Validity is the appropriateness, meaningfulness and usefulness of the inferences made from test scores, and validation is the process of accumulating evidence to support such inferences. Validity in an employee-selection context refers to the ability of a test to select individuals to perform a specific job. Pre-employment selection testing is the primary focus of most test validations. A pre-employment test is not valid in and of itself; rather a testing process is valid in the context of a specific job. Valid testing allows one to select the best employees, which has a direct impact on profitability. Research has consistently shown that the dollar value of an increase in test validity is many times larger than the cost of even the most expensive tests. Thus, any possible increase in cost associated with a more valid test would be small in comparison to the revenue saved or gained by pretesting employee proficiency.

First Advantage tests were validated by a third party company, Linkage Inc., which is one of the world's most respected HR consulting organizations in testing and assessment. The whole validation process was under the supervision of trained psychometricians from The Donath Group (a leading provider of psychometric and evaluative research services).

Content-Oriented Validation

A test is content valid when it is a representative sample of the content of the job for which applicants are being selected. That is, a test is said to be content valid if it is known to measure the knowledge and skills required for a job or if it is known to simulate the tasks that make up a job. Typically, a content-oriented validation study will begin with a “job-analysis” to identify the tasks that make up a job as well as the knowledge and skills required for a job. Then, on the basis of this job analysis, the test developer will design the test so that the test items match the various parts (knowledge/skills or tasks) of the job. Often, critical knowledge/skills and tasks are represented by more items than those that are less critical.

Criterion-oriented Validation

Criterion-oriented validation consists of demonstrating that there is a useful relationship between test scores and one or more measures of job relevant performance (criterion). That is, a criterion-valid test is able to predict employee performance on some criterion, and the criterion must be relevant to performance on the job. Typical criteria include measures of job performance, training performance, etc.

To validate a test using a criterion-oriented approach, one might first administer the test to a group of job applicants, then randomly select employees from these applicants, and finally (without considering test performance) review the new-hire’s performance on the job. The test is valid if the new-hire’s job performance could have been predicted by the pre-employment test. The relationship between test scores and performance is usually summarized as a statistical correlation; a significant correlation is demonstration of validity. Many variants of criterion-oriented approaches to validation exist. For example, if the test is given to current employees and the test’s results are found to be correlated with employee’s current job performance, the test can be considered valid.

This latter approach is referred to as “concurrent validation” because the employees are tested at the same time as their performance is measured.

Construct-oriented Validation

Construct validation is a demonstration that the test measures the human characteristic (or construct) that it is designed to measure. If the test measures a well-defined construct (and only one construct), it is said to be construct valid. To demonstrate the construct validity of a test, one must first theoretically define the characteristic (construct) that the test measures. This definition should include a list of measures to which the test should relate as well as measures to which it should not relate.

Then, one must conduct a series of studies to determine that the test does indeed relate (or does not relate) to the appropriate measures. No single study can demonstrate the construct validity of a testing procedure. Construct validation is not typically used for demonstrating that a test is a meaningful and useful predictor of performance for a specific employment context. Rather, construct validation is typically used to demonstrate that a psychological test measures the construct that it is designed to test (for example a test of intelligence must be related to other tests of intelligence, and must not be related to personality).

"...The content of all First Advantage tests have high levels of content validity as measured and reported by subject matter experts. Criterion-related validity studies have yielded (and are expected to continue yielding) high relationships with job performance since tests are designed with a specific focus on and analysis of the job tasks, skills and abilities required to perform the job. Studies in this area have been conducted and are presently being implemented for many of the tests.

Reliability, which is a necessary condition for the validation of test scores, is within and above the range of most skill-based tests in the industry. This assures the test user of accurate measurement and very useful information for decision-making.

Users of First Advantage tests can feel confident that the information produced from a First Advantage test will supply reliable and valid information that may be used as an important part of their screening, assessment and human resource decision-making processes..."

- from a report by The Donath Group,Psychometric and Evaluative Research Services

If your system is able to make standard http (TCP/IP port 80) and SSL/https (TCP/IP port 443) connections to external servers, then you should be able to run First Advantage Online Testing.

The First Advantage Internet system is Internet-based and can be accessed via browsers within your organization's network, but you cannot install your own copy of First Advantage Internet product on your internal network.

Sometimes wireless networks are less reliable than wired networks due to noise, interference and other factors. Try taking the test again through a wired connection.

First Advantage Internet Testing does not support exam delivery inside hotmail. To start an exam from an eTicket delivered to hotmail, copy the link and paste it into a standalone browser.

No. First Advantage's interactive assessments, the highest-quality, most-accurate performance-based assessments in the world, are all available over the Internet. First Advantage also maintains a broad range of knowledge-based tests in areas such as accounting and computer programming, as well as behavioral assessments that use a variety of question types, including interactive performance-based, multiple choice, matching and other question types.

כObviously any assessment given outside of a proctored environment (like a staffing company's office) opens up the risk of cheating. Most First Advantage customers who use First Advantage's Internet testing solutions for home-based testing also maintain an in-office assessment work-station for re-testing applications as a security backup to ensure that candidates did not use unauthorized help material while being tested at home. If candidates are warned that they may be re-tested when they show up at your office, the risk of cheating is greatly reduced.

Also, certain First Advantage tests are less susceptible to improper use. First Advantage's behavioral testing products, for example, ask job candidates to provide honest answers to questions regarding attitudes and aptitudes. There is no greater risk in giving these types of assessments over the Internet than there is giving them in any other environment.

In addition, First Advantage's TalentScout adaptive tests have additional built-in test security since an adaptive test will vary for each test taker, depending on his or her skill level. Adaptive testing does not eliminate the risk of job candidates using documentation or other unauthorized help material when taking a test at home. It does, however, add a new layer of security to Internet-based performance testing.

Other materials on the effective use of testing over the Internet for staffing services and other organizations is available from Evaluate free of charge here.