Did this written test in an interview discriminate against people with dyslexia?

Question

I recently had an interview for a job as a CNC machinist. They sat me down for a written test. It was a timed test with more questions that could be answered in the time period.

Most of the questions seemed reasonable for a manufacturing job. Familiarity with measurements, some simple math and trig. Then the questions became rather ridiculous. There was a series of questions like this: two columns each had a list of 10 addresses, but a few addresses in one of the columns would have some letters reversed. How many addresses aren't exactly the same? I can figure this out, but it takes me substantially longer than most people.

Why would they even ask such questions unless they were intentionally discriminating against people with dyslexia? What should I have done? Tell them that I have dyslexia and need more time for the test?

---

It would definitely be easier for me to identify the differences if they were the numbers. Say it takes you 5 seconds to distinguish between: "19874 Evergreen Terrace, Chicago Illinois" and "19874 Evergreen Terrcae". It takes me 15 seconds. If the one question has 10 addresses to compare, it literally took me 100 seconds longer to answer the same question. The question isn't measuring anything I'd do on the job. It's not measuring how intelligent I am. Why ask it unless they didn't want dyslexics to work there?

Answer 1

It sounds more of a test to spot minute differences and typos in long material sheets and such, which is probably a routine job, rather than specifically designed to exclude dyslexic people. I understand how you may feel that its targeted towards you, but its very likely its a misunderstanding, and making a test like that is discrimatory and possibly illegal if its not related to the job. I suggest bringing it up with your interviewer politely, pointing out that you have no trouble with numbers but with words, and still enthusiastic about the job, but do not accuse them right off the bat.

Answer 2

Why would they even ask such questions unless they were intentionally discriminating against people with dyslexia?

Because it seems to be a good test to see whether people can spot small differences quickly. Unlike comparing shapes, it's also easy to administer and grade. I could imagine, spotting small variations from the pattern that is supposed to be produced is something important to the job.

What should I have done, tell them that I have dyslexia and need more time for the test?

If you need special accommodations for a disability, then yes, absolutely, you should have told them. I have no idea why you think a written test is perfectly fine, but this part is explicitly inserted to discriminate against people with dyslexia. Does not make any sense to me.

A company cannot administer tests that accommodate every possible disability at the same time. In many countries, they are legally required to accommodate your special needs on request. The easiest option would be to give you more time on a written test.

But they cannot possibly do so when you don't say something.

The best way to do so would be in advance. For logistical reasons, so they can prepare a different test scenario for you. For marketing reasons, so you can tell them in the same conversation how it never affected your job performance. For bureaucratic reasons, because you probably have some paperwork to submit for your disability. And for credibility reasons, because if you mention your dyslexia only after doing 8 pages of a 10 page written test and hitting a certain type of task, the interviewer will be suspicious.

From the interviewers point of view, the interview is a test on how you will perform as an employee. We all know it's more of an educated guess than science, but it's the best we can do.

I will not hire employees that produce poor results but have a reason for their failure. I will absolutely hire employees that tell me what they need (and "more time" is really the simplest of those requests) and upon getting it produce good results.

Answer 3

Given that a typical CNC program looks like this (though much longer):

N40 G90 G00 X0 Y0
N50 G01 X-10 Y-20 R8
N60 G01 X-50 R10
N70 Y10
N80 X-19.97 Y25.01
N90 G03 X7.97 Y38.99 R18

The ability to quickly assess written data and spot errors in it is a core function of the job, and the test described is meant to check just that. I don't think that you have a leg to stand on in this case.

Answer 4

It would definitely be easier for me to identify the differences if they were the numbers. Say it takes you 5 seconds to distinguish between: "19874 Evergreen Terrace, Chicago Illinois" and "19874 Evergreen Terrcae". It takes me 15 seconds.

Your "terrcae" example is biased because you know that "terrace" is a word and "terrcae" isn't. The argument completely falls apart when you compare "19874 Evergreen Terrace, Chicago Illinois" to "19784 Evergreen Terrace, Chicago Illinois", since both are valid numbers.

You're conflating a spelling test with a reading comprehension test. They're not asking you to know how to spell words, they're testing how accurate you are when reading complex information.

It's not measuring how intelligent I am.

Correct. The test doesn't actually measure intelligence, it measures observance, i.e. the ability to distinguish differences in written information.

The question isn't measuring anything I'd do on the job.

I disagree. You're going to be handed written specifications for the products you're developing. If you're unable to reliably read the specifications correctly, you're going to build things using wrong specifications, which wastes time, money and resources.

Why ask it unless they didn't want dyslexics to work there?

Because maybe this job that relies on reading precise specifications isn't cut out for those who struggle to do so?

Dyslexia is a disability, and reasonable provisions must be made for disabilities.

However, when that disability significantly interferes with a core requirement for the job, then you're still not a great candidate for the job.

It doesn't matter why you're struggling with parsing the information. What matters is that this is an important skill for the job you're applying for, and failing to meet the standard is going to make you a less desirable applicant.

You're not being withheld from the position because you have a disability, you're being withheld from the job because you lack core skills (which may or may not be caused by a disability).

---

I also noticed something interesting when you mentioned:

It was a timed test with more questions that could be answered in the time period.

There are two possibilities here. It depends on whether the company explicitly told you that there wasn't enough time.

If they did tell you that, they are likely trying to see if you're prone to doing half-assed rush jobs when under pressure, or whether you prefer to do things right even if it ends up being slow.

If they didn't tell you that, it's possible that there was sufficient time allotted for the test but you don't think so because you struggled with the subject matter. If so, this is a good indication that this particular job isn't cut out for you.

Answer 5

If the test tests your ability to do the job, it's fair, and not discrimination.

I have dysgraphia, and if a job requires actually writing something down, I cannot do it. Someone in a wheelchair isn't going to be accepted by the NBA, and you're not going to have someone who is legally blind as an airline pilot.

The key word in the ADA is REASONABLE accommodation. It is not reasonable to believe that a job which would include having to correct errors of transposition should not have such things on an exam.

Answer 6

I am assuming this test was administered in the US, given that they seem to have used addresses in American cities.

In the US, the relevant law governing hiring is the Americans with Disabilities Act of 1990. As far as employment is concerned, the main enforcing body is the Equal Employment Opportunity Commission. This is what the EEOC has to say about testing: the ADA "makes it unlawful to use employment tests that screen out or tend to screen out an individual with a disability or a class of individuals with disabilities unless the test, as used by the employer, is shown to be job-related and consistent with business necessity." Employers are obliged "to make reasonable accommodations, including in the administration of tests, to the known physical or mental limitations of an otherwise qualified individual with a disability who is an applicant or employee, unless such accommodation would impose an undue hardship."

What constitutes a "reasonable accommodation" is defined by case law, but it goes beyond the mere convenience of an employer or the ability to do a task at the maximal speed. For instance, the Law School Admissions Council has been compelled to provide extra time, readers, scribes, etc. for people with relevant disabilities; and courts ruled against UPS for having a more stringent policy on driver hearing than is required by the Department of Transportation. If an accommodation--as simple as having a second person double-check the CNC codes, which would seem like a pretty obvious practice for anyone, regardless of dyslexia (something about 'measure twice, cut once'?)--would mean that the candidate could do the job, then it would be discriminatory to exclude that candidate on the basis of dyslexia. Or to use a test that tends to disqualify people who could be successful employees with accommodation.

Moreover, given that sources as diverse as the Houston Chronicle, Indeed.com, study.com, and books on living with dyslexia specifically call out "machinist" as a good career for people with dyslexia--one at which people with dyslexia may even excel due to being visually-oriented reasoners--claims from commenters here that dyslexia would disqualify you for employment as a machinist are not credible.

All of which is to say, it sounds like this test did have a discriminatory impact, which is in conflict with the language of the law and the enforcement pattern. (Though only the courts can actually decide in any individual case.)

And that's what it comes down to in the answer to your question. The law's the law, but what matters is what you can get enforced; and it can be years, even if you win.

If you want, it would not be unreasonable at this point to talk with the local EEOC office about the situation you experienced. I wouldn't expect much from it, but what do you have to lose? The test administrators probably did not specifically intend to discriminate against you, or people with dyslexia generally, so much as just being ignorant of their legal obligations. From the perspective of the law, this does not matter. That said, if it were me, I would not complain to the company directly at this point, because it would only give them the opportunity to come up with an alternate excuse as to why you were not hired, if there actually were an investigation.

---

For next time, what you should do going forward is to be ready to respond immediately. You've learned something from this interview: this kind of discriminatory testing exists and you might have to address it. It sucks and it's unreasonable that you should have to do this, but more preparation will maximize your chances of success. I would suggest:

• Have a specific set of criteria for what questions would prompt you to ask for an accommodation. Write them out so they're clear to you, though you don't necessarily need to bring them to the interview.
• Have specific accommodations in mind that seem reasonable and appropriate to you. Again, write them out, though it's up to you if you want to have the written copy in your back pocket.
• Prepare a script for how you would explain the situation and ask for the accommodation. This will make you more confident, clearer, and less awkward.
• Speak with a doctor in advance and get some sort of documentation that you could provide on the spot if challenged. I'm thinking a letter diagnosing your dyslexia and describing the specific kinds of tasks that are disproportionately challenging (e.g. recognizing transposed letters, but not numbers, etc.)
• Then, next time you see a discriminatory test, go ahead and ask for that accommodation.

Don't bring it up unless you have to. But when you encounter something testing you unfairly like this, you'll be prepared to say something like:

I don't like to bring attention to it if I don't need to, but I have dyslexia. As a result of this medical condition, questions like [SPECIFIC QUESTION EXAMPLE] are disproportionately challenging for me, and I am concerned that this test won't accurately reflect the skills and capabilities I would bring to this job. Is there a more directly job-relevant way I can demonstrate my skills for you? [You could offer to go over some CNC code with them if that's in your skill set, etc.; the goal is a) to show them you want to work and want to show it in a very job-relevant way and b) remind them that they have an obligation to accommodate you, but simultaneously seem helpful in giving them a solution for how to do that.] Alternatively, could I ask that you omit this section of the test from my evaluation/give me extra time on this section/[other accommodation option you've thought of beforehand]?

This would also be a good opportunity to segue into some of the ways your coping strategies have helped you be a better employee and a good influence on your team. The point is that you should always be communicating confidence, positivity, and an eagerness to find solutions--these are traits that people will admire regardless of the circumstances.

If they come up empty, you add "I'm just looking for a reasonable accommodation [using those words specifically is important] that would allow me to demonstrate as best I can the value I would bring to your company." If they're dismissive, that's when you might show the doctor's note/documentation that you're carrying with you just in case. And if they still refuse at that point--well, they're probably breaking the law. It's up to you what you're willing to go through to try to see it enforced--but if they won't follow this law, who knows what other employee protection laws they'll disregard? Nothing left to do at that point but carefully document the experience and try again.

Answer 7

I would go back to that employer, describe that you do have dyslexia, and ask them frankly what they think that this means for you as a CNC machinist. Be completely open with them and expect them to also be so with you.

Answer 8

This answer expands on my deleted comment referenced in another answer. The other answer has the legal side so maybe they should be merged or something.

TLDR: This test does not measure aptitude for the job at hand, and the idea that it could probably comes from mistaken ideas about how skills work. Dyslexics might well do worse at the task without this implying they would do badly at the job. Aspects of the job which other answers have suggested might legitimately be tested by the address test (so that dyslexics performing poorly means they would be poor at the job) are incorrect. Indeed, there is reason to believe dyslexics might be especially good at just those aspects of the job so flagged up in the other answer (handling code).

As defined by the National Health Service, dyslexia is a specific learning difficulty, which means it causes problems with certain abilities used for learning, such as reading and writing. Unlike a learning disability, intelligence isn't affected.

Furthermore, according to an article published in Nature, although the predominant view is that phonological awareness is impaired in developmental dyslexia, readers with dyslexia also have a selective visual impairment, which may impact on reading. In addition to visual processing disorders, it may well be that there are different kinds of dyslexia reflecting the different deficits which can impair the learning of reading. The OP has described the specifically visual difficulties that he had with the sub-phoneme-level manipulations of letters in the employer's test.

For the employer's test to be legitimate performance on the test should predict performance on the job. Originally this was somewhat unfathomable to me because I conceived the job as mainly about machined parts (and precision thereof), so if the test just purported to measure "attention to detail" say, then this presumes transferability of results that is not the case. I would go so far as to say that whole picture of human ability is mistaken; in general we are sufficiently adaptible learners that skill ends up being much more domain-specific. The domain experience dominates performance outcomes much more than nebulous global character traits. (In other words, don't let someone tell you that you can never do X because you are Y - you just need to put in the 10,000 hours practice).

In that sense detecting misspelled addresses tells you nothing about CNC machining. But @Draco-S makes a good point about reading (and detecting errors in) written CNC programs being a core part of the job. Surely detecting mistakes in addresses is sufficiently similar to the same in CNC programs that performance in one may be presumed to transfer to performance on the job?

Maybe a little, but I remain very doubtful. The example in @Draco-S does not look like a densely stacked list of addresses. There is structure in computer code, especially of this archaic tabular type, that is not present in addresses which are closer to freeform text. This matters to (visual) dyslexics, whose reading problems are likely to involve low-level visual pattern encoding problems as early as the precortical pathways from the eyes via the thalamus. (To my mind, the low-level nature of these encoding problems is why the British Dyslexia Association recommends simple interventions, like different-coloured paper, which have found success. IMO as an ex-vision scientist, dyslexia is ultimately not a higher-level 'cognitive' deficit about reading and language but a perceptual one about sensory encoding. Hence presumptions about higher level cognitive traits such as attention to detail are inappropriate).

Dyslexics, like autistic people, appear unusually often in some roles. Programming is one role in which people with dyslexia appear unusually often. That is to say, rather than impairing dyslexics, their phonological deficits have engendered coping strategies which turn out to be very applicable to coding. The involvement of written text is mitigated by auto-correcting IDEs but also mainly, as I said above, that code is much more structured than free text.

Answer 9

In a comment on the question (which a moderator removed: https://chat.stackexchange.com/rooms/110317/discussion-on-question-by-jeff909-did-this-written-test-in-an-interview-discrimi), user "benxyzzy" says:

As an ex-vision scientist, the people here arguing there isa big difference between defects in machine parts versus text gotchas are absolutely correct. The people here believing that any old test purporting to challenge your "attention to detail" is an appropriate way to assess the same for a specific scenario are completely wrong. It is domain specific and the pattern recognition issue that dyslexics suffer from is a specific visual deficit. They're not just "worse at reading" or even "worse at attention to detail" (ugh). Their failure on a test like this tells you nothing about CNC ability.

So, if that is correct, and furthermore if the test was intended as a proxy to measure candidates' ability to correctly identify component codes, and/or to assess the correctness of G code (as several commenters and answers think), then the test does discriminate against those with dyslexia.

Just in general, this test disfavours anyone whose ability to spot spelling mistakes is significantly worse than their ability to do the job of a CNC machinist by comparing codes. Since dyslexia may very well cause a big difference of this kind, and since dyslexia is a protected characteristic (unlike just general poor proof-reading ability), this could be unlawful discrimination.

However, we've all assumed here that the test has to do with the codes used in CNC. It is at least conceivable that there is some other part of the job which requires checking addresses, and they are testing that. Then, you might still question whether that's a necessary part of the job, or whether they could reasonably accommodate CNC machinists with dyslexia by re-assigning those tasks elsewhere. But at least in that situation they would only be testing the skills that are part of the job they want to fill.

Furthermore, and this is important if you plan to follow up at all: even if the test in fact does discriminate against those with dyslexia, it does not follow that the test was designed to discriminate against dyslexia. The answer to, "Why would they even ask such questions unless they were intentionally discriminating" may be as simple as, "because they don't know much about dyslexia, and do not realise that their test is bad". Several people commenting and answering on this question think the test is a good proxy for the real task, whereas people with dyslexia and those with relevant scientific experience think it's a bad proxy. That makes it pretty believable to me that the people setting the test might genuinely (but incorrectly) believe that it's a sensible way to test accuracy on the real task.

Some people writing technical tests have, for some reason, a tendency to try to make things "fun", or "relevant to everyday life", or non-obvious precisely how the question relates to the real job. Whatever their reasons for doing this, that urge is not in itself a deliberate effort to keep people with particular disabilities out of the job. However, the urge will lead to them sometimes inventing needlessly bad proxies for the real task, so it does have that effect. I would argue that, as a matter of urgency and of their legal responsibilities as employers, they should stop it!