Psychometrics is defined as the psychology branch dealing with design of measurements and tests within the field. It also involves the administration and interpretation of the results. The Myers-Briggs Personality inventory, the WAIS III IQ test and the Wonderlic Personnel test are all examples of psychometrics.
Measurement is a key component of psychometrics. Picking the right measurement types will help people understand the results and make the best determinations for future steps. The Myers-Briggs uses numbers to measure what type of personality trait a person has, while The Wonderlic Personnel Test is used to help determine the general intelligence level of a person in order to place him in the best work position. The Myers-Briggs could involve over 100 questions, while most Wonderlic tests use fifty questions in 12 minutes. Although the two tests measure different qualities with different scales, they could be combined to help a manager choose someone with the right personality and intelligence skill set to fill in needed company positions.
Validity in psychometrics shows how well a test or inventory expresses the theory in question and is expressed in decimals, such as .85 validity. The WAIS-III, for example, is considered a generally valid measurement of intelligence. It uses a mix of surveys, personal interviews and hands-on exercises while making allowances for people with disabilities. Scoring shows not just how intelligent a person is but where their strengths and weaknesses may lie. Any large disparities in certain categories can lead to discoveries of learning disabilities or traumatic brain injury.
Validity of a test implies reliability, which means how accurate the test or psychometric inventory measures what it measures. It is also measured in decimal. One way of reliability measurements is the test-retest method in which higher reliability rates can be achieved if the test in question can repeat the same or similar scores. However, a test can be reliable in its methods but not necessarily properly measure a theory. Thus, reliability does not always imply validity.
Psychometric results often have multiple streams of interpretation based on the raw numbers. Correlation is an expression of how two sets of data relate. Many times a correlation can indicate a need to further examine a hypothesis or theory. If, for example, a psychometric inventory was given to measure levels of depression among city workers, the results might show a negative correlation between city workers' level of depression and economic downturn. It doesn't mean that economic downturn caused depression, but it could be related. Another city facing similar situations may not show as negative of a correlation with a possible variable that city workers received better mental health care in that region comparatively.