Message to Candidates: Cheating Works … Sometimes!

How many applicants fake test results and assessments?
Does cheating work? Is it worthwhile?
What can you, the employer, do about it?

Personality Assessments
I have always been suspicious of self-rated assessments, as candidates know the job they are interviewing for and can guess what to say or not say. Many studies, such as the one recently published in the International Journal of Selection and Assessment (They Don’t Do It Often, But They Do It Well: Exploring the relationship between applicant mental abilities and faking, Julia Levashina, Frederick P. Morgeson and Michael A. Campion), have shown that self-assessments are indeed faulty:

This research [on fake personality measures] consistently demonstrates that candidates are able to fake personality measures by recognizing the correct, job-related, or preferred answers, and artificially inflate their scores.

Scary, right? Well, it depends on who is doing the cheating. Many candidates who have gone without a job for six months or more will tell you that it is good to know how to play the system in order to get a job.

Biodata Assessments
Some organizations may agree that personality tests can be faked, yet still believe in the strength of their biodata assessment. Are they correct in doing so?

First, what is biodata? Biodata is a commonly used term in industrial and organizational psychology for biographical data. Biodata is defined as “… factual kinds of questions about life and work experiences, as well as to items involving opinions, values, beliefs, and attitudes that reflect a historical perspective.” The basis of biodata’s predictive abilities is the axiom that past behavior is the best predictor of future behavior.

Biodata has an advantage over personality or even interest inventories, as it tells you the past behavior of a person, and from there it can predict one’s future actions … assuming one tells the truth!

How Many Cheat?
A newly released study from Julia Levashina, Frederick P. Morgeson, and Michael A. Campion on real candidates in real job application situations will give us the answer.

And this is a serious study, as 17,368 applications were analyzed across many different job categories (general management, economic and political analysis, public relations, etc.) with an innovative but strong way to detect the fake. Also, it is important to note that “candidates were warned that their responses could be verified and that any attempts to falsify information could be used as a basis for not employing them.” Thus, it was not a laboratory experiment.

So, how many are fakes? How many among those 17,368 applicants were trying to fake their way in?

The researchers divided the applicants into three groups, which we have taken the freedom to name:

Complete liars: 173 candidates (1%)
Fakers: 1,389 candidates (8%)
Stretchers: 4,168 (24%)

In short, a third of the people you will see will pretend to have done many more things than they actually have. In practice it could look like this:

These examples look obvious, and are for the sake of fun and illustrating the point, but they are probably what you can read on a resume or hear during an interview. Some strategies can help you uncover the hoax. We will cover them at the end. But the question still remains: Does cheating work?

Article Continues Below

Does Cheating Work?

Statements used in the research assessments were not as obvious; they were experiences or behaviors important to successful job performance. These included interactions with others, adaptability, initiative or persistence, leadership. These are less easy to fake. For instance, when you first move into a new place, how much time do you spend exploring your new surroundings (5 = a great deal of time; 1 = very little time)? They were capable of cheating, but how well did that work in favor of the fakers?

The research on this is clear: all groups of fakers “obtained higher scores on the biodata measure.”

Interestingly, the research showed that people with higher levels of mental abilities fake less often, but when they do it they get significantly higher scores. In short, the clever fakers are the ones benefiting the most.

So, we can safely predict that once job seekers learn that stretching the truth on applications and interviews works in their favor, they will continue to do it more. Thus, if today we see one in three people stretching the truth, tomorrow we may see one in two.

Talent Acquisition Response

Of course I/O psychologists will combat these statements by saying that they use empirical rating versus rational rating procedures. In short, more is not always best and other techniques prevent the fakers from winning. Incorporating other testing strategies should therefore be the first step, but it’s best to not take a chance, so I advise complementing such techniques with the following three simple and cheap strategies:

  1. For verifiable facts (i.e. Harvard MBA) perform a verification (academic, employment, etc). Not only will you avoid a bad hire, but you’ll prevent potential brand erosion and embarrassment.
  2. For results or behaviors that require one to have expertise (i.e. “recoded and secured the whole encryption software”), if straight technical assessments aren’t possible, make sure that a technical person (on your staff or outside if it is very unique) is part of the interview team to cross check the candidate to validate the expertise. At a minimum, a telephone interview or video conferencing should be performed if a face-to-face meeting is not possible.
  3. For results or behaviors where you can learn the jargon quickly (i.e. manage the on-time on-budget new ATS implementation), I recommend colleagues rate the candidate and or perform a reference check 2.0. These can be used as well for the previous section if you question the achievement level or the personality fit of a candidate, as technical competence is not always synonymous with performance and integration.

Armed with these tools, the next time you have three finalists in front of you, you will have the certainty of not picking the fake one.

Yves Lermusi (aka Lermusiaux) is CEO & co-founder of Checkster. Mr. Lermusi is a well known public speaker and a Career and Talent industry commentator. He is often quoted in the leading business media worldwide, including Fortune, The Wall Street Journal, Financial Times, Business Week, and Time Magazine. His articles and commentary are published regularly in online publications and business magazines. Mr. Lermusi was named one of the “100 Most Influential People in the Recruiting Industry” and his blog has been recognized as the best third party blog.



11 Comments on “Message to Candidates: Cheating Works … Sometimes!

  1. But do the clever fakers who get hired perform better?
    If the test is validated against successful behaviors (as opposed to biodata) the answer might be yes. lol

    Good article and great advice. Your passion for the subject since you founded Checkster is evident. I would also look for support for your premise among the job coaches who could easily be described as teaching one how to “stretch” in resume writing, interviews and even onboarding.

  2. Very interesting; and especially relevant in this job market. I still say that nothing can replace strong reference checks by a trained hiring manager or a person actually acting on belalf of the hiring manager asking specific questions related to past behaviors and skills and then adressing the new role and relating it to their ability and experience. Most background checks are too vague.
    Funny point Gerry made that the clever fakers may perform better, but I doubt it- maybe for certain types of jobs…

  3. For hiring companies to get honesty from candidates and employees, I submit they too need to practice it. How often does a company promote their career opportunities and then lay off staff after a bad quarter? Have employers ever posted inflated job ads that have not lived up to their billing? How many firms have described their culture as ‘one, happy family’ when the reality is that it’s a brutal political landscape filled with land mines?

    Corporate America is reaping a disengaged, dishonest workforce because those are the seeds they’ve sown.

  4. I found this post to be a house of cards based on false (faked?) ideas. I reject the term “faker” in the first place except for the 1% termed “complete liars”, and even that phrase is an oxymoron: a lie is either told, or it is not told since we only have one form of the word “lie”.

    Or do we ?

    A lie in the legal sense is a false statement or the witholding of information unknown to another in order to gain or contribute to gaining something of value. There is no legal or moral basis for complete liars, fakers, and stretchers, unless we are talking about three different things.

    Lots of people speak falshoods and manage other’s views of shared reality all the time for small or incomprehensible reasons- to live is to “lie” on those levels, and if you don’t understand that, you may think that fakers and stretchers are liars too, but to hold that view is to also state that everyone is a liar, and that’s not a position that works for job assessment.

    IOW, to respond cogently to this post, you have to dig at the nature of truth, and if you are successful in that minor exploration, you have to apply the results to business practices.

    Isn’t a clean and perfect interview suit a lie ? You won’t dress like that on the job. You almost never look like that to begin with. Shouldn’t you dress in a median outfit, your everday gear ? Some applicants do….

    The custom of polishing up for job interviews, court appearances, going on TV, etc. is basic: it’s done so that the content of your character may not be obscured by your tribal and class based styles, because it’s so easy to trick the human mind with symbols and cues alone. Likewise with personality assessment; it would be absurd not to “wear” your best personality when seeking a job. Is that faking ? Hardly.

    There are a few other logical inconsitances and of course the big flaw with how we do and talk about assessment today.

    Here is one: “Many candidates who have gone without a job for six months or more will tell you that it is good to know how to play the system in order to get a job” Aren’t those exactly the people who don’t know how to play the system, since they remain jobless ?

    Another: “For results or behaviors that require one to have expertise (i.e. “recoded and secured the whole encryption software”), if straight technical assessments aren’t possible, make sure that a technical person (on your staff or outside if it is very unique) is part of the interview team to cross check the candidate to validate the expertise. At a minimum, a telephone interview or video conferencing should be performed if a face-to-face meeting is not possible.”

    Does anyone hire any expert without any demonstration or verification of expertise whatsoever ? I doubt it. And for any role not needing expertise, what’s the point of assessment ?

    The fatal flaw: people and teams do not exist as independent entities. Every team is a mix of people, and you can only truly assess the performance of a team by the aggregate peroformance of it’s players in the exact context of the contest at hand and the leadership available. Every winning coach knows that some games can be won by at times by deploying measurably less talented players who matchup better against certain opponents or with certain situations, and you can have a group of brilliant star players who simply cannot perform as a group.

    This is not to reject any form of pre-qualification, of course. As the author noted, smarter people are better liars and so may be better employees, as Gerry instantly realized. The more creative, abstract, and group-based the job, the greater the difficulty in pre-assessment, even when there are easily measurable skills involved- like hitting a baseball or knocking down a three-pointer.

    My suggestion for your takeaway: you will never, ever, ever hire a person who is not a “faker”. Try to hire people who will mesh into your teams, adopt your goals, and are likely to earn their paychecks. Normally these folks will not be “complete liars”, but take every reasonable step to ensure that they are not.

    Depending on reality-distorting tools whose relationship to cause-and-effect is inherently unknowable, while discounting the vast power and complexity of knowledgable human intuition may not be the best way to succeed.

  5. You nail the BIO Faking part but let Personality Test Faking off easy. Congrats on a solid revue of BIO faking. You should check out Dr. Amy Hooper’s Ph D Dissertation at the U of Minnesota (2006-8)on the pervasiveness of faking in personality scales. Pretty scary stuff, and confirms my observations when on the inside of unamed companies that administered personality tests to hoardes of candidates. We could show the tests correlated usefully with job performance when creating the test– but those were correlations between test scores and performance scores for incumbents (people who already work for the company). Then when the tests were administered to candidates and used to decide who to hire, the correlations with their job performance were much lower (lower than are normally found for unstructured interviews– well under .2). My conclusion: people who have jobs will reveal themselves when answering personality questionnaires and those revelations do predict job performance. People who need a job answer in ways they think will get them a job, thus those answers do not reveal who they are and do not relate well to future job performance. But check out Amy’s dissertation for a more comprehensive empirical review of what is know about this topic.

    RE: Your second recommendation. You can have candidates describe specific past examples of tackling a challenging technical problem related to the targeted expertise, and then ask them to provide the name and email address of a credible third party who can confirm their success in that specific situation. I have implemented this in an Online Behavioral Interview and it has worked quite well. The reply rates to the request for confirmation are in the 40% range here and at least 20 points higher in Europe.

  6. It was a very interesting article, but there are definitely some gaps in the conclusions drawn. The above post by Chuck poses a rather interesting consideration–companies “lie” all the time about a job, the corporate culture, etc. Even in the interview process candidates will hear how there is advancement potential, great opportunities, etc.–all of which turns out to be false. Why is it acceptable for a company to “lie”–yet require its candidates to be completely truthful? If companies are creating cultures of deception–but want “honest” candidates, it would appears that rather than “testing” candidates about their honesty, some ethical corporate house cleaning is in order. Rather than creating hypocrisy, create ethical cultures. Perhaps then the need to find “honest” candidates would not be so difficult.

  7. Great article, and I would agree with Dr. Janz – a solid stance on Bio Data. However, lots and lots of holes in this, too. I am with PreVisor, the largest independent provider of assessment and talent measurement tools in the human capital space. Our organization maintains a staff of more than 70 professionals with PhD credentials in Industrial/Organizational Psychology and the most robust library of assessment content available.
    Mr. Lermusi talks about a number of things here: bio data testing, skill checking, reference checking, and technical interviewing. All are solid practices and we advise our clients to do all those things. Bio data itself is one item-type in a broad spectrum of assessments, and the most effective way to measure a candidate consistent with best practices is to include job-relavent, content-valid, multiple item types. We include bio data, personality, cognitive ability, situational judgement, hard skills, and culture fit measurements in rollup that is job-specific. We are also at the absolute cutting-edge of testing method, including our leadership in the space in creation of Computer Adaptive Testing, or CAT engine testing, that allows competencies to consistently be measured without any two candidates seeing the same static test. This makes the assessment impossible to fake, and impossible for candidates to post the “correct answers” on facebook or other social networking sites. If an organization feels that a candidate can “fake” their abilities or qualifications in an area where terms are easy to throw around such as IT, we recommend testing for specific skills, such as programming languages, network, and database languages, all areas we have validated testing for. We would never tell our clients to stop doing background checks, reference checks from former employers, or to conduct behavioral-based interviews. However, the only way to accurately predict if a candidate can do the job, if they are a fit for the job, or even if they have the cognitive ability to do the job, is to put tools in place that are backed by the normative data of lots and lots of experience talent measurement. I will be at ERE Expo in a couple weeks – anyone interested in this topic – I’d love to buy you a drink and chat about it!

  8. This is not a validated formula but I have found it has worked over the last 20 years.

    30% of the Grade is past work experience
    30% of the Grade is the interview
    30% of the Grade is testing
    10% of the Grade is references (if they come up bad it equals 100%

    If any of the top three do not agree you need to find out why. Test are not always correct, past work experience may not fit what you need, and interviewing does not always give you a true picture.

Leave a Comment

Your email address will not be published. Required fields are marked *