The structured behavioral interview has several strengths that contribute to reliability, validity, legal defensibility, and perceptions of fairness. On the other hand, the unstructured interpersonal interview can be one of the most unreliable and invalid methods of selection available. The validity of the unstructured interview has been reported to be lower than most other types of selection systems. Due to the potential for subjectivity and bias, an unstructured interview process leaves an agency particularly vulnerable to legal attack. The structured behavioral interview also greatly enhances the quality and honesty of information gathered from employment interviews. Listed below are the strengths of the Structured Interview:
Bias is reduced because candidates are evaluated on job-related questions, which are based on an analysis of job duties and requirements. Subjective and irrelevant questions are not asked.
All candidates are asked the same questions so everyone has the same opportunity to display knowledge, skills, and abilities.
Pre-determined anchored rating scales are used to evaluate answers to interview questions. This reduces disagreements among interviewers and increases accuracy of judgments.
A panel of interviewers is used to record and evaluate answers in order to minimize individual rater biases. Therefore, the use of a panel is a plus.
Research has demonstrated that properly developed structured interviews can have high reliability among interviewers and predictive validity for future job performance.
Job-related procedures used to develop structured interview questions increase content validity.
Procedures used to develop structured interviews are consistent with the advice of professional and governmental guidelines, and thus more legally defensible.
Structured interviews allow managers to take part in the selection process in a role with which they are familiar.
Job-relatedness and consistency of the process may increase the perception of fairness among candidates. The job-relatedness may also help candidates obtain a realistic perspective of the job, which can aid in self-screening..
Types of interview questions
Following the job analysis, interview questions should be developed from behaviors determined during the job analysis to be critical and essential to the performance of the job. There are four types of interview questions: job knowledge, background, hypothetical situational and actual past behavior.
Job knowledge questions may ask interviewees to demonstrate specific job knowledge or provide documentation of job knowledge.
Background questions focus on the work experience, education, and other qualifications of the candidates.
Hypothetical situational questions present the interviewee with hypothetical situations that may occur on the job and ask how the interviewee would respond to the situations. The use of situational questions in an interview is based on the assumption that a person’s intentions are related to behavior; thus, how a candidate says he or she will handle a problem is most likely how he or she would actually behave in that situation.
Actual past behavior questions require candidates to describe the activity of past jobs that relates to the job for which they are being interviewed. The use of actual past behavior questions in an interview is based on the assumption that a person's past behaviors are related to future behaviors. Therefore, how a candidate has handled a problem in the past is most likely predictive of how he or she would actually behave in that situation in the future.
Asking open-ended questions, as opposed to questions that can be answered with a yes or no, will allow the candidates to reveal more about themselves. If a question is developed to determine if a candidate does or does not meet a specific requirement, then a close-ended question could be appropriate; for example, “Do you have a valid driver’s license?” or “Do you have experience with Microsoft Word?” Otherwise, open-ended questions usually gather more information; for example, “Describe any experience you have had in using computer-based word processing programs.” Psychologists recommend using a variety of these types of questions.
Job Knowledge Questions
After repairing a piece of machinery, why would you clean all the parts before reassembling them?
(5) Particles of dust and dirt can cause wear on moving parts. Need to have parts clean to inspect for wear and damage.
(3) Parts will go together easier. Equipment will run better.
(1) So it will all be clean. I don’t know.
Please sell me this product using basic selling techniques.
(5) Candidate simulates selling the item to the interview panel by incorporating the following selling techniques: (a) identifies and presents the product, the customer needs, and the benefits of the product; (b) demonstrates the product; (c) handles resistance; and (d) closes the sale by asking for an order.
(3) Candidate uses only three of the techniques or performs one poorly.
(1) Candidate uses only two of the techniques or performs them very poorly.
Suppose you were going to miss an important business meeting due to unforeseen circumstances (e.g., illness or family emergency). What would you do?(5) I would contact the person in charge of the meeting to forewarn of my absence, and I would arrange for a responsible person to attend in my place.
Some jobs require climbing ladders to a height of a five-story building and going out on a catwalk to work. Give us examples of when you performed such a task.(5) Heights do not bother me. I have done similar work at heights in the past [and gives examples].
Developing rating scales and benchmarks
A decision must be made regarding the scoring system or rating scale to be used in the interview. The rating scale can be as simple as “acceptable” or “unacceptable,” or it can be a three, four, or five-level, point-based scale. It is difficult to define more than five levels that can be meaningfully and consistently assessed. The most critical element of the rating scale is not how many levels it has, but rather how those levels are defined.
Rating scales should be defined by benchmarks for each question. Benchmarks are suggested answers to the questions that are linked to the rating scale. Benchmarks provide a frame of reference for assessing the candidate’s responses objectively and consistently. There should usually be at least three suggested answers for each question: a superior, a satisfactory, and an unsatisfactory response.
In the examples a five-point rating scale was used. Benchmarks were developed for five points (superior answer), three points (satisfactory answer), and one point (unsatisfactory answer). To develop benchmarks, using the guide below, ask subject matter experts (SMEs) to create answers that fit the different levels of the rating scale. If the questions have been used in interviews previously, SMEs may use actual answers they have heard from candidates.
It is not essential to describe the 4 or 2-level answers, because the 5, 3, and 1 answers give adequate anchor points for making a rating decision on any of the levels. The 3-level benchmark is usually the easiest to develop, so try describing that answer first. Example answers should fit the requirements of the job. Superior answers should not far exceed the requirements, and unsatisfactory answers should not be so low that they do not help distinguish between candidates. Also, try to avoid making the superior answer a more sophisticated or simply reworded version of the satisfactory answer. Organizational jargon, acronyms, and slang should be avoided. Developing benchmarks is also a method of evaluating the interview questions. If it is too difficult to determine the benchmark answers for a particular question, the question should be reviewed for possible revision or elimination.
The interview panel should meet to review the job description and job analysis, design the interview questions, and set benchmarks for answers to the questions. The panel should also choose a coordinator to lead the interviews. Interview panels should have at least three persons. Having the immediate supervisor of the open position serve on the interview panel is recommended since he or she may be the best expert on the duties and responsibilities of the position. Other panel members might include the division director, a human resources representative, a co-worker, representatives of other departments, or a representative of the customers served by the position. All members of the panel should be familiar with the duties and responsibilities of the position being filled. Every effort should be made to have the panel reflect the race and gender makeup of the candidate pool, which may reduce the potential for bias.
Using a panel to conduct the interviews may reduce the impact that personal biases of individual interviewers may have on the selection of an employee. It is also important to use the same persons as interviewers for all of the candidates. Different interviewers are likely to evaluate answers differently, but if the interviewers are always the same persons then there is consistency in the ratings of candidates. Training the interviewers will increase consistency.
The candidate should be informed that notes are being taken. Taking good notes is extremely important in conducting a structured interview. This note taking is known as the observation phase of the interview. Interviewers should not rely on memory for two reasons. First, what seems perfectly clear during an interview can quickly be forgotten or confused, especially after interviewing several candidates. Second, all employee selection decisions should be documented. When informing the candidate that the interviewer will be taking notes throughout the interview, validate in the candidate’s mind the reason for taking notes as being in the best interest of the candidate — you want to be sure to give full credit for all the knowledge, skills, and abilities demonstrated during the interview. A new, clean interview guide should be used by each panel member for each candidate. Close the interview with an open-ended question, such as: “Is there anything else you want us to know about you?” and “Do you have any questions for us?” In closing the interview, explain the notification process again, even if you did it earlier.
Evaluating the Candidate (Scoring)
Each rater should score each question from their notes immediately after the interview is completed. Raters should independently take notes regarding a candidate’s comments on each question as it is answered. Raters should also independently, score each question immediately after the interview is completed.
When one interview has concluded, raters should give themselves approximately ten minutes before beginning the next interview to review their individual ratings as a group and make sure that there is a general consensus on each question. For example, on a scale of 1 to 5, if one rater gave a 4, one a 3, and the other a 1 on the same question, this might indicate some rater bias or misunderstanding of the candidate’s comments. All raters should be no more than one point away from all other raters’ scores. For example, it would be acceptable to have two 3s and a 2 given on the same question. If raters find that there is not a general consensus on a question, they should discuss the reasons for their ratings and attempt to reach a consensus.