Blog 10: Adolescent Delinquency

Do high school and middle school friendships actually matter? Are the people in this photo below really able to make a lasting impact on one another?

The following two articles offer a resounding yes to this question.  They use SNA methodologies to investigate crime and deviance in adolescent friendship networks. Haynie (2001) introduces the idea that the structure of friendship networks may be associated with adolescent delinquency. Then, Haynie & Payne (2006) build on this work, using SNA to consider how friendship networks explain levels of delinquency across different racial ethnic groups.

Haynie (2001):

What is the research question? 

  • In this study, Haynie is seeking to explore if structural properties of friendship networks have any impact on the known associations between delinquency and peer delinquency for adolescent students.

How is the data collected? 

  • Adolescents were asked during an interview to identify their 5 closest male and female friends from a roster of students.

What is the sample population? 

  • Taken from the National Longitudinal Study of Adolescent Health, the sample population in this study was all adolescent students attending an identified 139 high schools, and then the adolescents who attend the feeder middle schools associated with these high schools.

What are the nodes? 

  • Adolescents during an in school interview

What are the links? 

  • Friendship between adolescents

What are the results?

  • As evidenced in Figure 1 below, Haynie found that in high-density networks, adolescent delinquency increased as their friends’ delinquency did. In other words, adolescents’ engagement in delinquent behaviors seems to be more impacted by their peers in dense networks. Additionally, they found similar results amongst network centrality and adolescent popularity. These results are particularly interesting from an educational perspective. They speak to the need for educators to be aware of students’ friends and the friendship networks that exist in their buildings.

How does SNA as a methodology help advance our understanding of these types of relationships?

  • In this instance, SNA serves a vital role in understanding the impact that peers have on students’ engagement in adolescent behavior. Though this study could have been conducted through correlational analysis, it would loose the second major finding about how network structure moderates the relationship between peers and engagement in delinquency.

Figure 1

Haynie & Payne (2006):

What is the research question? 

  • Though no explicitly stated research questions, this article seeks to investigate the role that friendship structure plays in explaining the different levels of adolescent delinquency between White, Black, and Latinx students.

How is the data collected? 

  • The same data collection methods were taken in this study — Adolescents were asked during an interview to identify their 5 closest male and female friends from a roster of students.

What is the sample population? 

  • Same as the prior study, the sample population included adolescents from middle and high schools who took part in the Add Health data collection.

What are the nodes? 

  • Adolescent students

What are the links? 

  • Friendship between adolescents

What are the results?

  • Their results provide evidence for the hypothesis that racial and ethnic differences in adolescent violent activity can be explained by friendship network structure.

How does SNA as a methodology help advance our understanding of these types of relationships?

  • If the authors were to investigate this without an SNA approach, they would have missed this major finding. In this case, SNA provides some clear rational to a phenomenon that was already made apparent through descriptive and statistical analysis alone.

The larger story within these two articles is that friendship structure matters to adolescent delinquency, and we, as educators should be cognizant of such.

Blog 9: SNA, friendships, and community

Article 1:

In this article, Kleit (2001) uses social network analysis to investigate the role that social networks play in job searches, specifically for individuals who live in scattered-site public housing communities.  Based on the theory (see Figure 1) that housing lower-income individuals with more affluent peers open them to more opportunities, Kleit hypothesized that living in more dispersed housing would provide lower-income individuals with access to more jobs.  Thus, to explore this, she sought to answer the following research question: Does living in small clusters of public housing in a non-poor area instead of in a dispersed housing pattern influence the types of social ties poor people use when they look for jobs? Using in-person survey data from 253 women living in either dispersed or clustered housing, Kleit asked participants to report the people who they had talked to or would talk to about jobs, as well as various other questions to account for the persons role in their life.  Kleit then identified which of the individuals listed were neighbors.  Thus, the nodes in this project were the individuals who lived int he neighborhood, and the links represent whether or not two individuals would communicate about jobs. 

She found that although there is greater diversity in dispersed communities, these communities are not as close as clustered communities, which she posits may influence their willingness to reach out for job connections as women in dispersed communities were less likely to reach out to neighbors than those in clustered communities.  In this sense, SNA allows us to view the structural impact of housing policies as they apply to entire communities as opposed to solely individuals. 

Article 2:

In their article, Valente and colleagues (2009) use social network analysis to investigate adolescent friendship choices and weight.  As highlighted in the meme below, and evidenced through decades of scholarship, peer relationships are vastly influential in adolescence.

Based on past research which suggests that weight gain seems to spread through family and friend networks, the researchers hypothesized that obesity may spread similarly through adolescent friendships.  Thus, the goal of the study was to investigate if overweight adolescents were more or less likely to have overweight friends.  Using in-person surveys with 685 students between ages 11 and 15, the researchers asked students to report about their friends in class as well as took measures of their weight.  Thus, the nodes in this project were the 685 students, the links represent those who were nominated as friends, and the students BMI were used as node attributes. 

They found that overweight girls were less likely to be nominated as a friend, even though overweight students nominated more friends. Additionally, their network analyses suggested that friendships were more likely between students of similar rather than different weight.  In this way, SNA allowed the researchers to investigate the various relationships of children through nodes and links as opposed to just singular relationships.  This provides for a more accurate depiction of the adolescent friendship network structure as it relates to weight.

Blog 8: Density and Friendship Development

Aggression is not a common trait that we seek in a friendship.  However, research suggests that aggressive children are still known to have friends.  In their article “Forms and Functions of Aggression in Adolescent Friendship Selection and Influence: A Longitudinal Social Network Analysis,” Sijtsema and colleagues (2010) investigate the impacts of aggression on adolescent friendship.  Through this work they seek to explore:

1. The development of friendship networks in adolescence

2. How friendship selection may be predicted by aggression?

To explore these questions, they collected survey data from 6th – 9th grade students over the course of 3 years.  Students were asked to nominate up to 18 friends in their grade at school as well as provide information on the length of their friendship.  During this data collection, they also collected a self-report survey about the students aggression.  The sample population for this study is a longitudinal sample of adolescents in middle school in a medium-sized urban community in the North East.  The sample was comprised of 65% White students, 17% Black students, 6% Latinx students  and 13% students of other race/ ethnicities.  About half of the sample identified as females and all participants were between the ages of 12 and 14. 

In this analysis, the students were used as the nodes, and the friendships between them were the links, whether one-sided or reciprocated.  The density metric was investigated to understand  the first research question about how friendship networks develop over middle school.  Density was defined as depicted in Figure 1, and as much could increase or decrease over time.  

Figure 1. Defining Density

By and large, Sijtsema and colleagues  found that the density of the network decreased over time, suggesting that students made fewer nominations of friends as they progressed through middle school.  Figure 2 demonstrates how this phenomenon occurred across 7 of the 8 different networks.  From this finding, they concluded that “adolescents were less inclined to just nominate classmates as friends. Instead they favored friendships that were mutual….” This contributed to the decrease in density of the network, but didnt necessarily suggest that adolescents had lost friendships between time point 1 and time point 3.

Figure 2. Density over time

In this way, the density metric provided us with an understanding of how the structure of the network as a whole changed over time.  This goes above and beyond what we could have learned from a nodal metric that only allows us to investigate how individual nodes function in a network.   Instead, structural metrics like this one inform us about the functioning of the friendship network as opposed to a few adolescents in the network.  

Sijtsema, J. J., Ojanen, T., Veenstra, R., Lindenberg, S., Hawley, P. H., & Little, T. D. (2010). Forms and functions of aggression in adolescent friendship selection and influence: A longitudinal social network analysis. Social Development19(3), 515-534.

Blog 7: The Changing Public Sphere

Habermas and Castelles are two scholars who think critically about how society and social networks overlap in our modern world. Habermas writes specifically about the public sphere, which he defines as the “nexus between public life and civil society.” Depicted in the figure below, the public sphere is a space where all people in society are provided the opportunity to engage in conversations and debates that are in the public interest, without being influenced by the state. However, these debates are ultimately an important and integral facet of democracy, and therefore influence the state.

Though the ideal public sphere is open to all regardless of social class, wealth, power, etc.; Habermas claims that access to the sphere varies, as does an individual’s degree of autonomy within it. Further, he warns of the deterioration of the public sphere as organizations are now taking the role of individuals, and media is turning information into a product to be sold. As a result, the public sphere is slowly becoming more closed and restricted.

Castells, takes a more optimistic outlook on the public sphere, crediting the rise of network society for the revitalization of the public sphere. From his point of view, a network society, is one in which communication is multi-dimensional and multi-directional, not limited by time or space. This is largely due to the role of technology, which decentralizes the power in society and gives access to a broader population. Additionally, it removes the requirement for people to be in a certain place at a certain time to hold power and have input. Due to these advancements in communication, he claims that more individuals have access to the public sphere, and actively engage in it. Examples of this active engagement are the online movements such as #lovewins, a movement which ultimately influenced policy during the summer of 2015. At a more local level, we see movements such as the Richmond Teachers for Social Justice who are similarly using online platforms to advocate for educational equity (see here).

Thus, largely as a result of enhancements in technology, the network society is known for new forms of time and space (timeless time and space of flows) that allow for democracy in action like the aforementioned examples. Given this idea, Castells claims that network society leads to a more connected, productive, open-minded and accepting society. However, his critics claim that his ideas are utopian. In my opinion, though utopian, I agree with Castells that these changes will improve the lives of various people. Though inequity in access still exists, and some voices will continue to hold more weight than others (celebrities, politicians, social media influences, etc.); I believe that any movement towards giving people voice has the potential to produce good.

Blog 6: Social Networks and Student Engagement

By and large we know that students struggle to stay engaged in the classroom and classrooms at any level can end up looking similar to Figure 1.

Whether from our own experience as a student, teacher, or parent; this is an often-experienced phenomenon. In fact, research has confirmed throughout time that many students are not engaged in school (Steinberg, Brown, & Dornbush, 1996; Yazzie-Mintz, 2007). Jennifer Fredericks, a leading scholar in the field of engagement warned of the consequences of disengagement stating that they “are especially severe. These youth are less likely to graduate from high school and face limited employment prospects, increasing their risk for poverty, poorer health, and involvement in the criminal justice system…” (Fredericks, 2011). For this reason, I find it extremely important to explore student engagement in a variety of contexts to not only understand the consequences of disengagement, but also to understand what is associated with higher rates of engagement.

For this reason, I plan to investigate the relationship between how connected a student is in the classroom social network, and their individual reports of engagement. Kindermann and colleagues suggest that students make friends with others who have similar levels of engagement (1993; 1996), and Ryan (2000) reports that peers can support each other’s engagement through information sharing and modeling. Given this evidence, I hypothesize that highly connected students in the social network will report higher levels of engagement. To investigate this idea, I plan to answer the following three research questions:

  1. Is a student’s level of connection to a classroom social network related to their levels of engagement in the course
    • Are highly connected students more or less engaged?
  2. Does a student’s level of connection to a classroom social network differentially impact the 4 different dimensions of engagement?
  3. Are there differences in level of connection by gender?

To investigate these questions, I plan to collect a network data (specifically a directed graph) using an online Google survey in an undergraduate human development course that I teach. I will ask students in my section of the course about their engagement as well as who they frequently work with. Demographic and network survey items will include:

  • Please provide your gender: _______________________
  • Year in school:
    • Freshman
    • Sophomore
    • Junior
    • Senior
    • Graduate student
  • List the people in this class that you have collaborated with. (Collaboration)
  • List the people in this class that you would study with or go to for help. (Dependency)
  • List the people in this class that you frequently talk with. (Neutral)

Additionally, since engagement is often conceptualized as a multi-dimensional construct including cognitive, affective, behavioral, and social aspects (see Figure 2), I will use the following 4 items to investigate students’ engagement.

  • How true are the following statements of you?
    • I try to connect what I am learning in this class to things I have learned before.
    • I put effort into learning in this course.
    • I look forward to this class.
    • I try to understand other people’s ideas in this class.

Given the lack of research I’ve found investigating the role of social networks in classroom engagement, I think that this will be a contribution to the field in understanding what it is that encourages student engagement.

Blog 5: Centrality Measures in Adolescent Research

Node centrality measures serve as an additional way, beyond the visual of a graph, to understand social networks. These measures provide more detailed evidence about how the network itself, and the players within it (the nodes) function. For example, the eigenvector centrality metric gives us an idea of which nodes are most influential in the network, serving as an indicator for spotting important nodes. Additionally, the betweenness centrality metric is basically the idea of how regularly a given node serves as a bridge or intermediary between other nodes in the network. In this way, centrality measures provide information not just on the number of edges connected to each node (as can be seen in the visual graph like the one below), but also explains the role that each of the nodes play in the network, and how their role impact larger network functioning.

Ennett and colleagues (2006) provide an example of how centrality measures such as betweenness centrality and a few other centrality measures that we haven’t specifically discussed in class. In their study of substance use in adolescent peer groups, they describe betweenness centrality as “the extent to which an adolescent in-directly links pairs of adolescents who are not directly linked as friends” (p. 168). They found that 45% of the adolescents in their sample served as bridges between neighborhood peers, and that betweenness centrality was significantly and positively correlated with the number of neighborhood substance users. However, other measures of centrality were more likely to be significantly associated with personal substance use, suggesting that the adolescents who are most visible in the network are more likely to use substances than the ones who serve as intermediaries. In this example, we see how the use of betweenness and other centrality measures allow for the understanding of how nodes in the network function, and the relationship between network positionality and an outcome of interest.

Similarly, Russell and Smith (2011) show the importance of using centrality measures such as degree centrality in their investigation of the afterschool network program in Dallas, Texas. For example, through exploring the degree centrality of certain programs, there were able to identify which programs were vulnerable due to the fact that they had just one or two sources of funding, and tended to work in isolation. The image below shows the differences across the network in relation to their funding structures. Programs in the clusters on the left generally receive funding from one or two organizations; whereas programs in the cluster on the right have more diverse funding mechanisms and are more integrated into the community of afterschool program providers.

Results from these two studies demonstrate the different ways in which measures of centrality can be used to understand not only network shape but also how nodes within a network differ in relation to each other, how edges in a network develop into clusters, and how these clusters and measures better allow us to understand our research questions.

 

References:

  • Ennett, S. T., Bauman, K. E., Hussong, A., Faris, R., Foshee, V. A., Cai, L., & DuRant, R. H. (2006). The peer context of adolescent substance use: Findings from social network analysis. Journal of research on adolescence16(2), 159-186.
  • Russell, M. G., & Smith, M. A. (2011). Networks Analysis of a Regional Ecosystem of Afterschool Programs. Afterschool Matters13, 1-11.

 

 

Social Capital, Bowling Alone, and #MeToo: Blog 4

Are we really bowling alone? Are we really pulling away from communal living in light of technological increases over the past few decades? In many ways, I think a superficial answer to this question is yes… Putman (1995) discusses the overall decreases we have seen in Americans participation in communal activities such as volunteering, religion, and bowling. Given his claim that participation in communal activities leads to the development of social trust and reciprocity, it makes sense that Americans disengagement with such activities would lead to a devaluing of social capital. However, social movements such as #MeToo and #BlackLivesMatter, which largely grew from technological platforms lend some evidence to the idea that Americans are still actively engaging in communal activities even though these activities are no longer bowling leagues.

The #MeToo movement serves as a prime example of how social capital interacts in networks. Lin (1999) discusses how social capital facilitates the flow of information, influences decision makers, and demonstrates ones’ social credentials. In the communal act of the #MeToo movement we see how engagement in this simple activity flowed information rather quickly through a gigantic online social network. The image below shows the spread of this movement in just 12 hours.

Additionally, this movement was the topic of discussion in policy circles and served to identify social credentials of the silence breakers (pictured on the cover of time magazine).   Thus, this movement serves as evidence that the emergence of the Internet has not devalued social capital, but rather altered the way we engage with social capital and in social networks,

For one, Lin (1999) sates that the Internet has broadened social networks, providing more access to free information and social capital than before. In this way, social capital isn’t as tightly controlled by the upper class and decision makers. In many ways, this allows for the emergence of new voices in the policy arena. However, it is important to note that social capital is not always a net positive for actors. Portes (2002) exemplifies this idea with the discussion of how social capital excludes outsiders and demands conformity. Using gangs as a prime negative example of social capital, he discusses how conformity can promote violence, ostracize new actors from engaging, and even place individuality at too high a cost to be achieved within a community. This results in stifled innovation, progress, and productivity.

In thinking about the value of social capital in research, I am drawn to the idea of community social capital. For example, given the newer emergence of an equity lens in educational research, it is important to consider which individuals hold more social capital and how that relates to educational experiences and outcomes. For example, are students with more social capital afforded better educational experiences and outcomes than their peers? Can SNA identify instances where this is happening and develop intervention to abate it? For my personal research, I am interested in considering who holds social capital both within afterschool program networks, and within the programs themselves. What resources do each of the actors in the network bring and how does this relate to the broader community social capital of the neighborhood within which the program is situated?

Leaving you with a question:
Kaushadin (2011) brings up the idea of social solidarity as a facet of social capital. This is basically a shared trust and reciprocity within a community. My most poignant example of this was experienced when I was a student at UVA. Given our strict honor code and a communal respect and adherence to it, it was very common for students to leave their valuables unattended for hours at end, trusting our community to abide by the code. Can you think of an example of such explicit social solidarity that you have experienced within a community?

It’s a Small World After All…

“What is the probability that any two people, selected arbitrarily from a large population, such as that of the United States, will know each other?”

This is the question that Travers and Milgram (1969) sought to investigate, and in essence defines the small world theory. The Small World Theory basically is the idea that our unexpected social links connect us to far more individuals than we realize, regardless of location. Kadushin (2011) demonstrates this idea in the first chapter of Understanding Social Networks by discussing how social media sites such as Facebook connect us to networks that make the world seem small. For example, through my 500 friends on Facebook, I am connected to various other individuals by either one or two degrees of separation. In other words, I am connected to my friends, their friends, and further their friends. The expansive nature of this system can be visualized in the image below.

However, these connections don’t solely exist online. This week’s lectures also demonstrated the small world theory, as it exists practically in our policy arena. The lecture mentioned Obama’s decision to change telephone surveillance of terrorists from 3 degrees to only 2 degrees of separation, demonstrating how wide our social networks go and the actual number of individuals who are still under surveillance given the change.

However, this example also highlights one of the very real risks of the small world theory. Kadushin (2011) warns that these weak ties have the danger of exposing us to more people than we desire to be exposed to. This seems to occur for a few reasons:

  1. Kadushin (2011) explains that weak ties are what connect networks that otherwise would be disconnected, connecting us outside of our insular networks.
  2. Granovetter (1973) claims that “the stronger the tie connecting two individuals, the more similar they are” (p.1362), which seems to suggest that weak ties connect us to those whom we are dissimilar from.

For these reasons, weak ties, such as those emphasized in the small world theory connect us to those who we are different from, and otherwise would likely not be connected to. This provides the opportunity for us to 1) be connected to more people than we desire and 2) be connected to individuals we may wish we weren’t connected to.

So, what is the impact of weak ties on the big world?

As previously mentioned, weak ties connect us people we otherwise would not be connected to. Further, Granovetter (1973) explains how weak ties allow us to explore connections between social networks, stating:

“Emphasis on weak ties lends itself to discussion of relations between groups and to analysis of segments of social structure not easily defined in terms of primary groups” (p. 1360).

Thus, the presence of weak ties allows us to consider the social networks of the world as cohesive as opposed to disjointed systems. We therefore can investigate the connection between various networks instead of solely focusing our research within insular networks. Due to this, we can begin to collect data and ask questions about how networks are connected, how we maintain such connections, and the importance of these bridges.

From this perspective, weak ties seem to be just as important as strong ties. In my opinion, they may even rival strong ties. Given that they are the source of newer ideas being introduced into social groups from dissimilar groups, they serve an important role of ensuring that our social networks do not become too isolated. This perhaps is most evident when considering the impact that weak ties likely played in the last presidential election.

So, who is Kevin Bacon, and why is he important in this discussion?

Kevin Bacon, the actor in the photo above, became the face of the small world theory when a few college students designed a game called “The Six Degrees of Kevin Bacon.” The goal of this game is to connect any actor to Kevin Bacon through no more than 6 other actors, thus demonstrating that even the world of Hollywood is subject to the small world theory.

Understanding our Modern Society

Given what we learned last week about the power of weak and strong ties on individuals, it is clear that social networks have the power to influence individuals’ behaviors, emotions, outcomes, etc. In this way, social networks provide us with the opportunity to explore human tendencies, and further understand how human beings exist in and engage with modern society. The social network perspective, according to Keim (2011) is “explains social actions not on the basis of individual attributes but in the context of social relations (Wellman, 1988: 31).” (p. 21). That being said, social network analysis steps beyond understanding modern society as a product of individuals’ traits or circumstances, and considers how the interactions between individuals or the interactions between systems influence such actions.

If however, we are considering interactions between individuals, it is important to consider the perspective from which we are viewing such networks. For example, in egocentric studies only the “network immediately surrounding participants is observed” (Robins, 2015, p.40). This means that our perception of a network is limited to what is immediately surrounding participants, and lacks information about how the network expands beyond these people or how it might be experienced by others. Thus, if we were to shift to the perspectives of a different sample of participants, the network likely would look different. Given this, it is important to consider the bias in SNA just as we would in any other traditional social science design.

So then, how does SNA data differ from traditional social science data?   SNA data differs from traditional social science data in that it necessitates two data sets, namely the nodes and the relations between them. For example, on our practicum, we had one data sheet with the characteristics / attribute of the nodes– similar to what we would see with traditional social science data (sheet A). However, this data sheet was paired with a data set that recorded the connections or relations between the nodes (sheet B). Thus, one of the main differences in SNA and traditional social science is the format of the data itself. Further, the main unit of analysis in SNA is the node or the tie as opposed to an attribute or outcome in traditional data analysis.  In this way, relational data must be treated differently than our average data set.

Though inherently different, relational data still necessitates descriptive analysis so that similarities and differences between nodes and ties can be considered. Through these descriptions researchers are able to consider the similarities and differences in groups within the network, further explaining how humans engage similarly and differently with modern society. However, diverging from traditional data analyses, one key reason that relational data makes predictive analysis challenging is because of causal ambiguity. Yang, Keller, and Zheng (2017) define causal ambiguity as the fact that we don’t know which data point has caused change in the other. Instead, we know that the two are related non some factor.

Overall, there seem to be some similarities between SNA and traditional social science research. However, SNA provides us with a new language and new techniques to further explore human beings’ interactions, emotions, and daily actions in modern society.

A thought-provoking quote:

  • “From a network perspective, one would argue that peers are not influential per se, but can become important if they are engaged in frequent contact, build cohesive networks of high density, transmit information quickly, and produce homogenous evaluations and normative pressures” (Keim, 2011, p. 21).