6/3/13 Quiz Feedback | Social and Economic Networks: Models and Analysis Feedback — Quiz Week 2 You submitted this quiz on Sun 2 Jun 2013 11:39 PM IST (UTC +0530). You got a score of 16.00 out of 16.00. Question 1 Which of the following is NOT evidence for Homophily? Your Answer Score Explanation Scholars with similar ethnic background are more likely to coauthor; Venture capitalists who graduated from the same universities tend to invest together; There are usually an equal number of boys and girls in a preschool. Total ✔ 2.00 2.00 / 2.00 Question Explanation Homophily refers to the fact that people are more prone to maintain relationships with people who are similar to themselves. Question 2 What is the (normalized) degree centrality of node 1? https://class.coursera.org/networksonline-001/quiz/feedback?submission_id=156390 1/6 6/3/13 Quiz Feedback | Social and Economic Networks: Models and Analysis Your Answer Score Explanation 2 1/3 4 1/4 Total ✔ 2.00 2.00 / 2.00 Question Explanation Degree centrality = degree / (n-1). Now degree of node 1 is 3, and n=13, so 3/(13-1) = 1/4 Question 3 What is the closeness centrality of node 1? Your Answer Score https://class.coursera.org/networksonline-001/quiz/feedback?submission_id=156390 Explanation 2/6 6/3/13 Quiz Feedback | Social and Economic Networks: Models and Analysis 1 4/7 3/5 4/5 ✔ Total 2.00 2.00 / 2.00 Question Explanation Closeness centrality = (n-1) / {\sum}_j l(i,j). Now n = 5, and the denominator is 1+1+1+2 = 5, so the answer is 4/5. Question 4 What is the decay centrality of node 1, with \delta = 0.5 ? Your Answer Score Explanation 1.25 1.75 ✔ 2.00 2 2.25 Total 2.00 / 2.00 https://class.coursera.org/networksonline-001/quiz/feedback?submission_id=156390 3/6 6/3/13 Quiz Feedback | Social and Economic Networks: Models and Analysis Question Explanation Decay centrality = {\sum}_j \delta ^ {l(i,j)}, hence 0.5+0.5+0.5+.25 = 1.75. Question 5 Compare nodes 3 and 4: which one has a larger betweenness centrality but a lower Bonacich centrality (b=1/3)? Your Answer Score Explanation Node 3 Node 4 Total ✔ 2.00 2.00 / 2.00 Question Explanation Node 4 is correct. Node 3 has a betweenness centrality = 0.5, and Bonacich centrality = 13 (when b=1/3); Node 4 has a betweenness centrality = 0.6, and Bonacich centrality = 11 (when b=1/3); The results are provided in page 50 in the slides of Week 2. Question 6 Consider all undirected networks with 3 nodes {1,2,3} - there are 2^3=8 networks in total. Which of the following is the property the every node has at least one link: A(N)={g | N_i(g) is nonempty https://class.coursera.org/networksonline-001/quiz/feedback?submission_id=156390 4/6 6/3/13 Quiz Feedback | Social and Economic Networks: Models and Analysis for all i in N} ? Your Answer Score Explanation {1, 2, 3} {12, 23, 13} ✔ { {12,23},{13,23},{12,13},{12,23,13} } 2.00 { {12},{13},{23} } Total 2.00 / 2.00 Question Explanation { {12,23},{13,23},{12,13},{12,23,13} } is correct. There are 4 networks in A(N): 3 networks {12,23},{13,23} and {12,13}, each of which has 2 links; and {12,23,13}, which has 3 links. Question 7 Consider generating Poisson random networks on 1100 nodes where each link has a probability p. Out of the options for p below, which is the lowest for which the networks are still likely to be connected? So, which is the lowest number that is still larger than the threshold for connection? [Hint: log(1100)=7 ] Your Answer Score Explanation 0.1 0.001 0.01 ✔ 2.00 0.2 Total 2.00 / 2.00 Question Explanation The threshold for the connectedness is t(1100)=log(1100)/1100 = 0.0064. So with p=0.01, 0.1 or 0.2, networks are all very likely to be connected. So 0.01 is the correct answer. https://class.coursera.org/networksonline-001/quiz/feedback?submission_id=156390 5/6 6/3/13 Quiz Feedback | Social and Economic Networks: Models and Analysis (See "A Threshold Theorem" in page 101 of the slides) Question 8 The Small-World model in Watts and Strogatz (1999) shows that by randomly rewiring a small but nontrivial fraction of links from a highly structured lattice: Your Answer Score Explanation Degree distributions observed in practice are matched. It is impossible to generate both high clustering and low diameter in one model. Networks with high clustering and high diameter are generated. Networks with high clustering and low diameter are ✔ 2.00 generated. Total 2.00 / 2.00 Question Explanation Watts and Strogatz (1999) provides a simple procedure to generate networks that have both high clustering and low diameter, both of which occur often in observed networks. The model is not great at generating degree distributions that match observed ones. https://class.coursera.org/networksonline-001/quiz/feedback?submission_id=156390 6/6