Social Network Blog - y11

This assignment asked us to conduct a twenty-person survey that sought to assess patterns in social capital and core discussion groups. There were many different variables of comparison for this analysis, but the most significant to me is comparing gender.

The first part of our analysis dealt with understanding one’s social capital through their personal network. To accomplish this, a position generator was presented to the survey participant, who then indicated whether they knew someone working in that position. They repeated this for 15 different jobs. Their social capital was calculated through assigning each occupation a numerical value from 1-15 (15 being most prestigious, 1 being least prestigious), and then calculated and averaged for each age and gender group.

As the results demonstrate, there are some significant differences between men and women when it comes to assessing social capital through their network. When averaging each group’s totals, it appears that women under 33 have the highest amount of social capital. This difference is notable especially when compared to men under 33, who have an average social capital prestige value of 55.8, while women under 33 have a value of 67.4. These findings also indicate that over 33, there is no real difference between men and women’s social capital from their networks.
To assess the differences in men and women’s social networks, I also recorded the frequency at which people identified others working in those occupations. The chart below indicates the findings for each age and gender group, as well as for each occupation.

These findings support the claim that women have higher social capital than men, since they know more judges, aircraft pilots, and professional journalists, which are ranked high in prestige. It is also interesting to note how often respondents indicated knowledge of people in stereotypically male or female occupations (for example nurse, secretary as female; mechanic, pilot as male). The chart indicates that there was a significant difference in how many men knew a mechanic, while more women knew pilots. Also, nurses and secretaries show consistency between the two genders. While reviewing my overall data, I was surprised to see that all respondents reported recognizing at least four of the professions and was usually spread out, indicating that this sample contained individuals with occupationally diverse network ties.
After analyzing each respondent’s network diversity, I wanted to see whether McPherson’s claim about declining strong-tie discussion networks was accurate. I averaged and compared data concerning sex (male vs female), relationship (kin vs non-kin), and distance (state or closer vs other state or farther). The table below summarizes the results:

As shown, there are various prominent differences between men and women when concerning discussion networks. Firstly, women had more average ties than men, however, this effect was not strong when comparing within the same age group. The most prominent overall difference was with the age groups, since those under 33 recorded more ties than those older. When looking at the percentage of people who included kin or spouse, women recorded more than men for all age groups, but particularly over 33. Also, there was no significant difference between men and women when it came to the gender of their close tie: most wrote women. However, there was a significant difference between men and women when it came to distance, since 38% of women under 33 mentioned that their tie was in the general area, while over 70% of men’s ties were.

Looking over the base of the results from my sample population of both men and women under and over 33, we can come to several conclusions, which can be compared against those confronted in prior class readings and discussions. To make it more comprehensible, each social network theme is presented separately.

Social support: In terms of social support, it appears that women are chosen (or referred to) more frequently as close ties, which indicates that they are regarded as individuals who have social support resources to offer. The fact that both sexes chose more women as discussion ties indicates that women are better equipped than men to be friends with both sexes. In Wellman and Wortely’s article on community ties and social support, they also found that women were more commonly referred to for emotional support through “expressing” rather than “repressing”. Thus, they would be more suitable to be regarded as strong discussion group ties.
Network size: While this survey did not cover weak or moderate ties, we can make several conclusions from our findings. Firstly, we do not seem so isolated. For those under 33, they reported approximately five strong ties, while those over 33 reported between 2.5 and 3, which is still higher than the 2.08 figure presented by McPherson and Smith-Lovin in “Social Isolation in America. However, their findings on a decreased relationship with those from voluntary associations and their community was found in this survey, as no one stated that their tie was from those two fields. This finding correlated to that reported in Hampton’s “Networked Sociability online, off-line”, where he reports a decline in community (in the traditional sense). However, since the use of new media was cited (especially in those under 33), we might expect that a new form of community is emerging that would allow the maintenance of strong and weak ties, as mentioned when reading about the impact of Facebook on bridging and bonding ties.
Community & privatization: Although we have a small and relatively homogenous sample population, there are numerous conclusions we can make about community. As mentioned when discussing network size, McPherson and Smith-Lovin stated that ties resulting from voluntary associations or community activities declined, which I found in my survey. This could indicate an increase in privatization, since many ties were spouse or kin, especially for those over 33. Another reason for privatization might be related to the findings of Kalmijn, who reported that as one gets married, their social network declined. This could help us understand the shift from 5 ties to 2.6-3 in those over 33, since they are more likely to be married.
Network density & diversity: As discussed earlier, I was surprised to find that my sample was as diverse as they were. They demonstrated that they knew people who were from both the more and least prestigious extremes of the list. Women had more social capital, which could be explained in several ways. First, women are considered stereotypically more social than men, which might be why they would do a greater breadth of different people than men. Also, since there are more men in the workforce, they might know more people who are in the same area or occupational prestige as him or her. While women are more likely to stay at home and socialize across a variety of different community levels, they might be more likely to meet more diverse people. The findings in my study also could be interesting when applied to the revised small world problem, where one could determine whether women’s paths were more occupationally diverse and how that affected the success of the folder to the target. Also, one area of the survey not already discussed was the relationship of the ties with one another. There were a variety of answers that leads to problems in making a generalizeable observation. While most people reported that their strong ties at least knew one another, some also noted that they were strangers. This might be attributed to the fact that there is some misunderstanding about what important matters are. For some, politics and current affairs might be important matters, which would lead them to talk to a select group of people who do not interact with those relied upon for favors or emotional support.
Strong ties, weak ties: Based on our analysis of people’s strong tie discussion network, we can conclude that people report having more strong ties than uncovered in the McPherson study, which indicates that people aren’t as emotionally isolated as we once thought. However, this study did not reveal much about weak ties, since the position generator did not ask the strength of tie. Doing so would allow researchers to understand just how important and beneficial it is to have weak ties in a variety of prestigious field. This fact was alluded in last week’s article, which stated that those who had more prestigious contacts were more likely to get more prestigious jobs.
Homophily: Based on gender and age, we find that my participants were similar to those in other papers. They were much more likely to include people who were closer in age to them, especially if they were non-kin, and when female, were more likely to have ties that were also women. In terms of education, there was not much difference, since almost all participants attended college or more education, as did most of their ties. Women over 33 also reported more kin and spouse ties, which reinforces their network homophily and also corresponds with Wellman and Wortley’s findings. My findings on homophily are also similar to what McPherson and Smith-Lovin discovered in “Birds of a Feather”, where gender, geographic, age, and geographic homophily are existing patterns in network formation.
Role of new media: There are numerous findings in the data that indicate that there are emerging patterns in new media use, according to both age and gender. Those under 33 reported that they used email and instant messaging to communicate with their strong ties, which this was much more rare with those older than 33. One reason for this finding could be that older people use email to keep in touch with those they work with, and do consider their ties at work to be close or part of their core discussion network. Also, those under 33 are more apt to use new media since they have had more of their life exposed to AIM and Gmail as the norm, thus facilitating the ease of use and making it more a part of their lives. My finding concurs with that reported by Baym in “Social Interaction Across Media”, where they found that the internet was a prominent and essential part of college-students’ communication patterns.
Issues of measurement: There were numerous issues with measurement that may have affected the outcome of our results. Even though I attempted to search out the most diverse group of participants, my sample is more representative of a convenience sample, where I surveyed friends of mine, family, and family friends. This could bias the results because of my own network characteristics. Also, the position generator was not a very reliable tool to use in its current state, since it missed a lot of occupations that many would know. Also, it appeared that, if replicated today, some of the prestige levels would be different, which would give us different indicators and values for assessing social capital. Another logistical problem with this survey was that people didn’t want to take it, particularly those over 33. They did not, generally, feel comfortable giving away this information, which might have led them to cut their ties short, or not be as open with us as possible. Those under 33 experienced a possible social desirability bias, where I observed that most respondents tried to think of six people or even more.

This week’s readings discussed the relationship between social status, poverty, and how one’s network affects these variables.

In “Social Isolation and the Underclass”, Fernandez and Harris discuss the implication of poverty on one’s social network. Their fundamental argument was that the underclass is “socially isolated” from the mainstream and community institutions, which makes people’s networks more homophilous, in a negative way. When discussing the maintenance and reproduction of the underclass, the authors claim that social isolation is more significant that the culture of poverty. This statement is important because it could help direct new efforts to alleviate the plight of the underclass through measures aimed at decreasing underclass network homophily. The authors cited that encouraging social mobility would be one of the best ways to combat underclass social isolation. To test the prevalence and importance of social isolation, the authors conducted a survey measuring three main claims: the underclass is socially isolated, the extent to which individuals in poor areas are more isolated than individuals in other areas of a city, and whether there exist any significant gender differences in the structure of social isolation. Another significant aspect of their analysis on social isolation was the inclusion of institutions. They wanted to see whether isolation was an individual or community pattern. When discussing network structure in the study, they focused on three familiar indicators of social networks: number to ties, diversity, and multiplexity. I think it’s also important to mention that the authors mentioned Burt’s weak tie argument; weak ties are more beneficial for a social network that strong ties.
In terms of the study, there were several things to note. First, the authors tested support from ties in two days: daily support vs. support in crises, which is significant since they can reveal two different networks. Also, this study was the first to explain another disadvantage to a small network. The authors explained that since people in the underclass have less ties, they are more likely to overburden their existing ties, creating a more stressful and uncomfortable relationship. Their results indicate that there are clear class and gender differences in patterns of isolation, with the nonworking poor being most isolated, and their levels of participation in community institutions. While reading their plea for solutions based around alleviating social isolation, I remembered something I read about a rent-control program in New York City, which would give building owners financial incentives if they offered an apartment and a controlled low price for people who make under $20-24,000 a year. Do you think such a program would work? Does it depend on the city?

The second reading covered the relationship between social resources and mobility, particularly when it concerns getting a job. In their examination of people’s transitions from their first to current job, they observed what network characteristics were most important in securing a better job. They initially claimed that people with better access to social resources will have better outcomes in instrumental action, which in this case is getting a job. Marsden and Hurlbert also claim that high SES and diverse networks are better sources for finding appropriate information, which could give them a better start to a job search. After conducting their analysis, they found several important network benefits in finding a job. First, the higher the contact prestige, the higher the respondent’s occupational prestige, which indicates an upward social mobility based on who you know. How does this apply to Penn, in that there is an alleged higher chance of people graduating into more prestigious positions? Will just going here ensure a better future through who we know? Relating to this question, researchers found that education and respondent’s prior prestige correlate to the prestige of their contacts. Third, they also found that tie strength is not so important in career networking. Since most of us are seniors, do you find these patterns similar to what you experienced when looking for jobs?

My target individual was Susan Yoon, and I am very pleased to report that my folder did, in fact, reach her. There were 2 links in total, with one link consisting of myself and Dr. Sipe, and the second with Dr. Sipe and Susan.

There are numerous reasons why I think my folder was able to reach the target destination, some of which relate to the hypotheses I set forth when making my predictions in Part I. I chose to pass my folder onto Dr. Sipe for numerous reasons, which include issues surrounding tie strength, proximity, and prestige/centrality within a specialized network like GSE. First, Dr Sipe and I can be classified as moderate tie, since we have know each other for two years, live in the same dorm building, and have at least weekly interactions taking place in meetings. While I do not choose to discuss important matters with him, I know that I could, especially since he is also ordained. I also chose to pass my folder to him because he lives in my building (high proximity), and would ensure that through seeing and talking to me often, he would be reminded to pass the folder to Susan. Lastly, Dr. Sipe is a faculty master of my building and a professor in GSE, which is the same school at Penn where my target, Susan Yoon works. Since he is a professor there, he has higher prestige, and hypothetically knows more people than Susan (she’s an assistant professor), a point that is especially true since he has been affiliated with Penn for two decades. In the case that he didn’t know Susan, his central role as both a faculty master and a long-term professor increases his network centrality, making it more likely he would know someone within GSE who also knew Susan. Through my predictions, I was also able to avoid making path inaccuracies, as noted in the Killworth study “The Accuracy of Small World Chains in Social Networks”. He notes that most people do not accurately assess the size and qualities of their networks, leading them to initiate longer chains than needed. Since I was able to get the folder to Susan in two links, I demonstrated path accuracy.

With a folder having successfully reached Susan, I can now look back and assess the four hypotheses I made in Part I.
Hypothesis 1: Susan and Dr. Sipe’s as both professors will be the strongest determinant of the path success.
Since the folder got there with only one intermediary (Dr. Sipe), I can conclude that their positions as faculty in GSE greatly aided the chances of a successful folder transfer. Their status as professors also indicates that they have a higher occupational prestige, which would lead to their possible statuses as prestigious actors. As explained in “Centrality and Prestige”, those actors are the object of extensive ties and form wide-spanning networks. Since they are the object of more ties, they would be easier to access, requiring less intermediaries. The success of the folder transfer might also be due to his role as an opinion leader, through his position as a professor. As explained by Burt in “The Social Capital of Opinion Leaders”, being a professor would allow one a greater opportunity to be an opinion leader, which would increase their social capital. As a result, they can more easily transfer information (or, in this case, a folder) across social boundaries.
Hypothesis 2: My moderate tie strength with Dr. Sipe will increase the chance of the folder reaching the target.
To my surprise, even though I did not hold a strong tie relationship with Dr. Sipe, he was very reliable and passed on my folder directly to the target. This may be attributed to our professionally-based relationship (he is a faculty master in the building I am an RA in), and the fact that we do not share just a weak tie.
Hypothesis 3: It would be more difficult to reach her since she is in another Penn network than myself, but that fact that both of us are female might decrease the number of intermediaries.
This hypothesis is both disproved and supported. Even though I assumed it would be more difficult to reach her, I was able to do so with a very small number of intermediaries. At the same time, we are both female, but I do not think this fact had any role in the success of the folder path. Even if we are both female, the intermediary between us was male, which should decrease, not increase the number of intermediaries, and should decrease success of the folder reaching the target. This last fact is supported in “Birds of a Feather”, which explains that men are more likely to act as gatekeepers, which could explain why Dr. Sipe was able to directly pass the folder to Susan.
Hypothesis 4: The folder will take one to two weeks to reach the target.
Based on an error from my part, the folder was not able to reach the target in this allotted amount of time, since I took sometime getting the folder to Dr. Sipe. It ultimately arrived in 21 days, which is seven days over my prediction.

Although my path to Susan Yoon was very direct, there are many differences between my observations and those experienced by the group as a whole. Overall, I was very surprised and impressed that 80% of the folders were able to reach Susan, particularly since this number is so much higher than that for Antonio (only 25%).

Here is an easy-to-understand analysis of my chain findings as compared to the rest of my group:

To summarize, I have less intermediaries than the rest of this group, with 1 vs. an average of 3.25. I also conflicted with the group findings, since I my transfer to someone of the opposite sex was in the minority of completed chains (15.4%). I also differed from the group in that I was more likely to pass it to someone in the same school, while not in the same department. Also, I took approximately double the time to reach the target than other completed chains for Susan. Additionally, my final chains did not share gender, years at Penn, nor department.

After looking at the mean results for chains directed as Susan, one can conclude several things regarding the strength of certain criteria in ensuring chain success. Firstly, it seems that the most significant indicators of a successful chain are transfers that share the same gender and same school (only applicable to final transfers), since those were the answers with over 50% of similarity. Homophily in these characteristics is similar to what McPherson noted in “Birds of a Feather”. In this paper, he mentions that status homophily plays a large role in determining the tie and future relationship of two people. In our case, we can state that shared school affiliation influences status homophily, since people in the same school are interested in studying the same thing, and may hold the same position. In terms of gender, McPherson notes that homophily exists within gender. Secondly, transfers with the same departments, same school (not true with final transfers), and same number of years at Penn is not significant when attempting to reach a target person. These results are interesting when compared to those uncovered by Milgrim in “The Small World Problem”. After conducting the original small world analysis of networks, he found that gender homophily played a significant role in folder transfers, as does an occupationally upward movement in folder transfers. These are two findings that are similar to those from our group, and demonstate, once again, that gender plays a very important role in how successful a folder transfer can be.

A further exploration into the degree of importance of gender and school affiliation is recommended, so we can conclusively predict which characteristic is most indicative of a successful chain transfer. Also, another interesting pattern to research is the important of race. This pattern has already been examined by Milgrim in “Acquaintance Networks Between Racial Groups”, where he finds that race plays a significant role in the small world problem, since it can affect what kinds of people one knows. In our case, Susan Yoon is Asian, and, when looking at shared final transfers, June Choo was the most popular. I know that she is also Asian, teaches a spinning class, and is very involved with Asian-American issues and groups on campus. This pattern signals a possible relationship between race and path success, which should be examined further.

Additionally, there were several connections between me and my group’s finding to those reported in “The Small World of the University”. In this reading, the authors conducted a similar analysis, in which they attempted to determine the number of links between an undergraduate student and an administrator on campus. They primarily focused on gender, class, and occupational patterns in folder path success. With a sample of 16 completed transmissions, they were able to conclude several things. First, 27% of folders reached the target, which is much lower than the success rate for Susan’s group (80%). Also, they found that 81% of folders were started by upperclassmen, which is an effect we could study, since all students in our class are juniors or seniors. Stevenson also found that students passed the folder to those in the same or higher grade. This finding is completely different to our group, where the majority (55.56%) of completed chains were transferred to students in a lower class. In terms of occupational status, the Boston College study was more likely to involve students passing folders to graduate students, faculty, or staff when they were unsure about the target person. This indicates that prestige can influence network connections. Like our group, Stevenson found that homophily on the basis of gender (particularly female) was a pronounced indicator of path success.

link to part I: http://www.mysocialnetwork.net/blog/481/y11/2006/09/ensuring_path_accuracy.html

This week’s readings focused on analyzing the effect that social network closeness and diversity have one one’s health and well being. In some surprising findings, authors of the various papers demonstrate that the diversity and reach of social networks can have a real impact on health, mortality rates, and the diffusion of disease. These findings are significant, since they take the study of social networks to a more practical and understanding domain, where many more can understand the real-life benefits of network diversity, where these findings can contribute to answering larger and more persistence social problems.

In “Social Integration and Health”, researchers at Carnegie Mellon examined that claim that the diversity of social networks can lead to reduced susceptibility of the common cold. The authors first examined the differing claims concerning the impact of social integration, role, and network size in health-related matters. A review states that greater social integration is associated with lower mortality rates and a lower chance for recurrence of cancer, as well as fewer depressive symptoms. They attributed these prior findings to several main ideas. Firstly, broader social networks can facilitate psychological well being through reinforcing role behavior and consequent expectations as well as reinforce a sense of meaning in community. These psychological states are reflected in physiological improvements in health through initiating changes in immune functioning and the release of certain hormones. After testing 276 adults on psychological and stress-related matters, and then exposing them to a common cold, they looked at who resisted infection. Researchers found that those with higher levels of network diversity caught fewer colds. It is important to note that it was merely diversity, and not tie strength or number, that supported this finding. They also found that network diversity does not impact the effects of stress on health. Were you surprised by this claim? What different explanations can you deliver to this finding? Do you think globalization and increased social mobility will effect disease diffusion since our networks will become more diverse?

The second paper discusses the implications of close social ties relating to depression in myocardial infarction. They ultimately examined 583 patients and assessed their rate of mortality or further cardiac events after myocardial infarction. They studied both the effect of existing depression and close-tie social network to determine which affected the rate of subsequent cardiac events more profoundly. I was incredibly surprised that the availability of close ties was more significant than depression as an indication for further cardiac events. They explained that this finding could be explained through the impact that close ties could have on patients, which is manifested through their regular care and surveillance. It also appears that the support of others is more valuable to the health than personal support. However, these findings are a little troubling due to the McPherson study claiming that our social networks are actually declining. Do you think one’s consideration of a close tie changes in more serious situations (being in the hospitable, poor health), which may lead to a temporary increase in social support to help the individual deal with that situation?

Lastly, Bearman, Moody, and Stovel conducted an examination of social network structure in a high-school setting, and used it to understand the patterns and diffusion of sexually active students. At first, I was a little wary reading the article, cringing at the thought that they’re going to relate romantic relationships to disease diffusion. But, their analysis included a strong argument for charting relationships, since they can help mould more effective campaigns and strategies for fighting STD spread. When explaining STD spread, they mentioned the “core”, which reminded me of the social network idea of network centrality. They also reinforced previous claims of homophily in romantic relationships. Ultimately, this study demonstrated the overwhelming benefits in observing networks from beyond an egocentric perspective, and fragility of the “spanning tree” in diffusion. After reading this paper, what do you feel are the most important points to consider when designing a sexual education program?

1. In a one week period of time, I recorded 86 new media interactions. These are the five people I interacted with the most:
- Candace: 13 (best friend at college, strong tie, lives in my building)
- Yacine: 11 (best friend for 5 years, strong tie, lives in TX)
- Dvorit: 8 (close friend at college, strong tie, lives in building)
- Erin: 8 (best friend since high school, strong tie, lives in NYC)
- Parents: 8 (my parents, strong tie, live in Romania)

By Media type, these are the individuals I interacted with the most:
Cell:
- Yacine: 9
- Erin: 5
- Candace: 4
- Kirsten: 3
- (for fifth place, multiple tied for 2)

SMS:
- Dvorit: 2
- Yacine: 1

Email:
- Dvorit: 3
- Josh: 3
- Erin: 2
- Noah: 2
- Marketing group: 2

Instant Messenger:
- Candace: 9
- Laura K: 2
- Parents: 2
- Nancy: 1
- Kirsten: 1

Skype:
- Parents: 4

Facebook Messaging:
- Dvorit: 1
- Jasmine: 1
- Cristina: 1

There seems to be a relationship between the medium use to interact with others and different tie variables.
In the cell phone category, 4/5 are strong ties which I have known for longer than one year, and consider to be close friends. Looking at my list of interactions, I have also noticed that cell phone use was primarily linked to both emotional aid and companionship, but was also used for job and club information, as well as small service and financial aid. Within this group of strong ties, they are all considered to be close friends, which explains why I considered them to be strong ties. Also, I have known each of them of several years. When looking at distance, two are considered long distance (one in TX, one in NY), and the rest are close ties located on campus, two of which are in my dorm building, and the other is a high-rise located a couple of minutes away. Even though they live in the building, I think I use the cell-phone to contact these individuals when I think they may be out of the building and don’t want to try their room. Four out of five individuals are female, which is a pattern I can generalize across my social network, since most of my friends are female. The one tie that is male is someone I dated for five years. All the individuals are between 21-23, and are very close in age to me, which I can attribute to the fact that we are the same year in college, or I met them in high-school.
For SMS, the two individuals I used this means of communication are strong ties, which I have known for at least three years. However, it is important to note that SMS use in general was very limited, and in one week, contacted only two people with it, which might show that I only deem SMSing a communication option when I have something short and punctual to send to someone I know well. The support exchanged in this mean is companionship (twice with Dvorit) and emotional aid with Yacine. There seemed to be no relationship with distance or gender, since one person was local, one long distance, and gender was split.
Email was a medium that was a lot more spread out. Although I seem to be checking my email constantly, it appears that I don’t write very many emails, and only respond if I really need do. Email is the method of communication I use in order to contact strong, moderate, and weak ties. I used it primarily for academic resources, club information, and job information. Although the majority of people I contacted are in college and are the same age and occupation (student), I was in contact with several people who are in the academic (my professors) or business (my jobs/future jobs) spheres. I noticed that, almost exclusively, I use email to contact these people, primarily since I am trying to be formal (thus, avoid cell phone, SMS, or IM), succinct, and do not have a strong tie with them. When looking at my email usage, I also noticed that this was the most diverse group of contact I had, and ranged from the ages of 20 to over 40, in different fields (both students, professors, and businesspeople). I used email to communicate with both males and females, and did so with individuals who lived on campus or in other states (most for job related purposes).
Instant messaging was a new medium that I used with people who lived near me, and were strong to medium ties. I used instant messaging for emotional aid, companionship, and in one case, small services. The people I spoke where all female and were students at Penn (same age, status), except for my parents who used this as a means of checking up on me and quickly telling me something. My parents were the only people who lived far away, and everyone else lived on campus. I was surprised that I spoke with a very close tie (Candace) so much over instant messaging, since we live in the same building, and see each other often. Most of the correspondences that took place were simple questions, jokes, or things we forgot to mention.
Skype was used exclusively with my parents, who live in Romania. We chose to speak on this medium since it is free, and we can thus avoid the expensive long-distance expenses associated with speaking. We are also very close (we speak at least once a day), so having a product like Skype allows us to maintain our relationship without worrying about racking up a big bill. And, since they are my parents, they are older than I am (over 50), male and female, and I consider them very strong ties. Additionally, the means of support offered through Skype were emotional aid and companionship.
Lastly, I used Facebook as a means of communication with both strong and weak ties, all of whom are affiliated with Penn (either current students or alumni). I used Facebook with three females who are either 21 or 23. The two girls who are strong ties I have known for over two years, and Jasmine is someone I met last year and currently lives in NY. Another person is on campus (in my building), and another currently lives in Kansas.

Almost exclusively, my new media interactions took place in the home or at work (which is located in my dorm building). I can attribute this to the fact that I often used email, instant messaging, Skype, and Facebook and need access to a computer with internet access in order to do so. Also, since I don’t have a working landline, I use my cell phone to talk when I am at home, and generally like to speak when I have information in front of me, or can actually hear the person on the other end, which is not always possible talking from the street. I also never used any means of communication while in class or at a friend’s house, since nothing I needed to communicate ever felt that urgent. Also, when I am away from my home is when I am enjoying leisure time, and is typically with my friends, which eliminates my need to communicate with them on another medium. When I communicated at work, I often used instant messaging and email, since talking on the phone is not regarded favorable. To people with whom I communicated outside my home were typically friends I was calling when they were late, or if I was unsure about where to meet them.

Based on the semester’s readings, I find that my findings generally agree with the literature we have reviewed this semester.
When reviewing my ties and seeing where they are located, the fact that most are in college and live so close to me indicates that I have a personal community. In “The Network Community”, Barry Wellman describes that community can be defined as “interpersonal networks that provide sociability, social support, and social capital to a member” (Wellman, p. 13), and as my usage shows, I use the same network of people in the area to provide me with valuable social support and sociability. Also, he states that community ties are narrow, specialized relationships, but have the potential to be broad and diverse (Wellman, p. 24).
One area of my results that significantly correlated with the course readings included the discussion of homophily. In some regard, I am very homophilous with the top five people I corresponded with. I attend the same university as some, am the same gender as the majority, often spend time together, and involved in the same organizations, or am related to them. In McPherson’s article on homophily, he notes that “homophily is the principle that a contact between similar people occurs at higher rate than among dissimilar people”, and my results support this definition, since those I was most similar to are the ones I communicated with the most. Along with age, gender, and location (for some), I am also homophilous with some people due to our shared educational status as students at Penn.
The fact that I received so much emotional aid and companionship resources from strong ties conforms Wellman and Wortley’s claim in “Different Strokes from Different Folks”. Although parents are often best for large services and financial aid, I did not require any during that week that allowed them to offer me that resource.
Another very important finding that these results delivered is that they notified me of who I can consider my core discussion group. In “Social Isolation in America”, McPherson states that people’s strong ties have declined, yet I noticed that I had five people who I could sincerely rely on to discuss personal matters with or provide me with supportive resources, which contradicts his finding. However, this can be attributed to the face that I am in highly social environment, which encourages discussion across various fields. Also, my new media use could explain this finding, since I use cell phone, instant messaging, and Skype to strengthen my relationships.
Lastly, the effect of new media on maintaining my friendships has been a fact that our class has discussed frequently. In the Facebook paper, the author noted that facebook helps bridge and broaden, and bond existing social networks, which I found to be true. The use of the internet has also played a significant role in my life and the way I communicate with others, as affirmed by Hampton and Wellman in “Neighboring in Netville”.

Overall, when I look back at my patterns of usage, I am little bit surprised that I noted so few (in my view) interactions. In my room, on a break from class, or at work, I feel I am constantly interaction with new media, particularly email. But as I skimmed through my sent folder, I noticed that I only reply to those that are pressing, and don’t often participate in an email listserve unless I absolutely have to. I am particularly surprised about my interactions since I am a member of many groups and hold various positions on campus, but same thing: I only contact people if they or I need to speak urgently. Maybe that’s not the best route to follow. Also, this week had some unintended distractions that prevented me from using new media. On Thursday and Friday, I had a two day Job Offer Exploration Day, where I was with future work colleagues, and was not available to return calls or check my email. Also, my patterns of communication do change periodically; for instance, when one of my groups holds an event, I am a lot more active on new media to help prepare.
Even though my cell phone interaction is also low, I predicted that because I am more likely to receive calls then call out myself. A number of my close friends live on campus, so I feel that a lot of my day-to-day interactions take place in good ol’ face-to-face. I am often around people, or in meetings, or in class, and I think the bulk of my socialization occurs in those arenas. Also, I am not a big fan of instant messaging people unless I am desperately bored (which might explain why so much instant messaging went on while at work) or have something time-important for work. I think that my decreased new media usage is also attributed to the fact that most of my meetings, my job, and my position as an RA all take place in Harrison, which is also where two close friends live, thus, I am able to take care of most communication matters in person. It would be interesting to see next year, when my friends and I are all spread out, how much new media I use. Also, I was a little miffed at my lack of high school social capital. Although I am an avid (maybe a little too avid) user of Facebook, I did not see it as a way of necessarily communicating with those from high school, but rather as a means of seeing how they’re doing.

I found the chapter on deviant behavior, ‘Deviance as a Search Process” incredibly interesting, as I was excited that someone used current social networking theory in a more contemporary and unusual manner. Tepperman found that those eliciting in deviant behavior follow a different route of social networking than previously discussed. While homophily is important in this case (to find others who engage in similar deviant behavior), the emphasis here is on secrecy (which slows down the networking process), the importance of intermediaries, and the role of the sought after. I found his point concerning perceptual maps to be very insightful and accurate. He states that the probability in success in pursuing deviant behavior depends largely on the searcher’s social-cognitive map, since it will guide him to start and narrow down his search. However, I did not agree with the author’s claim that when one doesn’t know whom to approach for deviant behavior, they will randomly choose a first step. I think that no matter what, people will somehow narrow their search for the right people, and if they aren’t familiar with deviant people in the area, they will use heuristics or stereotypes to guide their search (for instance, they might use their stereotyped knowledge and visit ghettos to find drugs). What role do stereotypes play in the pursuit for deviance? Do you think a person’s first step in an unknown area if effectively random?

In the extensive chapter on diffusion networks, the author discussed the various ways that information spreads across and within social networks, which also includes the role of network players such as opinion leaders. He finds that opinion leaders play a large and very important role in the dissemination of information in a network, with their power increasing as they demonstrate more opinion leadership. Also, he goes into more depth about the actual profile of an opinion leader. Since networks are often heterogeneous (to some extent), opinion leaders are often those who are more educated, more financially secure, more cosmopolitan, and more involved with the media and current affairs. This raises an interesting point about inequality, since those without these characteristics are often minorities, the poor, or women. How do you think the process of diffusion changes (if it even does) across movements aspiring for social change? Also, do you agree with their profile of an opinion leader?

In the last reading, Burt discussed the social capital of opinion leaders. In this paper, he states that opinion leaders are essential actors in the two-step flow of communication due to their prestige in the community. This model asserts that communication flows from the media to opinion leaders, who then disseminate it down to the greater public. However, they are not necessarily considered central actors, since they are more likely to bridge and broker between groups, not inside them. So, instead of opinion leaders, these actors should be considered to be opinion brokers. Within this two-step flow of communication method, Burt also discusses the role of contagion and its relationship with equivalence or cohesion. He finds that, in information spread, contagion occurs within groups due to equivalence, while cohesion occurs across groups. I find this article to relate closely to my marketing class, especially when discussing new and more cost effective ways to promote new products. We have discussed that instead of launching expensive and elaborate media campaigns, companies should target community opinion leaders who will be able to spread the word about a product. Do you think it would be more effective for a company to search for and use the help of opinion leaders to promote their product, or should they stick to using traditional media (think of the Alpha Pups example from Diffusion Networks)?

The four readings of the week focused on the internet and its ability to act as a social network enabler vs. disabler, and the issues surrounding the internet’s effect on social capital.

In “The Structure of the Web”, the authors discussed the implications of mapping out the structural characteristics of the internet. They discussed several theories relating to the structure of the internet and discuss that the internet is composed of hubs and authorities. In particular, they mentioned that the internet is changing the definition of community. Instead of community being built around geographical location, community is expanding to be based on the shared interests and topics. This is very important to forming new ties, since weak ties formed in this virtual community could end up being more fruitful that those being based geographically. For instance, if I am really interested in classic cars and their restoration, I might not find many venues offline to find people who might have information about getting into the business. However, with a virtual and densely-knit community, establishing weak ties that revolve around topic might provide me with more resources to draw from, in order to get into the restoration business. However, this proposition might be conflicting, since interest-based groups tend to be more homophilous. So, which would be more advantageous – weak ties based on geography, or weak ties based on shared interest?

The article published in the New Scientist definitely struck a familiar cord with me. As a social network user myself, I view Facebook and Myspace as an innocuous and fun way to strengthen existing ties and connect with a myriad of friends. However, in light of the recent phone tapping debate and the strengthened efforts to prevent future terrorist attacks, I never imagined my silly and trivial Facebook profile could attract NSA attention. This article focused on how information from social networking sites (i.e. Facebook, Myspace, Friendster) are being transformed into an Resource Description Framework, and associated with our financial, legal, and communication profiles. The RDA aims to transform the internet’s wealth of data and information into a semantic web, which would allow entities such as the NSA easy access to what used to be incomprehensible information. Although I find that this might be a very valuable resource when exploring the minds and lives of terrorists, I find that there is just too high of a chance that this information could fall in the wrong hands. If information is manipulated or if there is a mistake in the system, it could cause permanent repercussions in a individual’s life. Do you feel safe knowing that information from social networking sites is being monitored by the government? Do we have anything to fear?

In “Spatially Bounded Online Social Networks and Social Capital”, I was very excited to see an academic paper covering the college phenomenon that is Facebook. I was also very surprised by their findings, especially as they affirm that Facebook has a purpose that extends past a tool for procrastination. As with the first article of the week, the authors see Facebook as a venue in which people can create communities and networks based on interests, not geography. However, they later state that college campuses are in themselves a specific geography. They also note that this site is successful at bridging and bonding social capital and achieving higher high school Also, Facebook is most typically used to maintain what initially began as offline ties, and is used for “killing time” rather than information seeking. For those of you who use Facebook, one issue that was mentioned about it was that it was exclusive to college students. Seeing as it recently opened up its service to anyone, what do you think the implications are to the online collegial community?

In Barry Wellman’s article focusing on a computer network as a social network, he discusses the implications of technology on our virtual and offline social networks. One very important consequence of internet technology has been that our control over them has greatly expanded. This has not only given more power of the individual network-seeker, but the household as well. This finding is also contradictory to the phenomenon of a more mobile environment, in which everyone is always moving and on the go. Also, we have greater connectivity with others, which has allowed us to build resources of support, ties, and potentially even social capital as we become more connected with those around us. His article neatly demonstrates that the internet is not an entity to itself, but a medium that has changed the way we live our offline experiences as well. Do you agree that the offline and online spheres are now combined, and no longer remain two separate entities?

This week’s reading focused on the ways in which the Internet has affected the construction of our social networks. A basic theme in all the readings was the differing opinion of the Internet as a virtual domain where social network boundaries are still unclear. In all the articles, it is clear that the Internet does play a significant role in the way we are able to construct, maintain, and expand our social networks. It also strengthens the point that achieving intimate relationships over the internet is challenging, since physical contact plays a large role in establishing shared interests, trust, and experiences.

While reading about the internet-based small world study, I was immediately struck by the enthusiasm in the responses. Even after she closed her email account, she still had people contact her that were desperate to help her succeed at her project. This project highlighted several unique aspects of the internet as a social networking tool. Firstly, people want to connect with others. Whether it’s in the hopes of forming a new tie, strengthening an old one, to learn, or to be kind, people are willing to exert the effort and time to invest in social networking behavior over the internet. Secondly, the intent facilitates information spread unlike any other interactive communication medium. Within hours and days, information spread to the remotest corners of the world. It’s interesting to note that Milgrim could never achieve this level of success with his small world experiment in the sheer volume that this 13-year-old girl collected. This raises the question of whether small-world internet studies can be more valuable for social networking and information spread research than the regular “pass a folder along” approach.

In our second paper for the week, Barry Wellman discusses the myth and legend surrounding the internet as both a constructive and destructive social networking tool through the examination of existing literature. One area of contention that struck me was the observation that our interaction with the world around us was a zero-sum game; if we spend more time on the internet, we immediately ignore physical interactions. Can real-life and internet coexist, and potentially benefit from one another? Wellman concluded that the internet provides specialized help from different internet resources, that it is often supportive, and that the lack of physical contact does have an effect on the quality of the tie. The pattern of support seeking is the same as through other mediums, but it is more accessible, diverse, and discreet, which then serves to imitate real communities. Also, the internet provides us with an array of weak ties, which increases structural holes in our network, and, as we all know, weak ties can be even more valuable than the strong ones! Wellman reports on the overall low-risk interaction also helps in our ability to create more diverse ties, and that the reciprocal nature of the internet is particularly strong in spite of a lower sense of obligation and more anonymity.

While reading “Social Interactions Across Media”, I found it interesting that the researchers used two different methodologies to both test and strengthen their findings. After conducting both studies, they found that with a socially-pioneering young audience, the internet is a very widely used social networking tool that is mostly used to complement other social networking venues, usually at home, is especially important in maintaining distant social circles. These findings were very insightful, since they indicate that both the internet and regular social networking media can co-exist, and actually benefit from one another. People use the internet to reinforce their strong ties through increased virtual interaction, and to establish weak ties that increase their overall personal network. Even if it is a solitary activity, surfers connect to other resources and people more easily and enthusiastically then meeting strangers over the phone or face-to-face. Do you agree with the claim that the internet can benefit, not destroy, the quality of our social networking?

The next article examines and compares the quality of online vs. offline relationships with young people in Israel. In their study, the examined the length of time they reported knowing a tie, and then looked at both active and content multiplexity in their relationships. Studying duration was not a suitable choice for this study, since the internet’s mass proliferation is relatively new and already puts internet tie duration at an unfair advantage. Researchers found that those with online friendships reported fewer topics of discussion than with offline ties. This finding connects to Wellman’s study that found that people use the internet in a specialized and selective manner (i.e. to access certain resources they need). However, it seems that the issue of physical contact is important when it comes to discussing intimate (relationship) issues. However, wouldn’t the increased anonymity of the internet and a user’s search for specialized resources offer them the opportunity to discuss these issues over the internet more, since they don’t risk judgment or disapproval?

The last paper dealt with how the internet affects community life. I found Hampton’s point on the internet as “a new realm for public space” compelling, since we are now liberated from physical barriers and constraints, and are able form active and thriving public spaces in a more accessible and diverse manner. His analysis on Netville revealed that, yes, the internet might be able to increase the quality of our physical relationship with ties in our communities, through increasing tie strength and multiplexity. However, he ends with stating that internet communities can become like “virtual street corners”, yet will not replace the socialization that takes place in physical locales. Do you think the internet, particularly web-conferencing and chatting, will decrease our desire to meet up in public spaces?

Finding ways to analyze and understand social capital is important for our understanding of social networks, in terms of how they form, maintain themselves, and grow. In this week’s readings, four papers discuss the various ways in which social capital can be measured.

In “Auditing Information Structures in Organizations”, the authors reviewed the six main network analysis research methods: sociometric questioning, diaries, observation, archival methods, ECCO questionnaires, and Milgrim’s Small World technique. The authors take this opportunity to discuss the strengths and limitations of each approach, as well as their implications for studying organizational networks. The authors made several interesting points when discussing the various techniques. First, they mentioned that sociometric questioning could be used for uncovering communications patterns within groups. Within an structured organization or corporation, I imagine that this information could be useful in determining hiring decisions for divisions, in terms of figuring out whose communication patterns would work best with a certain team or department. Also, I found the authors’ point about how a respondent interprets a question very interesting, since it could explain some of the more depressing “Lonely World” findings we saw in previous studies. If people consider important matters to be work, politics, or personal, then their answers will differ accordingly as will the people they think of when answering these surveys. .

The next three readings evaluate different generators. In Marin and Hampton’s study, they discussed the four different single name generators, as well as evaluate the MMG and MGRI multiple generators. This article was striking in that it reinforced the general consensus amongst social network theorists – it is becoming increasingly harder to find a reliable and consistent methodology. In their results, they found that the single-name generators did not “approach the reliability of the full multiple generator” (13). In order to combat this problem, the authors propose the incorporation of the MMG or MGRI, which would improve consistency due to a more specific research scope (for example, the use of only “discuss” and “socializing” generators”). This article was very interesting since it dealt with a social network theorist’s real problems in research design, and proposed much improved and less-taxing recommendations.

“The Position Generator” looks to solidify a social capital theory. The authors discuss the position generator, which differs from the previous article’s focus on name generators. They decide to focus on position since it would offer them more valuable insight into network structure. Based on the Taiwan Social Networks Study, researchers were able to determine that social capital is different among the sexes, especially in terms of obtaining it (through social status) and this ultimately women’s exposure to leadership positions, both at work and in their community. I found this study very compelling, since it offers researchers the opportunity to delve deeper into whole network structure, and determine how and why a network has formed based on where ties are within the network. It also offers valuable information on prestige, centrality, and even homophily. In this specific instant, using the position generator offered more conclusive data on comparing women and men’s social capitals. Using such generators are especially beneficial in pinpointing societal problems (such as inequitable social capital) and perhaps help find ways to change the pattern.

The last article, “The Resource Generator”, focused on studying network structure through testing a newly developed generator. This tool combines the position and name generators and looks at “general” social capital indicators as well as places a stronger emphasis on access to social capital. The methodology of this paper was different from the previous ones since it focused on “latent traits”’ along with general demographic information, they included labor market position, income, and occupational prestige. I found this technique to be one of the strongest social network methodologies I’ve come across so far, since it combined two existing (and reliable) measures into a more comprehensive model. The questions administered were able to answer an array of research questions, since one can deduce income, SES, education, and occupational information from it, as well as tie information.

In the interview, Lynne Smith-Lovin and Robert Putnam discuss a myriad of reasons that may attribute to the decline in core discussion networks in the last 20 years. Smith-Lovin initially makes a claim that this specific change is part of long-term social change concerning our allocation of time. She argues that we spend more time involved in paid-labor and are more dispersed, thus, farther away from our strong ties. Peter Monge explains that frequency is an important element in establishing strong ties with others, and so logically, if we have less time to spend with others, it would attribute to the decline in our core discussion network. Putnam adds to this argument by stating that the long-term pattern in lower group affiliation and leisure activities leads to a general decline in our networks. I find this to be the most important element of the change in our core discussion groups, since we simply do not have the time nor have the venues to create strong ties. Also, Putnam later adds that religious institutions have a large role in developing strong ties. As a nation that has become less religious (he notes that this changing), it can have a significant role in the development of strong ties from an early age.

One element that they didn’t explicitly discuss in their interview was the important role of suburbs in changing social networking. Although the mass move to the suburbs was mostly prevalent after WWII, the last twenty years have marked an even stronger move to cities due to rising urban costs and increased urban sprawl, which has contributed to the finding that more people are dispersed. Elizabeth Bott finds that more dispersed family units have less strong ties with one another. Since Wellman and Worley found that we get a lot of our support services from family, then we can conclude that this pattern of dispersement directly contributes to our declining number of strong ties.

Also, to additionally understand the more grand-scale structural societal changes taking place, Kalmijn would mention the changes in our patterns of dyadic withdrawal due to changes in marriage rates, cohabitation, and divorce. He would note that since our society has become more liberal when it comes to establishing serious relationships (i.e. more acceptable for people to cohabitate without being married), and combine this with the fact that most of McPherson’s findings that most strong ties are spouses, we can assume that network size is declining since more people are cohabitating, and thus replacing their larger network and support of close friends with the similar resources their spouse can offer them. Also, since the divorce rate appears to have risen from the 80s to 2004, one can also assume that a loss of a strong tie in our core discussion group could be due to losing a spouse as a confidant.

The basis of Punam and Smith-Lovin’s discussion revolves around strong ties in core discussion networks. They classified these individuals as people with whom one would discuss important things. Based on the findings of Wellman and Worley in “Different Strokes from Different Folks”, the authors explained the different resources that ties offer someone. They note that supportive functions have a direct relationship with the strength of the tie. Their research concluded that people get the majority of their social support from their small number of strong ties, and includes emotional aid, small services, and companionship. Thus, if our core discussion groups are declining, which means that our strong tie network is declining, then we can expect a direct decline in the availability of services. Also, is Wellman’s paper on “The Network Community”, he cites that most community ties are specialized, where members within that community offer specific kinds of social support. So, since the two talk-show guests mentioned the overall decline in community involvement and civic duty, we can additionally conclude that, overall, we are losing other specific kinds of social support that came from our community networks. In Wellman’s study, he additionally adds that communities “are not just a way in which people spend some of their leisure time but key mechanisms by which people and household obtain resources.” Thus, based on the findings that people have less leisure time and lesser role in community and civic affairs, one can assume that their ability to obtain resources (services, information, support) will also be negatively affected.

In McPherson’s study on the decline of social networks also raises the interesting issue of homophily and predominately kinship ties. In “Birds of a Feather”, McPherson and Smith-Lovin also note that homophily is dominant in people’s social networks, in regard to forming new ties, and retaining old ones. If we are predominately strongly connected to those who are like us, chances are that we will face redundancy, where the same support services are offered my multiple people. Consequently, we constrict ourselves from gaining other support services that may not be prevalent within our existing homophilous network. Also, friendship ties that are less similar dissolve at faster rate. Thus, this could lead us to lose more diverse support offerings. Our ability to form new ties is also limited in that we network to those that are similar (in terms of race, gender, values, activities), so, as a result, we directly limit the amount of new information available.

The issue of social capital is another interesting perspective to approach the decline in core discussion networks. In Burt’s research about social capital, he notes that people obtain information, referral, and access from other ties. Thus, with the decline in a core discussion group and Putnam’s observation of an overall change in the size of networks (including weak ties), it is possible that we are experiencing a decline in social capital. This is also reinforced by Burt’s observation that social capital is rooted in the social capital of organizations and communities, so if we have less leisure time to partake in those associations, then that also contributes to declining social capital.