Online delivery of high quality assessments
By Stephen J. Conway and Christopher Paterson

There has been a considerable amount of discussion in recent months concerning the feasibility and efficacy of psychometric assessments being delivered on the Internet. With the increasing importance of online recruiting -- some 18 million CVs are expected to be on the net by 2002 with an estimated 70% of UK companies making use of online recruiting sites -- it seems logical to assume that online delivery of assessments will follow. Time, distance, efficiency and cost factors all argue in favor of this trend. For example, the head of KPMG's online graduate recruitment effort estimates that the firm will save hundreds of thousands of Pounds and reduce administrative time by 40% over the next few years.

And yet, many HR professionals have questioned the quality and reliability of online assessments and have expressed concerns over a lack of standards. We would agree that the presence of cheap online assessments from unknown or unproven sources is not consistent with best practice. However, the solution that some commentators have suggested - avoiding online assessments entirely - is tantamount to throwing the baby out with the bath water.

A thorough consideration of the issue leads to a very different conclusion. In the first place, the Internet should not be viewed as an end in itself, but rather as a means to an end. In the case of online assessments, that end should be to accurately and efficiently place the "right" people in the "right" jobs. With psychometric assessments, as with all other software or web-based products and services, the old axiom of "garbage in, garbage out" pertains.

The many benefits of Internet delivery of products and services need not be laboured over here. In considering any psychometric assessment, three major factors should be analysed: reliability, validity and appropriateness. By reliability we mean that an assessment should produce consistent results over time, i.e., if you take an assessment today and take the same assessment a year from now, the results should not vary to any great extent. Validity refers to the accuracy with which a given assessment measures what it purports to measure, i.e., if it purports to measure assertiveness, it does so. Appropriateness means that the instrument selected measures aptitudes, interests and personality/behavioural traits that are consistent with the requirements of a given job.

Publishers of psychometric assessments should be willing and able to provide users with a technical manual that details how the assessment was constructed and validated, with reliability and validity coefficients included. A recent study published by the U.S. Department of Labor ("Tests and Assessments: An Employer's Guide to Good Practices") states that effective psychometric assessments should have a reliability coefficient of at least 0.70 and a validity coefficient of at least 0.35. The study concludes that assessments not meeting these minimum standards are unlikely to produce the desired results.

Appropriateness is achieved in two ways. First, the assessment should be constructed so as to eliminate gender, age, racial, ethnic and cultural biases. Rigorous validation studies, including adverse impact studies based on the 4/5ths Rule, will accomplish this and should be documented in the technical manual. Second, today's most advanced assessments permit customised benchmarks to be created for each company and each position. An individual candidate's results can then be matched against the specific benchmark to determine if there is a good job fit or not.

Having identified assessments that are reliable and valid and that are appropriate for the particular job, it then remains to determine the best method of delivering and administering the assessment. Early generation assessments were strictly paper and pencil affairs. More recently, many assessments have been computerised and are available on-screen with extensive reports being generated by the software. The logical next step in this evolution is delivery of assessments online. Obviously, having assessments online provides 24/7 accessibility as well as rapid scoring and transmission of results. State-of-the-art online assessments go even further -- they provide a complete web-enabled system that permits clients to register candidates, create customised job patterns, administer the assessments, score them instantaneously, and maintain records, including tracking all candidates. Such systems are user-friendly and business oriented, providing reports that do not require special interpretation or analysis.

Online delivery of assessments is cost efficient and saves countless hours of administrative time, thus freeing up valuable HR resources to be concentrated on finding and hiring the right people for the job. This is true for both large and small organisations. It is our view that the trend toward online delivery of psychometric assessments is here to stay and that it will grow as more and more HR and Recruitment professionals recognise the benefits. Having said that, we would also remind readers that the use of assessments (online or otherwise) should be viewed as only one part of the hiring process. Interviews, background checks and the like will continue to be important; however, the use of quality online assessments can greatly streamline the entire recruitment process.


Stephen J. Conway is Managing Director of
Christopher Paterson is Managing Director of IPS