Work in this cluster aims at collecting and constructing appropriate data sets and at the development of methodological tools, mainly measurement instruments, measurement models, and statistical analysis procedures that are particularly important for ICS research. Researchers in this cluster also provide methodological support to substantive-oriented ICS researchers.

An important element of this research cluster is the design and collection of appropriate data sets. These include complex and large-scale multi-actor, multi-level, and multi-event data sets in various fields (e.g., families and households, neighborhoods, work organizations, inter-firm relations). Internationally comparative data sets continue to be very important. Individual-level data are increasingly combined with detailed secondary sources (e.g. via postal code information and geographical information from Statistics Netherlands) in order to model the impact of the meso- and macro-social context. Advocating a multi-method approach, ICS research also employs designs that usefully complement large-scale surveys, such as experiments, vignette studies, and expert interviews, often yielding useful data for specific tests of predictions derived from theoretical models.

Statistical research in this cluster is based on general multivariate statistical methodology and includes topics such as scale construction, categorical data analysis, and structural equation modeling. The focus is on measurement models in socially structured situations (like social networks and local contexts), and on statistical models for analyzing data sets with a complex structure (multi-actor, multi-event, multi-level) that are being applied in empirical ICS research. This includes procedures for measuring and analyzing social networks; multilevel analysis and its use for the analysis of contextual data; and longitudinal methods, in particular event history analysis and methods for repeated cross-sectional surveys. Particular attention is paid to statistical models that reflect theoretical modeling of purposive individual behavior in the social context.

The development of models for social network analysis is an important topic in this cluster. There recently have been important advances in data analysis procedures for social networks based on stochastic models, made possible by the use of computer-intensive statistical methods, mainly Markov chain Monte Carlo methods. This cluster contributes to this line of research, in particular to methods for dynamic network analysis, including the joint modeling of the evolution of relational networks and of behavior and performance of individual actors. The integration with theoretical work is pursued by the use of random utility models and close collaboration with the other clusters.

The ICS maintains a support cluster Data archiving. The aim of this support cluster is to facilitate and professionalize the collection of primary data by ICS researchers, to document and archive primary data collected by ICS researchers, to facilitate the use of secondary data from Dutch and foreign sources by ICS researchers, and to promote the use of standardized ways of measurement in ICS research. Via this support cluster, ICS cooperates closely with archives for social science data in the Netherlands and abroad, for example, the Steinmetz Archives of NIWI.