In which research design, the independent variable is selected rather than…

Research designs are broadly classified by whether the researcher can manipulate the independent variable (IV) or must instead select it as it already exists. In an experimental design, the researcher actively manipulates the IV under controlled conditions and observes the effect on the dependent variable (DV). In an ex post-facto (causal-comparative) design, the IV is a pre-existing characteristic, condition, or event (for example, gender, smoking status, or an accident that has already occurred) that cannot ethically or practically be manipulated — so instead of manipulating it, the researcher selects groups based on the naturally occurring levels of the IV and studies its relationship with the DV through statistical comparison, without any experimental intervention.

The stem asks specifically for the design where the IV is ‘selected’ rather than ‘manipulated’ — that is the defining feature of the ex post-facto (causal-comparative) research design.

Contrast with the other offered designs:

• An experimental design manipulates the IV directly under controlled conditions — the opposite of selecting a pre-existing IV.

• A historical design systematically reconstructs and interprets past events from documents, records, and relics; it is not built around selecting or manipulating any IV–DV relationship at all.

• A descriptive survey design describes the current characteristics, opinions, or status of a population using tools like questionnaires; it does not test an IV–DV relationship through selection or manipulation either.

This matches the standard research-methodology definition confirmed across multiple sources (Sage Encyclopedia of Research Design and worked UGC NET Paper I archives for this exact item), which is why ex post-facto research design is the design where the IV is selected rather than manipulated.