Traditionally in the social sciences, our studies often start with a population of interest, design a sampling strategy for reducing that population to a manageable (& representative) subset, who we then recruit; we then measure various attributes of that sample with any variety of data collection approaches. In this book, I choose to start with network measurement, before turning to network sampling. When we get to network sampling, I will describe three primary approaches; the two ideal types—ego and complete network designs—and a third known as partial network sampling, which sits somewhere in between. Saving sampling for later in the chapter was not an arbitrary choice made just to confuse and frustrate readers looking to expand their methodological repertoire to include social network approaches.21
I start with measurement because what are described later as “ego network” designs are readily incorporated into standard survey research designs (or most other individually-oriented research strategies). In my years of teaching social networks—analysis and research design—I’ve found that familiarity at the outset can help facilitate folks learning the ideas that are new to them. So, we’ll start by discussing how some key elements of measurement in network data can be tacked onto any standard individual-oriented research design, essentially identifying another class of variables one might want to gather about sampled individuals.
So for the moment, let’s set aside the considerations of how we will determine which ties to focus on, and simply focus on what we’re going to gather about a set of nodes and ties that comprise a network of interest.22