Sampling

The Social Study panel is designed to be representative of persons aged 16 and over, resident within private households in Belgium, regardless of their nationality or citizenship. Persons resident in collective households such as retirement homes, student homes, religious institutions or penal institutions are not part of the target population and are therefore a priori excluded. Because of fieldwork constraints, persons resident in the German-speaking community are also excluded.

Sampling procedure

The panel is based on a probability sample drawn from The Belgian National Register of Natural Persons. To obtain the random sample, we employed a stratified, two-stage cluster design where boroughs (NIS6) were the primary sampling units, and residents the secondary sampling units.

At the first stage, 800 sampling points were randomly selected by a systematic random sampling procedure with replacement clustered within boroughs, explicitly stratified by regions and provinces and implicitly stratified by target population size. Sampling points corresponded to a fixed number of expected panelists (n = 5) and a varying number of gross sample persons, depending on expected response rates. We estimated expected response rates based on historic and current response rates achieved by several face-to-face, random-probability surveys of the general resident population in Belgium.^*
Because response rates are typically lower in more urban areas, we oversampled based on urbanicity class of the borough (higher-order Belfius classification). This involved selecting larger (gross) sampling points in more urban areas and smaller ones in rural areas.

At the second stage, persons within the selected boroughs were randomly selected from the Belgian National Register by Statistics Belgium (Statbel). This was done using a systematic random sampling procedure without replacement and with equal inclusion probabilities, while implicitly stratified by age and sex.