Validity of an experiment certifies its fairness. The factors that could affect the validity of a test are represented by the followings: appropriate measuring tools and equipment, the variables that are measured, the established range of values. There exists two types of validity, internal and external.
There are many means to check the internal validity of an experiment, such as predictive, concurrent and face validity.
Predictive validity shows if a new measure can forecast the future results. Face validity indicate if the measurement is appropriate and suitable for what is intended to measure. Concurrent validity indicates if the results of a new experiment are comparable to the results obtained by performing other experiment, in the same field.
Improvements to internal validity are brought by means of investigations using single- or double-blind study techniques.
There are also several means to check the external validity of an experiment, such as ecological, historic and population validity. The improvements to external validity are brought about by means of investigations using diverse samples. The more diverse the sample used are, the more applicable to a wider population the results will be.
Three simple ways to improve the validity of an experiment:
1. Change only one independent variable at a time.
2. Measure dependent (responding) variable accurately.
3. Did you keep all other variables constant, were there variables like wind speed that you could not control and didn't measure that impacted your results?