Connecting the Dots: Constructing sociometric networks from personal network data using SPIDER software

Instructor: April Young

Using practical examples from public health research, this module will provide information about the following:

• How to collect personal network (egocentric) data in a way that allows you to construct whole (sociometric) networks, and

• How to cross-reference personal network data to construct sociometric networks using SPIDER, a new software package for de-duplicating network data.

Standard protocols involve asking participants, or egos, to give the name and basic demographic information (i.e., gender, race, age) of their network members, or alters. Study staff then cross reference these data with that of other participants and named alters to construct a sociometric network representing all direct and indirect connections among participants. Due to information bias, low literacy, and the transience of many relationships, the data used in cross-referencing can be inconsistent. Thus, the process of constructing a sociometric network can be laborious and subjective. In this mini-module, we will discuss complications that arise when trying to construct sociometric networks from personal network data, drawing on examples from research in HIV, sexually transmitted infections, and substance abuse. Then, we will suggest strategies and discuss technologies that allow researchers to mitigate these complications and more easily and reliably construct sociometric networks; this includes a preview demonstration of the new software package, SPIDER. SPIDER (to be released later this year) will provide users with a system that enables efficient, semi-automated network construction using a library of robust, statistically rigorous algorithms, rich desktop-based annotation tools, and secure web-based technologies. The customizability of SPIDER allows for multi-disciplinary utility in studies using varying designs.