You are here : Home > 2007 evaluation results
2007 evaluation results
The results of the 2007 evaluation are presented in three ways: as a whole, by category, and individually for each site evaluated. Regardless of how the data is grouped, the first column in the results table displays the mention obtained in the form of a title, a score out of ten, and a letter of appreciation. The subsequent columns display the error rate in each of the four success criteria for priorities 1 and 2.
The detailed results of each page evaluated (three pages per site) and the types of errors found are not divulged here. They can be found in the detailed individual site reports and can be purchased online by clicking on the download icon located in the last column of the results table, next to the item of interest.
The 200 Web sites evaluated were chosen from ComScore’s Top 500 list. We deliberately excluded sites of a pornographic nature and certain sites which could not be identified from the Web site owner’s name. In fact, the list takes into account 199 Web sites, since a contest site is no longer online.
Overall results
This first table shows the average score for all 200 Web sites evaluated, all categories combined. Please refer to the legend at the bottom of this page to better understand the results and the meaning of column headers.
| Mention | Priority 1 errors | Priority 2 errors | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Rating | Score | Grade | P | O | U | R | P | O | U | R | |
| Top 200 Web sites | Weak | 6,04 | D | 86 % | 28 % | 30 % | 96 % | 96 % | 98 % | 87 % | 100 % |
Note : Since the objective of the evaluation is to find areas of improvement for accessibility, the scoring system is based on the number of errors found and not on the number of correct elements on a given page. A score of 0% is therefore considered perfect, whereas a score of 100% means a multitude of errors were found.
To obtain more detailed information on the results of a particular Web site, you can purchase a report through our online ordering service by clicking on the download icon in the “Report” column of the results table, next to the item of interest. To obtain additional information on how to purchase a report, click on the image "Buy your 2007 report" in the right hand column of this page.
Results by category
This second table displays, from best to worst, the average score by category for all Web sites evaluated in each category. Please refer to the legend at the bottom of this page to gain a better understanding of the results and the column headers.
| # | Category | Mention | Priority 1 errors | Priority 2 errors | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Rating | Score | Grade | P | O | U | R | P | O | U | R | ||
| 1 | Public administration | Good | 7,34 | C | 58 % | 18 % | 11 % | 77 % | 72 % | 93 % | 79 % | 100 % |
| 2 | Education and employment | Weak | 6,60 | D+ | 87 % | 18 % | 15 % | 95 % | 95 % | 100 % | 79 % | 100 % |
| 3 | Social networking | Weak | 6,39 | D | 77 % | 17 % | 32 % | 97 % | 100 % | 97 % | 88 % | 100 % |
| 4 | Financial services | Weak | 6,11 | D | 89 % | 37 % | 7 % | 89 % | 96 % | 93 % | 96 % | 100 % |
| 5 | ICT | Very weak | 5,98 | E | 76 % | 29 % | 36 % | 97 % | 97 % | 100 % | 89 % | 99 % |
| 6 | Goods and services | Very weak | 5,77 | E | 95 % | 24 % | 35 % | 99 % | 97 % | 98 % | 89 % | 100 % |
| 7 | Entertainment | Very weak | 5,75 | E | 93 % | 35 % | 53 % | 98 % | 99 % | 100 % | 87 % | 100 % |
| 8 | Travel and transportation | Very weak | 5,74 | E | 96 % | 24 % | 27 % | 98 % | 100 % | 100 % | 78 % | 100 % |
| 9 | Media | Very weak | 5,48 | E | 94 % | 45 % | 19 % | 100 % | 100 % | 99 % | 96 % | 100 % |
| Error rate | 85 % | 27 % | 26 % | 94 % | 95 % | 98 % | 87 % | 100 % | ||||
Note : Since the objective of the evaluation is to find areas of improvement for accessibility, the scoring system is based on the number of errors found and not on the number of correct elements on a given page. A score of 0% is therefore considered perfect, whereas a score of 100% means a multitude of errors were found.
To obtain more detailed information on the results of a particular category, you can purchase a report through our online ordering service by clicking on the download icon in the “Report” column of the results table, next to the item of interest. To obtain additional information on how to purchase a report, click on the image "Buy your 2007 report" in the right hand column of this page.
Weighting system
The above scores are calculated out of 10 and take into account Priority 1 and Priority 2 checkpoints. During the evaluation, each question is weighted differently to reflect its impact on accessibility. A negative answer to the question gives no point. A positive answer gives the maximum number of points. Percentage-based questions provide a value between 0 and the maximum number of points. The final score is calculated out of 10.
More information?
If you still have questions on how to consult the results, please do not hesitate to contact us!
Our 7 main recommendations
Following the example of 2003, this year’s study allowed us once again to identify the most common Web accessibility issues for people with disabilities. These seven recommendations were selected for their ease of implementation and the major impact they can have on the user experience :
- HTML or CSS coding errors
99 % of sites evaluated contained invalid code and style sheets (errors were found). This can have a significant impact on users of assistive technologies that are less tolerant of these types of errors. The cost of correction is greater because each page must be evaluated and corrected. It should be noted, however, that these errors are often repetitive, and many of them can be corrected by means of a global replacement. Use valid code and style sheets. Implementing this recommendation greatly benefits the technical performance of the site: a substantial economy of bandwidth, a reduction in maintenance and operating costs, the portability and interoperability of information, better search engine indexing, and a better level of basic accessibility. - Unusable content or features without JavaScript
99 % of sites evaluated contained unusable content or features without JavaScript. This has a major impact on people unable to use a mouse because of a motor or visual impairment, making a site partially or completely unusable. Offer an alternative means of navigation for any navigation system dependent on JavaScript or on any other client-side programming technology. This can be done by offering redundant textual links at the bottom of the page or a complete, up-to-date site map using textual links. - Unused or misused headers
96 % of sites evaluated contained unused or misused headers. The impact is significant, as it enables users to save time and quickly explore the content of a page, at a glance, just as unimpaired users do. Applying this recommendation can also improve the site’s ranking in search engines, as certain engines give more importance to keywords highlighted in this manner. Structure pages with headers to allow users with no overall view of the screen to rapidly assess the contents of the page and easily get to where they want to go. It is therefore simple to modify the code to transform a given paragraph intoH1,H2, etc. - Form labels are wrongly associated or missing
93 % of sites evaluated contained forms whose labels were wrongly associated or missing altogether. The impact is significant for blind users who are unable to fill out certain forms because the screen reader software is unable to correctly match up the fields of the form with the right labels. This also has an impact on people with motor disabilities who are limited in their ability to select a checkbox or a radio button. Associating the labels allows them to select the field by clicking anywhere on the label, thereby requiring much less precision. Explicitly associate form labels and fields. This problem affects almost all forms on the Web. This correction is somewhat more expensive to make, as the change has to be applied to every field of each form. - Image links or image map regions without text equivalents
90 % of sites evaluated did not have text equivalents or alternate text (ALTattribute) for certain image links and image map hotspots. This has a major impact on blind users because it can render a site partially or completely impossible for them to use. This also affects search engines, which can index this content (ALTattribute) and improve the site's ranking. Provide text equivalents or alternate text (ALTattribute) for all image links and image map hotspots. - Fixed font sizes
82 % of sites evaluated contained fixed font sizes (pixels, points, etc.). The impact is significant if we consider the number of visitors affected, i.e. users over forty, and those with a slight or moderate visual impairment. The cost of making this correction is minimal because these pages already use style sheets, which only need to be modified once. Use resizable font by specifying relative values. - Images and other design elements without text equivalents
78 % of sites evaluated did not have text equivalents or alternate text for images, photos, and other graphical elements. This problem has a major impact on users who are blind because it can render some of the content impossible for them to perceive. The same holds true for search engines. Provide text equivalents or alternate text for all images, photos, and other graphical elements. The cost of correction is higher because each page needs to be corrected individually.
Legend and context
Definitions
- P / Perceivable (priority 1 or 2) : Ensure that any user can perceive the content. For example, an image without a text equivalent cannot be perceived by a blind person. Also, a sound file without a transcription cannot be perceived by a deaf person.
- O / Operable (priority 1 or 2) : Ensure that all interface elements are operable by any user. For example, a person unable to use a mouse must be able to navigate through content using a keyboard. Data tables and unstructured content make navigating difficult for people who have no overall vision of the screen and have to explore the content line by line.
- U / Understandable (priority 1 or 2) : Make content and controls understandable for as many users as possible. For example, indicating changes in language allows screen reader users to hear content descriptions in their language of choice. Simple language and coherent navigation systems make content more understandable for persons with a cognitive impairment.
- R / Robust (priority 1 or 2) : Web site content must be accessible by current or future technologies, including adaptive technologies (software or hardware allowing persons with disabilities to use a computer independently in order to access information).
Mention
The overall mention of a site is obtained by calculating the average mention of all individual pages :
- 10,0 = Meets requirements (AA)
- 9,5 à 9,99 = Excellent (A+)
- 9,0 à 9,49 = Excellent (A)
- 8,5 à 8,99 = Very good (B+)
- 8,0 à 8,49 = Very good (B)
- 7,5 à 7,99 = Good (C+)
- 7,0 à 7,49 = Good (C)
- 6,5 à 6,99 = Weak (D+)
- 6,0 à 6,49 = Weak (D)
- 3,0 à 5,99 = Very weak (E)
- 0,0 à 2,99 = Extremely weak or nil (F)
- Shared licence
. Certain rights reserved.
Study completed by AccessibilitéWeb. Development : WebConforme. Design : QuiboWeb. - Last updated : December 6, 2007.
- Accessibility.