2.2 Sampling Designs

In the courses I have taught on SNA, invariably a substantial number of participants in those courses came in assuming all network studies were essentially going to be ego network in nature. So, these students have often remarked they couldn’t figure out why we needed an entire workshop (or even more, a whole semester!) to incorporate these ideas into their methodological toolkits. But, while the capabilities of even ego network research are rapidly expanding (Perry, Pescosolido, and Borgatti 2018), in terms of data collection, this captures but a small sampling of network studies’ capabilities.38 To differentiate between these sampling designs and other key elements of the GSS and a few other exemplar studies used throughout this book, see Table 2.2.

Table 2.2: Design Characteristics in 3 Exemplar Studies.
GSS Add Health Project 90
Name Generator(s) Important Matters Friends Social Ties
Romantic Partners Sexual Partners
Drug/Needle Sharing
Response Format Free Listing Roster Free Listing
Caps 5 (implicit) 5 male & female friends (explicit) 16, across all ties (implicit)
Time Constraints w/in last 6 mos w/in last 18 mos (romantic) "w/in last 6 mos (own ties)
Values NA Rank-ordered (friends) NA
Name Interpreters Tie Characteristics Tie Characteristics Tie Characteristics
Alter Characteristics Alter Characteristics
Alter-Alter Ties Alter-Alter Ties
Sampling Design Ego Complete Partial
Sample Characteristics Random Sample Census w/in 128 schools Link Tracing
Mode(s) of Administration In-person Interview Self-Administered (some ACASI) In-person Interview
Waves Cross-sectional 3, 6mo + 1y apart (16 schools) 5, yearly
Dates 1985, 2004 1994-96 1988-92

This table summarizes how several of the principles described in the text are applied in three studies drawn upon as examples throughout the book. (NOTE:These characteristics refer only to the network aspects of the respective studies.)