Social Network Analysis for Ego-Nets: Multilevel Approaches - Online
|University of Glasgow|
Dr Beth Fylan and Professor Mark Tranmer
07/06/2021 - 08/06/2021
Online run by Glasgow University
View in Google Maps (G12 8QQ)
Social network data are often in the form of ego-nets where, for example, an ego (a person being asked) nominates their contacts (alters) on the basis of some relation; for example, social support. Ego-net data usually include characteristics of ego and alter – such as demographic variables. Ego-net data also include tie value information such as the (ego-assessed) amount of support provided to by the alter to the ego. In some cases, several ties are recorded to measure the structure and function of support.
This two-day course begins with a session on ego-net data collection, based on the first-hand experience of the collection of medicines management and social support data for research within the National Health Service (NHS). The practicalities and ethics of data collection are considered here. We also explain how ego-net datasets are coded and prepared for analysis.
Next, we consider how to analyse ego-net data for the case of each ego having its own unique set of alters – with examples from real-data. This is the case where alters of one (external) ego do not overlap with the alters of the next ego: the ego-nets are distinct. We explain how such data can be analysed in a multilevel modelling framework, where the unit of analysis is the value of the tie between ego and alter. We explain how the modelling can be done using the statistical software, R. Furthermore, we explain how the results of the quantitative data analysis can be related back to qualitative ego-net data, in a sequential mixed-methods approach.
In the next session we consider the case of overlapping ego-net data, where some of the alters of one ego can overlap with the alters of the other egos; i.e. two or more different egos can both nominate the same alter. For example, alongside family and friends, perhaps distinct to each ego, alters could also include doctors or medical staff that care for several of the patients (egos) and thus overlap. In this example the egos (patients) are external to the alters (doctors, medical staff, families and friends). An alternative, overlapping, case is where each person in a network “takes a turn” at being an ego, and is also an alter in other egos’ networks. Here, the egos and alters are both internal and the tie values are recorded for each ego-alter pair. We explain how cross classified multilevel models can be used for both the external and internal cases. Moreover, we explain how other levels such as families or areas could be added to the analysis, if such information was also available for the ego-net data.
Finally, we reflect on engagement – in terms of the study design, and the dissemination of the results of these analyses – with non-academic organisations and with members of the public. We focus on the case of the NHS, based on first-hand experience.
Some knowledge of multilevel modelling would be an advantage, though we review multilevel models from first principles on Day 1. Knowledge of multiple / logistic regression would be an advantage though we do not focus on the technicalities of these approaches, beyond interpretation of parameter estimates. Some knowledge of R would be an advantage - but we will provide R scripts for all practical sessions. For the practical sessions, the participant should have a current version of R/R studio on their computer - available free from https://cran.r-project.org / https://rstudio.com No other software is required for this course, apart from R / R studio. Prior to the course we will provide more details of the R packages needed for this course, and also details of how to install them on your computer.
Session 1: Lecture: Ego net data collection and coding
Session 2: Lecture: Ego net data analysis: distinct ego-nets.
Session 3: R Practical: Ego net data analysis: distinct ego-nets (valued ties)
Session 4: Lecture: A mixed-methods approach to ego net analysis.
Session 1: Lecture: Overlapping ego nets and cross classified multilevel models
Session 2: R Practical I: Cross classified models for overlapping ego net data: egos external from alters (binary ties)
Session 3: R practical II: Cross classified models for overlapping ego net data: egos and alters internal (binary ties)
Session 4: Engagement: study design, data collection, dissemination and policy.
Crossley, N., Bellotti, E., Edwards, G., Everett, M. G., Koskinen, J., & Tranmer, M. (2015). Social network analysis for ego-nets: Social network analysis for actor-centred networks. Sage. [Especially Chapter 6 ]
de Miguel Luken, V., & Tranmer, M. (2010). Personal support networks of immigrants to Spain: A multilevel analysis. Social Networks, 32(4), 253-262.
Fylan, B., Tranmer, M., Armitage, G., & Blenkinsopp, A. (2019). Cardiology patients' medicines management networks after hospital discharge: A mixed methods analysis of a complex adaptive system. Research in Social and Administrative Pharmacy, 15(5), 505-513.
Vacca, R., Stacciarini, J. M. R., & Tranmer, M. (2019). Cross-classified Multilevel Models for Personal Networks: Detecting and Accounting for Overlapping Actors. Sociological Methods & Research, 0049124119882450.
Bates, D., Sarkar, D., Bates, M. D., & Matrix, L. (2007). The lme4 package. R package version, 2(1), 74.
Hadfield, J. D. (2010). MCMC methods for multi-response generalized linear mixed models: the MCMCglmm R package. Journal of statistical software, 33(2), 1-22.
Intermediate (some prior knowledge)
The fee per teaching day is: • £30 per day for UK/EU registered students • £60 per day for staff at UK/EU academic institutions, UK/EU Research Councils researchers, UK/EU public sector staff, staff at UK/EU registered charity organisations and recognised UK/EU research institutions. • £100 per day for all other participants In the event of cancellation by the delegate a full refund of the course fee is available up to two weeks prior to the course. No refunds are available after this date. If it is no longer possible to run a course due to circumstances beyond its control, NCRM reserves the right to cancel the course at its sole discretion at any time prior to the event. In this event every effort will be made to reschedule the course. If this is not possible or the new date is inconvenient a full refund of the course fee will be given. NCRM shall not be liable for any costs, losses or expenses that may be incurred as a result of the cancellation of a course. The University of Southampton’s Online Store T&Cs also continue to apply.
Website and registration
Mixed Methods Approaches (other), R, Ego-net data collection
Related publications and presentations