Miranda Lubbers Jos Luis Molina The dynamics of

Miranda Lubbers & José Luis Molina The dynamics of personal networks With three substantive applications

The Spanish team � Miranda Lubbers (PI), Autonomous University Barcelona � Isidro Maya Jariego, University of Seville � José Luis Molina, Autonomous University Barcelona � Ainhoa Federico de la Rúa, University of Toulouse II – Le Mirail (and associated with the University of Málaga)

Our project � We focus on the dynamics of personal networks �Sociocentric networks are valuable for understanding the recurrent influence between a network and a behaviour of its members for which the organizational boundary is relevant (e. g. , pupils’ peer relationships at school and their academic performance). � Our subproject is, however, interested in predicting individual outcomes that are not primarily produced within a single context, but for which all social circles in which a person moves need to be considered (e. g. , social well -being, health, or adaptation of immigrants).

Personal networks � Personal networks represent an individual's social context, the intermediate level between the individual and society. � It is at this level that processes of socialization and social integration in society take place.

Past studies have limited our understanding of this social context: � (1) They tended to capture only a small part of the network (e. g. , the 5 -10 most intimate and supportive ties). � Weak ties have their own virtues, as they are more numerous, more heterogeneous, and less connected among themselves.

Past studies have limited our understanding of this social context: � (2) They often measured network composition but not structure � The structure of relationships is assumed to affect the flow of information through a network (mobilization of social support, social control, conformity of opinions), which can have consequences for individual outcomes ego

Past studies have limited our understanding of this social context: � (3) Many measured personal networks crosssectionally. �A longitudinal measurement helps us understand the processes of socialization or social integration, and accelerating or disruptive effects of life events on such processes

Thus… � If we want to understand the processes of socialization and integration in society, we need: � larger networks, � network structure, � longitudinal measurement � This type of data is expensive to collect, but the number of studies with a more elaborate measurement of personal networks is growing

But how to analyze such data? � Typically, multilevel analysis is used to predict persistence of ties over time or changing contents of ties… Ego Alter �… but it assumes that relationships among alters are not important for the relationships they have with ego

Design for this ECRP � To use SIENA to model the “co-evolution” between… � Networks: The structure of relationships among alters (dichotomous) � “Behavior”: The strength (/persistence) of relationships between ego and alters (valued / dichotomous) � Outcomes Tie strength ego Network of alters at a higher (ego) level.

This would enable us to investigate… �… for example, � whether changes in network structure affect changes in the supportiveness of ego-alter ties � whether alters with whom ego feels closer over time become more embedded in their networks � whether relationships with structurally embedded alters are easier to maintain

But we would need some changes in SIENA �A Bayesian model that works with ML and symmetric data � (…)

Test this on three projects � Two empirical projects proposed for ECRP � José Luis � Miranda � Isidro � Ainhoa � One Personal networks of immigrants, Catalonia Personal networks of high school graduates, Alcalá / Sevilla empirical project attached: � Collaboration Claire Bidart, Ainhoa & Miranda on the analysis of changing networks of youths in the Caen study (4 waves with 3 -year intervals)

Illustration: Chris Mc. Carty et al. (2007) The study

Research questions � At the individual level: � How do the personal networks of immigrants change over time (post-migratory)? � How are these dynamics associated with the process of immigrant adaptation? � At the relationship level: � What drives these changes? � RSIENA: e. g. , Which characteristics of respondents, alters, and networks explain whether ties with Spanish alters become stronger over time?

Sample � In 2004 -’ 06 (t 1): Interviews with 504 immigrants in Catalonia (N = 301) and New York (N = 203) � Funded by NSF; PI: Chris Mc. Carty, University of Florida � In 2007 -’ 08 (t 2): Second interview with a selection of 77 of the immigrants from the first project who still lived in Catalonia � Funded by ESF (previous ECRP); PI Spanish project: José Luis Molina � In 2012 (t 3): Third interview with as many of the participants of the second wave as we can trace, complemented with persons from the first wave only � Funded by the Ramón y Cajal grant of MICINN of Miranda Lubbers

Four groups of origin (in the Spanish project) � Argentineans � � Dominicans � � Labour migration Secondary labour market (construction, …) More or less half of them report that relatives in origin depend on them economically Moroccans � � Diverse motivations for migration; typically individual migration Very diverse occupations (reporter, psychologist, musician, architect, waiter…) Easily accepted by the native Spanish, relatively few experiences with racism Most numerous group with longest residence in Catalonia Chain migration, mostly labour migration (families) and family reunification The majority works in the secondary labour market (cleaning, clothing industry, farm labour…) Senegalese / Gambians � � Most recent migration of the four groups, predominantly male; 42% already had histories of migration Chain migration, mostly temporary labour migration (individual) The majority works in the secondary labour market (agriculture, construction, …) Relatives in origin depend on them economically

Procedure of data collection � Mixed-methods design: � 1. Quantitative part: Survey with personal network instrument Computer-assisted personal interviews with the software Egonet http: //sourceforge. net/projects/egonet/ � 2. Qualitative part: Semi-structured interviews Audio-recorded

Illustration: Jürgen Lerner et al. (2008) Two examples of changing networks

A young Moroccan woman t 1, last year in high school Red: Spaniards Green: Moroccans in Morocco Blue: Moroccans in Spain Gray: The others SIZE: Closeness (1 -5, the larger the closer) High School

A young Moroccan woman t 2, trying to find a job His family The rest of her family Red: Spaniards Husband brothers/ sisters Green: Moroccans in Morocco Blue: Moroccans in Spain Gray: The others SIZE: Closeness (1 -5) Friends and neighbors

An example of “ethnic segregation” “ In the past, I felt I belonged here. At school they always said “You are from here, you are Catalan”. But now… many people [Moroccans] who don´t know anything about Morocco and who only have a Moroccan name, they can´t find a job. “I will call you”. They make you feel Moroccan” “There are people who do not let you feel catalan. They have never said anything to me directly but you know they talk bad about Moroccans. And about the culture. And I am Moroccan. It doesn´t feel good, you know? “I don´t feel I belong here nor there”

Argentinean woman t 1 Husban d Inlaws Catala n class Best friend s Red: Spaniards Green: Argentineans in Argentina Blue: Argentineans in Spain Gray: The others SIZE: Closeness (15) With margin: Stable People known via the inlaws Family, childhoo d friends and people known via them

Argentinean woman t 2 Husband friends People known via the inlaws Catala n class Inlaws Red: Spaniards Neighbours in Argentina Green: Argentineans in Argentina Family Blue: Argentineans in Spain Gray: The others SIZE: Closeness (1 -5) With margin: Stable Childho od friends People known via family

An example of “regression” “I felt Argentinean and it changes bit by bit… now that I know a bit about the [Catalan] language, I feel a bit more that I belong here, but not completely so. ” “It is difficult to go [to Argentina]. Every time I see how my mother is a bit older and it is more difficult to leave her again”. “I like the people there. If I had to go back, I would go easily. I would miss the commodities, the organization is better here, everything is cleaner, but at the sentimental level…. I miss the feeling, the tact, the contact, and here I have nothing”

Illustration: Miranda Lubbers (2010) Some conclusions from the first two waves

Some conclusions: How do the networks change over time? Although there was quite some turnover in the networks, we do not observe a tendency toward “interactive integration” in the host society… no increases in: � the number of Spaniards � the strength of ties with Spaniards � the centrality of Spaniards � Also, the average strength of ties didn´t increase. � But networks became more local over time (more co-ethnics in Spain plus higher centrality, less co-ethnics in origin) � And the proportion of Spaniards was almost twice as high among the ties that were most recently formed (than among all ties), and these ties do not differ in emotional closeness (nor in stability) from the youngest ties with coethnics in Spain. It´s a slow process… �

Some conclusions: How do the networks change over time? � Immigrants maintained many active relationships with people in the country of origin – and not only did they maintain ties, they also reactivated sleeping ties, formed new ties in origin… transnationalism � The overall trends in fact were not representative of any network: there was a large variety among individuals of trajectories of network change of which integration is only one.

Are the network dynamics associated with immigrant adaptation? � We found that changes in network composition, even when they were temporal fluctuations (e. g. , due to recent travels to origin), covary with fluctuations in ethnic identifications - we cannot conclude this (yet) for sociocultural adaptation, because it differs too strongly between the four immigrant groups)

What drives these changes? � New relationships were acquired in defined contexts (work places, neighbourhoods, a language course which create meeting opportunities with groups depending on the context) and via others (transitivity) � … We have the data prepared for SIENA …

Ilustración: José Luis Molina The third wave In preparation

Questions for the third wave. . . � To compare the dynamics at a short time interval with dynamics at a larger interval � We also want to collect new information (and omit some old variables), among others… � Include two new measures of psychological adaptation � More data on what flows through the networks that makes that networks facilitate or hinder adaptation (social support, social control)

Thank you mirandajessica. lubbers@uab. es joseluis. molina@uab. es