The process of evalulating and developing the theory behind and the measurement of social capital is in its early stages . This weeks readings proved challenging due to the nature of operationalizing social capital. It is interesting that network analysis is used in two vastly different processes, namely the communication audit process as described and evaluated by Zwijze and Jong; while also being used in evaluating social capital in the Van Der Gaang and Snijders piece. Rather than creating measures that specifically fit for each of these scenarios it seems as though researchers have mostly just reached for the available and assigned different constructs to the same measure. This at its core is problematic, therefore, I was glad to see some analysis of effectiveness on the part of Zwijze et al. and Van Der Gaang et al.
Zwijze and Jong attack the quite complex issue of isolating auditing techniques to evaluate their strengths, weaknesses and advantages. Network analysis is great for mapping communication structures, tracking density, frequency of interaction and measures such as these, yet network analysis alone doesn't achieve the goals of a communcation audit. Assessing the quality of communication cannot be adequately understood through analysis of patterns of communication and ties etc. Though sociometric questioning can be posed through interview, free recall, specifc time period recall etc, validity questions come to the fore partially from self report and in part due to the fact that participants may be motivated to skew their responses in order to protect their network of ties (as to not be exposed) or in order to appear more conncected than they really are. Being that communication audits are ceterainly conducted by the company in question (or hired contractors) people are very careful about the information they disclose and its implications. I particularly liked EECO analysis and the Diary approach since they more allude to the quality, style, distortion and effectiveness of communication. In fact I believe that it is essential that in addition tonetwork anaylsis and sociometric questioning that ECCO analysis or Diary studies be examined since people are notoritously bad at recalling their social networks in a given time frame.
The Martin and Hampton paper was great in that it took a closer look at current methods of conducting network analysis and showed that there is a marked difference in reliability and validty when multiple name generators are used. While the single name generators of "discuss" and "socializing" were close for some measures, they did not accurately discribe network characteristics. I liked the MGRI method for the simple fact that it maximizes construct validity. Without construct validity it isn't clear exactly what has been measured, or what the results actually mean. The only question I have is whether the exchange approach to questioning fully encompass the idea of social support? In my opinion only one aspect is covered sufficiently and that is instrumental aid Ironically the "discuss" and "socializes" question (which are questions hinting at emotional aid and companionship) were the single name generators. Perhaps more questions should be targeted at the emotional and companionship aspects of social support.
Question:
What is a good possible question for a MGRI which alludes to social support and companionship?
Should there be more questions focusing on this aspect of network analysis? Why or Why not?
By far, my favorite article this week was the piece on social capital for the simple fact that it further confused and complicated my understanding of how to measure social capital. In fact, I even found it difficult to think of better alternatives to operationalizing social capital and really showing the returns on social capital. The idea of capital revolves around returns and being able to isolate the effects in a compelling manner. Although many agree that there is such thing as power inherent to structural position, it is notoriously hard to make causal claims. The most talked about example would be job placement and the mobilization of network contacts to be place in a position, but the troubling construct of human capital makes it almost impossible to definitively point to obtaining a job as the mobilization of social capital as opposed to the effects of human capital. This is also ehtically problematic as well. How do individuals access these resources? More troubling though is how do we assign network characteristics meaning or ranking in terms of social capital? The Var Der Gaang article similarly pointed to the necessity of having multiple measures in order to assess social capital. The most interesting and problematic thing about the Van Der Gaang piece was the idea of a generalized social capital measure for nations. How would we make sense of, assign value to and analyze the social networks of a nation? Is this even possible? What in fact are the resources being mobilized and how can they be measured? These are all question that come to mind when reading about social capital.
Question:
What do you think are the best indicators to use of social capital? Can we ever isolate social capital from other tyoes of capital?
DO you think it is possible to have a general social capital measure for the US? How long would this take?