Validity is the difference between what a selection test actually measures and what it aims to measure. Validity is defined as 'the agreement between a test score or measure and the quality it is believed to measure' (Kaplan and Saccuzzo, 2001). The validity of a particular test used for assessment is really important since it has a huge impact on the results. These tests can be both for selection or appraisal and the continued use of tests - which are not valid for the organisation - reflects poorly on the talent management of the organisation. Test validity helps the manager to understand the market that he is looking at that in turn helps him design the employer branding. Test validity also helps him to objectively quantify the results.
Each test has certain tools that try to predict the performance of a potential employee at the time of selection. Predictors are information points that an employer gets from a potential employee about himself or herself. These information points can come in from application forms, online assessment tests, psychometric tests and various other methods.
These predictors try to verify the criterion measures stated below:
a. Will the core competencies of the potential employee map to the job description?
b. Will he be able to adjust himself to the work culture
c. Job products
d. Job outcomes.
The test validity is the degree to which there predictors will actually be accurate. This process to verify this degree is known as validation. The process shows a significant statistical relationship between the predictor that the employer is using and the criterion measure the employer wants to use as yard stick. It is important to note here that criterion measures for job success have different meanings for different employer and it is really important that these are in sync with the talent management philosophy of the organisation.
Types of Validation:
a. Criterion Related Validity
A significant relationship should exist between the information at hand and some measure of work behaviour or performance. For example there can be significant statistical relationship between one year of educational gap and the turnover rate.
b. Construct Validity
A construct is a psychological trait such as - leadership, resilience, intelligence, verbal ability etc. Here in, instead of using information points that are provided by the potential employee the tests somehow aims to assess the constructs. They are a part of the job description and specification. For example there can be significant statistical relationship between the ability to solve a comprehension passage and the ability to understand a set of instructions at work.
c. Content Validity
A test will have content validity if it actually tests certain skills that form a significant part of the job. For example - there can be a preliminary excel test for a receptionist or clerk.