# ANLY512 Assignment 5

ANLY512 Assignment 5
Anonymous
1/27/2019
Purpose
Methodology
Part 1: Data Processing
Part 2: Social Network Analysis
Part 3: Social Network Analysis Statistics
Part 4: Results
Part 5: Conclusion
Purpose
This dataset contains information collected from the students of the ANLY 512 class of Harrisburg University through a survey and it consists of 24
observations. The dataset contains information regarding the surveyee’s name, names of students that the surveyee has worked with, number of
times the surveyee has worked with those students as well as an estimation of how many seconds their classmate would be able to hold their
breath. Through this analysis, I aim to see the relationship between various students through the variables collected and determine if there are
clusters that exist within the various groups of students. This report includes the usage of data visualizations as well as the results obtained through
my findings.
Methodology
The dataset contains 24 observations across 17 variables. Using the igraph package in R, I generated various network plots in order to analyze the
relationship between students in terms of gender, estimation of length their classmate would be able to hold their breaths, frequency of
collaboration between students and whether these students are in similar fields in terms of their career. Additionally, I calculated various statistics in
order to further analyze the survey dataset.
Part 1: Data Processing
In this section I read all required libraries needed for analysis, imported the required dataset, removed duplication and renamed columns.
Part 2: Social Network Analysis
CODE
CODE
CODE
Figure 1: Social Network Analysis Plots to Analyze the Relationship between Career Similarities and Gender as well as Frequency of Times Worked and
Length of Time Students Can Hold Their Breath
Figure 1 was generated to show the relationship between similarities in terms of students’ career and gender as well as the frequency of times
collaborated and the estimate length of time students can hold their breath. The plot on the left shows that there are more females in this class
compared to their males counterparts. From the same plot, it is also evident that a majority of the students are not in similar careers as their peers.
Besides that, the surveyee is also asked to estimate the amount of time that their classmate will be able to hold their breath. From the graph on the
right, it is evident that a majority of students aproximates that their classmates are not able to hold their breath for longer than 100 seconds (1
minute and 40 seconds). In the same graph, the red arrows represent a frequency of 3 or more times worked with a classmate. The graph shows that
there are several students who have worked more than 3 times with their classmates on various assignments and projects throughout this course.
However, the red concentrated line in the middle represents an outlier in which a student identified that they have worked 60 times with another
student. It is also important to note that the white circles in the plot are due to the fact that these two students information were not identified by
other students through the survey.
Figure 2: Social Network Analysis Plots Showing Largest Cluster as well as Smaller Clusters
Moving on, Figure 2 above shows the different clusters or “cliques” that is present in the class. The plot on the left shows the different smaller
CODE
clusters that exists within the class. It can be seen that these clusters consists of two to eight students. This shows that these smaller groups worked
more closely with each other. It can even be hypothesized that the students within the clusters are working together in their final projects.
Furthermore, the plot on the right shows a more broader, larger cluster that exists in class. From the plot, it is evident that two thirds of the students
of this class fall in this category. This also shows that although the class does have smaller groups or “cliques”, students with a yellow circle worked
more closely with each other as compared to those that are in gray.
Part 3: Social Network Analysis Statistics
Figure 3: Dendogram of Different Clusters
The dendogram or tree diagram shows a more diagramatic representation of the clustor plots in Figure 2. This breaks down the relationship of
various students in the class and illustrates the different clusters in a more systematic manner.
CODE
CODE
Figure 4: Histogram of Degree
In addition to that, I also analyzed the degree of connection of students in this class with their classmates. When the ‘degree’ function was ran, I can
see that in general, the range of degree of connection among students is from 1 to 8. With the help of a histogram, it is easy to see that the mode of
degree of connection for students in this class is from 4 to 6. This tells us most of the students have between 4 to 6 number of connections with their
fellow teammates.
Not only that, I also utilized the ‘distance’ function to examine the number of edges between two nodes. The results tell me that without taking into
account the fact of whether or not the students are currently working together or have worked together before, the average number of
edges/connections between two students is 3.3.
Besides that, by looking at the closeness centrality of the nodes, I learned that the contrast between students’ closeness with each other is not
significant as most of the students seem to share the same level of closeness. Moreover, I also decided to examine the eigenvector centrality of the
nodes, which illustrates the influence among students, in terms of how students who are well-connected are linked to other well-connected
students.
Part 4: Results
The results of my analysis indicate that there are many relationships between various students and the variables collected. For one, there are more
female student as compared to male students in this class. Besides that, a majority of students who are working together in this class does not share
the same careers. This may be a good thing as diversity in careers bring out different ideas and backgrounds which each team can utilize when
working with one another. From my analysis, most of the students have worked less than 3 times with their classmates. Figure 2 shows the clusters
that exist in this class. It can be seen that students generally work together with classmates who may or may not be within their immediate clusters .
Having said that, in the smaller clusters, smaller groups are present whereby students in each cluster worked with one another more closely than
others. It can be hypothesized that these smaller clusters make up the groups for the team project.
Part 5: Conclusion
Through my explanatory analysis of social networks within the class, it can be concluded that there are various relationships that occur within this
class. The average number of connection between 2 students is 3.3 students, which shows that students in this class are generally well-connected.
Having said that, this social network analysis can be further enhanced with the inclusion of more variables in the survey, so that different aspects of
the social structure within the class can be observed and analyzed.