.

I put coordinates into a format spreadsheet like the one below (Table:12), which I then

added in ArcGIS to run the XY to Line tool, used to create radial flow maps.

Table 12: XY to Line Tool Excel Format for Radial Flow Maps

66

The POINT_X and Y columns of the spreadsheet above represent where those

people are originally from and the DESTINATION_X and Y columns represent the

columns where the people currently live. The Pizza column represents their pizza-style

preference. Radial Flow maps are a great way to depict people’s identity through pizza

because they are able to show movement. In this case, each pizza-style was represented

by a particular color line featured on the map and in the legend. These particular maps do

not represent all respondents. They only represent people who were born in the region or

state and then moved to a different location. The point of these maps is to see if a

regional identity exists around pizza and does that identity remain even after moving

somewhere else. In other words, my interest in the statistics are twofold: 1) How and

where do people move, and 2) Do their food preferences remain the same? Each of the

lines on the maps are a person’s response and their preferred pizza-style. The maps have

many different color lines moving out of the particular state or region I chose to study.

One color line represents that state or region’s location-specific pizza-style and the other

color lines represent the remaining pizza-styles that are part of this study. When a

different color line flows out of the state that is being represented, this could be

interpreted as a person who changed their identity (food preference) by moving out of

that state and preferring another pizza-style in their present location.

I’ve created six radial flow maps for the states that feature the popular location-

based pizza-styles: California, Connecticut, Illinois, Michigan, Missouri, and New York.

I also created a 7th radial flow map that was based off the All-American Pizza-style

Percentage Map from the previous pages. I’ve selected a block of states that are in the

middle of the U.S. that have a have a high percentage preference for All-American pizza.

67

These states include North Dakota, South Dakota, Nebraska, Kansas, and Oklahoma. I

created this 7th radial flow map for the purpose of seeing where the people of middle

America were moving to and how (or if) their pizza preference has changed.

Figure 28: Middle America Pizza Preference Radial Flow Map

Table 13: Middle America States Radial Flow Map Stats of Survey Responses

A total of 84 respondents came out of these five states for the Middle America

Radial Flow Map and All-American pizza is the number one preferred style with 21

68

votes. New York-style had 15 respondents and Chicago-Deep with a 12. Middle America

has respondents for every pizza style except for Sicilian / Grandma style. This map is a

great example of seeing the changes within peoples food preference. In one case, I

noticed right away a light blue line that represents New Haven-style pizza leaving

Oklahoma and moving to New England. It is just fascinating to see how place can change

a person’s food preference.

Figure 29: California Pizza Preference Radial Flow Map

Table 14: California Radial Flow Map Stats of Survey Responses

69

California reported an extremely large amount of responses (209) and really

represented a wide range. Many people who were born in California and preferred

Neapolitan style remained true to that preference, even after they moved. Many people,

also, though, reported preferences for a vast array of pizza styles. All in all, perhaps

because of the vast size of the state and its lack of a pizza “identity,” those born in

California are literally all over the map.

Figure 30: Connecticut Pizza Preference Radial Flow Map

Table 15: Connecticut Radial Flow Map Stats of Survey Responses

70

In the case of CT, most people who moved, moved within the small state of

Connecticut or to nearby New England states. These people maintained their preference

for New Haven style pizza. The ones who moved further either maintained a preference

for New Haven, or tended to pick New York style or Neapolitan style. It seems that a

majority of respondents maintain their Connecticut identity by remaining loyal to New

Haven Style pizza.

Figure 31: Illinois Pizza Preference Radial Flow Map

Table 16: Illinois Radial Flow Map Stats of Survey Responses

71

The Illinois radial flow map is one of my favorites because the response rates look

like a giant star across the U.S. and the amount of people selecting Chicago Deep and

Chicago-thin pizza dominates with a percentage of 77% out of the total responses of 94.

It’s safe to say that the people of Illinois are proud of their pizza and maintain this

preference no matter how far they move. More than any other style, these styles seem to

represent “home.”

Figure 32: Michigan Pizza Preference Radial Flow Map

Table 17: Michigan Radial Flow Map Stats of Survey Responses

72

Though on a much smaller scale, Detroit-style pizza seems to remain a preference to

locals after they move to other states. Respondents either remain true to Detroit-style or

their neighbors, Chicago-style pizza. If going by these maps, it seems the Midwest breeds

people proud of their pizza roots.

Figure 33: Missouri Pizza Preference Radial Flow Map

Table 18: Missouri Radial Flow Map Stats of Survey Responses

A total of 45 people who were born in Missouri and then moved took the survey

and 19 of them still prefer St. Louis Style pizza. The 19 responses is equivalent to 42% of

73

the total count. This is a pretty impressive number and seems to match the trend of the

loyal Midwesterner. New York-style pizza is the 2nd most preferred in MO with 10

responses or 22%. Chicago-Deep came in 3rd with 9 responses or 20% of the total.

Figure 34: New York Pizza Preference Radial Flow Map

Table 19: New York Radial Flow Map Stats of Survey Responses

This New York radial flow map is another one of my favorites, because you can

clearly see the domination of responses that prefer New York-style pizza. A total of 105

people responded through the survey instrument from New York and 76 responses or

74

72% of the total count prefer New York-style pizza. Many people in this map were

originally from NY and migrated west or to Florida. In addition to being the most

preferred style country-wide, the slice of pizza also remains most preferred to those who

move away from NY.

75

Chapter 6 - Discussion

In less than half a century, Pizza has gone from a specialty food known only in

the scattered Little Italy’s in America’s biggest cities, to a beloved All-American food

staple, standing tall alongside hot dogs, hamburgers and Apple Pie. Unlike those other

All-American favorites, pizza has developed a great variety of regional variations. This

project has uncovered a number of interesting findings regarding the regional variations

in pizza style and the taste preferences for pizza across the US.

One of the major findings of this study is the somewhat surprising national

preference for New York City’s Slice style pizza. It should be noted though that this

finding may be slightly biased because of the nature of social networking, and the effect

of the author’s birthplace on the East Coast. The pizza-style preferences data may change

as more people from a wider geographic range contribute VGI through the survey

instrument in the future.

Another interesting, but perhaps not surprising result is that many respondents

appear to like pizza styles originating in or near their place of residence. New Yorkers

prefer their style, but also the other popular location-based pizza-styles, which include

Connecticut, Illinois, Michigan, and Missouri, all logically and predictably preferred their

local regional styles.

Another important finding uncovered by this study is the utility of social

networking sites as data collection tools. Facebook and Twitter definitely worked well in

terms of their capacity to solicit responses from respondents in far flung locations. The

weakness of social media as a tool for collecting survey data is that it has a tendency to

replicate existing human networks. As a result, there was a spatial bias in the responses

76

collected by the survey instrument. California and the greater New York area were

oversampled, and much of the interior of the United States were under sampled.

Additional effort was needed to help craft a more representative sample. The

Target emailing campaign that focused on individual Facebook friends by location and

various Geography Departments throughout the states with missing data, was very

effective way to gain survey responses. I especially remember how helpful the

Geography Departments in Kentucky and Kansas were, with a very high participation

response rate on February, 12th 2013 with a total of 104 responses combined.

Perhaps the most surprising finding was promoting this project on Twitter was

that I could reach out to anyone such as celebrities, famous pizza places, and popular

Geography Blogs and GIS companies. A retired professional hockey player, Darren

McCarty, of the Detroit Red Wings re-tweeted about my website and helped gain a

survey response rate in the Michigan area. A handful of popular and well respected

pizzerias in the United States tweeted about my project as well. It’s amazing how fast

word travels online and it was exciting to see my Google map grow with red pins of

responses each day from locations around the world.

MetaFilter, a very popular community weblog online posted a hyperlink to my

website that features a feed of ongoing comments of real opinions of the people’s pizza-

style preferences. It’s amazing to read through all the comments of these people, because

they are from all over the world that clearly had different beliefs and thoughts about

pizza.

Volunteered Geographic Information has enormous potential as a data resource

because it is contributed data from real people around the world. The fact that I was able

77

to create a survey instrument and collect data from people about their pizza preferences

by locations was amazing. There are so many subjects that can be touch on by VGI and I

hope to see more in the future. The Waze app is a great example that VGI really works

and we should consider using more of it to help make informative decisions.

My radial flow identity maps are very preliminary and also very flawed. You need

almost a separate, and more in-depth survey, but they do seem to show that there are

some areas of the country who strongly maintain their identity. They are also fascinating

to see right away where people move to from one point to another.

I will continue to update Geography of Pizza in the future and hope to have more

responses and more accurate maps. Geography of food and identity, in this ever mobile

world, is a key to learning about how food connects us to place and to larger concepts

like the self, and home.

78

REFERENCES

Adams, P.C. 2009. “Geographies of media and communication.” New York: Wiley-

Blackwell.

Akella, M. “Creating Radial Flow Maps with ArcGIS.” 2011.

http://blogs.esri.com/esri/arcgis/2011/09/06/creating-radial-flow-maps-with-arcgis.

Backstrom, L., Huttenlocher, D., and Kleinberg, J., 2006. Group formation in large social

networks: membership, growth, and evolution. Proceedings of the 12th ACM SIGKDD

International Conference on Knowledge Discovery and Data Mining, 15, 44-54.

Benedetto, A. Di.”Italian Immigration to the United States.” U.S. Institutions Survey

Papers (2000) pgs 1-5.

Buonassisi,V. “Pizza Boston”: Little Brown Press, 1984.

Cinotto, Simone. “Leonard Covello, the Covello Papers, and the History of Eating

Habits Among Italian Immigrants in New York.” The Journal of American History

(2004):pgs 498-505.

Cinotto, S. 2004. “Now That’s Italian!”: Representations of Italian Food In American

Popular Magazines, 1950-2000.

Cinotto, S. “The Taste of Place: Food in the Narratives of ‘America’ and ‘Italy’ by

Italian Immigrants of New York, 1920-1950,” in America and the Mediterranean, eds.

Massimo Bacigalupo and Pierangelo Castagneto (Turin: Otto, 2003), 145-154.

Coleman, J. D., Y. Georgiadou, J. Labonte. 2009. Volunteered Geographic Information:

the nature and motivation of producers.

Goodchild, F. M. 2007. “Citizens as Voluntary Sensors: Spatial Data Infrastructure in

the World of Web 2.0.”

Gordon, P.J. , J.L. Sterrett. 1999. “Geographic Preference for Fast Food,” Proceedings

of the Academy of Marketing Studies, 1-4.

Gurney, E. 2011. “From Spaghetti and Meatballs to Pizza: The Invention of Italian-

American Restaurant Staples by Immigrants, for Immigrants.

Hardy, D. 2008. Discovering behavioral patterns in collective authorship of Place- based

information.

Lindenfeld, L. and L. Silka. 2011. “Growing Maine’s Foodscape, Growing Maine’s

Future.”

Lord, Eliot, The Italian in America. Freeport: Books for Libraries Press, 1905.

79

Meier, P. P. 2011. “Verifying Crowd sourced Social Media Reports for Live Crisis

Mapping: An Introduction to Information Forensics.”

Miller, Hanna. “American Pie.” American Heritage Magazine 57, NO 2 (April-May

2006).

Reinhart, P. American Pie My Search for the Perfect Pizza. Ten Speed Press, 2003.

Shortridge, B. 2003. “A Food Geography of the Great Plains,” The Geographical

Review.

Singer, S. American Pizza Styles and Components. Webster's Digital Services, April 27,

2011.

Sui, D. & M. Goodchild. 2011. “The Convergence of GIS and social media:

challenges for GIScience”, International Journal of Geographical Information Science.

Varasano, J. Varasano's Pizzeria. http://www.varasanos.com/PizzaRecipe.htm. 2006.