The Absolutely Unbiased Algorithmic Job Interview

Interview sketchSome large enterprises like Fannie Mae, Bank of America, General Motors, or A.I.G. might occasionally be lax in their hiring processes. With such unwieldy organizations, a mistake or two in hiring can happen and isn’t always of much concern, especially knowing that a government bailout is only a phone call away. But smaller companies don’t have the luxury of making hiring mistakes. Unlike a too-big-to-be infallible company that can absorb the impact of a bad hire, smaller companies can’t afford to bleed money hiring employees who aren’t expert in their trade, don’t take their responsibilities seriously, or don’t know how to schmooze clients.

Before they are about to hire, the leaders and hiring managers at these small companies comb through applications carefully, and by the time they are ready to interview job candidates, they may have lost a few night’s rest. Can you blame them? They are about to bring someone they don’t know into a company that they’ve fanned into a robust flame from kindling and dedicated their life’s breath to. Will this new person help the company or extinguish all profit? Will he or she treat the client and fellow employees with respect or make everyone’s lives an icy hell? There’s no way of knowing, but the company needs to expand, and there’s not much choice but to bring on a new recruit.

To make sure the employer gets the right person, recruits must participate in a rigorous hiring process, a most dangerous game in which hunter and prey are not always sure which one is being hunted and which one is being preyed upon. From the employer’s perspective, candidates should be tested like Army Rangers, running through obstacle courses and forced to dive to great depths with short supply of oxygen. By this, I mean that candidates should submit an application, undergo a pre-interview screening, and then attend a structured interview. Before the employer conducts this structured interview, she has chosen five or six applicants, made insensible judgements about the status of their secondary schools and colleges, and checked their social media accounts for embarrassing photos.

As the interview is a complex process, taking into account not only a potential hire’s job history and skills but also his personality and attitude, and as the resulting hire may have to be lived with for many months or years afterwards, it’s best if hiring managers don’t take on all of the evaluation responsibilities. Employers should encourage other company employees to sit in on the interview, say nothing, and make their own evaluations. Whenever there’s a chance that a bad decision might be made, as many people as possible should be present so that later on, when it comes time to place blame, blame can be shared.

In the traditional structured interview, the employer picks a process by which each candidate is evaluated in a uniform manner. The candidate arrives at the office at a scheduled time, and after a brief hour spent in the lobby filling out a fresh application and waiting, he is invited into a conference room. Sitting across from a panel of evaluators, the recruit receives a detailed explanation of job responsibilities, including applicable regulations and nomenclature with which he is expected to already be familiar. In the meantime, the interviewer pays attention to whether the recruit exhibits active listening skills, paraphrasing what the interviewer has just said before expounding on it. Hearing interviewees repeat back what has just been said is a sign that the candidate has been paying attention and this can boost the interviewer’s ego, increasing a candidate’s chances of getting the job.

Job responsibilities explained, a round of behavioral questions follow. These questions are designed to test the candidate’s skills in leadership, teamwork, interpersonal relations, problem solving, and ability to handle stress. Sample questions include “Tell me about yourself?” “Why should we hire you?” “What is your greatest strength and greatest weakness?” “Where do you see yourself in five years?” “And why do you want to work for this company?”

Questions not to be asked include those about the candidate’s race, age, sex, sexual identity, or sexual preference. If the candidate comes in wearing six necklaces—Christian cross, Islamic crescent, Jewish star, Hindu Om, Buddha lotus flower, and Chinese Ying Yang amulet—the recruiter should not give in to temptation and ask about the candidate’s preferred necklace and thus religion. If the candidate drops his wallet and family photos fall out, the employer should not make the mistake of commenting that the candidate must be married and have children, eh? Such questions, when reported to the government’s Equal Employment Opportunity Commission, have been known to cost companies $200,000 on average with a few settlements spiking into the millions.

Setting the many fears of interviewing aside, as the panel of evaluators listens to candidate answers, they should take note of the style of answering as much as the answer. Does the candidate keep everything positive? Does he stumble over words? Does he veer off on tangents unrelated to the position? Does the candidate know why he is applying to this company outside of needing money to pay for cable?

Not all job candidates are whom they appear to be.

Not all job candidates are who they appear to be.

At the conclusion of the interview, the evaluator asks the candidate if he has any questions, and the candidate should have prepared a few. This is a good time for the interviewer to be alert for a last minute lapse in judgement. Some candidates may ask questions like “Remind me, how do you do this job again?” “When will I get my first promotion?” “What sort of flextime arrangements can we work out?” And “does your company’s insurance include mental health benefits?” These are all inappropriate questions and suggest selfish, ulterior motives.

Instead, recruiters should look for recruits who ask one or two of the following: “When do you expect to be calling candidates in for a second interview?” “When do you expect to make a hiring decision?” “Would you like a copy of my references?” Or “if I don’t hear from you, do you mind if I follow up in a week?” These questions signal candidate interest in the job and are good enough reasons to present the recruit with a business card and end the interview with a brisk declaration such as “We’ll be in touch.”

When the candidate returns home, he should immediately get to work on the thank you note he will write by hand and snail mail it to the hiring manager. Meanwhile, the hiring manager and recruiting panel will evaluate the candidate pool one last time, spreading the five or six resumes out on the conference room table, consulting interview notes and photographs culled from the Internet by implicit search, and calling the other members of the interview panel into the conference room for counsel.

Once the selection is made, references can be called, but this is often a useless step if you are expecting an honest evaluation. Past employers rarely say anything bad about a former employee. If things didn’t work out at a past job, the former employer is likely to stick with the facts and make a comment such as, “yes, he or she worked here.” In recent years, defamation lawsuits have been filed based on job candidates receiving bad references, so when honesty becomes a costly policy, pure positivity and sticking to the facts serve as a good substitute.

References checked or unchecked, the good news is delivered by phone call, and the person believed to be the best candidate for the job and the best fit for the company culture is hired. On Monday, the new hire shows up for work and everything seems great until Tuesday when the employer discovers that the newbie doesn’t have all the skills he claimed during the interview. Nor does he like to show up on time for meetings. Nor does he work on his assigned projects but prefers to spend the first half of the day switching out players on his fantasy football team.

Giving the new employee a warning or two but observing no change, by Wednesday, the employer knows it’s time to call the new hire into the conference room and tell him things aren’t working out. This may lead to tears and threats and leave the employer swearing never to make another hire or at least wondering, “How could I have done better?” “How could I have saved my company time and money and chosen the right candidate?” “Where did I go wrong?”

The result of answering hundreds of multiple-choice questions has almost always led to a selection of highly talented, creative, and diverse peoples.

Although news to some, employers nationwide should be happy to know that once again, ever-inventive science has come up with a solution to avoid the many pitfalls encountered in the structural interview that I’ve described above, and these days, short-term hires can be avoided. In recent years, researchers at Harvard have developed a system that combines the benefits of big data with a carefully constructed hiring algorithm.

Using the power of math along with yottabytes, the researchers have been able to create profiles of ideal employees for particular jobs. Simultaneously, researchers also developed a list of multiple-choice questions that could be administered to prospective candidates for similar jobs. Once the questions, covering both hard and soft skills, are answered, answers are compared to the profiles already developed, and the choicest candidates fall readily from the trees.

Thus far Google, Xerox, and a few other high-tech companies have used such hiring methods and report that their hiring successes are up by 25%. These companies tell us that the tests have proven more accurate in predicting a successful hire than the standard structured interview with all of its behavioral questions and collective evaluations. Researchers say that one benefit to the algorithmic hiring method is that it is color and sex blind. It doesn’t make judgments about those issues because the multiple-choice questions have been tailored to avoid them.

Everyone knows that this is also one of the benefits of the Scholastic Aptitude Test (SAT) given to students in high school. Rather than relying on a student’s entire educational record, grades, and extracurricular activities—all of which can reveal too many of the specifics of character—most colleges prefer to place greater emphasis on the unbiased SAT score. The result of answering hundreds of multiple-choice questions has almost always led to a selection of highly talented, creative, and diverse peoples.

With such past triumphs in the use of standardized testing, it’s understandable that employers would defer to an algorithm. No longer will they have to sit in an interview and worry about saying the wrong thing, blatantly revealing their prejudices against minority races, women, old people, transgender people, or people with disabilities. Such a test would even eliminate the possibility of exhibiting microaggressions, ticks that signal unconscious prejudices that the interviewer didn’t believe she had until the interviewee pointed them out. Having a test that eliminates all prejudicial possibilities, the employer would not risk a lawsuit from the EEOC and could save on liability insurance, investing that money in morale-boosting office-pool Powerball tickets.

Using algorithms, employers could also rejoice that they will never have to ask another behavioral question and listen to a long-winded answer, such as the following response to the strengths and weaknesses’ question: “I like to make sure that my work is perfect,” the candidate will say, “so I tend to spend a little too much time checking it. However, I’ve come to a good balance by setting up a system to ensure everything is done correctly the first time. For example, I was once working at a warehouse facility when the files where stored in brown corrugated boxes, and I thought it would be better to use gray filing cabinets…” and so on and so forth.

What recruiter really wants to take the time to listen to stories about corrugated verses metal filing systems, following up with questions like, “so what made you go with the 24-gauge instead of the 32-gauge metal file?” and writing down all of this information. If such detail is attended to, when it comes to making the evaluation, the employer will have to sift through several legal sheets of notes, and I am speaking here of pages not only for a single candidate but for five or six? The time it takes to analyze those notes could be much better spent attending a Rhianna concert or visiting the cigar store and that humidor that has lately gone unattended.

But most of all, the great advantage of such a standardized test is that the employer will never again have to rely on her own judgement or the competencies of fellow employees to make a hiring decision. The algorithm will do all the work for her, and if anything goes wrong with the employee, then the algorithm, not the employer or human resources department, can take the blame.

Algorithms and big data could mean the end of flawed ancient wisdom, the practitioners of which have often selected and trained corrupt job candidates.

Algorithms and big data could mean the end of flawed ancient wisdom, the practitioners of which have often selected and trained corrupt job candidates.

A few old-school human resources managers might object that a big part of an interview is about evaluating the applicant’s interpersonal skills. “Is it possible,” they might ask, “to determine from multiple-choice testing how a candidate will interact with others?” The Xerox Corporation insists that the standardized test and algorithm makes hiring for interpersonal skills possible, having used the method to hire customer service representatives. But if the old-fashioned hiring manager is not convinced and still wants to conduct a personal interview, she should at least admit that during the interview process, many biases can emerge and every effort should be made to eliminate them.

As a way to eliminate bias in the structured interview, I’d suggest taking a tip from the new algorithmic method and keep candidate identities anonymous. Interviewers should never actually meet the candidates. Interviews should be conducted behind curtains like violin auditions for the local symphony. A voice-masking device could be used to disguise the sex and possible dialectical ticks of a candidate so that ethnicity and social status could also be obscured. Alternatively, a voice recognition computer might serve as an intermediary, taking the candidate’s answers and repeating them to the interviewee through headphones in an isolation booth. In general, when computers are incorporated into the hiring process, or any process, like political elections for instance, the results are always increased proficiency and fairness.

The more evidence gathered from big data, the more we are beginning to understand that interpersonal skills are highly overrated. Despite the lip service that many employers pay to the concepts of people skills and teamwork, in my experience, most offices are organized from the top down. Employers would much rather tell an employee what they want and have them do it rather than define parameters and then assemble a brainstorming group to talk things over before making a decision. What would happen if the military operated in this touchy-feely, groupthink fashion? Too much time would be spent arriving at consensus, and the seamless efficiency of conducting warfare would suffer.

Those who worry about the repercussion of hiring by tests that are tailored to a specific profession should stop living in the past. With the rise of computer technology and the development of sophisticated algorithms, more and more human decisions can be avoided and many would agree that this is a good thing—as humans aren’t very good at making decisions. Increasingly, activities that used to be handled by people, such as surgery, making art, or even writing blogs can be adapted to robots or computers and accomplished with better effect.

Isn’t it comforting to know that all workplace problems of an interpersonal nature will soon be handled by algorithms and big data, leaving humans to carry on only relationships of a recreational nature? However, if you happen to be one of those people who are not interested in pursuing human relationships of a recreational nature, then I’m sure the scientists at Harvard will soon come up with an algorithm by which you can find people who share that same interest.

Note: This article was written with the assistance of Rhonda Serendip, a marketing and management consultant at “Serendip Business Marketing and Management Consulting.” Ms. Serendip has over 25 years of experience helping businesses deliver valued and scalable products and services to an elite clientele while actualizing their business profit potential with forward thinking actionable strategies. Ms. Serendip is also a valued employee at Macy’s.

This entry was posted in Business Communication, Satire, Science and Technology, Uncategorized and tagged , , , , , , , , , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s