A researcher develops a questionnaire to assess the personality trait of impulsivity among adults. In a journal article, she presents evidence that college students tend to get essentially the same score if they take the test twice, two months apart. She also presents the average score, the highest score, and the lowest score obtained by two large samples: one of 2,000 college students and one of 750 community-dwelling non-college adults. However, when you look at the sample questionnaire items she included in the article, it seems to you that they relate more to whether a person is sociable, outgoing, and fun than to whether an individual is impulsive. In this scenario, you are questioning the of the researcher's questionnaire.

The correct answer is validity

The perfect validity signals that there is no measurement error, that is, that “the differences observed in the measurement reflect the real differences between the objects (or individuals)” .

In general, content validation consists of “showing that the test items are samples of the universe in which the researcher is interested”, proving that this sample is indeed representative. A measurement of any kind is valid when it actually measures what it is proposing to measure. The researcher, when defining a method and a sample for a given study, must pay attention so that external factors do not exert any influence on the variables involved, under penalty of having their final results deviated in a biased way. Some bias or trend can occur in several aspects of the research, not only in the definition of the method and the sample.

The idea of validation also involves the concern with a neutral way of writing the questions of an instrument, the way of practicing the interviews for data collection, recording, information processing and dissemination of results.

In the previous figure, it can be seen that in the zone of low validity, although the pitches are well concentrated among themselves, the focus is shifted. This, in the practical case of a survey, could mean the existence of some tendency, since all records are being displaced from the central objective (as an example, one can quote an interviewer who induces the respondent to a certain answer).

barack obama he actually listens