Geography of Instagram-Based Business Network Study

Geography of an Instagram-Based Business Network Dane Estok Advisor: Chris Fowler, PhD 1 Penn State University GEOG 596A SU1 2015

Outline Hypothesis Quick history of Company X Proposed methodology Project timeline Anticipated results 2

Capstone Hypothesis The closer a person lives to a featured family, the more likely they are to buy a shirt from company X . Tobler s first law of geography Everything is usually related to all else but those which are near to each other are more related when compared to those that are further away. Investigation of how the physical and social elements combine 3 Geographic concentration of sales vs. geographically-boundless internet -- what does that look like? What is happening? Tobler W., (1970) A computer movie simulating urban growth in the Detroit region . Economic Geography, 46(2): 234-240.

Granovetters, The Strength of Weak Ties (1973) Strong vs. weak ties A combination of the amount of time, the emotional intensity, the intimacy, and the reciprocal services which characterize the tie. Weak ties have most influence for diffusion of information between two networks Whatever it is to be diffused can reach a larger number of people, 4 and traverse greater social distances when passed through weak ties rather than strong. Granovetter, M. S. (1973, May). The Strength of Weak Ties. American Journal of Sociology, 78(6), 1360- 1380.

Rapoport and Horvath, 1961 A Study of a Large Sociogram Junior High students in Michigan List eight best friends in order Random samples taken/Network mapped First and Second choices = smallest network (strong ties) Seventh and Eighth choices = largest network (weak ties) 5 Rapoport, A., & Horvath, W. J. (1961). A STUDY OF A LARGE SOCIOGRAM. Behavioral Science, 6(4), 279-291. Retrieved from http://search.proquest.com/docview/60560487?accountid=13158

Judith Blau, Unpublished (1980) Children s psychiatric hospital in NYC Integration of various departments Low turnover and high moral why? Weak ties 6 Granovetter, M. S. (1983). The Strength of Weak Ties: A Network Theory Revisited. Sociological Theory,1 201-233.

Breiger and Pattison, (1978) Three types of Ties Social = strong Community affairs = stronger than business, weaker than social Business-professional = weak 7 Granovetter, M. S. (1983). The Strength of Weak Ties: A Network Theory Revisited. Sociological Theory,1 201-233.

Mok et al. (2009) Does Distance Matter in the Age of the Internet? Distance impact on social ties pre/post internet Previous study in East York, Toronto (1978) Focused on email and face-to-face contact to maintain relationships Email is generally insensitive to distance Face-to-face Within 5 miles strong contact 50-100 miles substantial decline in contact No major change in contact from 1970 2000s Phone contact has slightly increased Diana Mok, Barry Wellman, and Juan Carrasco Does Distance Matter in the Age of the Internet?Urban Studies November 2010 47: 2747-2783,

Brown and Reingen (1987) Social Ties and Word-of-Mouth Referral Behavior WOM Information flow for three piano teachers Macro level Weak ties as bridges Micro level Strong ties more influential than weak Jacqueline Johnson Brown and Peter H. Reingen Journal of Consumer Research Vol. 14, No. 3 (Dec., 1987), pp. 350-362 Published by: Oxford University Press Stable URL: http://www.jstor.org.ezaccess.libraries.psu.edu/stable/2489496

Saxton and Wang (2014) The Social Network Effect: The Determinants of Giving Through Social Media Donations are different online than offline Social Network Effect Small donations Fundraising related to organization s Web capacity Social pressure and attention-getting ideas Improve one s standing in social network Impulse donating More prone to donate to health causes Donation discrepancy regory D. Saxton and Lili Wang The Social Network Effect: The Determinants of Giving Through Social MediaNonprofit and Voluntary Sector Quarterly October 2014 43: 850-868, first published on April 24, 2013

Takhteyev et al., Geography of Twitter networks (2012) Social contact benefits from physical proximity. Basic fact of social life Internet = distance is dead? (Cairncross, 1997) This study examined the influence of the following on the formation of social ties on Twitter: Geographic distance National boundaries Language 11 Frequency of air travel Cairncross, F., 1997. The Death of Distance: How the Communication Revolution is Changing our Lives. Harvard Business School Press, Boston. Takhteyev, Y., Gruzd, A., & Wellman, B. (2012). Geography of Twitter networks. Social Networks 34, 73-81.

Takhteyev et al. Study Findings 60% of Twitter ties are unidirectional Distance IS NOT dead 39% of ties are shorter than 100km Ties up to 100km are more common than if formed at random Ties decrease with distance Number of airline flights is best indicator of non-local Twitter ties Expected since travel maintains face-to-face contact, which then leads to social media ties. 12 Takhteyev, Y., Gruzd, A., & Wellman, B. (2012). Geography of Twitter networks. Social Networks 34, 73-81.

Hypothesis The closer a person lives to a featured family, the more likely they are to buy a shirt from company X . Tobler s first law of geography Everything is usually related to all else but those which are near to each other are more related when compared to those that are further away. Investigation of how the physical and social elements combine 13 Geographic concentration of sales vs. geographically-boundless internet -- what does that look like? What is happening? Tobler W., (1970) A computer movie simulating urban growth in the Detroit region . Economic Geography, 46(2): 234-240.

Capstone Hypothesis Data might contradict hypothesis If so, why? Does the internet diminish/eliminate the effect of distance for company X ? 14

Additional Capstone Questions Further investigating during project: How do posts by influential bloggers affect the sales of Company X ? Are sales growing around the featured family each month, or are they being influenced by other factors such as: 15 Employee location Influential blogger location Auction vendor location?

Data Sales data from June 2014 to June 2015 Customer address Date of sale Amount spent Family locations Influential blogger locations 16

More Data To test my hypothesis: Geocoded Shapefiles Family location Sale location Employee location Influential blogger location 17 NHGIS.ORG 2010 Census Tract Population Data

18

19

Proposed Methodology I To test my hypothesis: Area-weighted share of the population making a purchase equation used for each family location. Create a graph where: x-axis = distance increment (10,20,50, 100, 500, 1000, etc ) y-axis = Area-weighted share of the population making a purchase 20 N = # of people within distance-band that bought shirt / Area of circle N / (Population of all census tracts / Area of all census tracts) = y axis Spending habits with distance?

Distance Decay Graph 21 Accounts for uneven distribution of people

Proposed Methodology II Within each distance band: Measure how the number changes Measure the rate of decline with distance Distance Decay Distance decay is a geographical term which describes the effect of distance on cultural or spatial interactions. 22 The distance decay effect states that the interaction between two locales declines as the distance between them increases. Once the distance is outside of the two locales' activity space, their interactions begin to decrease. Distance decay coefficient to be determined. Wikipedia contributors. "Distance decay." Wikipedia, The Free Encyclopedia. Wikipedia, The Free Encyclopedia, 13 May. 2015. Web. 2 Aug. 2015.

Proposed Methodology III Python Script to automate processes For each distance-band for each featured family Make it dynamic, so it can gather data with specific dates, specific families, purchase amounts, distance, etc 23

Conferences Western Regional Science Association February 14-17, 2016 Submission deadline (rough draft): October 15, 2015 24 Association of American Geographers March 29 April 2, 2016 Abstract submission deadline: October 29,2015

Proposed Timeline August 4, 2015 September 27, 2015 GEOG 596A presentation Complete rough draft of paper August 9, 2015 October 11, 2015 Have all shapefile data correct and ready to go Complete final rough draft for Complete IRB Human Subject Research submission application October 14, 2015- August 23, 2015 Submit to WRSA Determine decay coefficient October 28, 2015 25 Complete Python Script Submit to AAG August 31, 2015 Complete initial testing/corrections to python script September 6, 2015 Have all graphs completed

Anticipated Results Uncertain There are many possible variables at play regarding the consumer and company x . Geographic concentration of family? Established following of family before company x introduces family 26 Hypothesis most likely supported Analysis might show anomalies Not all of company x activity is factored in Auctions

27 Questions?!