SOCY60361 Social Network Analysis: concepts and measures
Computer based assessment
2024-2025
DEADLINE: 26th Jan 2025, 2pm
One assignment of 3000 words, +10% (i.e. 3300 words max). There is no lower limit for this assignment, but you must make sure your analysis is fully described and justified.
Note: You must include an accurate word count on the front page of your essay. Failure to do so will lead to an automatic 2 mark deduction.
Your word count should include all text in the essay (including any footnotes), but does not include tables, pictures and the bibliography.
For this assignment you will be required to perform. analysis in Ucinet and to produce visualizations. Each performed task’s output and each visualization need to be added to a final document which represent your assignment (in a word/pdf file).
All datasets are in Ucinet dataset files.
The assignment is composed by 2 parts. Please answers all questions of all parts. The final mark is calculated as the average score of all the parts.
PART 1.
For this part, you will be using the dataset CAMP92, and the attribute file CAMPATTR
These data were collected at the 1992 NSF Summer Institute on Research Methods in Cultural Anthropology. This was a 3 week course given to 14 carefully selected participants and 4 instructors (total 18 people in the network). Network data were collected at the end of each week. These data were collected at the end of the second week. The data were collected by placing each person's name on a card and asking each respondent to sort the cards in order of how much interaction they had with that person since the beginning of the course (known informally as "camp"). This results in rank order data in which a "1" indicates the most interaction while a "17" indicates the least interaction.
The dataset CAMP92 contains network data for week 2 and week 3. The attribute file contains information about the gender of the participants (female=1/male=2) and their role (student =1/instructor =2).
1. Create a matrix which contains only the first 3 choices for each respondent in week 3 (include the matrix in your answer). Produce a network graph from your matrix. Colour the nodes according to gender, shape the nodes according to roles. 10/100
2. Are there any cohesive subgroups in the network? If so, how many? Explain and justify your answer, describing the method you have used to address this question and explaining why you chose that method. You can use more than one method. 30/100
3. Using any three measures identify the most central actors? Briefly explain the measures you have used and what they might tell us. 30/100
4. Is there any evidence of gender and/or role homophily in this network? Briefly explain the method you have used to answer this question. 30/100
PART 2.
For this part, you will be using the dataset Krack-high-tech and its attribute file High-tech-attr.
These are data collected from the managers of a high-tec company. The company manufactured high-tech equipment on the west coast of the United States and had just over 100 employees with 21 managers. Each manager was asked to whom do you go to for advice and who is your friend, to whom do you report was taken from company documents. In addition attribute information was collected. This consisted of the managers age (in years), length of service or tenure (in years), level in the corporate hierarchy (coded 1,2 and 3; 1=CEO, 2 = Vice President, 3 = manager) and department (coded 1,2,3,4 with the CEO in department 0 ie not in a department). This data is used by Wasserman and Faust in their network analysis book.
1. Produce one visualization with friendship ties, one with advice ties and one with report-to ties. Colour the nodes by level and size them by tenure. 30/100
2. Do ‘reporting’ and friendship ties help to explain who asks whom for advice? Justify and explain the method you have used to address the question, presenting all relevant outputs. Could you use a standard regression model to address this question? Explain your answer. 70/100