Expand Readability recommendations

[lslalbai]

The advice on the Readability page suggests using certain tools to assess readability. However, all the tools suggested use an Automated Readability Index, which attempts to assess readability based on characters or syllables per word and per sentence.

I recommend you add a "common words" test as well such as Dale-Chall. This can be an initial or supplemental test to complement the others. It is more generous to common words that may incidentally have more characters or syllables than the simple automated tests above. An online implementation is available. Using both an automated and a "common words" test can be helpful especially on texts where the two diverge. In that case, user testing can help you understand which one might be more representative of readability in your context with your audience.

[nonprofittechy]

Thanks, this is a great suggestion. FYI, we have been brainstorming a more informative measure than "readability" that helps capture whether a form is easy or hard to use. We are thinking along "complexity" lines.

[lslalbai]

Thanks! In our user testing goals we describe it as "comprehensibility". It's a qualitative, not a quantitative, measure, and given it's obtained through user testing, it's fairly expensive and slow to obtain. I'd be curious to see if there is a faster, cheaper, and more informative measure.

That said, we are starting to gather enough data to be able to better correlate "readability" with "comprehensibility". We just haven't done that work yet.

[nonprofittechy]

Who is the "we"? Would be interested to learn more since we're actively developing this.

[lslalbai]

My team at Lone Star Legal Aid. This is part of our work developing Texas Disaster Legal Help and the LSLA self-help resources.

[BryceStevenWilley]

I missed this conversation, but I too have been doing some thinking about readability scores, and glad to realize we came to similar conclusions (some mix of scores, Dale-Chall is good, but scores aren't the end-all-be-all)! It's difficult because things like Hemingway or WriteClearly have the right idea in that their suggestions are mostly independent of the reading scores, but the reading scores they use aren't the best IMO. I might try to take on rewriting our advice here to match those ideas.