How can the objectivity of research be enhanced?

Concept: Objectivity in research means that the findings are determined by the evidence itself and not by the investigator’s personal beliefs, preferences or expectations — any competent investigator working with the same data and the same procedure would arrive at the same conclusion. Objectivity is therefore not a single act but a property the whole research process has to be built for, and it rests on three mutually supporting requirements: impartiality of the investigator, reliability of the measurement, and validity of the measurement.

Application: Each of the three requirements closes a different route through which the investigator’s person could enter the result.

1. Impartiality keeps the investigator’s own preference out of the choice of problem, the selection of the sample, the recording of observations and the interpretation of results — it removes the bias that originates in the observer.

2. Reliability keeps the measurement itself stable: an instrument that yields the same reading on re-administration, across its own items, and across independent raters produces results that belong to the phenomenon rather than to the occasion or to the person who happened to score it.

3. Validity keeps the measurement pointed at the intended construct: an instrument that actually measures what it claims to measure prevents a personally preferred meaning from being read into a number that in fact represents something else.

Cross-check: Test whether any one requirement on its own would be sufficient. A perfectly impartial investigator using an unreliable instrument still obtains readings that change from one administration to the next, so the conclusion depends on the occasion. A highly reliable but invalid instrument reproduces the same figure every time while measuring the wrong construct, so a stable result is still open to a subjective reading. A reliable and valid instrument in the hands of a partial investigator can still be applied to a loaded sample and reported selectively. Because none of the three alone secures objectivity, and each removes a distinct threat to it, the requirement is the combined set rather than any single member of it.

Result: Objectivity is enhanced by impartiality, reliability and validity taken together — that is, by all of these.