Chapter 6 Selection

The Selection Process

Selection is the process of analyzing applicants to identify the candidates who best meet the requirements of an open position. “Best” often refers to how closely a candidate’s knowledge, skills, abilities, and other characteristics align with the job specification for the position. A good selection process should result in workers who are successful in their jobs and stay with the organization over time. Figure 6.1 outlines the main steps in the selection process and the key questions to be answered at each step. Sometimes these steps will be taken consecutively and sometimes concurrently, depending on the nature of the job, timelines, and chosen selection strategy. The order of the steps can also vary.

Whether they know it or not, organizations have a hiring strategy for each position. Making that strategy explicit allows HR practitioners to ensure that it makes sense and is compliant with the law. An explicit strategy also allows HR practitioners to evaluate the effectiveness of the selection techniques used and to make improvements. In this chapter, we outline each step in the selection process, including the key considerations, the tools available, and when to use them. We conclude the chapter by offering some ways to evaluate the effectiveness of the selection process.

Relevance, Reliability, and Validity in the Selection Process

At its core, the selection process is about gathering information on a candidate, evaluating and analyzing that information, and using the results to predict whether the candidate will succeed in the position in order to inform the hiring decision. Information gathering begins during recruitment, when an organization decides what preliminary information to require candidates to submit. The selection process provides the opportunity to collect more information from a smaller number of candidates as the pool is narrowed. Interviews, work samples, tests, and reference checks are all common methods for gathering additional information during the selection process.

A key question is which tools provide an organization with the information that it needs to select a candidate to hire? In deciding what information to collect and consider, it is important to know whether or not it is relevant to the hiring decision. That is, will the information help you to predict whether a candidate will be successful if hired? It is also important to know whether or not the information is accurate. The concepts of validity and reliability can be helpful when considering what information to collect on candidates and use during selection.

As discussed in Chapter 3, reliability is the degree to which the information that HR practitioners collect is consistent over time. For example, if they administer a test, then it should produce a consistent result each time it is administered. If they test the same person twice and get different results, then the test might not be a good measure to use. The desire for reliable information can affect how they collect information on candidates. For example, if they are interested in knowing about someone’s educational attainment, they might get a more reliable answer by asking for transcripts than they would by relying on a candidate’s resumé or statements in an interview. A transcript will provide the same answer each time they look at it.

Measures must also be valid (i.e., measure what they claim to measure). For example, if HR practitioners are interested in knowing how fast a potential cashier can work, they might time how quickly the candidate can scan and bag groceries during a simulation. This test is only valid, however, if the candidate’s actual performance matches the test result. If candidates consistently perform better on the test than they do on the job (perhaps because there are no customer distractions during the test), then the simulation might not be a valid test.

Validity takes three forms relevant to selection.

• • Criterion validity is the degree to which a measure’s results are correlated to actual performance of some aspect of the job. A skill test, for example, has criterion validity if candidates who score higher on the test also perform better on the job.
• • Content validity is the degree to which the content of a measure matches the content of the real-life situation. In other words, does the tool evaluate relevant aspects of the job in question? Is it missing pieces (in the scanning and bagging example above, the impact of customers was missing), or does it include content not present in real life?
• • Construct validity is the degree to which the tool measures something real and relevant. What is being measured must be real. Sometimes this is self-evident, such as measuring typing speed. Other times it is not. For example, an HR practitioner might want to measure a candidate’s honesty. But how does one measure a concept as complex as honesty? What is being measured must also be relevant to the job. Examining the job specification can be helpful in determining the relevance of the information that an HR practitioner is considering gathering.

To be useful, the information collected and used in the selection process must be valid, reliable, and relevant. If the information is not valid and reliable, then it is unlikely to be useful in predicting a candidate’s performance and thus of little use in making a hiring decision. Complicating matters, some jobs lend themselves to more valid, reliable, and relevant measurement than others. It can be easier to ascertain which selection criteria are relevant in routine jobs (e.g., working on an assembly line) than it is for jobs in which the duties are highly variable and/or rely on workers who exercise significant discretion in completing tasks (e.g., residential plumbing repairs or customer service). For this reason, it is important to develop selection processes appropriate for each job or position.

Candidate Screening

The first step in the selection process is candidate screening. During this process, each application is analyzed with the goal of removing from consideration candidates who do not meet the minimum requirements and are unlikely to be good fits. The end result should be a manageable number of high-quality candidates who can be more thoroughly vetted. Screening can require reviewing the applications several times. If there are many applicants, then an HR practitioner might first review each application to determine whether the applicant meets the minimum requirements for the job set out in the job specification. Often the minimum requirements are objective and quantifiable and related to education, years of experience, professional certification, and so on. The need for clear criteria should shape the job posting (see Chapter 5). For example, if education is an important factor, then stating a minimum requirement of “a university undergraduate degree in X field” is preferable to a more general statement such as “possessing postsecondary education.” The former creates quick and clear demarcation during selection, whereas the latter leaves too much room for judgment. If the candidate screening criteria are vague, then that might suggest the job specification (Chapter 3) requires additional work.

Once candidates who do not meet the minimum job specifications are removed from the applicant pool, it might be necessary to conduct a second round of evaluation to identify which of the minimally acceptable candidates will proceed to the interview stage and/or the testing stage. Having a clear numerical target during this stage is helpful (e.g., we want to interview a maximum of four candidates and hire one). This target can be informed by past experience with hiring as well as the costs and benefits of expanding this short list of candidates to evaluate further.

This second phase of screening can be more difficult because all candidates are minimally qualified. During a second round of screening, the criteria used should be closely linked to the job specification so that the decisions are anchored in the actual needs of the job. It is important not to evaluate the applicants in an informal, “gut feel” way. Doing so can introduce an array of issues regarding bias and undermine validity and reliability. A better approach is to establish a matrix to allow for a multi-variable evaluation. An evaluation matrix could involve a grid in which each criterion that a candidate meets is checked.

Table 6.1 shows an example of an evaluation matrix based on the job posting for a labourer presented in Feature Box 5.2. Since this is a second-level screening, the organization knows that all applicants have met the basic qualifications. This screening assesses whether or not each applicant has experience with each of the six main job duties. A matrix is useful because a visual representation of each candidate’s match with the screening criteria can create a shared understanding of the candidate’s qualifications and facilitate decision making. In this case, Table 6.1 suggests that Candidates C (four of six) and D (five of six) seem to have the widest range of experience with the job duties.

A matrix can be made more complex when warranted. For example, Table 6.2 adapts the matrix presented in Table 6.1 to allow each reviewer to assign applicants points for each duty based on the applicant’s level of experience. This approach creates a slightly more nuanced assessment of each applicant. When comparing Tables 6.1 and 6.2, for example, note that Candidate D is the highest rated in both grids. But the second highest rated in Table 6.1 is Candidate C (who has the broadest level of experience), whereas in Table 6.2 Candidate A is rated higher (because of much more experience).

Neither matrix is better than the other; each just measures different aspects (breadth versus depth) of the candidates’ experience. Which one you choose will be determined by which aspect is most important. It is also possible to add complexity by weighting each column in Table 6.2 differently. For example, if having long experience using small machinery is highly desirable, then you might multiply whatever score candidates get on this column by two. This would shift the scores in Table 6.2 such that Candidates A and C tie with eight points each.

Scoring: 0 = no experience; 1 = <1 year of experience; 2 = 1–3 years of experience; 3 = >3 or more years of experience.

Ideally, the screener should have no awareness of the sex, ethnoracial background, age, or other characteristics of the candidate. This is because research has consistently shown that unconscious bias among screeners affects their choices. For example, resumés with “white”-sounding names are more likely to get candidates interviews than identical resumés with “non-white”-sounding names.2 A similar effect is found regarding sex,3 and, as an intersectional analysis would predict, women whose names suggest that they are persons of colour experience a compounding bias.4 Blind screening controls for bias and ensures that screening decisions better reflect candidates’ abilities.5 A different approach to reduce bias in screening is to adopt identity conscious screening, in which the screener actively searches for candidates from marginalized groups to select for interviews. This practice is controversial but can be effective in increasing diversity in the workplace.

Some organizations use technology to automate all or part of the screening process. The software, often called applicant tracking systems (ATS), searches for key words, phrases, or characteristics identified by the organization to filter the candidate pool. This technology improves the efficiency of screening by evaluating hundreds of resumés in seconds but also comes with trade-offs. Clever applicants can fool the ATS by tailoring their resumés to “match” the key aspects for which the organization might be searching. The internet is full of websites offering tips on how to “beat” ATS. Also, ATS can entrench existing hiring biases. The person programming the system establishes the criteria and keywords that it will search for and therefore can embed biases into the program against certain social characteristics. For example, in Chapter 5, we discussed Amazon’s artificial intelligence recruiting system and how its algorithm taught itself to discriminate against women.

A final, but less common, tool that can be used for screening, especially at the second step, is a screening interview. It allows the screener to gather information on the candidate that might not be available on the resumé or application form. Screening interviews are typically short, often take place by telephone or video conference, and focus on specific qualifications or qualities needed for the job. The advantage of screening interviews is the richness of information that they can provide in making a difficult screening decision. They are a resource-intensive way, however, to gather information, requiring logistical arrangements and personal engagement. Also, they risk introducing bias into the screening process. They might be most appropriate when the candidate pool is relatively small, the job requires KSAs that can be assessed easily through a conversation, or the job is at a high level in the organization (thus warranting the extra cost).

Interviews

Once the applicant pool has been winnowed to the desired size through screening, organizations can begin to collect more detailed data on short-listed candidates to make a hiring decision. By far, interviews are the most common technique as well as one of the most maligned. An interview consists of a series of questions posed to a candidate to draw out information on the suitability of the candidate for the job. Candidates feel a great deal of pressure in interviews since they know that their answers are being closely evaluated. They also complain about strange questions, such as “if you were an animal, what kind of animal would you be?” On the employer side, interviewers can struggle to discern an authentic answer from someone prepared to tell the interviewer what the candidate believes the interviewer wants to hear. Some people perform better in an interview situation than others. Research shows that poorly designed interviews end up being both invalid and unreliable in predicting employment outcomes.6 Subjectivity and personal bias easily creep into decisions based on interviews, including biases related to ethnoracial background, sex, disability, and other characteristics. There is no way to “blind” interviews.

So why bother with them? Because, when well designed, interviews can provide a richness of information that no other selection tool can. Interviews provide insights into the candidate as a person and allow for an evaluation of skills and abilities that goes beyond keywords and formal qualifications. As a result, interviews are an excellent tool for evaluating how well a candidate will fit into the existing organizational culture. Interviews are also the only moment when a candidate can ask questions of the employer, which can also enhance the likelihood of finding a good fit (see Feature Box 6.1). Plus, people have a legitimate need to “see” the person who might be joining their organization. As imperfect as they might be, interviews satisfy many selection needs. But interviews need to be designed and executed carefully to address their inherent weaknesses.

Interviews can take many forms, from an impromptu session with a manager when a candidate drops off a resumé to formal, structured proceedings with set questions and a panel of interviewers. The nature of the job, the organization, and the resources available for the selection process help to shape the type of interview that is appropriate. There are four decisions to make when setting up an interview process.

1. 1. Medium: Most interviews are conducted in-person. If resources are scarce and/or candidates are out of town, then a telephone interview might be conducted. It deprives the interviewer(s) of visual cues, such as body language, but it can blind the candidate’s ethnoracial background, disability, and potentially gender. Another alternative is an online interview through videoconferencing software.
2. 2. Formality: Unstructured interviews do not have an established set of questions. The interviewer(s) can craft questions in response to the flow of the interview. Usually, evaluation criteria are similarly informal, with interviewers going with their overall impressions. On the other end of the spectrum are structured interviews, which have pre-established and standardized questions and evaluation standards. In between are semi-structured interviews, which have standardized questions but allow room for interviewers to respond to information that comes out in the interview.
3. 3. Format: Who participates in an interview and when can vary. One-on-one interviews consist of one interviewer and one candidate. Panel interviews consist of multiple interviewers who interview one candidate. A sequential interview process has the interviewee move from one interviewer to another in sequence, usually answering different questions at each stage.
4. 4. Content: Three types of questions can be asked. Knowledge and background questions probe factual information about candidates and their skills and knowledge. These questions can supplement information provided in the application, or they can be designed to confirm certain information. Situational questions pose scenarios that candidates might be confronted with in the position and ask for their responses to the situation. Behavioural questions look backward, asking candidates to reflect on an experience in the past, talk about their actions, and possibly indicate how they might act differently today. Interviews often have a mixture of each type of question, depending on the nature of the information sought.

Decisions on the characteristics of an interview depend on the nature of the position being filled and the organizational resources available. In general, the research has found that unstructured interviews are more likely to have issues with subjectivity and bias.7 During such interviews, the interviewer is more likely to evaluate candidates based on overall impressions. Research has shown that interviewers are more likely to develop positive impressions about candidates similar to them,8 which can perpetuate existing inequalities in a workplace as well as increase the risk of a human rights complaint.

Nevertheless, unstructured and one-on-one interviews have their place. If there is a need for a quick decision and the job permits a straightforward evaluation of fit (e.g., hiring for a temporary labourer job), then an unstructured, one-on-one interview can be worthwhile. Bias in such an interview can be reduced through a consistent set of questions and the use of rigorous scoring criteria. Having more than one interviewer, preferably with different perspectives, to provide a range of viewpoints can also reduce selection bias. As for the types of questions to use, behavioural and situational questions have similar levels of validity and reliability.9

HR researchers often point to an overall lack of predictability for interviews, citing people’s flawed ability to evaluate other people.10 These researchers prefer selection methods that remove human subjectivity and fallibility from the calculation. Yet interviews remain the most popular form of selection process. Research often overlooks the human aspect of selection. People have a need to know and understand the people with whom they will be working. Individuals tend to trust their own judgment, even if it is flawed, over that of computers or test results. One strength of interviews is that they are opportunities to see the whole person and not just test scores. Because of human bias, seeing the whole person can create as many problems as it solves, but there are ways to structure interviews to minimize the impact of that bias. Interviews will likely continue to be the most popular form of selection tool for a long time to come.

Interview Mistakes and Training

Mistakes are easy to make during an interview. Asking the wrong question or incorrectly evaluating an answer can have consequences for the success of the selection process. Table 6.3 examines some of the most common interview mistakes, regardless of which type of interview is selected.

Although there is no way to eliminate fully biases, mistakes, and subconscious tendencies in interviews, their impacts can be reduced with appropriate training. Training interviewers should be a part of the selection planning process. Interviewing is a skill and, as such, can be taught. Training can include three aspects of interviewing.

• • Interviewers need to know how to ask questions in a neutral manner as well as how to evaluate the answers and apply a scoring guide (see “Making the Decision” below). Some tips in effective note taking can also be of assistance.
• • Training should address interview management, including making the applicant feel at ease, remaining in control of the interview, and engaging in good listening practices.
• • Most importantly, training should address issues of bias and non-verbal cues. Training people to spot their biases is a difficult task and can require repeated efforts. That said, the only way to prevent bias from interfering with the selection process is to make interviewers conscious of their biases and how to overcome them.

Selection training can include opportunities for simulation, role playing, or job shadowing to make the interviewer comfortable with the process. Training should also include basic knowledge of the position to be filled and its job specification so that the interviewer can identify candidate responses relevant to the position. Interviewers should also be trained on the evaluation scheme.

When they have discretion in the questions that they ask, it is imperative that interviewer training includes discussion of interview questions to avoid, for practical and human rights reasons. Vague questions provide little useful information on which to base a selection decision. An example of a vague question is “what is your biggest weakness?” It is not connected to the job specification, does not speak to identifiable qualities of the candidate, and creates too much room for providing an inaccurate answer. Of course, most candidates will say that their major flaw is that they are perfectionists or work too hard. Metaphorical or trick questions are more popular recently but no less problematic. Asking a candidate what kind of animal they would be might seem to be creative but does not elicit a response that can be meaningfully evaluated (e.g., how do we know whether a “lion” or a “bear” is better suited for a position?).

Preventing litigation on human rights issues is another reason to teach interviewees about which questions to avoid. Certain questions, even innocently framed ones, can lead to accusations of discrimination and concerns about breaches of privacy (see Chapter 2). All interview questions should be related directly to the position, specifically the job specification. Even when a question is related directly to a job requirement, it is important to be careful how it is framed. Table 6.4 offers some examples of questions about legitimate job requirements that can still run afoul of human rights and some alternatives. It is useful to note that the “bad questions” outlined in Table 6.4 also fail to extract the information needed to evaluate a candidate properly. In contrast, the “good questions” avoid possible litigation and extract the information required for the selection process.

Training should also address the potential for human rights issues to emerge from candidate disclosures. For example, a candidate might ask about on-site child care or flexible hours and, in doing so, disclose that they have four young children. The candidate’s family status is of no relevance to the selection process. Having this information poses a small organizational risk: if the candidate is not hired, then they might attribute the decision to their family status. The best option here is for an interviewer to answer the specific question and move on. No record of the question should be made so that it does not influence subsequent decision making.

A final component of training is to teach and remind interviewers that, during the interview (including the informal “chit-chat” as it gets under way), the interviewer is in a position of power over the candidate. The candidate needs the job and is competing against others for it. The interviewer will decide who gets the job. It is more imperative for the candidate to react and respond to the perceived needs of the interviewer than vice versa. This creates a power dynamic that must be openly recognized and managed by interviewers. Off-hand comments, even a compliment about a tie or shoes, can be interpreted by the candidate as subtle messages about how they are perceived. The power dynamic can turn an innocent remark (e.g., “nice blouse you have on”) into a subtle power move, reminding the candidate that the interviewer is the one doing the evaluating. It can also make the candidate feel uncomfortable or unsafe.

Applicant Testing

Some employers use tests to measure aspects of a candidate’s skills, aptitudes, and personality. Like interviews, the purpose of testing is to elicit information to inform the selection decision. Tests are often asserted to provide more objective measures of candidates than interviews because they remove human bias. Tests are often (but not always) quantitative, making it easier to calculate their validity and reliability. Testing can be expensive and resource intensive.

In broad terms, there are two kinds of tests used during the selection process: demonstration and predictive. Demonstration tests measure a candidate’s knowledge, skills, or ability to do tasks by putting the candidate through a series of tasks, scenarios, or questions that can reflect the knowledge or skills required to perform a job. They are demonstrative because they ask the candidate either to demonstrate what they know about the job currently or to perform some aspect of what the job will entail. There are a few kinds of demonstration tests.

• • Work knowledge tests pose questions or scenarios to the candidate that relate to the job. These questions or scenarios might focus on technical aspects of the work or facts that someone in the job would need to know. The key is to ensure that the questions are linked directly to the job specification and the work that needs to be performed. The candidate can be evaluated based on how well they answer the questions.
• • Work sample/simulation tests ask the candidate to engage in a part of the work. The candidate might be asked to perform a series of duties related to the work or given a simulated situation in which they need to take action. For example, a candidate for a communications job might be asked to write a media release and a newsletter article within a certain time frame. A bus or truck driver might be asked to navigate a route and successfully park the vehicle. Simulation tests can involve live simulations involving other workers, such as a mock OHS accident investigation.
• • Physical ability tests require the candidate to perform a series of physical tasks or undergo physical tests to determine their physical capacity to perform the duties of the position. They might also involve sensory or perception tests. As shown in Chapter 2, though, employers need to be careful to design physical ability tests to test capabilities that are real proxies for the position or BFORs.

The value of demonstration tests is that they provide useful information about a candidate’s ability to perform the duties of a position and allow an organization to compare the performances of several candidates. The position-specific nature of demonstration tests means that they must be designed for the job. This can make such tests resource intensive to conduct. Simulations require coordinating many people, which is logistically challenging and requires much advance planning. The creation of demonstration tests also requires expertise by the tester in both the job and the test design.

Predictive tests measure aspects of the candidate’s personality or preferences. These tests are not directly related to the duties of the position. Rather, they attempt to measure qualities of the candidate that an organization believes predict their success in the position. Predictive tests can take several forms.

• • Cognitive ability tests measure candidates’ mental capabilities. General cognitive tests, commonly referred to as IQ tests, measure verbal or written comprehension, quantitative ability, and reasoning or logical ability. There are also specific cognitive ability tests that measure only those aspects of cognitive ability relevant to the job. Many cognitive ability tests are available, and they can vary significantly in content. They share the goal of creating a quantified measure of a person’s cognitive capacity.
• • Personality tests examine elements of a person’s basic traits, such as introversion or extroversion, agreeableness, conscientiousness, openness to new experiences, and emotional stability. A key assumption of such tests is that certain personality traits are more likely to lead to success in certain positions. Similarly, interest inventories seek to measure how well the candidate’s personal interests align with the job requirements. Like cognitive tests, myriad personality tests are available commercially.
• • Integrity tests, sometimes called honesty tests, assess a candidate’s predilection toward honesty. Questions focus on scenarios related to breaking rules, stealing, or lying. Integrity tests can be stand alone and overt, or they can be embedded in personality test questions.
• • Medical exams involve a physical examination and possibly other health tests (e.g., conducted on blood or urine samples) to establish a candidate’s ability to meet the physical requirements of the job and to identify any problematic health-related issues (e.g., drug use, uncontrolled diabetes). The invasive nature of medical exams suggests that they should be conducted only near the end of the selection process and only on the leading candidate. Drug tests, a popular but controversial form of medical exams, are discussed in Feature Box 6.2.

None of these predictive tests measures anything directly connected to a position, so it is reasonable to ask how they help an organization to make a better selection decision. The utility of these tests flows from their track records (as measured by researchers) in predicting future job success (usually defined as a worker who remains with a new employer for a period of time or how well the worker performs the job). If studies find a statistically significant correlation between these tests and increased success, then the tests are deemed useful. It is worth noting that, at times, the effects found can be fairly small.13

One of the challenges of evaluating the predictive power of these tests is that they can only show statistical correlation (i.e., A and B occur together) and not causation (i.e., A causes B). When studying human behaviour in a real environment, it is not possible to control all of the variables, so HR practitioners can never be certain what caused the success. Furthermore, studies of predictive tests need to create a simple and measurable definition of success. What a successful selection looks like will differ from organization to organization and from occupation to occupation. This caveat is not meant to dismiss the validity of research on these tests but to remind us that HR practitioners need to look at a range of tools and evaluate, based on the evidence of their efficacy, which ones work best for their organization and its goals.

Some also argue that these tests are problematic for a different set of reasons. They are invasive since they ask personal questions and make judgments about a worker’s intelligence, personality, and tendencies. Workers might argue that employers should not be entitled to know this kind of information about them because it is beyond the scope of the employment relationship. This argument suggests that the instrumental value of the tests in marginally improving selection success is not worth the violation of workers’ privacy and dignity.

These types of tests also need to be understood in the context of power dynamics in the employment relationship. Employers are not just looking for workers who can competently perform the job. They are also looking for workers easy to manage. Personality and integrity tests can also be used to select workers who will be more compliant and cooperative with management and to weed out workers who might resist authority, agitate for a union, or stand up for their rights. For example, an employer might test for agreeableness (i.e., trusting, cooperative, collaborative). Employers desire this trait in workers because such workers will likely integrate more easily into a team. They are also less likely to challenge authority or criticize an employer’s decisions.14

Predictive tests are usually utilized as a supplement to other selection tools, including interviews. They can provide useful additional information not accessible through other tools. This explains their growing popularity. Nevertheless, the shortcomings and problematic use of these tests should be considered when deciding whether they are a practical addition to the selection process.

Reference and Qualification Checks

One of the last steps in the selection process is to confirm the qualifications of the candidate and to check with people who have experience with the candidate. For some jobs, a criminal record check might also be needed. This step is usually performed only once the preferred candidate is identified or when only a couple of candidates remain in contention because it is resource intensive. Each check serves a different purpose.

• • Qualification checks involve confirming that candidates possess the necessary registrations, certifications, credentials, or qualifications that they claim to have. It can involve contacting schools, professional regulatory bodies, or other relevant bodies that can confirm the information provided by the candidates. Qualification checks are part of an organization’s due diligence to ensure that its workers are indeed qualified to perform their jobs. Qualification checks are also an honesty test.
• • Reference checks involve contacting individuals to solicit their opinions of candidates and their suitability for a job. Since references are selected by a candidate, checks need to be conducted carefully. The goal is to move beyond general platitudes and solicit information related to the demands of the position and the workplace context that the candidate will be entering. Questions should address situations that the candidates might experience in the new workplace (e.g., how do you think the candidate will respond to an open space workplace?) or specific elements of the job (e.g., this is a fast-paced job; how well can the candidate think on their feet?). It can be useful to ask references to provide examples of how the candidate handled specific situations or performed specific tasks in the past. Candidates can also provide letters of reference. They are typically less useful than a telephone call because they do not provide any opportunity to ask job-related questions. The research indicates that reference checks should be used to verify findings from other parts of the process and not to predict how well a candidate will perform in the job.15 Some groups of workers, such as those new to Canada, might have difficulty providing references. This requirement can then discriminate against them.
• • Criminal record checks will reveal whether a candidate has been charged with or convicted of a criminal offence. Successfully passing a criminal record check (or a child welfare check) might be a bona fide occupational requirement for certain positions (e.g., working with a vulnerable population). Usually, the candidate needs to apply for the criminal record check after being hired and then provides the result to the employer before beginning work.

Before conducting a reference or qualification check, it is important to seek consent from the candidate to make the necessary contacts. The checks should also be conducted in accordance with relevant privacy legislation to ensure that private information is handled appropriately (see Chapter 2). Despite the growing practice of checks, organizations should refrain from checking applicants’ social media accounts, as explained in Feature Box 6.3.

Making the Decision

After gathering a range of information on candidates, an organization must decide whom to hire. A few candidates might be eliminated during the selection process as unsuitable. Usually, though, an organization will have two or more candidates among whom to choose. A decision strategy can help those charged with selection to weigh the information and determine the best candidate(s). Two broad decision-making strategies are commonly used.

Many organizations adopt an intuitive strategy. Those making the selection will look at each of the minimally qualified candidates and balance several factors.

1. 1. Organizational needs: There might be pressing organizational needs that the hiring must meet. For example, production pressure might require a candidate who can get up to speed quickly. Or the organization might need to prioritize hiring under-represented groups.
2. 2. Key qualifications: The job specification outlines the minimum qualifications necessary for the job. Some candidate qualities will be more important than others in the performance of the job, and candidates with greater strengths in these areas might be preferable.
3. 3. Suitability: Candidates need to fit into the organization’s culture and environment. For example, some will function better than others in a loud, crowded workspace with a collaborative atmosphere. Similarly, some will have a working style more consistent than others with an organization’s norms. That said, too much emphasis on “fit” can reinforce existing biases within the organization and cause appropriate candidates to be overlooked because they do not fit the “image” of the organization. Emphasizing fit can also result in candidates who will not challenge the dominant culture of an organization, which might not be desirable.

On this basis, an organization will make a hiring decision. The key drawback of this approach is the potential for inserting bias into the decision process. Bias can take one of two forms. First, human decision making is a complex matrix of reason and emotion. Individuals are often persuaded by factors not relevant to the decision, many of which are subconscious. These unrecognized factors can lead to incorrect weighting of qualities or to blind spots. The result can be a suboptimal decision. Second, intuitive approaches are more likely to replicate existing employment patterns in the organization. For example, similar-to-me bias and a focus on fit (see Table 6.3) can result in candidates from under-represented groups being overlooked.

To address such concerns, many organizations turn to more quantitative approaches, which can involve constructing a scoring system to apply to each candidate. The criteria for the position are listed and weighted, reflecting their importance based on the established goals. Candidates are scored based on those criteria, and those with the highest scores are selected. Some types of information are easier to score than others. Test scores are easily applied. Interview results require a bit more discretion in assigning scores. One advantage of this approach is that it forces decision makers to focus on each criterion, and a candidate’s fit based on that criterion, rather than a general, overall assessment.

Table 6.5 provides a simple example of a selection score card based on the job specification for a construction labourer (road work) set out in Chapter 3 (Feature Box 3.3) and extended through Chapter 5. Those charged with selecting a candidate assign a point for each required quality for each candidate.

• • The number of points available for each quality reflects its importance. For example, experience can provide a candidate with up to three points (out of a total of 10).
• • The number of points scored increases with the candidate’s qualifications. For example, candidates with more experience receive more points in the experience category.
• • Some qualities or levels of performance are listed as automatic failures (AFs). Candidates who do not possess these qualities are automatically deemed unsuitable for the job and discarded.

When making multiple hires, these scores can be used in different ways. Top-down selection sees an offer made to the candidate with the highest score. Additional offers are made by working down the list until all of the positions are filled. As an alternative, an employer can adopt a banding approach, by which candidates are clustered into groups (e.g., by quartiles).

Within each quartile, the sequence of offers is made based on other, non-scored considerations. For example, if an organization wants to hire more women, offers might be made first to women in the top quartile of candidates and then to men in that quartile. A bottom-up selection can also be used, in which a minimum score is set, and candidates who fall below that score are eliminated, whereas those above that score are either selected or move on to a second step in the decision process (e.g., a second interview, consideration of other criteria, etc.).

The development of scoring criteria entails trade-offs that HR practitioners should consider. For example, in Table 6.5, having job experience is not a requirement, but it is awarded a significant number of points. Using a top-down selection approach (i.e., hiring the candidate with the most points) means that the selection system might work against younger workers or newcomers to the occupation. This kind of built-in bias might be justifiable because (at least theoretically) it reduces training costs. It is worth considering, however, whether hiring younger workers might be desirable for other reasons. An organization with an older workforce, for example, might wish to hire younger workers in order to attenuate the risk caused by many workers retiring at the same time. The extra costs associated with hiring less experienced workers might be justified in this case and warrant a reweighting of the selection scorecard in Table 6.5.

The Job Offer

It is important to move through the entire selection process quickly because candidates are likely to apply for multiple jobs, and good candidates can be snapped up quickly. For this reason, once a selection is made, usually an initial offer is made via a phone call. The call should be to the point, offering key information such as wage rate, starting date, and so forth. The candidate might request some time to consider the offer or discuss elements of it further (e.g., make a counteroffer on the wage rate). It is important for the person making the offer to know which elements are negotiable and be prepared to negotiate them, including knowing the permissible range of acceptable terms.

The phone call is usually followed with a written offer, which becomes the official offer of employment. This is not the employment contract, signed once the offer is accepted, but for both clarity and convenience the written offer should include the following elements.

• • Job title
• • Compensation, including bonuses
• • Benefits, including items such as a moving or car allowance
• • Starting date
• • Probationary period
• • Union membership (if applicable)
• • Conditions to which the offer is subject (e.g., successful drug test, criminal record check)

It is possible that the candidate will turn down the offer. In such a case, an offer is then extended to the next most qualified candidate. For this reason, it is usually advisable to identify more than one potential hire and to hold off on notifying unsuccessful candidates until a hire is confirmed. When a candidate accepts the job offer, an organization moves to settle any outstanding conditions and prepares to finalize the employment contract and orient the new hire to the organization (see Chapter 7).

The final step of selection is to notify unsuccessful candidates. The time to do that is when the hire is finalized (i.e., the candidate has accepted and met any outstanding conditions). Typically, only those candidates interviewed or asked to take tests are contacted. The notification should be brief and polite. Unsuccessful candidates might ask why they were not chosen. The response should be honest and measured. A good strategy is to focus on the strength that distinguished the successful candidate. This approach allows for a clear indication of why the choice was made while not pointing out shortcomings of unsuccessful candidates. It is important to remember that second and third choices might be appropriate, and you might want to consider them for future openings, so maintaining a positive relationship can be advantageous.

Failed Searches

Although most selection processes result in hiring the required number of workers, sometimes a selection process does not result in a hire. A failed search happens most often when no qualified applicants applied for the job, an organization cannot make an offer that a qualified candidate will accept, or delays in or changes to a selection process render the competition moot. In some cases, especially when hiring to a position of key importance, a failed search is a better outcome than hiring someone simply to fill the position.

It is useful to identify why a search failed. This understanding can suggest ways to prevent a repeat in the future. There are several potential explanations for a lack of qualified candidates.

• • There might be few potential candidates who meet the qualifications. This might require altering the job design so that more candidates are qualified or expanding the geographic scope of the recruitment effort.
• • Word of the job opportunity did not reach potential applicants in time for them to apply. This might require a more extensive and/or longer recruitment effort.
• • The job opportunity was not sufficiently enticing to potential applicants. This might require revisiting the compensation, duties, or working conditions of the job. It might also suggest that a review of the employer’s reputation is warranted.

An organization unable to negotiate an acceptable offer might suggest that the compensation or working conditions associated with a position are not competitive. In contrast, unexpected internal changes and delays are simply a fact of life and not usually within the control of an HR practitioner.

Evaluating Selection Efforts

Evaluating the effectiveness of the selection efforts allows an organization to understand and improve its performance. An important question, though, is how to define what is effective. Simply asking “did the position get filled?” does not generate information that can be used to guide future selection efforts. It is also important to look at both the recruitment efforts (see Chapter 5) and the selection efforts. Although there is clearly overlap, they have distinct goals that should be considered separately.

To determine how to evaluate a selection process, HR practitioners should start by identifying the goals and context. Was it important to hire candidates at a low cost? Or alter the composition of the workforce in some way? Or is success about finding candidates who meet certain performance criteria? Once they know what the goals were, they can select an evaluation method that will give them the answers that they seek. Two common approaches to assessing selection are looking at the time that it took to fill a vacancy and a new hire’s quality of fit.

Time lapse measures the time that it takes to fill the vacancy, including selection time. The number of days from when the postings/ads first went out to when the successful candidate accepts the job offer assesses the efficiency of the process. This measure more often examines the selection process since it is easier to control timelines on the recruitment side by establishing closing dates et cetera. Time is not always the most important factor, and for some positions slower might be better if it allows an HR practitioner to find the right person (e.g., hiring a new chief executive officer who will chart the direction of the organization). Nevertheless, this measure is useful if the concern is filling vacancies quickly to prevent downtime, reducing the need for overtime, or overworking existing employees.

Quality of fit considers post-selection outcomes of the process. Looking at turnover rates among new hires after a certain period of time, productivity of new hires, new hire job satisfaction, or other performance ratings can provide useful information about how effective the process was in finding the appropriate candidates for the vacancies. These measures require a period of time to elapse before they can be used.

In theory, an organization that seeks to increase the diversity of the candidate pool might evaluate the degree to which its selection process altered the percentage of employees from each targeted group. The data might then be compared with the demographic make-up of the organization or the labour market. In practice, collecting demographic data with which to categorize applicants can be problematic. Although candidates can be asked to self-declare whether they fit into a specific equity category, some will be reluctant to self-disclose because of fear of discrimination. Ascribing certain characteristics to candidates through observation is a fraught process and best avoided.

Organizations interested in improving their selection processes might also query candidates about their experiences with different aspects of the process (e.g., clarity of communication, timeliness of a response, perceptions of an interview). Such inquiries gather information only from candidates who moved through the selection process. This exclusion suggests that the utility of information collected from such inquiries is limited.

Overall, evaluation of a selection process is difficult. Some metrics, such as cost and time, are easy to measure and can increase efficiency but do not speak to the effectiveness of the process. Most of the evaluation measures only indirectly address process effectiveness. Turnover and job performance speak to the success of finding employees who can do their jobs. But none of them can reveal whether the “best” candidate was hired since there is no way to measure that. In the end, selection evaluation needs to be iterative, a process of looking back and forth between the process and the outcomes and reflecting on how selection methods can be improved.

Conclusion

The controversies at the Canadian Human Rights Museum point to how inadequate hiring practices can lead to more complex problems at an organization. The poor hiring practices led to the creation of a workplace culture steeped in discrimination and unfairness. New hires perpetuated that unfairness since they were selected through a flawed process. The situation spiralled out of control, and eventually the organization was thrown into crisis. Along the way, dozens of workers were mistreated. Not all of the problems at the museum can be blamed on the hiring process, but it played a significant role.

One of the lessons from the museum vignette is that selection processes need to be carefully designed and implemented. The process of selection needs to be effective at its core function: that is, identifying candidates who are appropriate fits for the positions and organization. But it must also be fair, transparent, and accountable. Ultimately, selection is not just about finding the right candidate but also a key process in producing and reproducing a workplace environment that operates by determining who is hired and how they are hired. This means that HR practitioners responsible for selection need to be mindful of the broader context in which their decision is made.

Most of the research on selection focuses on finding workers who are the best fits for the jobs. Much less of it addresses how selection feeds into broader workplace dynamics. The good news is that it is possible to do both things. Designing a rigorous selection process that aligns with broader organizational goals will help to create a healthier workplace environment. A well-designed process of selection reduces the impact of individual bias and thus reduces the risk of replicating unhealthy workplace dynamics. And though no process can guarantee that the “best” candidate is selected, it can minimize the mistakes that lead to a “bad” candidate being selected.