1 / 41

The Analysis of Individual Attitudes within Complex Relational Structures Footballers, teams and attitudes towards masc

Social phenomena and standard statistical testsAttitudes and informal structures within an AFL team:Research questions, measures and dataCompeting explanations: individual and structural effectsA brief diversion into exponential random graph modelsModel form, estimation, specificationAttitudes

alta
Download Presentation

The Analysis of Individual Attitudes within Complex Relational Structures Footballers, teams and attitudes towards masc

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


    1. The Analysis of Individual Attitudes within Complex Relational Structures Footballers, teams and attitudes towards masculinity Garry Robins Dean Lusher University of Melbourne Peter Kremer Ballarat University ARCRNSISS National Conference 2006 Theory, Methods and Applications of Spatially Integrated Social Science Melbourne, May 2006

    2. Social phenomena and standard statistical tests Attitudes and informal structures within an AFL team: Research questions, measures and data Competing explanations: individual and structural effects A brief diversion into exponential random graph models Model form, estimation, specification Attitudes and informal structures within an AFL team: Models, results and conclusions Final comments Modelling social phenomena: Complex data dependencies

    3. A standard statistical test (the t-test)

    4. A standard statistical test (the t-test)

    5. A standard statistical test (the t-test)

    6. A standard statistical test (the t-test)

    7. The test depends on independent observations

    8. The test depends on independent observations

    9. The test depends on independent observations

    10.

    11.

    12. Critiques of the decontextualisation of social contexts “(A social) system does not reside in the individuals taken separately, though each individual contributes to it; nor does it reside outside them; it is present in the interrelations between the activities of individuals … One could say that all the facts of the system can be expressed as the sum of the actions of individuals. This statement is misleading, however, if one fails to add that the individuals would not be capable of these particular actions unless they were responding to (or envisaging the possibility of) the system.” (Asch, 1952, p. 252) “The very assumption of statistical independence … detaches individuals from social structures and forces analysts to treat them as parts of a disconnected mass. Researchers following this tack … are forced to neglect social properties that are more than the sum of individual acts.” (Wellman, 1988, p.38).

    13. Critiques of the decontextualisation of social contexts “… the rapidly advancing discipline of sampling … separated individuals from their social context of friends, acquaintances, and so on, but also deliberately ignored an individual variable’s context of other variables in the name of achieving “more complete” knowledge of the variable space. Sampling not only tamed contextual effects to mere interactions, it also thereby produced data sets in which the levels of contextual causation were deliberately minimized. This would later enable a whole generation of sociologists to act as if interaction were a methodological nuisance rather than the way social reality happens.” (Abbott, 1997, p. 1162: emphasis in original) The preoccupation with individual cognition has left researchers “ill-equipped to do much more with the so-called cognitive revolution than apply it to organizational concerns, one brain at a time”. (Weick and Roberts, 1993, p.358).

    14. An empirical example: Masculine behaviour and social networks in team structures 1.Research questions and measures 2. Competing explanations: individual and structural effects

    16. Research Issues What sort of attitudes towards women do AFL players have? Are attitudes towards women associated with the social structure of AFL teams? What effect does the club culture have on attitudes?

    17. Individual level questions

    18. Games played Playing status Position Age Marital status Cultural background Working or studying outside of sport Individual level questions

    19. Social Network Data All of the incidents reported in the media occurred after hours

    27. The basic question:

    28. But …

    29. And structural effects …

    30. A brief diversion into exponential random graph models Model form, estimation, model specification

    31. Exponential random graph (p*)models (Frank & Strauss, 1986; Wasserman & Pattison, 1996; Snijders, Pattison, Robins & Handcock, 2005)

    32. Exponential random graph models P(X = x) = (1/c) exp{?Q ?QzQ(x,y)} normalizing quantity parameter network statistic the summation is over all configurations Q zQ(x) = ?Xij?Qxij signifies whether c = ?xexp{?Q ?QzQ(x,y)} all ties in Q are observed in x

    33. Markov Chain Monte Carlo MLE for exponential random graph models P?(X = x) = (1/?) exp{??z(x)} = exp{??z(x) - ?(?)} (1) Since (1) is an exponential family of distributions ML estimate ?*(x) is the solution of ?(?) = z(x), where ?(?) = E?{z(X)} asymptotic covariance matrix of ?* is given by (?(?))-1 = [cov(z(X))]-1 But ?(?) and ?(?) are not computable, hence need for MCMCMLE Two different approaches to MCMCMLE: Handcock (2003), following Geyer and Thompson (1992) statnet Snijders (2002) SIENA (also pnet)

    34. Model specification for network structure without attributes Markov random graph models (Frank & Strauss, 1986) New specifications (Snijders, Pattison, Robins & Handcock, 2005)

    35. Model specification for network structure with attributes Many possible effects but the simplest involves edge level effect within and between node classes

    36. An empirical example: Masculine behaviour and social networks in team structures 3. Models, results and conclusions

    37. Exponential random graph model for the AFL study: Structural effects in the model

    38. Exponential random graph model for the AFL study: Attribute effects in the model

    39. Parameter estimates: Major structural effects

    40. Parameter estimates: Major attribute effects

    41. Conclusions of the research

    42. Final comments

More Related