In sociology and research terms, internal validity is the degree to which an instrument, such as a survey question, measures what it is intended to measure while external validity refers to the ability of results of an experiment to be generalized beyond the immediate study.
True validity comes when both the instruments used and the results of experiments themselves are found to be accurate each time an experiment is conducted; as a result, all data that is found to be valid must be considered reliable, which means it must be capable of being repeated across multiple experiments.
As an example, if a survey posits that a student's aptitude score is a valid predictor of a student's test scores in certain topics, the amount of research conducted into that relationship would determine whether or not the instrument of measurement (here, the aptitude as they relate to the test scores) are considered valid.
The Two Aspects of Validity: Internal and External
In order for an experiment to be considered valid, it must first be considered internally and externally valid. This means that an experiment's measuring tools must be able to be used repeatedly to generate the same results.
However, as University of California Davis psychology professor Barbara Sommers puts it in her "Introduction to Scientific Knowledge" demo course, the truth of these two aspects of validity may be hard to determine:
Different methods vary with regard to these two aspects of validity. Experiments, because they tend to be structured and controlled, are often high on internal validity. However, their strength with regard to structure and control, may result in low external validity. The results may be so limited as to prevent generalizing to other situations. In contrast, observational research may have high external validity (generalizability) because it has taken place in the real world. However, the presence of so many uncontrolled variables may lead to low internal validity in that we can't be sure which variables are affecting the observed behaviors.
When there is either low internal or low external validity, researchers often adjust the parameters of their observations, instruments, and experiments in order to achieve a more reliable analysis of sociological data.
The Relationship Between Reliability and Validity
When it comes to providing accurate and useful data analysis, sociologists and scientists of all fields must maintain a level of validity and reliability in their research-all valid data is reliable, but reliability alone does not ensure the validity of an experiment.
For instance, if the number of people who receive speeding tickets in an area varies immensely from day to day, week to week, month to month, and year to year, it is unlikely to be a good predictor of anything-it isn't valid as a measurement of predictability. However, if the same number of tickets are received monthly or annually, researchers may be able to correlate some other data that fluctuates at the same rate.
Still, not all reliable data is valid. Say the researchers correlated the sale of coffee in the area to the number of speeding tickets issued-while the data may appear to support one another, the variables on an external level invalidate the measurement tool of the number of coffees sold as they relate to the number of speeding tickets received.