How to check 3.1.5: Reading Level?

[jake-abma]

Wondering what approach to take if I have to check multiple pages in multiple languages, or even one page in one language, to get a valid result, and what that result actually means.

So I want to check the following text: http://webtestingcourse.dequecloud.com/

Welcome to Gefälscht CompuTech Gefälscht - German for "Fake". This is a fake website, created by Deque Systems, for use during training. Our clients are pointed to this site so that we can train them using an inaccessible site without having to point to the site of a real company. This site has a number of intentional accessibility problems. It is not an example of how to do things but rather how not to do them. If you're interested in knowing how you can make an accessible website, need help making your website accessible, or want to have your website tested for accessibility, please contact Deque Systems.

From Deque demo page: http://webtestingcourse.dequecloud.com/

I do not know exactly what "reading ability more advanced than the lower secondary education level..." means, so I turn to a checker (which one to use (?!) and in this case I take:

https://readabilityformulas.com/free-readability-formula-tests.php

Add the text and click on "Check Text Readability" and get results, namely:

Purpose: Our Text Readability Consensus Calculator uses 7 popular readability formulas to calculate the average grade level, reading age, and text difficulty of your sample text.

Your Results:

Your text: Welcome to Gefälscht CompuTech Gefälscht - Germ ...(show all text)

Flesch Reading Ease score: 58.8 (text scale) Flesch Reading Ease scored your text: fairly difficult to read. [ f ] | [ a ] | [ r ]

Gunning Fog: 10.5 (text scale) Gunning Fog scored your text: hard to read. [ f ] | [ a ] | [ r ]

Flesch-Kincaid Grade Level: 9.4 Grade level: Ninth Grade. [ f ] | [ a ] | [ r ]

The Coleman-Liau Index: 9 Grade level: Ninth Grade [ f ] | [ a ] | [ r ]

The SMOG Index: 7.6 Grade level: Eighth grade [ f ] | [ a ] | [ r ]

Automated Readability Index: 9.2 Grade level: 13-15 yrs. old (Eighth and Ninth graders) [ f ] | [ a ] | [ r ]

Linsear Write Formula : 9.5 Grade level: Tenth Grade. [ f ] | [ a ] | [ r ]

I see: fairly difficult to read, hard to read, 9 Ninth Grade, 7.6 Eighth grade, Tenth Grade

My conclusion: "It is hard to read, difficult, but also fits a time period of 6 years + 2 or 3 years period, fitting the definition...

... so I PASS (but still difficult to read...?!)

Other ways? Is this a valid way? What way to go from here?

[alastc]

I suspect this is AAA because it is not that easy to test... I would agree that it fits under the threshold, and for readability that's what you can do (pending further testing procedures in WCAG 3.0).

[jake-abma]

@alastc wrote:

I suspect this is AAA because it is not that easy to test

But what is 'a' way? I really have no solid clue... Whenever a SC is created is should deliver a proper way on how to test, isn't it?

The Sufficient techniques may help, for sure, but they don't seem to align with the requirement of the normative text...

[jake-abma]

see for example: https://www.w3.org/WAI/WCAG22/Techniques/general/G153

This is not a prove of "lower secondary education level", and should not be Sufficient...

[alastc]

For the purpose of WCAG 2.x, pick your test (or aggregate test as you found) and work on the content until you think it meets the threshold.

For a language based requirement in the 2.x framework, that's what we have. That's also why it is AAA, because the answer doesn't really fit a binary answer.

[jake-abma]

hmm, probably as you say, but then the test rules change big time for AAA (and how to deal with it), and also the compliance rules (as your strict test rules of A and AA don't apply) IF you want to add AAA SC in your statement, don't think that loose end serves the purpose well.

training / working with different people who want to know the rules of WCAG make this difficult, as suddenly the AAA are different from A and AA, but are setup and pretty much look like they should follow the same path...

[alastc]

For us, so few people request that we test or train to AAA, and for those that do it's generally a case of running a workshop to establish which aspects make sense to implement, and then work out how.

Some are relatively easy (e.g. colour contrast depending on your brand colours), some are things for their editorial handbook, some things are just not feasible for some orgs (e.g. pronunciation). But overall, I think taking them on a case-by-case basis is most appropriate.

It's worth having a look through GreggV's feedback on Silver for the rational for having them in the same doc.

But unless you have a practical suggestion for change in 2.x, I think we can close this.

[jake-abma]

will close, but conclusion is still = "can't test, can't measure, but do your best to make it simple..." as we have no solid way of a valid result.