Controlling the factors that affect Validity
“Collecting data and its validity is your (the researcher’s) responsibility.” My professor.
What validity is there when a researcher does a study like he is making a romance novel? To better control the factors that affect validity, collection of data must be handled by the researcher – hiring others to get the data is not a good choice. A researcher should know he has the control. He is the one calling the shots. He makes the choice whether or not to put and what filters or segregators to put in obtaining data he wants to get. Before such control is applied, the researcher must know which, what, and how to control. Having scarce information on this matter renders more questions to the result of a finding and to the study as a whole.
As a responsible researcher, one should look into the factors affecting validity and apply the principles for achieving internal and external validity. Keeping this in mind and putting it into action means having a more reliable research and a more reliable result.
Factors affecting validity
- Contemporary history – the background of participants affecting the research.
- Maturation process – (approaches, treatments, etc… used usually in experimental research having been used for a long time.) Time affecting the result of the research/test.
- Pre-testing procedures – sometimes pre-tests aren’t well structured. Sometimes the pretest and the treatment have an interaction with the Independent Variable.
- Measuring instrument – one must make a valid, correct instrument. A question not understood but answered is a researcher error.
- Statistical regression – results should show natural or normal expectations. An example is that the subject who gets the lowest score in the pre-test will get a lower score than the highest in the post-test.
- Differential selection of subjects – a biased selection of conditioned sample. If a researcher selects subjects and not uses random sampling, it is an error. It is a cause of internal invalidity.
- Experimental mortality – the loss of subjects from comparison groups could greatly affect the comparisons because of unique characteristics of those subjects.
- Interaction of variables – some variable may interact and their interaction may affect the internal validity.
Principles for Internal and External validity
- Maximization of experimental systematic variance – variance is a variation that shows the differences of variables. A researcher should maximize systematic experimental variance. An example of this is using different theories, paradigms, and different treatments for different subjects. With this, it is easier to see the effect of the treatment.
- Minimization of error variance – error brought by measurement and sampling errors. A researcher should criticize or observe closely where the error comes from.
- Control of extraneous variable – undesirable variables that influence the relationship between that an experimenter is examining should be controlled.
The experimenter can eliminate it by investigating which is an extraneous variable and intervening variable, by randomization of sample, or by building into the design an extraneous or an intervening variable as an independent variable.
- Physical Control – the physical environment, schedule, place, equipment for the physical treatment of groups. A researcher must make sure that subjects and groups are within the same equal environment.
- Statistical Control – the use of statistics. Analysis of covariance is a good way to control extraneous variables. A collection of statistical models and their associated procedures may be used in experimental research.
As a researcher, one should use every method to take away extraneous variables. When Validity is established, the study will have less resistance against its approval. A researcher can’t possible rule his research unless he has rule over its validity.
