Performance-Based Selection versus Profiles
Every company faces the same challenge: How to find and hire employees who perform well, fit the company’s values and culture, and stay. We call them “quality hires.” Other than guess work, the two most popular ways of identifying a quality hire are profile and performance-based selection. While each method has its advocates, they come from contrasting theoretical perspectives, and provide very different end-user information. The objectives of this paper are to provide a brief overview of each approach and review their respective strengths and weaknesses.
Profile-based Selection
Profile-based selection, often referred to as a hiring or employment profile, describes what a composite of current high performing employees “looks like” in terms of knowledge, experience, or other personal characteristics. The details of a hiring profile are in terms of specific type(s) of work history, educational experience(s), knowledge, overt skills and characteristics, as well as other work-related factors common to current high performing employees. In addition, hiring profiles often incorporate one or more personality, cognitive ability, or other “flavor of the month” tests.1
This profile is then used as a benchmark for comparing the profile of a job applicant to the profile of the composite high performing employee. If the candidate matches the performance profile, he or she is eligible for hire. The following figure summarizes the process.
Figure 1: Development Process for Profile-based Selection
The simplicity of the figure speaks to the most alluring aspect of profile based selection: it is crayon-simple to do! It is also economical and quick to do, and makes sense to most managers and employees. For the most part, such hiring profiles do a good job of identifying the characteristics common to high performers within a company.
However, these benefits are counterbalanced by several shortcomings. The first one is that high performers must be designated prior to the development of a profile, placing considerable weight on the ability of managers to accurately identify and assess performance and keep it separate from personal bias or personality issues.
Biases such as political issues among supervisors, departmental loyalties, buddies choosing buddies and people nominating people with whom they are most comfortable all affect manager nominations of “high performing” employees. These biases are always problematic, but can also be highly counter-productive to a company that is trying to increase employee diversity or move to an even higher level of performance. In addition, it is highly conceivable that employees chosen as “high performers” may in fact be just average or poor performers due to the restricted make-up of a company’s workforce. In short, existing high performers create an artificial ceiling on performance and unseen barriers to the entry of quality hires that are systematically different from the existing performers.
The net of these problems is that a company that wants to raise performance standards, change the definition of effective performance, or select employees on a more inclusive basis, will not be able to do so using a profile alone. Instead of selecting a quality hire, you may very well select an adequate hire. This is because of the most significant assumption of profile based selection: there is only one profile.
As a consequence, a company will, at best, hire employees like it already has. There are several problems with this approach. First, the employees profiled have typically been hired without a rigorous selection process in the first place and may not be at all representative of the potential for successful employees. Second, most people cannot accurately identify performers beyond the five to ten percent of employees who are at either extreme of performance. That is, subjectively ranking employees often results in the extreme high or low rankings remaining stable while other rankings tend to shift around significantly (Cascio, 1998). Third, single measures of general performance are typically used, failing to reflect the specificity and diversity of performance that is actually required on the job. Further, performance rating errors (e.g., inadequate discrimination errors; general impression errors) may be magnified.
The implication, of course, is that many candidates who would actually be successful if hired will not conform to the current profile. In fact, our own experience has been that, when asked, managers typically indicate that as much as eighty percent of their current employees perform well enough to be rehired. In a tough labor market, this “false negative” error can seriously hamper a company’s ability to staff itself.
Last is the issue of false certainty. When a measure such as a personality profile or a cognitive ability test is given to the “best performers” and then used as a basis for selecting all future employees, there is the assumption that the employees nominated as the best score on the measure in a way that is systematically different than employees who were not nominated. In our experience, this simply is not true.
For example, we use a personality measure that has twenty-five dimensions and frequently score all twenty-five of them when conducting a validation study. It is rare to find more than three or four scales that differentiate between high and low performers. Others have found even less promising results. For example, only one of sixteen personality factors predicted the success of pilots in a recent study conducted by Bartram (1995). This suggests that using multiple dimension profiles (as is common) may result in people being hired who are not quality hires; that is, perform, fit and stay at a rate superior to candidates that do not fit the hiring profile.
Performance-based Selection
Performance-based selection is often referred to as criterion-based selection or performance-based testing. It is similar to profiling in that the aim is to identify the characteristics of successful employees that can then be applied to job candidates, but is different in one major respect: While profile-based selection focuses on the characteristics of the job incumbent, performance based selection focuses on actual job behaviors and success (e.g., how effective performance, fit, and retention are defined) without regard to the personal characteristics of current job incumbents. The process of developing a performance-based selection test is shown in Figure 2.
Figure 2: Development Process for Performance-based Selection
The process starts with the identification of what a quality hire does, rather than what he or she looks like. At this step, job incumbents are not included in the process. Rather, groups of job experts, results from job analysis, and the needs of the company are mined for the actual definition of performance, fit and retention.
We sometimes start the process by having job experts consider generic performance dimensions such as the sample shown in Table 1 on the next page. In conjunction with job descriptions and knowledge, the job experts revise the performance dimensions to fit their definition of successful performance. In sum, a performance-based selection test is a research-based tool used to identify the knowledge, skills, abilities and other characteristics (KSAOs) that are important for success in a particular job. This test assesses candidates on those KSAOs that have been found to be related to actual performance on the job. There are five general steps to developing an effective performance-based test:
- Identify the specific knowledge, skills, ability and other characteristics (KSAOs) of job success.
- Develop test(s) that measures the success-related KSAOs.
- Test a group of job incumbents while their supervisors simultaneously evaluate each employee on the dimensions of job success identified.2
- Statistically link test results to performance measures.
- Use the tests or scales that are related to job success to develop a final employee selection test battery.
As a result of this process, only performance predictors that are representative of real, important, and measurable criteria of job success are identified. Performance-based selection, in contrast to profile-based selection, lets you know what someone can actually do. In addition, because the measures of performance are created independent of job incumbents, they can include aspects of performance that are currently important as well as aspects that will be important to the company in the future. This solves one of the challenges inherent in hiring profiles; that is, they are not suited for inclusion of future-oriented performance requirements. Although the steps in the performance-based selection development process may appear more cumbersome as compared to profile-based selection, the end result is a hiring process that contains the critical elements for success on the job. In short, the efficiency is realized in the final product.
Table 1: Sample Performance Dimensions
As a result of this process, only performance predictors that are representative of real, important, and measurable criteria of job success are identified.
Performance-based selection, in contrast to profile-based selection, lets you know what someone can actually do. In addition, because the measures of performance are created independent of job incumbents, they can include aspects of performance that are currently important as well as aspects that will be important to the company in the future. This solves one of the challenges inherent in hiring profiles; that is, they are not suited for inclusion of future-oriented performance requirements.
Although the steps in the performance-based selection development process may appear more cumbersome as compared to profile-based selection, the end result is a hiring process that contains the critical elements for success on the job. In short, the efficiency is realized in the final product.
Further, any indicator that is not strongly related to job success— performance, fit, and retention—is not included in a performance-based selection system. Performance-based tests rely on concrete statistical evidence. The steps included in the process of creating a performance-based test ensures a relationship with specific job outcomes and behaviors (e.g., managing costs, working well with others, staying for an extended periods of time). Equally as important, a selection system designed using a performance-based approach is legally defensible and will assist a company in taking steps to increase diversity among its employees.
The performance-based selection process does not require managers to nominate a group of “high performers”; the measures of job success do the nominating. That is, on-the-job performance is linked to performance on a particular test. In addition, KSAOs that are not specifically related to job performance are not included in the final selection system.
Which Hiring Method Offers the Real Hiring Advantage?
Although the goal of both approaches is the same, performance-based selection results in more effective and fair selection of quality hires. In addition, performance-based hiring tools are more easily defended against charges of discrimination and can be used to raise the bar on performance and increase diversity. The bottom-line is that a candidate who performs well within a performance-based selection process is far more likely to be successful on the job, now and in the future. Still, many practitioners today assume that selecting new employees based on characteristics of high performers alone is an acceptable, if not advantageous, approach. The fact is that doing so is simply bad practice. As far back as the 1960’s, selection experts have strongly cautioned against the use of such profiles in the hiring process (Blum & Naylor, 1968; Kirchner, 1965). A main reason for such caution is that in order for a selection system to be effective, it is critical that its tools offer the ability to discriminate among varying levels of job performance (Hansen & Conrad, 1991). In large part, person-based profiles fall far short in this regard. In fact, it is often impossible to determine what components (if any) of a profile are actually related to job performance. This is a problem ethically, professionally, practically, and legally.
Profile-based selection offers a general profile of job performance based on the characteristics found in subjectively identified high performing job incumbents. One reason for the popularity of these hiring profiles is their apparent common sense: “If this is what our performers look like, then new employees that look like them should also be performers.” The reality is that this makes no sense at all and can hurt the ability of a company to make quality hires and offer opportunity to people who would not otherwise have been considered for employment. While such profiles can serve as useful prescreening tools, or as a complement to performance-based selection, detailed, job specific behaviors are not typically included in the development process. Again, this leaves open the possibility that factors not related to actual job performance are being used to select employees. If hiring decisions are in fact being made on the basis of non-performance related information and this information contributes to adverse impact, legal defensibility is highly improbable. It is our recommendation that profiles may be used in conjunction with, but not in place of performance-based tools. Table 2 offers a short checklist of the primary advantages of each hiring approach. As the table shows, performance-based selection offers real hiring advantages in terms of actual job performance and long-term organizational needs. Profile-based hiring’s advantages are minimal in terms of real organizational impact. The choice is simple. If you want real advantages in hiring, use performance-based selection.
Table 2: Comparing Profile vs. Performance-based Selection
The bottom-line is this: In the realm of the human resource professional, there is no room for procedures, regardless of how appealing they may be, that detract from the contribution of the profession to the quality of a company’s workforce or that have the effect of reducing rather than increasing employment opportunity. For these reasons, we strongly recommend performance-based selection.