Reply 8-1 RK Reply to:
In an experiment, one of the most important things a researcher must focus on is the principle of validity, or the assurance tha
In an experiment, one of the most important things a researcher must focus on is the principle of validity, or the assurance that the research is actually studying what it was intended to study (Myers & Hansen, 2012). When setting up an experiment, there are two forms of validity in particular that the researcher must focus one. Firstly, the researcher must focus on internal validity. This type of validity is concerned with the assurance that the changes that were measured in the dependent variable (DV) were actually caused by the manipulation of the independent variable (IV) and not caused by any other types of extraneous variables (EV). Secondly, researchers must be concerned with their study’s external validity. External validity refers to how well the results of the study generalize to the population of interest from which the sample was pulled (Myers & Hansen, 2012). When setting up an experiment, there are several ways in which the researcher can increase both of these validity types.
For internal validity, one of the main concerns is the study’s amount of control. Control refers to the number of safeguards that were put in place in order to control for EV’s, or variables other than the IV that may have influenced or changed, the DV (Myers & Hansen, 2012). The best time to plan for these controls is during the set up of the experimental procedures. For example, researchers should ensure that they have properly defined their variables, properly set up their measuring techniques, and included enough levels of the IV to properly test their hypothesis. Additionally, in order to control for EV’s caused by subject variables, or individual differences between subjects, researchers should ensure they utilize random assignment (Myers & Hansen, 2012). Researchers can also choose to measure their internal validity after their study is over. For example, researchers can provide questionnaires, or surveys, to their subjects after the study concludes in order to investigate how the subject felt during the examination, or if they were confused during the study. These are important factors to look at as these could introduce possible EVs into the study that may have resulted in the measured behaviors (Myers & Hansen, 2012).
As mentioned above, the primary concern for external validity is the generalizability of the identified experimental results (Myers & Hansen, 2012). External validity has a type of double-edged sword type of relationship with the previously discussed principle of internal validity. For example, as internal validity increases with the number of controls put in place, external validity decreases with large amounts of controls as it is thought to limit realistic behavior. There are several ways in which research can increase their studies external validity. For example, researchers can use the principle of aggregation, or the grouping together of data, subjects, stimulus, or trials (Myers & Hansen, 2012). The idea behind this is that by aggregating these factors, they are generalizing the behaviors over a wide range of subjects, stimuli, and occasions. This provides for the most realistic measure of behavior. Additionally, researchers can choose to incorporate multivariate designs, or designs that include multiple dependent variables (Myers & Hansen, 2012). The idea behind multivariate designs is that rarely do behaviors exist independently. By measuring multiple DVs, or behaviors, researchers can gather a more accurate representation of how the manipulated IV actually impacted behavior. Lastly, researchers can maximize external validity by designing experiments to minimize reactivity, or the tendency for subjects to change their behavior when they feel as though they are being watched (Myers & Hansen, 2012). It is a major belief that lab settings, or settings in which the behavior is overly controlled, does not elicit realistic behaviors. With that in mind, researchers often try to maximize the measure of behavior in more natural settings such as in field experiments or in naturalistic observations (Myers & Hansen, 2012).