Manipulation (the deliberate variation of a variable) is always a part of…

A research method is classified as "experimental" only when the researcher deliberately manipulates (varies) one or more independent variables and then measures the resulting change in a dependent variable, all under controlled conditions. No other classification of research makes such deliberate manipulation a defining requirement.

In this question, the four listed labels sort research along different classification criteria — some by method (historical, descriptive, experimental) and one by purpose (fundamental). Because manipulation of a variable is exactly what turns a study into an "experiment", a research study that does not manipulate anything cannot be called experimental, no matter what topic it studies.

• Historical research: reconstructs and interprets events that have already occurred, working from existing documents, records, and artifacts — the researcher does not create or alter any condition.

• Fundamental (basic) research: is defined by its purpose of extending theoretical or scientific knowledge, and can itself be conducted using experimental, descriptive, or historical designs — so it has no fixed method of its own.

• Descriptive research: records the characteristics, frequency, or status of a phenomenon exactly as it naturally occurs (for example through surveys or case studies), without altering any variable.

• Experimental research: is built around deliberately varying one factor and observing its effect on another, which is why manipulation is always present in it.

Hence, of the four options, only experimental research always involves manipulation.