Social Networks Assignment 1: University-Industry Collaboration and Social Networks U. Matzat

1
Social NetworksAssignment 1
University-Industry Collaboration and Social
Networks U. Matzat
2
Course design

• Aim knowledge about concepts in network theory,
  and being able to apply them, in particular in a
  context of innovation and alliances
• Intro
• Network theory and background
• Network methods
• Small world networks
• Business alliances as one example of network
  strategy
• Assignment 1 analyzing university-company
  networks (ego-centered networks)
• Assignment 2

3
A Recap Two Approaches to Network Analysis

• Complete Networks
• The relations of all actors are taking into
  account
• Existing data (archival)
• Case study (f.i. a family or this class)
• Where is the boundary of the network?

• Ego centered network
• Network from the perspective of the actor
• Extension of traditional survey instruments
• Random sampling possible
• Statistical analysis possible

4
The Assignment

• We analyze what are the barriers hindering
  collaboration between university researchers and
  companies from being (more) successful?
• With collaboration we mean a working
  arrangement between at least one university
  researcher and one or more employees of a
  commercial firm with the intention to develop
  further an idea or a preliminary product of
  university researchers so that it can be brought
  to the market.

5
The Assignment

• Think of ways in which the ego-centerd network of
  the researcher may influence
• the success of collaboration with a company
• Derive two hypotheses. Make use of ideas of
  Burt's (2001) theory of structural holes or of
  Coleman's (1988) closure theory.

6
The Assignment

• Test those hypotheses using the dataset we
  provide.
• SPSS data set collaboration_2005.sav
• File contains 150 cases (rows)
• information about the researcher, his/her
  university position, his/her faculty, his/her
  ego-centered network data, and characteristics of
  the collaboration.

7
The report

• Structure and content of the report must resemble
  a scientific article
• Introduction
• Brief discussion of the topic university-industry
  collaboration and social networks.
• Theory
• Present a theoretical framework and derive two
  hypotheses.
• Method
• Describe the data, measurements, and statistical
  methods.
• Results
• Describe the networks in detail (e.g., size,
  density) depending on your theory
• Present the results of the analysis, by means of
  table/picture, and a verbal clarification of the
  table/picture.
• Discussion/Conclusion
• What do the results say about your theory? Would
  it make sense to run other analyses? How could
  one improve this study? What did we learn?

8
Logistics

• DEADLINE Monday, November 3, 1100 am
• WORD-ATTACHMENT (.docx) EMAILBOX Uwe Matzat
  (umatzat_at_gmail.com) and Chris Snijders
  (c.c.p.snijders_at_gmail.com )
• I am available for questions
• IPO 1.32/ umatzat_at_gmail.com

9
Problem Theory- Empirical Research

• 1.) The problem we work on is
• The success of collaboration (variable v56-v59b)
• 2.) Formulate a (ego-centered) social network
  theory that explains why some researchers have
  more success in their collaboration with firms
  than others.
• 3.) Test the two hypotheses that were developed
  in the theory section and present the results.

10
The data set

• Information about collaborations of 150 Twente
  University researchers of different disciplines
• position of the researcher

11
In general collaboration with a commercial
company is strongly appealing to me.
12
Faculty of the researcher
13
To what extent do you consider the collaboration
with the business firm to be successful?
14
To what extent are the goals of the collaboration
achieved?
15
If there would be a possibility to cooperate with
this business firm in the future, would you then
do this again?
16
With hindsight to what extent has the business
firm tried to take your technology and run?
17
With hindsight to what extent has the business
firm succeeded in taking your technology and
run?
18
Methods of data analysis

• multiple linear regression analysis
• dependent variables v56-v59b
• think about how to combine a selection of them
  into one single indicator depending on your
  theory
• select independent variables
• a) network characteristicsb) control variables

19
The data preparation

• You only have to work with SPSS!
• However, you definitely have to transform the
  given data set, for example by creating new
  variables with the compute command within SPSS.
  So a considerable amount of work consists of data
  recoding and data transformation to compute the
  appropriate network characteristics that you
  think are important.
• Have a look at the hints for measuring network
  characteristics on the course website (in the zip
  file). Dont take the SPSS commands literally.

20
To Do

• Have a look at the data
• Understand how to calculate the network
  characteristics
• Re-read the stuff about social network theories
• Try it out!
• Send the report to me (and Chris Snijders as a
  backup) until Nov 3