Social Network Blog - g18

My results from our Social Network Survey were very interesting, especially in regard to the readings that we have done throughout the semester. The first noteworthy finding of my study was the mean numbers of discussion partners among my subjects. For class we read McPherson et al.’s article “Social Isolation in America: Changes in Core Discussion Networks over Two Decades.” In this study, the authors found that the number of people who said that they had no one with whom to discuss important matters nearly tripled between 1985 and 2004. They also found that the mean number of discussion partners for such matters had decreased from 2.94 to 2.08 during this time. In light of these findings, I found my results to be extremely interesting. I found the average number of confidants for all of my subjects to be 4.7. Even when broken down by age and sex, my means were significantly higher than McPherson’s findings. For the Young Male group the average was 3.6, for Old Males it was 4.6, for Young Females it was 5.2 and for Old Females it was 5.4. This is of interest because it goes against McPherson et al.’s assertion that the size of our discussion networks is decreasing. Furthermore, not a single one of my respondents said that they had no one with whom to discuss important matters. In addition to the average number of discussion partners, I found several other areas of interest in my results.

Gender
The first is the differences between male and female respondents. On average, the females in my study had 5.3 people with whom they discussed important matters, whereas the males had only 4.1 on average. In spite of this a great difference between the number of total ties males have and females have, the difference in the types of relationships that make up these ties is minimal. The McPherson et al. study found that men and women’s kin networks and non-kin networks were converging and that men no longer had significantly more non-kin ties than women and women no longer had significantly more kin ties than men. My results supported this finding. In my results, the percentages of ties that were kin were relatively similar between males and females, as were the ties that were non-kin. An additional area of interest when comparing the genders of my respondents is the diversity of their core discussion networks in terms of the type of support that they provide. This is particularly noticeable between the two younger groups. Young females had confidants in 8 out of the 11 possible relationship types, whereas young males only had confidants in 5 out of the 11 types. Since each of these different types of relationship provides different types of support, as was explained by Wellman and Wortley, younger females receive more different types of support from their core discussion networks than younger males do. These differences in support were not nearly as pronounced among the older groups, which brings me to my next area of interest: age differences.

Age
My results were very different from those discussed in McPherson et al. with regards to age. The 1985 study discussed by McPherson et al. found that, with age, network size decreased greatly. Conversely, McPherson’s 2004 study found that age was not strongly related to network size. However, I found that not only was there a correlation between age and network size, but the relationship between age and network size was the opposite of that found in the 1985 study. For males, the younger group had an average size of 3.6, whereas the older group had an average of 4.6. The difference for females was less significant, but still existent. The younger group had an average of 5.2 and the older group had an average of 5.4. This increase in discussion network size with age goes against both the 1985 study and the 2004 study. In addition to the differences between my respondents based on age, I found it interesting to look at the differences between the ages of the ties that they listed. In many of the surveys there was little age homophily between the respondent and the people with whom they discuss important matters. I did not find this result particularly surprising because many of the subjects discussed important matters with parents and siblings who are unlikely to be one’s exact same age. While homophily did not have significance in age, it was significant in other areas.

Homophily
One area in which homophily was present was gender. I found that, with the exception of family members, respondents almost never discussed important matters with people of the opposite sex. The only other instances where this occurred were when the person was the respondent’s girlfriend or boyfriend. In addition to gender, there was homophily in the education level of the subjects. Nearly all of my subjects either were currently attending college or had completed college. Similarly, most of their discussion partners had achieved or were planning to achieve the same level of education as the respondents. This homophily in discussion networks is not surprising in light of McPherson et al.’s discussion of homophily, from which they concluded that there was strong homophily for both strong and weak ties. Presumably, if one discusses important matters with someone, they are strong ties. If this holds true, then my findings go against Granovetter’s notion that tie strength can be measured by frequency of interaction.

Frequency of Interaction
I found that the frequency of interaction, as reported by my respondents, ranged from very occasional (2 days a month) to very frequent (30 days a month). If a respondent is discussing important matters with someone, it seems logical that he or she would consider that person a strong tie. There were cases in my surveys where a respondent interacted with a tie fewer than 10 times in a month, but had known that person for 20 years, whereas they interacted almost daily with someone who they had known for only 2 years. In both cases the ties were very strong, but their frequency of interaction was very different. Therefore, while a tie might be considered weak on Granovetter’s scale, the respondent still clearly considers them strong if he is listing them as one of his few discussion partners.

Position Generator
In addition to gender differences and homophily, I was interested in looking at how network diversity influenced the number of discussion partners of each of my subjects. I looked at both the number of jobs listed on the position generator and the diversity of these jobs with relation to their scale. In Nan Lin’s study of position generators, she discusses the importance of position generators with relation to social capital. She also discusses inequality between male and females in their social capital. She says that men have more social capital because of their presence in the work force. However, based on the position generator, I found that women actually had more social capital than men. On average, females listed 9.2 jobs from the position generator that had a range of 12.1. Males, on the other hand, averaged 7.4 jobs with a 10.4 range. Women, who according to this position generator have more social capital, also have more average discussion partners than males. However, there were also several cases in which people who had more diverse networks had fewer discussion partners and people who had less diverse networks had more discussion partners. There is no way to prove causation here, but it is clear that there is some sort of correlation between social capital and the number of discussion partners. Based on my surveys, there was no significant correlation between network density and the number of people in one’s close discussion network.

Media Use
The use of different types of media was interesting here. None of the younger respondents used landlines to talk to the people who they listed and the older respondents used landlines equally or less frequently than they used cell phones. This is interesting because one might expect the discussion of important matters to be limited to more private forms of media. However, as I found in my diary of my new media use, this is frequently not the case anymore. Postal mail was nearly obsolete in my surveys and e-mail and IM use was significantly less common than cell phone use. This showed that respondents still prefer to talk to their close ties through media that is viewed as more “real”, as Baym and Zhang discussed that internet is still viewed by many as a less “real” medium for interaction.

Problems
While many of these findings seemed interesting to me, I think that there are many factors of this particular study that could have affected my results. First, as McPherson et al. discussed, there is a great deal of ambiguity in the use of the term “important matters”. Several of my respondents specifically asked me what I meant by “important matters”? Since everyone can interpret this question differently, it is difficult to really compare their results. Additionally, I don’t feel that this position generator accurately measures the diversity of one’s network. I know several instances in which the respondent knew someone distantly, but happened to know their first name and therefore included them in the position generator. However, for example, I know that my mom knows my gardener by first name, but she does not have social interactions with him to the point where I would include him in her social network. However, these were both issues that were present in the original study in 1985 and the 2004 study as well, so they did not affect my comparison between my results and theirs. However, there were also several things that did differentiate our study from theirs. First, our sample size was extremely small, so it is really difficult to draw any conclusive information based on these results. Additionally, our subjects were not randomly selected. I distributed my surveys to people who were most accessible to me (family, friends and friends of my family members). Therefore, my results are much skewed. Additionally, since I was reading these surveys to these people, they likely would have felt pressured to respond in certain ways. For example, friends of my mom might have felt bad not listing my mom as one of their close discussion partners, which could have influenced the results. Similarly, people might have been embarrassed to admit to having few people with whom to discuss important matters, so they might have included people who they do not actually discuss important matters with. In several instances, I had respondents list 5 people and while listing their 6th person they would say “I could list a bunch more, but I guess I will just say this one.” Therefore the numbers of people listed on the surveys are not necessarily accurate representations of the actual networks. Additionally, I think that the length of the survey skewed my results. I had respondents (particularly boys 18-22) who stopped in the middle of taking the survey and told me that it was too annoying and that they wanted to stop. I am sure that many of the subjects who did complete the survey felt similarly and may not have answered accurately, which is always an issue with self-reporting surveys. Additionally, this survey touches upon fairly personal and private issues, which could have led respondents to be uncomfortable and to self-report even less accurately. Overall, I think that this survey was interesting as were many of my findings, but I do not feel that one can draw definitive conclusions based on this very small, skewed sample.

Download file

I found my results of the Small University Experiment to be quite similar to what I had predicted in part 1 of this assignment: http://www.mysocialnetwork.net/blog/481/g18/2006/09/small_university_experiment_pa.html. As predicted, my first link (a fellow student at Penn) passed the folder on to a student at the Graduate School for Education. As I predicted in the first part of this assignment, all of the recipients of the folder were females. Additionally, the folder crossed over from the College to the Graduate School of Education very early in the process. I was unsure as to how many links would receive the folder before reaching Susan Yoon, but I predicted approximately 5 links. As it turned out, my folder only took 3 links to reach Susan. Another area in which I predicted incorrectly was the length of time it would take for my folder to reach Susan. I suspected that it would take about two weeks for my folder to get to Susan, when it actually took only three days. There are a few possible explanations for the differences between my predictions and my results.

Passing my folder on to a strong tie (Heidi) meant that I was able to predict how she would proceed with the folder. Since I know her very well, I knew that she knew someone in the GSE (Morgan). I also knew that she, being one of my closest friends, was trustworthy and would make every effort to pass the folder along in the most advantageous way. While I was able to predict two links into the chain, I did not know much about my friend’s tie in the GSE. Therefore, I did not know how many more links it would require for Morgan to reach Susan Yoon. In my prediction, I aired on the side of caution and guessed that the folder would reach two more people before finally reaching the target. While I expected the number of ties to be relatively small, I overestimated the size and underestimated the connectedness of the Graduate School for Education, causing my prediction to be too large. In addition to the number of links, I misjudged the amount of time it would take for my folder to reach Susan. I think that there are several factors that led to the rapid arrival of my folder. First, as explained above, my folder went through fewer links than expected on its path. The fewer people a folder is passed to, the shorter time it should take for the folder to reach its target. Therefore, the above explanations would also help to justify the rapidity with which my folder traveled. Additionally, both of the intermediary links were strong ties. As a result, the folder’s recipients might have felt guilty not passing on the folder in a timely manner. As my close friend, Heidi would have felt obligated to do me this favor. Similarly, she passed the folder on to her cousin’s fiancé, who would have probably felt a similar sense of obligation. In fact, in this case the obligation might have been even stronger because not only was this a close tie, but also a soon-to-be family member and therefore Morgan probably wanted to make a good impression on my friend by doing her this favor. In addition to the closeness of these ties, general timing also played a role in why my folder traveled so quickly. It just so happened that my first link already had plans to see the second link the weekend after receiving my folder. Since Heidi and Morgan had already planned to meet, passing the folder along was less of a burden to Heidi. Had these two links not had plans already, Heidi would have had to specifically seek out Morgan in order to pass the folder along, which would have undoubtedly taken a longer time.

My personal folder was interesting in relation to the overall class data. The mean number of links for completed chains in our class, when the data from our two targets was combined, was 3.5 links. While this number is very slightly higher than the 3 links my folder went through, it is still lower than my predicted number. This is interesting in light of our readings by Milgram and by Stevenson et al. In Milgram’s Small-World study, he found that it took an average of five links to connect any two given people. It does not surprise me that our average number was lower than Milgram’s because his study was done between people across the country, while ours was limited to a much smaller geographic area (within a ten block radius). Because of the campus nature of our experiment, it seems to make sense that we would require fewer links to reach a given person. However, the Stevenson et al. study found that only 1.25 links were necessary to reach a target person on a college campus. While this number is much lower than ours, the author also discussed previous studies that took place within organizations. In Lundberg’s study, he found that the mean number of intermediary links for organizations was slightly more than 3 (very similar to our results). In Stevenson and Gilly’s study of hospital settings, they found the mean number to be 2. There is one major difference between our study and Stevenson’s that I think could have greatly contributed to the large difference between our results. This difference has to do with the selection of the target people. Stevenson explained, “we picked the undergraduate dean of the school of management as the target. The undergraduate dean was picked because he was located in the building where most of the classes are held and would be easy to physically access,” (3). For our experiment, Professor Hampton specifically chose targets that would be difficult to access. Had he chosen, say, the undergraduate dean of the communications school, we might have reached the target in fewer average links.

In addition to the comparison of our number of links to those of Milgram and Stevenson’s studies, it is interesting to compare our results to theirs in other areas. For example, Stevenson et al. found that folders generally stayed within the student population until being passed to an administrator. While the majority of our transfers were to people of the same affiliation, this is only a slight majority. We found that 58.5% of the transfers made were between two people of the same affiliation. This could be due, again, to the relative obscurity of our targets in comparison with Stevenson’s study. Perhaps students thought it would be more effective to pass their folders to people of different affiliations in order to reach their somewhat random and difficult to find target. In contrast to Stevenson’s findings that students never passed their folders to someone in a younger class, our results showed that students did pass their folders down. While it was the most common for folders to be passed within the same grade (47.8%), 34.8% of students passed their folders to younger students, while only 17.4% passed their folders to older students. However, I think that this could potentially represent a flaw in our particular method of analyzing our data. We based the “grade” of the students by the number of years that they have been at Penn. However, a graduate student who has been at Penn for 1 year is older than a senior undergraduate. Therefore this would show up in our data as having been passed to a student in a lower class, when really it was passed to a student in an upper class. Also, most of our classmates are seniors, making it difficult to pass the folder to students who have been at the school longer than they have. Similarly to Milgram’s and Stevenson’s findings, the majority of our folders were passed to people of the same gender (72.55% of folders). Furthermore, 90% of our final links were the same gender as the target person. I did not find this to be surprising, as I predicted in the first part of this assignment that my folder would remain with females. Based on McPherson et al.’s study of homophily and our stereotypical views of who would be the most responsible with passing the folder along, it makes sense that most of our folders stayed with females.

There were also interesting results in our study that were not really addressed in Stevenson’s study. First, we found that only about half (54.9%) of our transfers were between people from the same school. We also found that only about a quarter (25.5%) of transfers were between people from the same department or major. Based on homophily, one might have guessed that these percentages would have been greater. However, since both of the targets were from both different schools and different departments than we were, it was necessary for these transfers to cross school and department boundaries when reaching the target. I also found it interesting to look at the use of strong versus weak ties for the 2nd alter. In spite of Granovetter’s discussion of the strength of weak ties, the majority of our classmates (though only a slight majority of 7 as opposed to 5 and 4…) passed their folders on to strong ties. I think that this shows that, even if it might not be the most effective route, we are more likely to trust our stronger ties to help us than we are to trust weaker ties. Overall, I think that our results were not unexpected. I think that many of the differences that we found between our results and those of studies we have read previous all seem to have logical explanations as far as differences in the studies themselves.

1. The five people with whom I interacted the most often were:
1. Mom - 24
2. Julie S. - 23
3. Hannah and EKG tie - 16
4. Heidi and Emily J. tie - 12
For each of the mediums used, the people with whom I interacted the most often were:
Cell:
1. Mom - 24
2. Dad - 6
3. Hannah - 6
4. Julie S. - 5
5. Tie: Heidi, Raine, Emily J. and EKG - 3
Text Message:
1. Julie S. - 8
2. Hannah - 8
3. Heidi - 7
4. Emily J. - 4
5. Monty - 4
Instant Message:
1. Julie S. - 10
2. EKG - 9
3. Shosfy - 5
4. Stackman - 5
5. Emily J. - 4
E-Mail:
1. 2009 listserve - 5
2. Tie: Julie P, Chloe and Lauren W - 3
3. Tie between 7 people/listserves - 2

Mom – Mother; known for 20 years; strong tie
Dad – Father; known for 20 years; strong tie
Julie S. – friend, classmate, sorority Big Sister; known for 2 years; strong tie
Hannah – close friend, former roommate; known for 2 years; strong tie
Heidi – close friend, classmate; known for 2 years; strong tie
EKG – friend, classmate, co-worker; known for 3 years; moderate tie
Emily J. – friend, co-worker, housemate; known for 2 years; moderate tie
Raine – friend; known for 1 year; moderate tie
Shosfy – friend, co-worker, housemate; known for 2 years; moderate tie
Stackman- close friend; known for 2 years; strong tie
Julie P. – close friend; known for 2 years; strong tie
Listserves – work-related; made up of friends, acquaintances and co-workers; known from 1-3 years; moderate and weak ties
Lauren W. – co-worker, friend; known for 1 year; moderate tie
Chloe – co-worker, friend; known for 1 year; moderate tie

2a. Based on the data that I collected in my diary, I found there to be a correlation between tie strength and the medium of communication used. For example, I only used my cell phone to talk to strong ties and, on occasion, moderate ties. Over the course of the whole week I never called any weak ties. Similarly, with text messages I almost always interacted with strong and moderate ties. There were a couple of occasions on which I interacted with weak ties via text message, but these were very rare. On the contrary, my e-mail interactions were almost all with weak or moderate ties. The strong ties with whom I interacted through e-mail were almost all friends who are currently studying abroad, making cell phone interaction very difficult. With instant messages, I conversed entirely with strong and moderate ties. I did not have any IM interactions with weak ties throughout the week. There were several factors that I found interesting about these findings. First, of the people with whom I interacted the most frequently in total, two of them were moderate ties. This was surprising because I had about seven strong ties with whom I interacted much less over the course of the week. This provides an example of how using frequency of interaction as a measure of ties strength, as was discussed in Granovetter’s article, is not necessarily accurate. For example, my older sister is currently working in New York and is very busy. Consequently, we did not interact much over the course of the week, but she is still an extremely strong tie. I also found my results for instant messaging to be surprising. I expected to have interacted with many weak ties via instant message. In reality, I didn’t interact with any weak ties through this medium. The slight anonymity of instant messaging led me to believe that I would have been more likely to use it when talking to people to whom I was not very close. However, my findings make sense in light of the Baym, Zhang and Lin article that we read, which explained that relationships are maintained through multiple media. The Mesch and Talmud article also mentioned the fact that we use the internet very often for maintenance of existing ties. My instant messaging interactions definitely supported this notion.

2b. In addition to the relationship between tie strength and medium, I found there to be a correlation between the type of support exchanged and the medium used. However, this relationship was slightly less significant than the above correlation. While in a few instances my e-mails were used for class information or for connecting with friends who are studying abroad, my e-mails were very predominantly work-related. These results seem to go against the Baym et al. findings that “e-mail was the main internet medium for social interaction.” In fact, e-mail is the only medium that I never used for social purposes, such as companionship. Instead, my view of e-mail as a more professional medium than the others I use led me to use it most frequently for work-related exchanges. While my e-mails were almost always work related, my exchanges of job information were not limited to e-mail alone. I actually used all four mediums for this purpose. I think that this is probably fairly unusual. My “job” entails me working with many people who are also my friends. Since our relationships are not strictly professional, our interactions were not limited to strictly professional mediums. While my exchange of job information was not limited to a specific medium, my exchange of emotional support occurred most often on my cell phone. Conversely, I noticed that when exchanging class information I never made cell phone calls, but instead used all of the other three mediums for discussion of school work.

2c. There was also a correlation between the type of relationship I have with a tie and the medium used with them. The only medium that I used to interact with my parents was my cell phone. This could be attributed to the fact that I most often used my cell phone when exchanging emotional support and I very often turned to my parents to provide this type of support. This could also be a result of their age, which I will come back to shortly. When I was interacting with people who were mainly co-workers (as opposed to my friends who I also happen to work with), our main means of communication was e-mail. However, we also interacted, on occasion, using several other mediums. Because, as I said above, I work with many people who are also my friends, there was some blurring in terms of relationship type in these situations. With the exception of my parents, nearly all my most frequent interactions were with multiplexed ties. As was discussed in the McPherson et al. study, it makes sense that many of my strongest ties and my most frequent interactions are with people with whom I have multiplexed relationship because multiplexity leads to homophily, which in turn leads to increased likelihood of strong ties. Because of this multiplexity, I do not find it surprising that my interactions with friends, co-workers and classmates spanned several mediums. My use of instant messages was limited to interactions with my friends. While many of these people may have also been classmates and co-workers, I did not use instant messages to interact with any family members or people who were merely acquaintances.

2d. and 2e. The duration of my relationships did not seem to be at all related to the mediums I used for interaction. The majority of the people with whom I interacted are friends who I have met here at Penn, and therefore many of my relationships have had very similar durations. This makes it difficult to see if duration of relationship has any impact on the type of medium that I would use for interaction. On the contrary, the physical distance between me and each of my ties did have a connection with the medium used to communicate. However, this correlation was only significant when the distance was extremely great. When my ties were further than one thousand miles away from me, our interactions were limited to instant messages and e-mails entirely. For anyone who lived fewer than one thousand miles away from me, there was no relationship between distance and medium. While the Mesch and Talmud study found that online ties are most often not between people of significantly different geographic locations, in my results I found that the internet was used in order to maintain ties that were geographically very distant. While their article focused on ties that originated online, which is not the case for my ties, I still find it important to note that their findings are not necessarily accurate for all online relationships. Because I have many friends who are currently abroad, communicating through non-online mediums is quite difficult, which therefore increased our need for online interaction. This returns again to the idea that we often use the internet to maintain existing ties, rather than to form new ones. Additionally, as McPherson et al. explained, “it takes more energy to connect to those who are far away than those who are readily available.” If this statement holds true, it would make sense that my interactions with physically distant ties would be mostly online because online interactions make overcoming great distances significantly easier.

2f. and 2g. Since the vast majority of the ties I interacted with were female, I cannot judge whether or not gender would have had a relationship with the medium I used for communication. However, the mere fact that most of my interactions were with females supports the concept of homophily, as discussed in McPherson et al.’s study. Returning back to Migram’s discussion of the Small World Problem, he found that people were three times more likely to pass a folder along to someone of their same sex. It therefore does not seem surprising that I interacted more frequently with people of the same sex. While gender did not play a huge role in my modes of interaction, age did have a large role. However, similarly to with distance, this only seemed to be the case with extreme differences in age. Most of the people with whom I interacted were between the ages of eighteen and twenty-three years old. In these situations, it was difficult to evaluate any potential differences in mediums used, especially because most of the ties were college students. However, in my interactions with my parents, who are fifty and fifty-one years old, age did have implications for my media choices. As discussed earlier, when interacting with my parents I always called them on my cell phone. The reason for this, I think, is that my parents are significantly less technologically savvy than my friends, co-workers and classmates between the ages of eighteen and twenty-three are. The cell phone most closely resembles older modes of communication, making it the favorable new media communication tool for my parents. Since both of my parents lack strong typing skills and have difficulty navigating their computer and text-messaging functions, using any of these other new media would have made interaction much more difficult.

2h. I think that all of these findings do speak greatly to the role of new media in our social networks. The first finding that I think is noteworthy is the fact that in all situations my new media use was a means of maintaining existing ties. Many of the studies that we read viewed the internet as a way to meet new people, but I did not have any interactions with new ties via the internet. Instead, many of my online interactions were with strong ties, while very few were with weak ties. I think that all of these correlations show how new media helps to maintain existing ties. I also think that new media seems to be better able to achieve this maintenance in homophilous, multiplexed relationships. For example, since my parents are significantly older than I am and I only have one relationship with them, our new media interactions are limited to one form (cell phone use), whereas my friends/co-workers/classmates and I are more homophilous and share several relationships and we interact through several different forms of new media. I also think that new media enables us to maintain a greater number of ties than we would be able to if we were limited to older forms of media. Through instant message and e-mail, we are able to maintain ties with people who we would probably be unlikely to continue interacting with if we could only do so via telephone. Online interactions and text messaging are less time consuming than talking on the telephone and also make it easier for one to multitask. For these reasons, we can maintain ties with more people and maintain a larger social network.

3. I found there to be very little difference, if any, between people with whom I interacted in private places and those with whom I interacted in public places. I interacted with both strong and weak ties both at home and in public. While there were differences in the mediums used in public versus private places (such as using text messaging in a restaurant instead of making a cell phone call), there were not differences in the types of support that were exchanged in public and in private places. It seems that new media use has decreased the gap between what is considered to be a private conversation and what is public. This could have an interesting relationship with McPherson et al.’s findings that we have fewer people with whom to discuss important matters. Perhaps people still have large social networks, but because many interactions are now able to take place in public we are less likely to discuss important matters at all. Maybe this decrease is less a result of changes in the size of our networks, and more a result of changes in the ways we interact with these networks.

This week’s first reading was about deviance and the search process. One of the parts of this article that I found interesting was the author’s discussion of all of the different distinctions that need to be made when discussing deviant behavior. For example, there are both legal and illegal forms of deviant behavior – behavior can disobey societal norms, and therefore be deviant, without necessarily being illegal. What I also found particularly interesting was the discussion of how it is necessary for an act to be difficult in order for it to be considered deviant. The author says that, for deviant behavior, “hardware, facilitators and providers of support are often hard to come by” and that, for this reason, most people with deviant intentions do not end up carrying them out. He then explains that when deviant actions are completed more often, they consequently are considered less deviant. I totally agree with this statement. For example, drinking is illegal for people under the age of twenty-one in the U.S., so it would seem that participating in underage drinking would be considered a deviant act. However, in college, since it is so prevalent, most people do not view underage drinking as deviant behavior at all.

In our second reading for this week, Burt talks about opinion leaders and their role in the spread and adoption of different things. He discussed how both closeness (cohesion) and competition (equivalence) can aid in contagion. Burt also explains that, while both can be predictors of contagion, there are circumstances in which cohesion would not work as a predictor and circumstances in which equivalence also would not work. If the ties between the two actors are strong, both equivalence and cohesion can predict contagion. However, if the two people are not equivalent, you cannot use equivalence to predict whether or not there will be contagion and if people are very equivalent but are not directly connected, then you can only predict using equivalence. In his discussion, Burt uses three examples of observed contagion: medical innovation, businesses, and politics. While I found these to be interesting examples, I also couldn’t help but wonder:
Can you generalize about the population based on these three examples?
I think that the medical field, corporations and the political realm all represent a certain tier of society and I wonder if the contagion of ideas or innovations would have been different if the population studied represented a different branch of society?

The third and final reading for this week was also about diffusion within networks and opinion leaders. In his discussion, the author used many examples ranging from the diffusion of modern math among school superintendents to the diffusion of cholera during an epidemic. I found these examples to all be really interesting and I think that in showing these examples the author really shows the reader why studying diffusion and opinion leaders can be really important. One point that I found especially interesting was the author’s point that opinion leaders conform highly to societal norms. Rather than being innovators and trend setters, like one might expect, the opinion leaders are actually not the first to do something. Often the first people to adopt an innovation might appear too different, making the general public hesitant to adopt. However, once opinion leaders adopt slightly later, other people will begin to follow. I thought that this was very interesting because I tend to equate “opinion leaders” and “trend-setters”, but those who we tend to view as “trend-setters” are not actually the very first people to adopt the trend usually. The author then discusses different ways of measuring this opinion leadership. He discusses four different methods: sociometric, informant ratings, self-designation and observation. The author discusses some of the strengths and weaknesses of each of these methods. Still, I wonder:
Do any of these methods seem especially problematic to you? Is there one that you think seems as if it would be the most reliable?
In three out of these four methods, the subject is the one reporting the results, whether the questions are about him/her or specifically or people who he/she knows. Either way, I think that, as we discussed with the name generator, it can be very difficult to recall one’s own network. Also, people might be embarrassed to say that the base their adoption of certain innovations on someone else, especially if this isn’t someone with whom they are actually close. I think that this sort of self-reporting while in some cases very accurate could also be problematic.

This week’s first article, “The Structure of the Web” by Jon Kleinberg and Steve Lawrence, discussed the way that the internet itself functioned as a web. Though more specifically discussing the paths between web pages and online “neighborhoods”, there were also many connections to human social networks. For example, the authors discussed how the web includes hubs, just as we have discussed in social networks. It was interesting to think of the web as made up of communities of web pages. As the authors described it, “a community can also be defined as a collection of pages in which each member page has more links to pages within the community than to pages outside the community,” (1850). I thought that this was really interesting because I had never really thought of the web in these terms. I had more thought of the communities on the internet being solely based on the interactions that people have on it, rather than existing between the pages themselves.

The “Pentagon Sets its Sights on Social Networking Websites” article was a really interesting look at online social networks and how they are being used today. The article points out the exorbitant amount of information that people voluntarily put on the internet about themselves. While I had heard all sorts of warnings about being cautious about what you post on Facebook because many companies now will look at your profile while considering you for a job, for example, I found the use of the information as discussed here to be quite surprising. While it may be quite effective to use online social networks to track terrorist networks, it is kind of frightening to know that the NSA could have the power to track all of our online interactions. While I understand its potential usefulness, I also think that this sort of surveillance might be bordering a little bit too much on reaching Big Brother status.
Question: Do you think that the level of potential information gain from tapping the web would be worth the lack of privacy that it would entail?

So finally, the moment we have all been waiting for: the connection of all of our social network talk to Facebook. The Ellison, Steinfield and Lampe article was very interesting and enjoyable to read. However, I also must admit that few of their findings came as a shock to me, as an “intense” Facebook user myself. One point that I initially found very interesting was in the authors’ review of literature about online social networks. The authors stated that “when online and offline social networks overlapped, the direction was typically online to offline,” (4). I found this very surprising. While on internet communities such as online dating sites this would seem to make sense, in most other situations I found it hard to believe that most people were forming relationships online first and then meeting these people in person. I found this especially surprising because I feel like we have all been so heavily trained to be cautious when meeting new people on the internet because they may not be who they say they are, etc. However, I was relieved to see that, later in the article, the authors discussed how their results showed the opposite relationship (relationships typically moved from offline to online. These authors attributed these results to the nature of the Facebook site, which seems to make sense to me. After reviewing previous literature on the subject, the authors discussed their own Facebook research. While I found their results to be pretty similar to what I would have expected, I think that the fact that all of their respondents were from one school could have influenced the results. I think that it would be really interesting to do this same study at different schools and see how the results might differ. The article also discusses how Facebook may not increase the number of strong ties that one has, but could greatly increase the number of weak ties that one has. I think that, if this holds true, this could be very significant in providing more job opportunities and gaining new information to people who have a lot of Facebook friends (assuming that they create their profiles keeping this possibility in mind).
Question: Do you think that it holds true that Facebook increases the number of weak ties that one has? Would you consider everyone who you are friends with on Facebook to be at least a weak tie? Or could Facebook potentially create a new category of social ties?

I really enjoyed Barry Wellman’s article, especially because at the end he answered pretty much every question that I had about his article beforehand. I found his response to his own third question to be very interesting. He asked if good online relationships could be considered equivalent to good face-to-face relationships. In his response, he made the very intelligent comment that this comparison works under the assumption that if we were not online at any given time, we would have having face-to-face interactions instead. I think that the truth is, in many cases, being online does not replace real life interaction. I feel like most people I know go online while doing their work, or watching TV or participating in other activities that are displacing face-to-face interaction, but I don’t think that being online itself attempts to replace our face-to-face contact. Therefore, I think that the internet can often help to strengthen our ties, rather than weaken them.

This week’s first article, “The Structure of the Web” by Jon Kleinberg and Steve Lawrence, discussed the way that the internet itself functioned as a web. Though more specifically discussing the paths between web pages and online “neighborhoods”, there were also many connections to human social networks. For example, the authors discussed how the web includes hubs, just as we have discussed in social networks. It was interesting to think of the web as made up of communities of web pages. As the authors described it, “a community can also be defined as a collection of pages in which each member page has more links to pages within the community than to pages outside the community,” (1850). I thought that this was really interesting because I had never really thought of the web in these terms. I had more thought of the communities on the internet being solely based on the interactions that people have on it, rather than existing between the pages themselves.

The “Pentagon Sets its Sights on Social Networking Websites” article was a really interesting look at online social networks and how they are being used today. The article points out the exorbitant amount of information that people voluntarily put on the internet about themselves. While I had heard all sorts of warnings about being cautious about what you post on Facebook because many companies now will look at your profile while considering you for a job, for example, I found the use of the information as discussed here to be quite surprising. While it may be quite effective to use online social networks to track terrorist networks, it is kind of frightening to know that the NSA could have the power to track all of our online interactions. While I understand its potential usefulness, I also think that this sort of surveillance might be bordering a little bit too much on reaching Big Brother status.
Question: Do you think that the level of potential information gain from tapping the web would be worth the lack of privacy that it would entail?

So finally, the moment we have all been waiting for: the connection of all of our social network talk to Facebook. The Ellison, Steinfield and Lampe article was very interesting and enjoyable to read. However, I also must admit that few of their findings came as a shock to me, as an “intense” Facebook user myself. One point that I initially found very interesting was in the authors’ review of literature about online social networks. The authors stated that “when online and offline social networks overlapped, the direction was typically online to offline,” (4). I found this very surprising. While on internet communities such as online dating sites this would seem to make sense, in most other situations I found it hard to believe that most people were forming relationships online first and then meeting these people in person. I found this especially surprising because I feel like we have all been so heavily trained to be cautious when meeting new people on the internet because they may not be who they say they are, etc. However, I was relieved to see that, later in the article, the authors discussed how their results showed the opposite relationship (relationships typically moved from offline to online. These authors attributed these results to the nature of the Facebook site, which seems to make sense to me. After reviewing previous literature on the subject, the authors discussed their own Facebook research. While I found their results to be pretty similar to what I would have expected, I think that the fact that all of their respondents were from one school could have influenced the results. I think that it would be really interesting to do this same study at different schools and see how the results might differ. The article also discusses how Facebook may not increase the number of strong ties that one has, but could greatly increase the number of weak ties that one has. I think that, if this holds true, this could be very significant in providing more job opportunities and gaining new information to people who have a lot of Facebook friends (assuming that they create their profiles keeping this possibility in mind).
Question: Do you think that it holds true that Facebook increases the number of weak ties that one has? Would you consider everyone who you are friends with on Facebook to be at least a weak tie? Or could Facebook potentially create a new category of social ties?

I really enjoyed Barry Wellman’s article, especially because at the end he answered pretty much every question that I had about his article beforehand. I found his response to his own third question to be very interesting. He asked if good online relationships could be considered equivalent to good face-to-face relationships. In his response, he made the very intelligent comment that this comparison works under the assumption that if we were not online at any given time, we would have having face-to-face interactions instead. I think that the truth is, in many cases, being online does not replace real life interaction. I feel like most people I know go online while doing their work, or watching TV or participating in other activities that are displacing face-to-face interaction, but I don’t think that being online itself attempts to replace our face-to-face contact. Therefore, I think that the internet can often help to strengthen our ties, rather than weaken them.

This week’s readings dealt with the issues of internet use and the role that it plays in our social networks. The first article, about Shannon Syfrett’s chain e-mail experiment, was very interesting to me. I was surprised that she received so many responses to her chain letter because I feel that most people I know have a tendency to delete such e-mails as junk mail and not pass them along or respond to the original sender. For this reason, I find it very interesting that her experiment was so “successful” and reached such a broad range of people. While I understand the internet’s capabilities of bridging people from such different places, I would have thought that people’s personalities and laziness would have caused a road block in this situation. I also found it very interesting/creepy that people were so concerned about replying to the chain letter that they searched for Syfrett’s phone number to call her and tell her that they were having issues passing the letter along. This made me wonder what exactly the contents of this e- mail chain were. Did the e-mail simply explain what her experiment was intended to find out? Or did she make up some sort of story to play off of people’s superstitions (as is often the case with such e-mail chains). Thoughts?

The Wellman and Gulia article that we read this week was concerned with internet use and its role in dissipating or extending our communities. One point that I found particularly interesting was the authors’ point that internet is not only used for information exchange, but also for various other resources. This article focuses on the community aspect and how internet groups can provide various different kinds of support. This support is often found in communicating with strangers. This is very strange, the authors notice and I agree, because in person we rarely communicate with strangers in a way that is at all personal. While I understand that it might be easier to interact with a stranger when there is a sense of anonymity, as is the case with the internet, I still find it interesting that so many people find comfort in forming communities with people who are complete strangers to them. The authors make the very valid point that, when communicating online, one can easily withdraw from a conversation by merely clicking “exit” if their conversation ever becomes uncomfortable. While this makes a lot of sense to me when concerning one’s online interactions with strangers who you might want to be easily able to cut off, I think that it is interesting to extend this notion the our online interactions with people who are not strangers.
Question: To what extent do you think that the ease of conversation termination in online interaction plays a role in what conversations we choose to reserve for online conversations?

The third article studied college students’ interactions and how they vary between face-to-face, telephone and online interactions. I liked that the authors discussed the fact that most people with whom we share online interactions are also people with whom we have face-to-face and telephone contact. It seems that many researchers assume that online interactions are a completely separate entity from all of our other interactions, when really the internet is often just an additional outlet with which to interact with the ties we already have. As the article states, “relationships are maintained through multiple media,” (302). While we may prefer to use certain media with certain ties, in general we tend to use a variety of different mediums of interaction with our friends, family, etc. These researchers did studies of interactions using two different methods. First they used a diary similar to the one we are completing for class, and second they asked subjects questions about their social interactions. Among other things, they found that people were more likely to use the internet to interact with people across long distances. This makes a lot of sense to me. Having many friends who are abroad this semester, I have been keeping in contact with them mostly through IMs and e-mails because calling Europe frequently would be very expensive. Rather than having to pay extra for long distance interactions, it would make sense that people would choose to use the internet instead because it does not take into account how far someone is from you. The option of using the internet can give someone the opportunity to interact with someone much more frequently without having to be concerned with financial obligations. Also, I found it very interesting that the authors found that those who use the internet the most are more sociable to begin with. I think that it makes sense that someone who enjoys having a large social circle and meeting new people would be more likely to use the internet, which gives them the opportunity to further expand their networks. Still, it seems a little bit strange to me that someone who is sociable and has a lot of contacts would need to use the internet to expand their circles.
Question: Why do you think this is? Or do you think that highly sociable people simply have more contacts to maintain, and this task is more easily accomplished on the internet than through face-to-face interaction and telephone use?

The Mesch and Talmud study that we read discusses the internet’s role in forming new relationships. When first reading this article, I immediately thought of the online dating craze that is currently sweeping the country. I think that, in light of this craze, many would argue that the internet is definitely a place to create new ties and expand your network. The authors also explain that the quality of a social tie is not contingent on where exactly this tie was formed, but rather on other factors. I agree with this statement. I think that, while certain interactions imply more closeness in relationships, it is still possible to develop meaningful friendships through unexpected media, as long as the two people involved get along well. An in-person interaction between two people who have nothing on common will not be of higher quality than an online interaction between two people who get along very well.
Question: Although it may be possible to form quality relationships through media like the internet initially, do you think that it is possible to maintain such quality relationships over time?

The final article, by Hampton, discusses various previous studies of computer-mediated communication. One such study found that “frequent internet users increased the size of their social networks, had greater face-to-face interaction with friends and family, and became more involved in community activities,” (223). I found this extremely interesting because the study essentially found that the internet does the opposite of all of the things that people had speculated it would do to our social interactions. Rather than destroying our sense of community and our involvement in such, this study discussed here suggests that the internet has actually improved all of these things. The author also discusses how computer based media tends to be local and reinforces our existing ties at home, in the neighborhood and in the workplace. In addition to its practical use for long distance ties, the internet is being used to increase the ease with which we contact even very local ties, which I find very interesting.
Question: While I understand how the internet increases the size of one’s network, I am having trouble figuring out how the internet would increase one’s face-to-face interactions and involvement in community activities. What do you think could cause these other increases?

The articles that we read this week all discussed various measures of social capital. The Zwije-Koning and De Jong article discussed the various different methods of network data analysis. They addressed the techniques of sociometric questioning (the most commonly used form), diary research, observation research, analysis of archival records, ECCO analysis and the small-world technique. I found the discussion of sociometric questioning to be particularly interesting. Although the authors tell us that this is the most commonly used method of network analysis, their discussion of the many flaws in this process makes the reader wonder why? I find it so interesting that despite the issues with validity in this process, researchers still continue to use it so heavily. The authors discussed how in any self-reported study, you run the risk of respondents being untruthful. They also talked about how in an office setting, employees might feel threatened and feel obligated to come across a certain way, and therefore to respond to the survey in a certain way. This made me think about the smoking study we read last week, in which the researchers gave students surveys in school. Like handing out surveys in a business environment, handing out surveys in a school environment could drastically influence the results. There was also discussion of the ways the respondents interpret the questions. As we discussed when talking about the number of people with whom we discussed important matters, question interpretation can also lead to much skewed results. Even when the authors compared sociometric questioning to the other methods of network data analysis, it fell far short of the others. In light of these issues it is difficult to understand why this method is still so widely used. While I understand that research of this nature is difficult, and, as the authors explain, this method is very efficient, I find it bothersome that we are still relying so heavily on this method of research.
Question: Since this method is considered to be the most efficient, can you think of any ways that it could be altered slightly so as not to drastically change the efficiency, but to improve the accuracy (particularly in terms of environment like the school and office situations discussed above)?

The Marin and Hampton article discussed issues of reliability with name generators and the differences between single and multiple name generators. One part of this article that I found very interesting was how the authors discussed the use of name generators within the four different approaches, but then how name generators are often misused. They discussed how, particularly with the role-relation approach, name generators can run into issues of reliability. The authors also discuss how name generators can be time consuming and difficult to administer. Similarly to the first article, these authors acknowledge that many researchers search for ways to reduce the time of such methods. It seems from these two articles that researchers often sacrifice reliability for efficiency. Given that shorter surveys may often lead to less accurate results, and longer surveys may lead to antsy respondents, which may also lead to less accurate results…
Question: Do you think that there is any possible way to balance these two? Which would you think is more important to the accuracy of a study: keeping the respondents’ morale up, or asking all of the more detailed questions?

The third article also discusses a generator, but this time they are discussing position generators, rather than name generators. First the authors discuss several more issues with name generators. They then go on to explain position generators and the benefits of using this method rather than name generators. Although the authors seem to find this method to be more accurate than name generators, I found it to be somewhat confusing.

The fourth and final article combines the ideas of the previous two. This article talks about resource generators, which are essentially combinations of name generators and position generators. Van de Gaag and Snijders argued that measuring social capital in a single way may cause us to lose a lot of information, so they aimed to measured social capital in a way that addressed each of its separate parts. The resource generator, they claim, achieves this by combining the good parts of name generators with the good parts of position generators. I found each of this week’s articles a little bit difficult to get through because, when read together, the different research methods that they discussed all sort of blended together. One thing that these articles definitely helped open my eyes to how difficult it really is to find a research method that can accurately measure social networks.

When I first read the title of this week’s readings (“Popularity, Centrality and Prestige”) I was excited an intrigued. I thought that the readings would be very interesting and spark a very long blog posting from me. And then I read the first two articles for the week…

The first reading, the chapter on Centrality and Prestige, discussed previous definitions and concepts related to centrality and prestige. Although I found it interesting to learn about all of the different ways in which these have been measured and discussed in the past, I thought that this article was so overly mathematical that it was painful for me to get through. I must admit that I enjoyed the use of the example of Florentine families, since I spent this past summer in Florence. I also found the authors’ three measures of centrality (p. 218) to be interesting. They listed them as 1. total effects centrality (“the total relative effect of an actor on the others in the network”), 2. immediate effects centrality (“the rapidity with which an actor’s total effects are realized”), and 3. meditative effects centrality (“the extent to which particular actors have a role in transmitting the total effects of other actors”).
Question: Do you think that there are any other measures that should be taken into account? Or do you think that these measures are sufficient?

I found the Freeman article to equally as boring and mathematical as the previous chapter was. This article also discussed how centrality has been viewed in the past and how there have been conflicting definitions of centrality. The author states that “the time has come…to stop, take stock, and try to make some sense of the concept of centrality,” (p. 217) and claims that this is the aim of the essay. However, the mathematical nature of this article just added to my confusion. One point that was clearly made that I found interesting was the author’s claim that centrality is really defined by an actor’s potential for control. It is interesting to think of how much control these well-connected people have. By being in intermediary on the paths between many actors, a central actor really does have a great deal of power of his or her network.

Just when I thought that I was going to take this week off from posting my blog because I had nothing to say about the previous 2 articles, I got to the “Uncloaking Terrorist Networks” article by Krebs. I thought that it was really cool how this author used social network analysis to look into Al Qaeda’s network. Although I thought that the author’s scales on which he determined tie-strength were a little bit iffy, I thought that this article was generally really interesting. I found it curious that the Al Qaeda network was generally dispersed. It would seem that this would have been risky for the terrorists because it would seem that for an event that was such a big deal you would want to work with people who you trusted, rather than weak (or even non-existent) ties. The author says that this might be because “keeping cell members distant from each other…minimizes damages done to the network if a cell member is captured,” but this seems counterintuitive to me. It would seem that a weak tie would be more likely to offer up information than a strong tie that is more invested in the people in his network. I also appreciated that the author admitted to weaknesses in his discussion.
Question: Can you think of any ways to better establish the strength of their ties other than based on who went to school together, etc.?

I thought that the article on popularity and smoking was also really interesting. I was intrigued by the findings that popular students were more susceptible to developing smoking habits. The author presents one possible reason for this – “students who were popular in the sixth grade may have felt that they needed to smoke in order to maintain their popularity,” (p. 328). Question: Other than this, what do you think could be causes for these results? I also found it interesting that the researchers found that, although other people listed smokers as friends more often, smokers listed fewer friends. This goes against what we discussed last week, when we said that smokers listed more non-smokers as friends and non-smokers listed fewer smokers as friends. Question: What about this population might have caused these results to be different?

I found the Moutappa et al. article on bullying and social networks very interesting. They discussed how bullies had larger friendship groups than other students. The authors studied 6th grade students, and I remember very clearly how, at that age, the “cool kids” were also the ones who were mean to the other students. For this reason, I think that this finding makes complete sense. I think that I would be really interesting, though, to look at older teens (like in high school) and see if the findings were still the same, because I feel like middle school tends to be a very interesting time in terms of social networks.

1. I think that there are several variables that could account for the changes found in the McPherson, Smith-Lovin and Brashears study. In the interview, Smith-Lovin discussed, fairly extensively, internet as a possible cause of these changes. Essentially, she ruled it out. She explained that, while not published in this particular study, she and her colleagues asked participants about their internet usage and found there to be no significant difference between heavy users and non-users. However, what this does not account for is the way that the internet may have changed our society as a whole, not just the individuals who use it. For example, online shopping has become a very popular craze of late, and as a result, it would seem that people would spend less time shopping in stores. In the interview, Putnam discussed a potential correlation between the decrease in civic involvement and the decrease in these intimate ties. I think that this could also be extended to say that a decrease in any public activity, (like shopping) where one could potentially meet new people and create new ties, could possibly contribute in some way to this decline. Although there may be no direct causal relationship (internet use causing fewer ties, for example), I think that it is very important to realize ways in which our society has changed as a result of the internet. Another factor that I think could be extremely significant is the entrance of females into the work force. I know that this is not an entirely new phenomenon, but it has increased. Bott discussed the ways that this affects conjugal roles and how it is correlated to social network connectedness. As one of the callers addressed in the interview, the busier we get, the most difficult it is to maintain bonds. Females entering the work force no only impacts how busy they are, but also how busy their spouses are because they must begin to help out with chores at home, for example. Therefore we all have less time to discuss important matters, which could account for why we have fewer people who we discuss such matters with.

2. The social ties discussed by McPherson, Smith-Lovin and Brashears are very strong, intimate ties. More specifically, the ties that are decreasing are non-kin ties (very strong, intimate friendships). According to Wellman and Wortley’s study, respondents seemed to get most of their social support (all kinds) through their small number of strong ties. Friends, the strong ties of concern in McPherson et al.’s study, are important sources of companionship. Though authors found that friends rarely provide support other than companionship, this is a very important form of support to provide because it is not provided for us by kin. Since people are having increasingly fewer non-kin ties, but a relatively consistent number of kin ties, people will continue to have access to emotional and financial support and provision of small and large services from parents, as well as emotional support and small services from siblings, as Wellman and Wortley also explained. The McPherson et al. study also found that, while all other strong ties have decreased, more people “discuss important issues” with their spouses. This is significant because Wellman and Wortley found that most significant relationships take place within groups, rather than couples (or relations between two people). Although Bott did not prove causation, she found that there was a relationship between conjugal roles and social network structure. According to her findings, if we rely more on our spouses and our roles become more joint, our social networks will be more dispersed. However, Granovetter explained that weak ties are stronger forces in extending our social networks than strong ties. Therefore, our decreased number of strong ties should not have much impact on our ability to form new social ties. Because we are dealing with strong ties here, rather than weak, I don’t necessarily think that this will have such a vast impact on society at large. Although, in the radio interview they discuss how these ties connect our society together and how losing these ties could have a major impact on society at large. I think that Granovetter and Burt would disagree with this concern and would feel that, as long as we have our weak ties (which we still do), society will be fine. On a more individual level, since we have fewer people from whom to get support in the form of companionship, we will begin to search for this in other ties. Perhaps this will make us even closer with our kin ties.