Skip to content

Experiments

Summary of experiments

This table summarises experimental studies for all Waves of the Study. Each experiment is described in greater detail with a summary of the design, a brief title and description, and the waves in which it was carried. Access the experiments via the following pages:

Find the results of all IP experiments in our Working Papers series, available in the publications library.

Submit an experiment or new survey content

Each wave of the Innovation Panel has experiments related to measurement or fieldwork procedures, or to address substantive social science questions. Researchers can apply in an annual competition to have their own experimental or methodological studies on the IP. Since Wave 16, the annual Innovation Panel competition has also included a specific call for proposals for studies comprising new questions without significant experimental or survey methods components.

A series of working papers documenting experiments and other studies are available in the publications library.

Researchers may have research questions related to the experimental manipulation or simply need to be aware of potential effects the experiment may have on their own research questions. This information may also be useful in formulating research ideas. The descriptions identify variables allocating cases to different conditions and other variables relevant to carrying out the experiment. Allocation variables are copied into the next waves as well.

In designing the experiments and allocating sample members to randomised treatments great care is taken to avoid contamination and confounding of experiments (Lynn and Jäckle 2019). Depending on the purpose of the experiment, treatments are randomised between individuals, between households, between interviewers, or between primary sampling units. The level at which randomisations are allocated is recorded in the documentation below. For experiments that might affect the same outcomes, the randomisations are fully crossed. In addition, all randomisations are stratified.

Email newsletter

Sign up to our newsletter