Confirmation vs. Consistency: Whodunnits and The Role of ‘Circumstantial Evidence’ in Vetting Job Candidates
Your immediate, unspoken and (in your mind) natural reaction is “Huh?”
To get past your “Huh?”, you should compare your manager’s question with the plot of a typical whodunnit detective novel: The butler was on duty the night his employer was shot and killed in his den, and his boss’s registered gun, which he kept in a closet, is still missing.
The Butler Did It?
Does that, as a stand-alone fact, confirm (to any degree, if not categorically) that “the butler did it”, or is itmerely consistent with that conjecture? If the former, the butler is much more reasonably to be deemed the prime suspect than if the latter is the case. (To get the analogy, equate “prime suspect” with “prime candidate”.)
Clearly, in the absence of any more substantial, direct, less circumstantial or indirect evidence, his being in the house and the gun’s having disappeared are merely consistent with the hypothesis that he shot his employer, i.e., it merely does not contradict and rule it out or dramatically reduce the probability he did it. And if it is merely consistent with his guilt, as his having a subscription to The New York Times, a jazz CD collection or a black ball-point pen is, can it in any way at any point serve as evidence of any suspected guilt?
Mere Consistency of Recruiting Evidence
That kind of question can—or should—come up in recruiting. Does the candidate’s degree in computer science confirm that (s)he can do the IT job, or is it merely consistent with that expectation? (On the other hand, doesn’t a lack of a degree in computer science disconfirm the same expectation to the extent that it reduces the probability of competency?)
These two examples illustrate the need to distinguish not only confirmation from consistency, but also absolute confirmation from statistical degree of confirmation. For example, the lack of a computer degree is not only consistent with being able to do the job (because of the possibility of some courses taken, independent study, innate genius, etc.), but also disconfirms it, although only to some degree, by making such competency less likely, i.e., less probable, mathematically speaking.
But, even if there is other, more weighty evidence, such as the employer’s blood on the butler’s jacket sleeve (or, analogously, a glowing letter of recommendation for the IT candidate), does the fact that the butler was in the house, given that additional information, to any degree, now actually confirm he whacked the boss, despite the fact that knowing “the butler was not in the house” is true would unequivocally prove that the butler didn’t do it (at least not by any direct means and alone).
Or is it still merely consistent with his guilt and therefore not admissible as “evidence”, just as his New York Times subscription, black ball-point pen or a jazz CD collection would not be? (There is, however, an important difference between not being in the house and not having a NYT subscription, despite both seeming to be “merely consistent” with guilt: absence in the former case, but not the latter, does indeed have a bearing on the case and likelihood of guilt.)
What Standard of Proof Should Job Candidate Vetting Require?
In any instance in which law and logic diverge on this issue, viz., in cases in which the law may specify that mere logical and factual consistency with guilt or responsibility is legal confirmation of it, which standard—e.g., the logical, scientific or the legal—should recruiters use in assessing job candidate “evidence”, e.g., presentation of credentials, such as that computer science degree?
[Note: the standards governing circumstantial evidence are stricter in criminal law than civil law, two cases in point being the O.J. Simpson criminal and civil proceedings in which he was acquitted and lost, respectively, and are encapsulated by the difference between “beyond reasonable doubt” and “more probable than not”, respectively.
Hence, a pertinent question to ask in the context of vetting job candidate facts as evidence (circumstantial or otherwise) is to which standard are you being held, as a recruiter or as an applicant?]
A sharp recruiter in the job market should not only be able to distinguish “direct” compelling evidence from merely “indirect” or “circumstantial” evidence, but also be able to judge to what degree such less compelling evidence is relevant, and by which standard (with respect to amount, quality and inter-relatedness of the evidence): logical, scientific, legal, theological and, most especially, professional—if it’s different from any or all of these.
Although with respect to many attributes of a candidate or aspects of the candidate’s background, circumstances, etc.,we instantly recognize “mere consistency”, with respect to other considerations, it may be very difficult to distinguish that from things about the candidate that strongly confirm the “hypothesis” about him or her.
For example, the candidate’s having a dog at home is clearly consistent with his or her having a grasp of supply-chain management principles, and neither confirms nor disconfirms that assumption. But what about something psychologically, if not logically, tougher to judge, e.g., having a skull and crossbones tattoo on his forehead?
I’m willing to bet that some recruiters will see it as consistent or irrelevant to the supply-chain management hypothesis, while many more, without admitting it, see it as prima facie disconfirmation, despite the seeming illogicality of that leap of judgment.
Logical, fair or not, one man’s mere consistency is another’s (dis)confirmation.
Indeed, one’s take on facts as mere consistency can be another’s confirmation, e.g., when through “confirmation bias”, stereotyping or simple prejudice, a candidate’s polished British accent—consistent with the hunch that the candidate will be a skilled PR spokesperson—is regarded as solid confirmation of it.
Such candidate and other assessment disagreements can be due to some flawed capacity for judgment, stereotyping or other bias on somebody’s part, or due to misinformation or misunderstanding about the specific “evidence”, especially misunderstanding attributable to a complete unawareness of the difference between mere consistency and (dis)confirmation.
Should Recruiters Follow a Legal Standard of Evidence?
So let’s get clear about the difference between mere consistency and (dis)confirmation and between the various standards available for weighting them, evidentially speaking. Prime among these is the law: When, if ever, is mere logical and factual consistency a kind of confirmation and when is it not—e.g., when would the butler’s merely being in the house constitute prima facie and admissible evidence (to some degree, however slight) of commission of or complicity in crime?
If it is, then mere possession of a computer science degree or a British accent could, by application of that legal standard, constitute confirmation of having the desired job qualifications.
The law and associated definitions and interpretations will vary in various jurisdictions, but, broadly speaking, “circumstantial evidence” is any evidence that is not eye-witness or otherwise virtually indisputable evidence, and that therefore requires an assumption, inference, implication or addition of another fact or facts to establish culpability or at least be considered relevant to the case and admissible as evidence.
Evidence in Bits vs. in Bulk
This interpretation invites the question of what happens if each bit of circumstantial evidence is presented and individually ruled upon as a “stand-alone” vs. in bulk. Generally but not always, in law, a set of purely circumstantial evidence, as well as individual evidence, is not regarded as conclusive. That may serve as a useful guideline in recruiting too.
In recruiting, this difference between stand-alone and aggregate would be that between looking at each credential separately from all others or as a set (what is sometimes called a “total evidence condition” for weighting individual bits of evidence), when deciding its relevance (and ultimately that of the set).
Common sense suggests evaluating the set as though it represents a “critical mass” of evidence—but only if qualitatively and collectively the sum is greater than the sum of its individual “parts”.
But one has to be very careful with this, because ten weak pieces of evidence (or merely consistent facts) don’t always amount to a set of strong evidence (as strongly confirmatory facts). That is, ten facts that are merely consistent with being guilty or being the best candidate or among the best candidates don’t always equal solid or sufficient proof, when taken together.
Recruiter vs. Courtroom Standards of Evidence
Here’s how that can happen. Look at and then score each fact below—first, individually, then in the context of the whole set of these. Individually judge the relevance of each fact as “merely consistent”, “qualifying” or “disqualifying” (the latter two corresponding to “confirming” and “disconfirming”). Then reach your conclusion about the degree of confirmation and conclusiveness: The candidate
1. has a university degree
2. has a resume with the right keywords and no grammatical or spelling errors
3. has provided appropriate references
4. has shown up on time for the interview
5. has dressed appropriately for the interview
6. has no weird postings on his Facebook page
7. has experience comparable to most of the other candidates
8. has displayed a positive attitude and has asked appropriate questions
9. is available as required
10. is physically very attractive and has an elegant British accent.
Let’s compare our individual item and set scorings. Mine: All are individually and collectively logically only consistent with the hypothesis that this is the best or among the best candidates (especially in the absence of a point-by-point comparison of these 10-item scores of all the other candidates).
(Analogously, such a comparison of the maid with the butler would also serve the cause of judicial and evidential thoroughness.)
To the extent that you may believe that either individually or collectively these ten items are also confirmatory, to what degree are they—again, individually or as a set?
If you believe you would have to think about that, compare that list with the following cited in the context of charging the butler with the murder of his employer. The butler
1. was the only other person known to be in the house at the time of the murder.
2. had a ferocious argument with his employer about holidays the week before
3. was the only other beneficiary, besides the maid, identified in the employer’s will
4. was having an affair with the maid, who was also having an affair with the employer.
5. was familiar with the use of a pistol
6. had the employer’s blood on his hands, which he claimed was the result of attempting to render assistance upon discovering him
7. had an ailing mother who had to have expensive, life-saving medical procedure within six months
8. had bought one-way tickets for two to Fiji for the following month
9. had one of the only other two keys to the house
10. did not have an elegant British accent
Notice that, as is the case with the first list, every one of these individual items is, on the face of it, merely consistent with the suspicion that the butler did it. Similarly, if any or all of them were not the case, grave doubts, proportional to the number or weight of the items discounted in this way, would be raised about the wisdom of singling him out.
Whether, in any given jurisdiction, all of that purely circumstantial evidence would suffice for a conviction is moot (even though some experts say that it should not be enough—that there should also be at least one item of direct evidence, for example or namely, the testimony of an eye-witness). But this much seems certain: a pile of purely consistent, circumstantial evidence seems far more compelling in a trial than in a candidate assessment.
So, when you are vetting your candidates and the “evidence” (individually or collectively) for or against hiring them, just ask yourself and answer the general version of the manager’s question, above, that sparked this entire analysis: “Does what you know about the candidate confirm (s)he’s the best choice, or is what you know merely consistent with that conclusion?”
If you do that, you should be able to consistently confirm that your hiring decisions are on the most solid ground possible.