13:58:56 <RRSAgent> RRSAgent has joined #eval
13:58:56 <RRSAgent> logging to http://www.w3.org/2013/08/01-eval-irc
13:58:58 <trackbot> RRSAgent, make logs world
13:58:58 <Zakim> Zakim has joined #eval
13:59:00 <trackbot> Zakim, this will be 3825
13:59:00 <Zakim> ok, trackbot; I see WAI_ERTWG(Eval TF)10:00AM scheduled to start in 1 minute
13:59:01 <trackbot> Meeting: WCAG 2.0 Evaluation Methodology Task Force Teleconference
13:59:01 <trackbot> Date: 01 August 2013
13:59:25 <shadi> zakim, this is eval
13:59:25 <Zakim> ok, shadi; that matches WAI_ERTWG(Eval TF)10:00AM
13:59:43 <shadi> zakim, who is on the phone?
13:59:43 <Zakim> On the phone I see Liz
14:00:03 <Zakim> +Shadi
14:00:17 <MartijnHoutepen> Just received e-mail from Eric, he's cancelled this meeting
14:00:36 <shadi> oh
14:01:35 <MartijnHoutepen> His phone and internet are both down
14:02:01 <shadi> thanks martijn
14:02:55 <Kathy> Kathy has joined #eval
14:03:05 <Zakim> +Kathy
14:03:16 <Zakim> + +1.510.334.aaaa
14:04:37 <shadi> zakim, aaaa is Peter
14:04:37 <Zakim> +Peter; got it
14:04:49 <Vivienne> Vivienne has joined #eval
14:04:58 <shadi> Liz: like the proposed text for confidence
14:05:28 <shadi> ...don't think can have more rigor
14:07:16 <Zakim> +??P22
14:07:31 <Vivienne> zakiml, ??P22 is me
14:07:38 <Vivienne> zakim, ??P22 is me
14:07:38 <Zakim> +Vivienne; got it
14:09:52 <shadi> Shadi: no meeting next two weeks
14:10:01 <shadi> ...but watching the list for emails
14:10:27 <shadi> ...putting together an updated Editor Draft
14:10:28 <shadi> ...please engage in online discussions
14:10:50 <shadi> Vivienne: have student working on random sampling generator
14:11:08 <shadi> Liz: Peter suggested pattern recognition
14:11:28 <shadi> Vivienne: forwarded suggestions to PhD supervisor
14:11:54 <shadi> Liz: for more complex applications it may need more "intelligence"
14:12:27 <shadi> Peter: did some research on statistical sampling methods
14:12:31 <shadi> ...looking at statistical rigor and certainty
14:12:49 <shadi> ...relating to sample size and variance
14:13:10 <shadi> ...several examples looked at polling groups of data
14:13:28 <shadi> ...including statistical significance and margin of error
14:13:59 <shadi> ...quite a lot of knowledge that could be productively applied
14:14:27 <shadi> ...that would be a good thing to bring in if a student has the skills
14:15:01 <shadi> ...tool could support "knowledge" about the website
14:15:02 <shadi> q+
14:15:28 <shadi> ...concerned about how this applies to web applications
14:16:12 <shadi> ...have an example of an application that has a save button that appears on several pages
14:16:35 <shadi> ...pressing that button inserts a link with an accessibility error
14:16:51 <shadi> ...not able to predict this behavior
14:17:03 <shadi> ...can only talk about what was identified and tested
14:17:47 <shadi> Vivienne: another example I have is where an application is actually a complete process
14:18:00 <shadi> ...so actually can't sample but need to test all steps
14:19:28 <shadi> Peter: agree that not accessible if step cannot be accomplished
14:19:34 <shadi> ...but what if one error appears somewhere on these steps
14:19:49 <shadi> ...does that prevent a person from using the application?
14:20:27 <shadi> ...WCAG does not distinguish the severity of violations
14:20:41 <shadi> ...but in pratice it may sometimes not be a problem
14:21:30 <shadi> ...need to make the report useful in the real world
14:21:40 <shadi> ...to deliver this type of information
14:22:13 <shadi> Vivienne: agree, sometimes errors have varying levels of impact
14:23:26 <shadi> ...Roger Hudson worked on a "calculator" to help prioritize and identify the major errors
14:23:57 <korn> korn has joined #eval
14:25:32 <Vivienne> Roger Hudson's tool can be downloaded from: http://usability.com.au/2013/01/accessibility-priority-tool/
14:25:46 <Zakim> -Kathy
14:26:27 <shadi> Shadi: model of "minimum process" to assess a website with sufficient confidence
14:26:43 <shadi> ...then additional requests by the Commissioner
14:26:53 <Vivienne> q+
14:26:59 <shadi> q-
14:27:28 <shadi> Peter: probably need to distinguish between static websites and web applications
14:27:37 <shadi> ...no clear line between these
14:27:51 <shadi> q- Vivienne
14:28:04 <shadi> Vivienne: agree with Peter
14:28:14 <Zakim> +Peter_Korn
14:28:20 <Zakim> -Peter
14:28:27 <shadi> ...more likely to miss specific interactions in web applications
14:30:00 <shadi> Shadi: back to the random sampling generator
14:30:29 <shadi> ...if the tool has more "knowledge" about the website, isn't that then structured sampling?
14:31:03 <shadi> Peter: the more you know about the website, the smaller sample size you need to maintain the same margin of error
14:31:33 <korn> See http://en.wikipedia.org/wiki/Sampling_(statistics)#Sampling_methods
14:32:04 <korn> See "Systematic sampling", which appears to be the term
14:32:15 <korn> Also "Stratified sampling"
14:33:28 <shadi> Shadi: random sampling was intended as a back-check the structured sampling
14:34:07 <shadi> Peter: question in stratified sampling is to know what you may have missed
14:34:27 <shadi> ...interesting article above
14:35:09 <korn> See "Practical example" at http://en.wikipedia.org/wiki/Stratified_sampling
14:35:46 <shadi> Vivienne: tool is designed to generate a purely random sample
14:35:54 <shadi> ...without prior knoweldge
14:36:05 <shadi> ...think Peter is suggesting enhancements
14:36:27 <shadi> ...but may not be possible within this student project
14:37:03 <shadi> Liz: what Peter suggests is not that hard
14:37:04 <korn> q+
14:37:28 <shadi> ...if one can come up with classifications for the pages
14:37:28 <shadi> q+
14:37:33 <shadi> ...for example pages with tables, forms, etc
14:38:38 <shadi> ack k
14:39:14 <shadi> Peter: classifying pages then sampling from those is probably achievable
14:39:28 <shadi> ...but might not be what we want
14:40:27 <shadi> ...for example for pages that have tables, many may have been produced using template A and others using template B
14:40:37 <shadi> ...need to ensure not to miss these few pages
14:40:47 <shadi> Vivienne: fabulous feedback
14:41:28 <shadi> Liz: instead of stratified, can call it "choice based"
14:41:48 <shadi> ...choose the classes then it will sample from them
14:47:28 <shadi> Shadi: remember a project looking at page elements and calculating "indicators" from that
14:47:44 <shadi> ...from those they created "clusters" of "similar" web pages
14:47:59 <shadi> ...and then selected from each of these clusters representatives
14:48:34 <Zakim> -Peter_Korn
14:48:36 <Zakim> -Shadi
14:48:40 <Zakim> -Liz
14:48:50 <Zakim> -Vivienne
14:48:51 <Zakim> WAI_ERTWG(Eval TF)10:00AM has ended
14:48:51 <Zakim> Attendees were Liz, Shadi, Kathy, +1.510.334.aaaa, Peter, Vivienne, Peter_Korn
14:48:56 <korn> korn has left #eval
14:49:01 <shadi> trackbot, end meeting
14:49:01 <trackbot> Zakim, list attendees
14:49:01 <Zakim> sorry, trackbot, I don't know what conference this is
14:49:01 <Liz> bye
14:49:09 <trackbot> RRSAgent, please draft minutes
14:49:09 <RRSAgent> I have made the request to generate http://www.w3.org/2013/08/01-eval-minutes.html trackbot
14:49:10 <trackbot> RRSAgent, bye
14:49:10 <RRSAgent> I see no action items