Highlight, copy & paste to cite:
Nevo, B., & Berman, J. A., (1994). The Two Step Selection Interview: Combining Standardisation with Depth, Research and Practice in Human Resource Management, 2(1), 89-96.
The Two Step Selection Interview: Combining Standardisation with Depth
In interviewing in the hiring of personnel has recently been reassessed. Formerly, interview techniques were criticised as being unreliable and invalid. Now, meta analysis indicates that a structured interview composed of systematic questions and objective ratings yields reasonable reliability and validity. However, the highly structured interview may prevent important personal issues from surfacing and from being discussed during the interview. A design is proposed for a “two step interview”, the first part of which is structured and the second, unstructured. The “two step interview” represents an optimal combination of a standardised, psychometrically oriented approach and a flexible, clinically oriented approach.
Employers have typically viewed the interview as an essential part of the employee selection process. Anyone who has been hired for employment has most likely experienced the selection interview. A wide range of interview formats is possible. Two main forms can be delineated: the non-structured interview and the structured interview. In its purest form, the non-structured interview is an open-ended, individualised, conversational dialogue between the employer and the applicant. The structured interview is a formal, test-like, question and answer session in which the employer assesses the applicant’s specific credentials in a predesigned, organised manner. We will examine the positive and negative features of each type of selection interview. Then we will advocate an integrated format that draws on the strengths of the two interview formats, while avoiding their limitations.
The Personnel Selection Interview
The unstructured, personnel interview is probably the most commonly used selection tool in North America arid Western Europe, and the sole selection device in many other countries. The personnel interview has at least five advantages, which may explain its popularity among business and personnel managers.
- The “personal touch”. An employer or manager may wish to evaluate the prospective employee personally. The employer can ask questions that an indirect resource could not pose, and can obtain a personal feel for whether or not the applicant would be suitable for the position and would fit into the social context of the company.
- Facility of operation. With a small number of applicants, interviews are relatively easy to operate. They can be given at any time or place, and do not need special materials for their execution.
- Perceived inexpensiveness. The interview appears to be a low- budget operation (again, when the number of candidates is small). No money needs to be spent on purchasing special tests or hiring test administrators.
- Face validity. The interview has high face validity. Interviewing seems to be the most natural method of selection because the impressions we form of others in everyday life also come from speaking and interacting with people.
- Flexibility. The interview is adaptable and is capable of serving multiple purposes. Interviews can be used at several stages of personnel management, including recruitment, screening, selection and negotiation. The interview is a way to introduce an applicant to the organisation in a warm, personal way. In addition, it gives the applicant the opportunity to satisfy his or her own questions about the company, and to get a feel for the group of people they might be working for.
Problems with the Non-structured Personnel Interview
Psychologists researching the non-structured personnel interview have examined its pros and cons (Arvey & Campion, 1982; Eder & Buckley, 1988; Hakel, 1982; Harris, 1989; Mayfield, 1964; Schmitt, 1976; Ulrich & Trumbo, 2965; Webster, 1982; Wright, 1969). The prevailing view among researchers for many years was that an open-ended interview is very problematic, and that managers would do best if they assigned zero or a very low weight to the interview results. Conclusions drawn by Mayfield in his 1964 review are typical of the views of others who wrote about the non-structured personnel interview until the mid-1980s:
- Consensus (interrater reliability) among interviewers is low,
- Behaviour domains are not covered in a consistent way,
- Unfavourable information influences interviewers more than favourable information,
- Interviewers make their decision quite early in the interview and spend the rest of the interview justifying (for themselves) this decision.
- Interviewers’ ratings are affected by a variety of cognitive biases (eg primacy-recency effect, first impressions, contrast effect, halo effect and social stereotypes).
- The interview has low predictive validity; performance during an interview does riot predict future job performance well.
Yet for those who use personalised selection interviews, the face- to-face format provides a sense of being informed and in control that they find hard to relinquish. And indeed, as described earlier, there are some benefits to an unstructured interview. Therefore, the goal should be to improve the interviewing process, and not to drop it or even to modify it to the point where it is no longer really an interview.
The Structured Interview
The instrument that has recently emerged as a proposed solution for the problems inherent in the personnel interview is the structured interview. This device combines the features of a personnel interview and the features of a standardised selection device.
Designing a Structured Interview
The proper design of the structured interview entails several phases that must be followed. The first of these steps is the determination of the particular dimensions that are to be covered during the interview. The employer must decide on the abilities, traits, attitudes and values that are believed to be necessary for success in both the particular organisation and the particular position that is being offered. The best process for this objective is job analysis. Comprehensive and accurate job analysis has been shown to improve the validity of a structured interview (Arvey & Campion, 1982). Job analysis is also necessary from a legal standpoint; the Uniform Guidelines for Employee Selection (Federal Register, 1979) in the United States require that job analysis be performed as part of the development, application, and validation of employee selection devices. Data from the job analysis should be used in the development of interview rating scales. In order to add structure to the interview, the predetermined dimensions must be “translated” into specific interview questions. The interviewer must be able to elicit information that will satisfy the criteria of the required dimensions.
In the second phase of the design, standardised ratings must be developed to systematise the scoring and interpretation of the interview. This process is what really makes the instrument “structured”. Scoring codes, accompanied by good examples and illustrations, allow for a uniform method of judgement by different raters, and across different applicants. These codes also determine the relative values of applicants’ abilities and accomplishments, ie how important the various attributes are for being hired.
In the third phase of the design, several interview conditions should be controlled and standardised in order to ensure reliable results. First is the physical location of the interview. An office or similar room should be used for all interviews. Second, proper interview training is essential. Interviewers should be trained in a way that will enhance the uniformity and fairness of the interview. Moreover, interviewers should be made aware of potential biases and how to avoid them. Third, interviewees should be given the same questions in the same word order and question order. Fourth, a single panel for rating the applicants against each other has been found to increase both the reliability and validity of the selection process (Arvey & Campion, 1982; Cronshaw & Weisner, 1989).
Table 1 compares the major features of structured and non-structured interviews. It is suggested that the “structuredness” of a selection interview is a continuum rather than a dichotomy. Table 1 characterises the two poles of this continuum.
|Feature||Structured Interview||Non-Structured Interview|
|Defining the specific dimensions to be covered by the interview.||Viewed as essential for the selection process. Specification of dimensions are guided by proper job analysis.||Not always viewed as necessary. Sometimes an assessment of “the quality and normality” of the interviewee is presented as the goal of the interview.|
|Questions presented to the interviewee.||For every dimension, a pool of predetermined questions is prepared. Questions during the interview are taken from these pools.||Questions differ from one interview to another depending on the dynamics of the interview.|
|Rating of the interviewee.||Rating is systematic. It is based on rating codes, anchored examples, and illustrations.||Rating either does not exist, or is performed by the interviewer intuitively.|
|Standardisation of the physical conditions of the interview setting.||Viewed as essential.||Viewed as advantageous but not essential.|
|Single vs. Panel of interviewers.||Panel of interviewers is preferred.||Single interviewer is preferred.|
|Training of interviewers.||Viewed as essential by both methods.|
A highly structured interview will carry all five features presented under the “structured interview” title. A highly non-structured interview will carry the other five features, each one of them to its extreme.
Recent meta-analyses show that the structured interview has moderate predictive validity. Table 2 presents the validities of structured and unstructured interviews, according to three most recent meta-analytical reports. Each of these three meta-analyses combines findings from several dozens of empirical “primary researches” (Hunter & Schmidt, 1990). The figures in the table express, via correlational scale, the strength of the relation between selection interview ratings (the predictor) taken at t0 and job performance ratings (criterion) taken at t1.
|Researcher||Type of interview|
|McDaniel et a11 (1987)||.36||.45|
|Cronshaw and Weinser1 (1989)||.31||.62|
|Wright, Lichtenels and Pursell2 (1989)||.143||.35|
- This study has been corrected (observed validities) for range restriction and criteria unreliability.
- This study has been corrected (observed reliabilities) for criteria unreliability only.
- This figure was reported by Hunter and Hunter (1984). It was employed by Wright, Lichtenfels and Pursell (1989) as a basis for comparison.
Similarly, some recent individual studies that employed a structured interview for selection purposes found moderate predictive validity. For example, Campion, Pursell and Brown (1988) reported a corrected validity of r = .56 in a study of entry level production employees; and corrected validities of r = .42 and r = .61 for two samples of sales clerks were reported by Arvey, Miller, Gould and Burch (1987).
In the light of all of the structuredness characteristics, it seems appropriate to view the structured-unstructured quality of an interview as a continuum rather than as a dichotomy. Unfortunately, it has not yet been established which features of the structured interview are essential in terms of enhancing the reliability and validity.
Problems with the Structured Interview
Although the structured interview provides some advantages over the unstructured interview, problems still remain. Structuring the interview may have a sterilising effect on the procedure. That is, the interview becomes less spontaneous and the interviewer may Jose track of potentially important follow-up information. In other words, there is a loss of flexibility that results from the inclusion of external loci of control (ie predetermined questions). Another dehumanising effect is the “cold” atmosphere that results from structuring. This effect may influence interviewees to be less open during the interview and it may discourage applicants from pursuing an employment position with an “impersonal organisation”, even if it is offered to them at a later stage. Interviewers and applicants tend to value an atmosphere where they can be spontaneous and direct.
Additionally, the introduction of structure brings an opportunity for applicants to “beat the system”. The use of standard questions and a fixed-response rating code in the structured interview can be dangerous because lists of popular interview questions are widely circulated and applicants are often well-prepared for typical questions.
The Two-Step Selection Interview
Here, we propose a specific type of interview that offers “the best of both worlds”. This “integrated” or “combined” interview consists of two parts. Part I is highly structured. In this part, the interviewer adheres to the format prepared beforehand. The interviewer presents a standard set of questions verbatim and in a predetermined order. Ratings are based on a standard rating guide. The probable duration of Part I is 20-40 minutes.
Part II is open-ended, flexible and unstructured. In Part II, the interviewer uses his or her training and experience to follow-up on important issues mentioned by the applicant. The interviewee is encouraged to bring up issues that he or she thinks are relevant to the hiring decision. This is done in a free-style, warm manner. The interviewer acts more as a clinical psychologist than as a personnel manager. The interviewer has to listen with his/her “inner ear” to any sign of pathology, personality disorder, difficulty in interaction with other people, behavioural disorders under stress, and so on. The score of the interviewee on Part II is based on the subjective, impressionistic judgements of the interviewer. Interviews ranging from 15—30 minutes are typical.
An experienced interviewer moves smoothly and in a natural style from the formal, structured part of the interview to the non-structured, more personal part of the interview. Weights for each part are determined ahead of time. For example, a decision may be made that the weight of the Part II score is 20% of the total score, and the weight of Part I is 809k. Knowing this ahead of time lessens the likelihood that interviewers will assign special weights to their “clinical” judgements. This tendency is a source of bias, and is considered to be one of the reasons for the low predictive validities of non-structured selection interviews. Therefore, once the relative weights of the parts are determined, interviewers should be consistent and fair across applicants.
The recent developments in selection interviewing research indicate that both structured interviews and personalised, unstructured interviews have advantages and disadvantages. The contribution of a structured design to the selection interview is a considerable boost in reliability, and more importantly, validity. Also, organisations are receiving greater scrutiny from many sources (eg government, unions), which has put pressure on them to objectify and standardise previously subjective procedures for hiring personnel. As a result, well-documented procedures for designing and improving the selection interview are now being utilised.
However, if too many constraints and regulations are placed on structured interviews, there is a risk that managers will revert to extremely unstructured procedures. Also, there is some likelihood that significant personal applicant information will remain hidden.
Thus, current research and reports from human resource professionals support the two-step interview process for personnel selection. Further research is needed to test and refine this approach. However, the value of a two-step selection interview seems clear based on experience and research to date.
Arvey, R. & Campiori, J., (1982), The employment interview, Personnel Psychology, 35, pp. 281-322.
Arvey, R.D., Miller, H.E., Gould, R. & Burch, P., (1987), Interview validity for selecting sales clerks, Personnel Psychology, 40, pp. 1-2.
Campion, M.A., Pursell, E.D. & Brown, B.K., (1988), Structured interviewing: Raising the psychometric properties of the employment interview, Personnel Psychology, 41, pp. 25-42.
Cronshaw, S. & Weisner, W., (1989), The validity of the employment interview, in R. Eder & G. Ferris (Eds.), The Employment Interview, Sage, Newbury Park, California.
Eder, R. & Buckley, M., (1988), The employment interview. In G. Ferris & K. Rowland (Eds.), Research in Personnel and Human Resource Management, JAL.
Federal Register, (1978), Uniform guidelines on employee selection procedures, 43, pp. 38290-38315.
Hakel, M., (1982), Employment interviewing, in K. Rowland & G. Ferris (Eds.), Personnel Management, Allyn and Bacon, Needham Heights, M.A.
Harris, M., (1989), Reconsidering the employment interview, Personnel Psychology, 42, 691-726.
Hunter, J.E. & Hunter, R.F., (1984), The validity and utility of alternative predictors of job performance, Psychological Bulletin, 96, 72-98.
Hunter, J.E. & Schmidt, F.L., (1990), Methods of Meta-Analysis: Correcting Error and Bias in Research Findings, Sage, Newbury Park, California.
Mayfield, E., (1964), The selection interview, Personnel Psychology, 17, pp. 239-260.
McDaniel, M.A., Whetzel, D.L., Schmitt, F.L., Hunter, J.F., Maurer, S., & Russel, J., (1987), The validity of employment interviews: A review and a meta-analysis, Unpublished Manuscript, (Quoted by Harris, 1989).
Schmitt, N., (1976), Social and situational determinants of interview decisions, Personnel Psychology, 29, pp. 79-101.
Ulrich, L. & Trumbo, D., (1965), The selection interview since 1949, Psychological Bulletin, 63, pp. 100-116.
Webster, E., (1982), The Employment Interview: A Social Judgement Process, SIP Publications, Schomberg, Ontario.
Wright, 0., (1969), Summary of research on the selection interview since 1964, Personnel Psychology, 22, 319-413.
Wright, P.M., Lichtenfels, P.A. & Pursell, E.D., (1989), The structured interview: Additional studies and a meta-analysis, Journal of Occupational Psychology, 62(3), 191-199.