RE: Feedback

>(it just slows down reading and increases error rate while reading).

Whilst I've always thought that this is the case for large blocks of text, can you cite a source\research for this?

From: Olaf Drümmer [mailto:olaf@druemmer.com]
Sent: 21 November 2014 15:30
To: Mark Barratt
Cc: Olaf Drümmer; Oscar Cao; w3c-wai-ig@w3.org
Subject: Re: Feedback

DEAR MARK,

On 21 Nov 2014, at 15:09, Mark Barratt <markb@textmatters.com<mailto:markb@textmatters.com>> wrote:


IBM (of all people) did the research to validate this in the last '50s, when the issue was labelling of instrumentation in fast jets.


research like this only applies to small numbers of short pieces of information that one is to **re**cognize (as in: match against a small set of to be expected text items) as quickly as possible. Uppercase only can also be more robust under difficult viewing conditions, like limited lighting, less than optimal viewing distance and such (as may be typical in a fast jet).  It does not apply to situations where the piece of text has information not necessarily known ore expected beforehand, and where are many such pieces of text (it just slows down reading and increases error rate while reading).

THANKS,

Olaf