Validity concepts and applications

Validity concerns are important in any study, providing a guideline of whether there is a strong correlation between what is found in an investigation and its applicability to its occurrence in general. This is therefore an important concern in scientific study to determine whether such a study is in fact applicable or whether its applicability is minimal. Three types of validity can be identified, including internal validity, external validity, and construct validity. Each provides a specific point of validity that can be applied according to the requirements of the study.

Internal validity refers to a strong causal link among one's independent and dependent variables. This also means the ability to rule out other variables, including extraneous or alternative causes for dependent variables. There is therefore an unambiguous link between causes and effects in the study, as well as a high level of causal control. The environment is highly controlled, and therefore major control can also be exercised over extraneous factors that might influence internal validity (Reliability and Validity, 2012).

External validity moves towards a wider viewpoint, with the inclusion of other people and other situations to which the study can be generalized. Strong external validity is determined by using random methods to apply to a defined population. A sample measurements and situations can then be generalized with confidence when these have proven consistent among a certain population (Reliability and Validity, 2012).

Construct validity concerns the corresponding values between the concepts and measurements within a study. In ideal conditions, the construct validity will be an accurate reflection of the abstract concept being studied. It is therefore highly important to obtain clarity in the conceptual definitions of variables within a study, since this clarity determines the student's ability to assess the correspondence of the measures obtained to the concepts beings studied.

Threats to external validity may occur from populations with different responses than dictated by the study hypotheses. This may occur when populations show wide variations in their responses to particular hypotheses, which are in violation of the expected response.

Construct validity may be compromised when alternative correspondences are achieved between a construct and an unexpected concept found within the actual environment being studied. In other words, when the construct does not correspond with the original hypothesis, its validity would be compromised.

In cloud computing and determining inside threats, various elements may need to be taken into account before taking action. One potential compromise to such an investigation might be the lack of validity when estimating potential threats. In other words, insider threats might be considered higher among employees who are knowledgeable in computer technology and spend more time at the company than other employees. However, such employees can be to the detriment of a full investigation, which might reveal a threat from another source, which might be unexpected.