Trellis and Breadboard: Software Tools for Social Networks Research
Thursday, October 7, 2021 – 3:30 – 5:00 pm
Dr. Thomas Keegan, Executive Director, Institute for Network Science, Yale University
Mark McKnight, Senior Software Engineer, Human Nature Lab, Dept. of Sociology, Yale University
There is extensive research on social connectedness as a protective factor from pathologies affecting both physical and mental health. Concerns about the validity of such findings center around the fact that this type of research relies heavily on self-reports, which may also potentially limit generalizability of findings (Weiner 2005).
This talk will describe two software tools than can be used to facilitate research on social connectedness and social networks research using methods that both improve the reliability of self-report methods and that do not rely on self-report at all. Breadboard (https://breadboard.yale.edu) is a software platform for developing and conducting human interaction experiments on networks online. It allows researchers to rapidly design experiments using a flexible domain-specific language and provides researchers with immediate access to a diverse pool of online participants. Breadboard experiments result in behavioral observations that are not readily tied to measures of social connectedness by naïve participants. Trellis (https://trellis.yale.edu) is a suite of software tools for developing, administering, and collecting survey and social network data that improves on more typical self-report measures by incorporating features conducive to capturing full sociocentric network data, even among individuals who may have challenges with more typical self-report measures (e.g, low-literacy populations, children, or groups with “weak ties”).
Breadboard: can be applied to many socio-behavioral research questions. For instance, we are currently using very generic experimental methods from behavioral economics to assess the differences in cultures of sharing and cooperation across villages in rural Honduras. Breadboard has received wide recognition as a superior tool for conducting online network research. It has been used by many researchers in the behavioral sciences.
Trellis: provides support for integrating social network data with custom-designed survey data. Also, you can use Trellis to conduct a photographic census of the respondent population and use such “photo IDs” to map entire social networks, avoiding artificially constricted network data. This last feature makes Trellis especially useful for many challenging settings, such as where there is low literacy or difficulty remembering names. This tends to be more of a "niche market" at the moment, but we believe it has great potential for more widespread applications.
We have been unable to find alternatives that are of comparable quality for conducting social interaction games online or for mapping social networks.
Prerequisites: Both tools require some programming ability, although we continually lower the burden in this regard. We have yet to develop completely user friendly versions for those with no coding ability.
Both are open source software. No license or fee is required for educational and nonprofit research use.
Further information is available on our web sites: https://trellis.yale.edu and https://breadboard.yale.edu
Breadboard has a tutorial video on the web site, and a full wiki at https://github.com/human-nature-lab/breadboard/wiki
- Lungeanu, A., McKnight, M., Negron, R., Munar, W., Christakis, N. A., & Contractor, N. S. (2021). Using Trellis software to enhance high-quality large-scale network data collection in the field. Social Networks, 66, 171-184.
- Shakya, Holly B., Derek Stafford, D. Alex Hughes, Thomas Keegan, Rennie Negron, Jai Broome, Mark McKnight et al. "Exploiting social influence to magnify population-level behaviour change in maternal and child health: study protocol for a randomized controlled trial of network targeting algorithms in rural Honduras." BMJ open 7, no. 3 (2017): e012996.
- Shirado, Hirokazu, and Nicholas A. Christakis. "Locally noisy autonomous agents improve global human coordination in network experiments." Nature 545.7654 (2017): 370-374. - Wilson, Robert C., and Anne GE Collins. "Ten simple rules for the computational modeling of behavioral data." Elife 8 (2019): e49547. DOI: 10.7554/eLife.49547 https://elifesciences.org/articles/49547
- Shirado, Hirokazu, and Nicholas A. Christakis. "Interdisciplinary Case Study: Understanding the Cooperation of Humans and Robots through the Collaboration of Social and Computer Scientists." Iscience 23.12 (2020).