A/S/L? Homophily of Online and Face to Face Social Ties Gustavo S. Mesch & Ilan Talmud Department of Sociology and Anthropology, University of Haifa

Theoretical Postulates  Social Relationships are embedded in social contexts.  Everyday activities (learning, work, leisure) bring people together, and attraction is based on social similarity.  Foci of activities structure opportunities and constrains of sociability.  Individual decisions of relationships formation is nested in structured foci of activities.

Social Homophily  Friendship formation, development and maintainance is based on homophily.  Homophily means that friendship occurs at a higher rate among similar individuals.  Shared status and life experience increases the likelihood of tie formation and tie stability.

Social Networks Online and Off- Line  Social capital is all ego’s ties, and the resources (virtual or real) that flow through ego’s network.  Internet is an integral part of contemporary social networks and a channel of interpersonal communication.  Evidence show that adolescents are creating and maintaining social contacts through the Internet.

Study and Research Questions Does Internet Activity decreases homophily? Does Internet Activity decreases homophily? The Study: Representative sample of Israeli adolescents (n=1000). Representative sample of Israeli adolescents (n=1000). Data on characteristics of ego-networks (up to 6 friends). Data on characteristics of ego-networks (up to 6 friends). Data on attributes of respondent. Data on attributes of respondent. Data on Internet activity. Data on Internet activity.

Methods Network data on persons defined as friends by respondent. Network data on persons defined as friends by respondent. Respondents named 6 friends. Respondents named 6 friends. Respondents asked about each alter’s age, gender, place of residence, how they met, main channel of communication, duration of friendship and topics of discussion. Respondents asked about each alter’s age, gender, place of residence, how they met, main channel of communication, duration of friendship and topics of discussion.

Analytical Strategy Two levels data: 1.Ego attributes and social ties 2.Alters’ attributes (from ego’s report) => Resulting ego-networks structure

Age Similarity Measurement  Age similarity: we subtracted ego’s average age from alters’ average age. Then a dummy variable for similarity was created : c 1 if the age is similar and 0 if age is different.

Similarity of Adolescents’ to their Friends According to Age {Low Similarity for Adolescents with Access}

Age Similarity According to Place in which the Friend was Met { Low Similarity with Friends met Online}

Findings: Age Similarity  On average, age similarity is lower for Adolescents with Internet access.  On average, age similarity is lower for adolescents that met friends online, in contrast with meeting at school and neighborhood.  Multivariate analysis: controlling for age, gender, parental education, nationality, number of siblings and self esteem, the likelihood of having friends similar in age decreases for adolescents that met friends online.

Gender Heterogeneity 1- Σ p 2 (1 indicates high IQV= heterogeneity) (k-1)/k {Highest gender heterogeneity for adolescents with access.29 to.40}

Gender heterogeneity according to place friend was met Highest heterogeneity for friends met online {.27;.32;.48)

Gender Heterogeneity of Social Network  Heterogeneity of social network is higher for adolescents with Internet Access.  Heterogeneity of social network is almost twice higher for friends met online in contrast to friends met at school.  Multivariate Analysis: Network gender heterogeneity is increased by adolescents’ age, meeting friends online controlling for gender, parental education, nationality, number of siblings, connection to the internet, self esteem.

Heterogeneity in Place of Residence IQV {Highest for Adolescents with Internet Access but differences minor(.147 to.254}

Heterogeneity in Place of Residence {Highest heterogeneity for adolescents that met friends online- but heterogeneity is still low.19, }

Heterogeneity in place of residence  Indication that heterogeneity is higher for adolescents with internet access and the ones that met friends online.  Heterogeneity in terms of place of residence is still low for all groups (less than.31).  Multivariate analysis confirms that age and internet friends are a source of heterogeneity in place of residence of friends.

Discussion  Individual ties are nested in foci of activity.  Demography, geography and technology affect individuals’ choices of friends and network heterogeneity.  Networks of adolescents that make friends online are broader in age, gender and place of residence.  Networks of online children are more heterogeneous according to age and place of residence.  Youth are being exposed to non similar ties.

Further Studies needed  How heterogeneous networks affect the strength of ties?  How network homophily is affected by age distribution of the population?  How heterogeneous networks affect the stability of social ties?  Does network heterogeneity creates diversity in terms of values, attitudes and tastes?  Are youth with heterogeneous networks in danger of abuse?