The working world of today is constantly changing. As a result of increasing globalization and digitization, professional profiles, areas of responsibility, work content and processes are changing and, as a result, so are the competencies employees must have in order to be successful in a fast-paced working world. For example, the increasing automation of routine tasks is helping analytical and interdisciplinary competencies such as adaptability and willingness to learn become increasingly relevant.
Against this background, it is surprising that assessment practice has changed little in the last 50 years. For example, in the majority of companies, application procedures are still based solely on biography-centered methods (Schuler et al., 2007). Such biography-centered procedures include, for example, the examination of the applicant's career on the basis of curriculum vitae or as part of a(n) (un)structured interview.
These are so-called backward-looking procedures — past performance is considered an indicator of future performance in the workplace (Schuler & Hoft, 2007). And this (so far) has mostly been successful. It is not for nothing that a central axiom of psychological measurement states that past behavior is the best predictor of future behavior (Sutton, 1994; Norman & Conner, 1996).
Definition: aptitude diagnosis
“Aptitude diagnosis is a collective term for psychological selection processes that are used to verify a fit between applicant and workplace.” (Schuler & Hoft 2007)
However, this assumption becomes problematic in a constantly changing world of work. Because if job profiles and work content change continuously, experience in a previous job naturally loses predictive power. Let's imagine, for example, that an innovative start-up in the financial sector is looking for a software developer to join a dynamic project team that uses the latest methods in the software sector. Applicants have already gained several years of professional experience as developers at a well-known banking institution, but worked there exclusively with a single established, but rather outdated, operating system. It may come as little surprise that good performance in the previous job alone will not be a good indicator of career success in the start-up. Rather, it will also be about hiring an employee who can quickly acquire new methods, is adaptable, and brings a high level of problem-solving expertise. All of these are competencies and qualities that can only be derived to a limited extent from a professional career.
The focus is therefore shifting from a very narrow, job-specific aptitude assessment to a wider concept of professional suitability, which, in addition to job-specific competencies, increasingly includes interdisciplinary suitability dimensions, such as adaptability and intelligence. This marks a shift away from a static concept of suitability, which considers only the applicant's current status quo, towards a potential analysis, which also takes into account the applicant's development potential.
Providing the tools for such a potential analysis is increasingly becoming the task of professional psychological assessment. Against this background, so-called construct-oriented methods are gaining in importance. These include psychological testing methods that record stable characteristics, such as general cognitive (e.g. intelligence, ability to concentrate) and non-cognitive abilities (e.g. personality traits, attitudes, interests).
Step 1: Requirements analysis
At the beginning of every professional assessment process stands a well-founded requirements analysis (DIN 33430, 2016). As part of this, criteria for career success are defined. The central question is: what does the applicant have to bring in order to be successful in the respective job? To define such requirement profiles, experience-guided, intuitive methods or person-empirical methods are applied (Schuler & Kanning, 2014).
Experience-guided intuitive methods usually rely on expert judgment, such as the manager's assessment. Based on their experience, they assess which competencies and personality traits are relevant for success in the job. Person-empirical methods, on the other hand, use statistical relationships between the characteristics of workers and their professional performance to derive requirement profiles. If it can be observed, for example, that tax officials with a high level of conscientiousness work particularly error-free and successfully, conscientiousness would be included in the requirement profile. Person-empirical methods have the advantage that they largely rule out distortions due to experts' subjective perceptions and can continuously adapt to changes in the working world. For example, managers tend to identify competencies that they themselves have as relevant, and often notice changes in the required competencies over time only with a delay.
A well-founded requirements analysis that constantly adapts to changes in a fast-paced working world is, however, the basis for the selection and design of suitable test methods. Because in order to be able to assess whether a person is suitable for a specific job or not, HR professionals must know what is really relevant to their job — and this is becoming more difficult than ever in a complex and constantly changing world of work. However, an empirically oriented requirements analysis is still rarely used in practice. As a result, we frequently observe test designs that are out of touch with practice: for example, abilities and characteristics are often tested using psychological tests that are only (still) barely related to professional success in the respective job.
Step 2: Selection and design of the test procedure
If there is agreement on the suitability dimensions to be recorded, the next step is the selection and design of the test procedure. The methods used should always be critically examined with regard to the quality criteria of classical test theory — in particular objectivity, reliability and validity. Here is detailed information on the design of test procedures with key quality criteria in mind.
Step 3: Selection decision
The third step is the evaluation and interpretation of the test results and finally the selection decision: you either decide for or against the applicant. The aim should be to keep the proportion of rejected applicants despite actual suitability (so-called false negatives) and the proportion of accepted applicants despite lack of eligibility (so-called false positives) to a minimum.
Step 4: Evaluation
In practice, the assessment process often ends at this point. This is not in line with the requirements for occupation-related assessments set out in DIN 33430, which call for a comprehensive evaluation of, and critical reflection on, the selection process implemented. While evaluation has become an integral part of other areas, such as production, it unfortunately still rarely takes place in personnel selection. Evaluation offers high potential for uncovering weak points, adapting them and thus gradually optimizing the selection process (Nachtwei & Schermuly, 2009).
If, for example, it turns out that a test process has led to mis-hires, the test can be adjusted accordingly to avoid this in future selection decisions. Such a process not only enables continuous development of the processes, but also awareness of potential changes in applicant requirements (cf. step 1).
Aptitude testing must not be a one-way street
The examples show that, in order to meet the requirements of a constantly changing world of work, assessment practice must not be understood as a one-way street that ends with the selection of an applicant. Rather, we must approach assessment as a continuous improvement process, over the course of which we learn more and more about the connections between the applicant's characteristics and professional success. However, the reality today is often different. For example, validated test methods usually have standardization cycles of more than 15 years, and requirement profiles are only sporadically checked for continued validity.
However, the variety of data generated as part of the selection process provides a wide range of opportunities to optimize personnel selection. With the help of intelligent evaluation and interpretation, combining psychological and statistical expertise, patterns can be identified that are barely comprehensible to people anymore. This is achieved, for example, with the help of self-learning algorithms. This is how modern apps in the assessment space use machine learning methods to continuously improve their test procedures. Instead of developing and standardizing a test at a specific point in time, data on the success of a selection decision is used to adapt requirement profiles, optimize test content, and thus determine the best possible statistical relationship between professional success and the characteristics of the applicant. As a result of this iterative process, the estimation of the applicant's suitability is ultimately improved and the probability of error is reduced. HR managers are less likely to be wrong, and the company benefits from cost savings thanks to a more objective and valid selection of personnel.
Sounds exciting? Find out here what such an interplay between psychological diagnostics and intelligent data analysis, integrated into a single app, could look like.
