SIENA news

RSiena: Siena in R

Appendix B of the RSiena manual contains a list of changes compared to earlier versions.

The manual for RSiena has been extensively updated on January 17, 2012. This hopefully now has removed the traces of Siena 3 which still used to be visible here and there. The links with the functions used for executing RSiena also have been given more consideration.

The Siena scripts page has been frequently updated with new scripts; large updates early September 2011, minor updates mid January 2012.

Important new options in RSiena since April 2010 are the sienaGOF and sienaTimeTest functions contributed by Josh Lospinoso, respectively for goodness of fit checking and for testing time homogeneity for data with 3 or more waves. The sienaGOF function currently is available only in RSienaTest. See the manual for further explanation of these functions.

Some new publications

Kevin Lewis, Marco Gonzalez, and Jason Kaufman (2011). Social selection and peer influence in an online social network. PNAS 2011; published ahead of print December 19, 2011.
Abstract
Disentangling the effects of selection and influence is one of social science's greatest unsolved puzzles: Do people befriend others who are similar to them, or do they become more similar to their friends over time? Recent advances in stochastic actor-based modeling, combined with self-reported data on a popular online social network site, allow us to address this question with a greater degree of precision than has heretofore been possible. Using data on the Facebook activity of a cohort of college students over 4 years, we find that students who share certain tastes in music and in movies, but not in books, are significantly likely to befriend one another. Meanwhile, we find little evidence for the diffusion of tastes among Facebook friends - except for tastes in classical/jazz music. These findings shed light on the mechanisms responsible for observed network homogeneity; provide a statistically rigorous assessment of the coevolution of cultural tastes and social relationships; and suggest important qualifications to our understanding of both homophily and contagion as generic social processes.
DOI: http://dx.doi.org/10.1073/pnas.1109739109.
The paper with the basic elaborate exposition about the "Siena approach" to studying dynamics of selection and influence:
Steglich, C.E.G., Snijders, T.A.B. and Pearson, M., Dynamic Networks and Behavior: Separating Selection from Influence.
Sociological Methodology (2010), 329-393.
DOI: http://dx.doi.org/10.1111/j.1467-9531.2010.01225.x .
Issue 32.1 of Social Networks is devoted to Network Dynamics and carries a tutorial and several applications of Siena.
Dijkstra, J. K., Lindenberg, S., Veenstra, R., Steglich, C., Isaacs, J., Card, N. A. & Hodges, E. V. E. (2010). Influence and selection processes in weapon carrying during adolescence: The roles of status, aggression, and vulnerability. Criminology, 48, 187-220.
Abstract
The role of peers in weapon carrying (guns, knives, and other weapons) inside and outside the school was examined in this study. Data stem from a longitudinal study of a high-risk sample of male students (7th to 10th grade; N = 167) from predominantly Hispanic low-socioeconomic- status schools in the United States. Longitudinal social-network models were used to test whether similarity in weapon carrying among friends results from peer influence or selection. From a goalframing approach, we argue that weapon carrying might function as a status symbol in friendship networks and, consequently, be subject to peer influence. The findings indicate that weapon carrying is indeed a result of peer influence. The role of status effects was supported by findings that weapon carrying increased the number of friendship nominations received by peers and reduced the number of given nominations. In addition, peer-reported aggressiveness predicted weapon carrying 1 year later. These findings suggest that adolescent weapon carrying emerges from a complex interplay between the attraction of weapon carriers for affiliation, peer influence in friendship networks, and individual aggression.
DOI: http://dx.doi.org/10.1111/j.1745-9125.2010.00183.x.
Maarten Van Zalk / Selfhout, William Burk, Susan Branje, Jaap Denissen, Marcel van Aken, and Wim Meeus (2010). Emerging late adolescent friendship networks and big five personality traits: A social network approach. Journal of Personality, 78, 509-538.
Abstract
The current study focuses on the emergence of friendship networks among just-acquainted individuals, investigating the effects of Big Five personality traits on friendship selection processes. Sociometric nominations and self-ratings on personality traits were gathered from 205 late adolescents (mean age = 19 years) at 5 time points during the first year of university. SIENA, a novel multilevel statistical procedure for social network analysis, was used to examine effects of Big Five traits on friendship selection. Results indicated that friendship networks between just-acquainted individuals became increasingly more cohesive within the first 3 months and then stabilized. Whereas individuals high on Extraversion tended to select more friends than those low on this trait, individuals high on Agreeableness tended to be selected more as friends. In addition, individuals tended to select friends with similar levels of Agreeableness, Extraversion, and Openness.
DOI: http://dx.doi.org/10.1111/j.1467-6494.2010.00625.x.

Further publications are listed at the webpage with literature and the webpage with further applications.

Degree-related effects; new terminology

Recent experience has led me (T.S.) to think more and more that degree-related effects are important in many data sets for obtaining a good model specification. These are effects reflecting that tendencies to send and receive ties can be influenced by the outdegrees and/or indegree of the actors involved; either by the degree of the sender - called degree-related activity, or by the degree of the receiver - called degree-related popularity, or by the degree of sender and receiver - called degree-related assortativity. This is described in the tutorial in Social Networks (2010, first issue).

The names of these effects have also changed. I am sorry about the confusion created by name changes, but in the end I prefer to have the best possible names. In the manual I try to mention earlier as well as current names in the descriptions of the effects (to help sort out the confusion...). What used to be called the popularity effect now is called indegree related popularity; what used to be called the activity effect now is called outdegree related popularity; what used to be called the outdegree^(1.5) effect now is called outdegree related activity (sqrt).

Known bugs in Siena 3

Bugs that affect only minor aspects of output, or that occur under very specific circumstances and are evident because they lead to the program stopping or even crashing, are not always mentioned here; see the downloads page for new versions where bugs are repaired as much as possible.

Means of changing dyadic covariates, as reported in the first part of the output files, are calculated including the diagonal values. What is worse, these means are not subtracted in the simulations.
For some time it has been impossible to estimate more than one network endowment effect. This was corrected in version 3.17w.
There was an error in the multi-project option when it makes use of unconditional estimation (which always is the case for networks and behavior). This was corrected in version 3.17w.
The degree - popularity (sqrt, for nondirected networks) and indegree - popularity (sqrt) effects as well as the raw (not sqrt) out-indegree assortativity effect had serious errors (for the in/degree-popularity effect, only in Siena versions 3.15 to 3.17u). Also the out-indegree assortativity, in-out degree assortativity, and in-indegree assortativity effects had errors. All these effects have been corrected in version 3.17v (22/05/09). I thank Krists Boitmanis for bringing these errors to my attention.
There were some inconsistencies in handling changing composition (discovered by Ruth Ripley), which have been corrected in version 3.17u (18/05/09). This presumably will not lead to large differences in results.
Rates of change depending on behavioral dependent variables are calculated based on the previous observation, not on the current value of the behavior. Usually this will have minor consequences.
There was an error in handling data sets with structural zeros and/or ones that are changing over time. This error was about the calculation of distances and estimation of rate function parameters, and it was corrected in version 3.17i (20/10/08); but then the correction was lost again in versions 3.17n - 3.17q. The correction is included again from version 3.17r.
The way of handling changing structural zeros and/or ones was further improved in version 3.17s (22/03/09) and then in version 3.17u (18/05/09).
There was a bug in the calculation of the transitive triplets effect. Specifically, this effect was combined with the transitive mediated triplets effect: in a transitive triplet (i -> j, j -> h, i -> h), also the tie (j -> h) was counted as contributing to transitive triplets for actor j. This error was corrected in version 3.17q (20/03/09).
These was an error in calculating standard errors of endowment effects. This was corrected in version 3.17q (20/03/09).
When running Siena05 and requesting only 1 simulation, some internal data files were overwritten. This started with version ???, and the bug was corrected in version 3.17p (10/02/09).
When specifying a (non-longitudinal) data structure for an ERG model with more dyadic covariates than actor covariates, the program crashed. This was repaired in version 3.15.
There was an error in the 'same covariate' (earlier called 'covariate identity') effect for the longitudinal (not the ERGM) case in Siena versions up to 3.13. It was repaired in version 3.14.
The ERGM estimation with structurally determined values had an error. For model codes 11 and 15 this was repaired in version 3.12a.
There is a bug in the Examine option of StOCNET in the calculation of the triad census. This bug is not present in the calculation of the triad census in Siena02.

Back to the main Siena page

	SIENA news





	RSiena: Siena in R Appendix B of the RSiena manual contains a list of changes compared to earlier versions. The manual for RSiena has been extensively updated on January 17, 2012. This hopefully now has removed the traces of Siena 3 which still used to be visible here and there. The links with the functions used for executing RSiena also have been given more consideration. The Siena scripts page has been frequently updated with new scripts; large updates early September 2011, minor updates mid January 2012. Important new options in RSiena since April 2010 are the sienaGOF and sienaTimeTest functions contributed by Josh Lospinoso, respectively for goodness of fit checking and for testing time homogeneity for data with 3 or more waves. The sienaGOF function currently is available only in RSienaTest. See the manual for further explanation of these functions.

	Some new publications Kevin Lewis, Marco Gonzalez, and Jason Kaufman (2011). Social selection and peer influence in an online social network. PNAS 2011; published ahead of print December 19, 2011. Abstract Disentangling the effects of selection and influence is one of social science's greatest unsolved puzzles: Do people befriend others who are similar to them, or do they become more similar to their friends over time? Recent advances in stochastic actor-based modeling, combined with self-reported data on a popular online social network site, allow us to address this question with a greater degree of precision than has heretofore been possible. Using data on the Facebook activity of a cohort of college students over 4 years, we find that students who share certain tastes in music and in movies, but not in books, are significantly likely to befriend one another. Meanwhile, we find little evidence for the diffusion of tastes among Facebook friends - except for tastes in classical/jazz music. These findings shed light on the mechanisms responsible for observed network homogeneity; provide a statistically rigorous assessment of the coevolution of cultural tastes and social relationships; and suggest important qualifications to our understanding of both homophily and contagion as generic social processes. DOI: http://dx.doi.org/10.1073/pnas.1109739109. The paper with the basic elaborate exposition about the "Siena approach" to studying dynamics of selection and influence: Steglich, C.E.G., Snijders, T.A.B. and Pearson, M., Dynamic Networks and Behavior: Separating Selection from Influence. Sociological Methodology (2010), 329-393. DOI: http://dx.doi.org/10.1111/j.1467-9531.2010.01225.x . Issue 32.1 of Social Networks is devoted to Network Dynamics and carries a tutorial and several applications of Siena. Dijkstra, J. K., Lindenberg, S., Veenstra, R., Steglich, C., Isaacs, J., Card, N. A. & Hodges, E. V. E. (2010). Influence and selection processes in weapon carrying during adolescence: The roles of status, aggression, and vulnerability. Criminology, 48, 187-220. Abstract The role of peers in weapon carrying (guns, knives, and other weapons) inside and outside the school was examined in this study. Data stem from a longitudinal study of a high-risk sample of male students (7th to 10th grade; N = 167) from predominantly Hispanic low-socioeconomic- status schools in the United States. Longitudinal social-network models were used to test whether similarity in weapon carrying among friends results from peer influence or selection. From a goalframing approach, we argue that weapon carrying might function as a status symbol in friendship networks and, consequently, be subject to peer influence. The findings indicate that weapon carrying is indeed a result of peer influence. The role of status effects was supported by findings that weapon carrying increased the number of friendship nominations received by peers and reduced the number of given nominations. In addition, peer-reported aggressiveness predicted weapon carrying 1 year later. These findings suggest that adolescent weapon carrying emerges from a complex interplay between the attraction of weapon carriers for affiliation, peer influence in friendship networks, and individual aggression. DOI: http://dx.doi.org/10.1111/j.1745-9125.2010.00183.x. Maarten Van Zalk / Selfhout, William Burk, Susan Branje, Jaap Denissen, Marcel van Aken, and Wim Meeus (2010). Emerging late adolescent friendship networks and big five personality traits: A social network approach. Journal of Personality, 78, 509-538. Abstract The current study focuses on the emergence of friendship networks among just-acquainted individuals, investigating the effects of Big Five personality traits on friendship selection processes. Sociometric nominations and self-ratings on personality traits were gathered from 205 late adolescents (mean age = 19 years) at 5 time points during the first year of university. SIENA, a novel multilevel statistical procedure for social network analysis, was used to examine effects of Big Five traits on friendship selection. Results indicated that friendship networks between just-acquainted individuals became increasingly more cohesive within the first 3 months and then stabilized. Whereas individuals high on Extraversion tended to select more friends than those low on this trait, individuals high on Agreeableness tended to be selected more as friends. In addition, individuals tended to select friends with similar levels of Agreeableness, Extraversion, and Openness. DOI: http://dx.doi.org/10.1111/j.1467-6494.2010.00625.x. Further publications are listed at the webpage with literature and the webpage with further applications.

	Degree-related effects; new terminology Recent experience has led me (T.S.) to think more and more that degree-related effects are important in many data sets for obtaining a good model specification. These are effects reflecting that tendencies to send and receive ties can be influenced by the outdegrees and/or indegree of the actors involved; either by the degree of the sender - called degree-related activity, or by the degree of the receiver - called degree-related popularity, or by the degree of sender and receiver - called degree-related assortativity. This is described in the tutorial in Social Networks (2010, first issue). The names of these effects have also changed. I am sorry about the confusion created by name changes, but in the end I prefer to have the best possible names. In the manual I try to mention earlier as well as current names in the descriptions of the effects (to help sort out the confusion...). What used to be called the popularity effect now is called indegree related popularity; what used to be called the activity effect now is called outdegree related popularity; what used to be called the outdegree^(1.5) effect now is called outdegree related activity (sqrt).

	Known bugs in Siena 3 Bugs that affect only minor aspects of output, or that occur under very specific circumstances and are evident because they lead to the program stopping or even crashing, are not always mentioned here; see the downloads page for new versions where bugs are repaired as much as possible. Means of changing dyadic covariates, as reported in the first part of the output files, are calculated including the diagonal values. What is worse, these means are not subtracted in the simulations. For some time it has been impossible to estimate more than one network endowment effect. This was corrected in version 3.17w. There was an error in the multi-project option when it makes use of unconditional estimation (which always is the case for networks and behavior). This was corrected in version 3.17w. The degree - popularity (sqrt, for nondirected networks) and indegree - popularity (sqrt) effects as well as the raw (not sqrt) out-indegree assortativity effect had serious errors (for the in/degree-popularity effect, only in Siena versions 3.15 to 3.17u). Also the out-indegree assortativity, in-out degree assortativity, and in-indegree assortativity effects had errors. All these effects have been corrected in version 3.17v (22/05/09). I thank Krists Boitmanis for bringing these errors to my attention. There were some inconsistencies in handling changing composition (discovered by Ruth Ripley), which have been corrected in version 3.17u (18/05/09). This presumably will not lead to large differences in results. Rates of change depending on behavioral dependent variables are calculated based on the previous observation, not on the current value of the behavior. Usually this will have minor consequences. There was an error in handling data sets with structural zeros and/or ones that are changing over time. This error was about the calculation of distances and estimation of rate function parameters, and it was corrected in version 3.17i (20/10/08); but then the correction was lost again in versions 3.17n - 3.17q. The correction is included again from version 3.17r. The way of handling changing structural zeros and/or ones was further improved in version 3.17s (22/03/09) and then in version 3.17u (18/05/09). There was a bug in the calculation of the transitive triplets effect. Specifically, this effect was combined with the transitive mediated triplets effect: in a transitive triplet (i -> j, j -> h, i -> h), also the tie (j -> h) was counted as contributing to transitive triplets for actor j. This error was corrected in version 3.17q (20/03/09). These was an error in calculating standard errors of endowment effects. This was corrected in version 3.17q (20/03/09). When running Siena05 and requesting only 1 simulation, some internal data files were overwritten. This started with version ???, and the bug was corrected in version 3.17p (10/02/09). When specifying a (non-longitudinal) data structure for an ERG model with more dyadic covariates than actor covariates, the program crashed. This was repaired in version 3.15. There was an error in the 'same covariate' (earlier called 'covariate identity') effect for the longitudinal (not the ERGM) case in Siena versions up to 3.13. It was repaired in version 3.14. The ERGM estimation with structurally determined values had an error. For model codes 11 and 15 this was repaired in version 3.12a. There is a bug in the Examine option of StOCNET in the calculation of the triad census. This bug is not present in the calculation of the triad census in Siena02.





	Back to the main Siena page