W3C home > Mailing lists > Public > public-wai-evaltf@w3.org > October 2012

RE: Input for survey about random sampling

From: Moe Kraft <maureen_kraft@us.ibm.com>
Date: Thu, 11 Oct 2012 08:57:30 -0400
To: evelleman@bartimeus.nl
Cc: "public-wai-evaltf@w3.org" <public-wai-evaltf@w3.org>, Vivienne CONWAY <v.conway@ecu.edu.au>
Message-ID: <OF19F9751A.B1990D85-ON85257A94.004677C5-85257A94.004730C1@notesdev.ibm.com>
I have to agree with Eric. It is difficult to put a percentage on large 
sites or applications. My team has recently been reviewing a large retail 
site with over 10,000 pages. The code is pretty static and the failures 
found are repetitive page to page. It is the content that is changing. It 
would be difficult to test 25% of this application which would result in 
2500 pages. 

Where this is a recommendation, I'm not sure we can quantify a percentage 
for the random sample. If we were, maybe it should be a percentage of the 
relative sample. Also, lets not forget about applications where the 
relative sample is really taking a look at user tasks; creating, 
searching, reading and editing documents or performing a basic task. The 
random sampling will differ site to site, application to application. 


Maureen Kraft
Accessibility Consultant Test Lead
Human Ability and Accessibility Center 
IBM Research 

Tel: 978-899-3114
E-mail: maureen_kraft@us.ibm.com 


From:   "Velleman, Eric" <evelleman@bartimeus.nl>
To:     Vivienne CONWAY <v.conway@ecu.edu.au>, "public-wai-evaltf@w3.org" 
Date:   10/11/2012 02:53 AM
Subject:        RE: Input for survey about random sampling

Hi Vivienne,

Maybe we should also add a question about how people conclude the number 
of pages of a website.
Personally I would be interested to know the exact number of pages people 
use. For the BITVtest in Germany it seems to be 3 - 5 pages, for other 
testing schemes it is mostly not more than 30 - 50 pages. If we would take 
20 percent of a website of 1.000.000 pages, that would mean checking many 
pages. In our discussion, there seems to be a tendency towards less pages.

This is also something to discuss in the Telco when we talk about 
confidence. How many pages do you need to have a certain amount of 
confidence in the results.

Kindest regards,


Eric Velleman
Technisch directeur
Stichting Accessibility
Universiteit Twente

Tel: +31 (0)30 - 2398270

Christiaan Krammlaan 2
3571 AX Utrecht

www.accessibility.nl / www.wabcluster.org / www.econformance.eu /
www.game-accessibility.com/ www.eaccessplus.eu

Lees onze disclaimer: www.accessibility.nl/algemeen/disclaimer
Accessibility is Member van het W3C

Van: Vivienne CONWAY [v.conway@ecu.edu.au]
Verzonden: donderdag 11 oktober 2012 5:02
Aan: Velleman, Eric; public-wai-evaltf@w3.org
Onderwerp: RE: Input for survey about random sampling

HI all
I think question 12 should reflect a percentage of pages rather than a 
finite number.  In one of the discussions with the WCAG group they 
mentioned that 25% of the total pages sampled should be randomly chosen. 
Can we re-phrase this question to include percentages?


Vivienne L. Conway, B.IT(Hons), MACS CT, AALIA(cs)
PhD Candidate & Sessional Lecturer, Edith Cowan University, Perth, W.A.
Director, Web Key IT Pty Ltd.
Mob: 0415 383 673

This email is confidential and intended only for the use of the individual 
or entity named above. If you are not the intended recipient, you are 
notified that any dissemination, distribution or copying of this email is 
strictly prohibited. If you have received this email in error, please 
notify me immediately by return email or telephone and destroy the 
original message.
From: Velleman, Eric [evelleman@bartimeus.nl]
Sent: Thursday, 11 October 2012 5:18 AM
To: public-wai-evaltf@w3.org
Subject: RE: Input for survey about random sampling

Dear EvalTF,

Following our discussion last telco, the minutes of that meeting and some 
remarks sent to me, this is an update of the survey. Please send your 
comments to the list. Before we start making this a scientific research 
project, please keep in mind that this is input information for a first 
draft text to add a random sample.

** Concept survey questions about sampling **

Note: You could fill in the survey more than once using different choices 
in the first question. Please note that the Methodology is about full 
websites (as defined in the Public Working Draft).

#01. Please choose the situation below for which you fill in this survey.
Your evaluation activity is:
  - (checkbox) In-house website evaluation for your own organization
  - (checkbox) Website evaluation for external evaluation commissioner
  - (checkbox) A re-evaluation of a website that you have evaluated less 
than 1 month ago
  - (checkbox) In depth evaluation
  - (checkbox) Basic evaluation
  - (checkbox) Detailed evaluation
  - (checkbox) Other - please specify below (text)

Please fill in the questions depending on the evaluation activity you have 
chosen in question #01:

#02. For which purposes do you evaluate websites:
  - (checkbox) Provide in-depth analyses to guide website owners
  - (checkbox) Monitor the accessibility of a website over time
  - (checkbox) Final check before releasing or purchasing a website
  - (checkbox) Provide label or other certificate of conformance
  - (checkbox) Do large-scale evaluation of many websites
  - [checkbox] Pre-sales activities (evaluating websites in order to 
initiate a contract)
  - [checkbox] Evaluate websites in early/medium development stages
  - [checkbox] Deliver training about Web accessibility evaluation
  - (checkbox) Other - please specify below (text)

#03. Do you consult with the evaluation commissioner regarding the sample 
  -(radio) Always
  -(radio) Sometimes
  -(radio) Never
  -(radio) Depend upon whether the website is being done without their 

#04. To what extent do the wishes of the evaluation commissioner influence 
the evaluation sample? 1= no influence (evaluator is independent in choice 
of evaluation sample), 5=total influence (evaluator is not independent in 
choice of of evaluation sample)
  -(radio) 1
  -(radio) 2
  -(radio) 3
  -(radio) 4
  -(radio) 5

#05. Do you select sample pages for evaluation based on:
  - (radio) Structured sampling only
  - (radio) Random sampling only
  - (radio) Automated sampling only
  - (radio) Both structured and random sampling
  - (radio) Both structured and random sampling supported by automated 

#06. Please specify any tools you use to help you select the sample, and 
how you use these tools for sampling. (text)

#07. If you use automated sampling only, how do you check the tool? (text)

#08. If you use structured sampling, how do you select the individual web 
pages? (text)

#09. If you use random sampling, how do you select the individual web 
pages? (text)

#10. How does re-evaluation of a website influence the sample? (text)

#11. Please specify if you sample pages differently or in different 
amounts for Basic Reporting, Detailed Reporting or In-Depth Analysis 
evaluation (text).

#12. What is the amount of pages you use in your sample when evaluating a 
  - (radio) 1 - 5
  - (radio) 6 - 15
  - (radio) 16 - 30
  - (radio) 31 - 50
  - (radio) 51 - 100

#13. If the amount of pages in your evaluation sample is variable, what 
are –in most cases – the factor(s) that have the most influence on the 
size of the sample:
  - [checkbox] Size of the website
  - [checkbox] Number of technologies on the website
  - [checkbox] Amount of money available for the evaluation
  - [checkbox] Number of templates
  - [checkbox] Other, please specify below (text)

#14. How does the size of the website influence your selection? (text)

#15. How do you adjust your sampling approach for websites that are 
heavily template-driven? (text)

#16. If you use automated testing tool(s), does that change your sampling 
  - (radio) Yes
  - (radio) No
  - Please explain (text)

#17. How do you adjust your sampling approach for web applications? (text)

#18. How do you sample mobile applications? (text)

#19. Do you look out for widgets, snippets, and other web page components 
to avoid re-evaluating them every time they appear on a website?
  - (radio) Yes
  - (radio) No
If Yes, how do you include them and the pages that they are integrated 
into the sample? (text)

#20. How do you identify and sample the functionality (processes) provided 
on a website? (text)

#21. Do you select further sample pages based upon results from an initial 
selection (eg. to balance web page types or success criteria)? (text)

#22. Please feel free to provide further thoughts or comments about how 
you approach sampling. (text)

Kindest regards,


This e-mail is confidential. If you are not the intended recipient you 
must not disclose or use the information contained within. If you have 
received it in error please return it to the sender via reply e-mail and 
delete any record of it from your system. The information contained within 
is not the opinion of Edith Cowan University in general and the University 
accepts no liability for the accuracy of the information provided.


(image/jpeg attachment: 01-part)

(image/gif attachment: 02-part)

Received on Thursday, 11 October 2012 12:58:04 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:40:23 UTC