Most assessment test research involves a short-term cross-sectional (one-time) study to obtain data and standardize scores during test development. These studies may have only a few hundred participants. It is an attempt to validate the test and publish results. For the test user, there are questions about the validity of the instrument for the people they test and questions about who the test was standardized on. These one-time studies provide scant evidence for test validity and there are rarely studies done in the future for follow up.

A common practice when validating tests is to administer a second (criterion) test that purports to measure the same thing. The test in question is correlated with the criterion test and the magnitude of the correlation provides validation.  Risk & Needs completes criterion study validation research on all of its tests early in their development.  This criterion validity research is summarized in each tests "Inventory of Scientific Findings."  With regard to many other tests, these correlation coefficients are of a magnitude of around .4 or less. While .4 is statistically significant it is not a very high correlation, and, rather modest support for validity. Often the criterion test was validated in the same way.

In everyday assessment settings it is not practical to administer two tests for the purpose of validation. For tests with multiple scales, a criterion test is needed for each scale. It is an imposition on both the client and staff. Yet, test validity is always a concern because of the changing nature of society, cultural influences, and civic response to crime. It would be advantageous to have a validation method that was ongoing or at least done on an annual basis. This could be done if data were readily available from test users.  Risk & Needs suggests consideration of "an innovative approach to test validation" on an ongoing basis.  This approach does not replace "criterion validity studies."  Indeed Risk & Needs does these criterion validity studies early in the development of all its tests. This "innovative approach" is only recommended for ongoing database analysis.

Risk and Needs has an innovative approach to test validation. Along with having test data from the users of the test via built-in test databases, Risk and Needs tests have suitable criterion measures contained within test databases. Suppose we wanted to validate an alcohol test or scale. A test to measure alcohol problems is valid if it correctly identifies people who have drinking problems. For this we would need to know that a respondent had a drinking problem and this information can be gotten from the respondent. One strategy would be to use an item off the test, such as, "I have been treated for a drinking problem," or "I have a drinking problem." This treatment item would provide evidence that the respondent had a drinking problem because there would be no reason to have treatment if a problem did not exist. By dividing the respondents into two groups, those who had treatment and those who did not, we can examine test scores for just the treatment group. Because they have had treatment, we would expect the treatment group to score in the problem range. A high percentage (90% or better) of respondents who had treatment and scored in the problem range would validate the test.

Other scales can be validated in a similar way. For example, drug treatment to validate drug scales, anxiety, depression, and distress treatment to validate anxiety, depression, and distress scales, respectively. Also criminal arrests records could be used, arrests for violence, domestic violence or the threat of violence to validate violence scales, sex-related arrests to validate sexual adjustment scales, etc.

This validation method could avoid administering another test (perhaps one that is not as good) as a criterion, and could be done on an ongoing basis. It is unique and innovative, and it avoids abusing staff time and unnecessarily intruding on the client. Sure there will be "ivory tower" purists who will only accept traditional criterion validity. Frankly, we are in that group when discussing a test’s initial validity research. However, on an ongoing annual basis this treatment versus non-treatment approach has some merit. Risk and Needs built-in test database’s makes this innovative validation method possible. The database contains data from test users and therefore test validity directly applies to the user’s population.

In instances where a client had been in treatment and considers himself cured, Risk and Needs tests incorporate questions to identify those clients who are in recovery or feel they have recovered. These test items are highlighted so that staff can verify that the client had a problem in the past but is abstaining.

Risk and Needs has another innovative validation method that is also possible with built-in test databases. In a method similar to comparing treatment and no treatment groups, we can compare test scores between first offenders and multiple offenders. A multiple offender is one who has been arrested two or more times, whereas a first offender has just the one or no arrest. It would be expected that multiple offenders score significantly higher than first offenders because they have a more established pattern of problems. Having a second and subsequent arrest is indicative of problems.

Comparisons between mean test scores for the two offender groups would determine whether or not the test differentiates between first offenders and multiple offenders. Significantly different mean test scores would validate the test. It would be expected that multiple offenders would score significantly higher on tests that measure problem severity. For example, offenders who have two or more alcohol-related arrests are expected to score significantly higher than first offenders on a test that measures drinking problem severity. Significant differences between first and multiple offenders would mean that the test distinguishes between these two offender groups and is valid.

Risk and Needs’ built-in database provides research capability that is not possible in other tests that simply do one-time studies during test development. Risk and Needs research is ongoing. Reliability, validity and accuracy are studied on an annual basis. This dedication to research is what sets Risk and Needs apart from the competition and has led to our current status as state-of-the-art in assessment tests.