Societal Inclusion: Evaluating the Accessibility of Job Placement and Travel Web Sites

Societal Inclusion: Evaluating the Accessibility of Job Placement and Travel Web Sites

Author(s) name(s): Jonathan Lazar, Brian Wentz, Dana Biggers, Jason Delair, Matthew Donnelly, Eludoyin Kashim, Andrew Henin, James Markakis, Angela Matos, Andrew McNicol, John Nixon III, Richard Osborne, Tatyana Postnova, Jeevakumar Raja, Roland Roberts, Harry Serra III, Vasilios Sfakianoudis, Vincent Tyler, Junhyeok Yun

Author affiliation: Towson University, Maryland, USA

To be published in the proceedings of the INCLUDE 2011 conference.

Abstract

People with disabilities face many barriers when using web sites. Frequently, web sites are inaccessible to those using assistive technologies, such as screen readers and alternative keyboards. Government web sites are often required by law to be accessible (in countries like the US, UK, Canada, and Portugal), but private web sites are typically not required to be accessible. However, some categories of inaccessible web sites can lead to forms of discrimination which are illegal, such as hiring discrimination and pricing discrimination. Job web sites, such as Monster, and travel web sites, such as Travelocity, are aggregators, providing access to information from other sources, even though the web sites themselves do not primarily offer jobs and provide no travel services. While individual employers and airlines may be required to avoid discrimination, it is rarely clear what requirements apply to aggregator web sites—those that summarize and provide data from multiple different sources. In this research, eight major job aggregator web sites and eight major travel aggregator web sites were evaluated for accessibility. Expert accessibility inspections were performed using the web accessibility guidelines from the US Section 508 regulations. This paper will present results from the expert inspections and discuss the implications for interface developers and public policy makers.

Keywords

Accessibility, assistive technology, blind, visual impairment, employment, discrimination, job web sites, travel web sites

Introduction

There are approximately 49.7 million individuals in the US with disabilities [15], and the World Health Organization estimated in 2003 that there were 600 million people worldwide with disabilities [21]. Unemployment is high for individuals with disabilities, and it is disproportionately high for particular impairments and disabilities. For example, the unemployment rate for working-age blind people in the US is between 70-75% [10], and approximately two-thirds of working age blind and partially sighted people in the UK are unemployed [13]. Some governments have focused on this disparity, with President Barack Obama (in the US) recently issuing an executive order that instructed federal agencies to increase the employment of those with disabilities [11].

People with disabilities often face problems with using technology, especially web sites, because of the way that most web sites are designed. Web sites should be designed so that they work properly with multiple forms of assistive technology, such as screen readers, keyboard-only input, and alternative pointing devices. Web accessibility is frequently defined by whether or not a web site meets legal requirements, such as Section 508 in the United States, or international standards, such as the Web Content Accessibility Guidelines (from the World Wide Web Consortium). While Section 508 (an addition to the Rehabilitation Act of 1973 that has been effective since 2001) has been the primary regulatory focus in regards to electronic accessibility for government technology, it has been widely adopted by private entities [3] and state laws [5] in the US. In the UK, the Disability Discrimination Act specifies accessibility standards for web sites, though its focus has also been for government web sites [2]. This has been updated by the Equality Act 2010, which attempts to strengthen the impact of the previous legislation, including pre-employment disability-related inquiries [4]. Many governments have based their web accessibility standards on WCAG 1.0 (the web content accessibility guidelines as proposed by the World Wide Web Consortium), but they often modified the guidelines to make them less subjective and easier to test and implement. Currently, national governments around the world are evaluating how to implement the new WCAG 2.0 as legal regulations.

Two areas of importance related to disability rights include employment and travel. At first, they seem like separate, unrelated issues, but both topics relate to independence. Both employment and travel are areas in which people with disabilities have legal rights. For instance, in both the US and the UK, a person cannot be discriminated against in hiring due to having a disability. And, in both the US and the UK, a person cannot be discriminated against in travel due to having a disability. But there is another similarity related to interface design: much of the employment data and the travel data available on the Web are available through aggregator web sites. So, while there may be laws or regulations covering how employers may advertise job openings, such as the Americans with Disabilities Act (ADA) [16], and how airlines must offer airfares [17], these laws do not specifically address aggregator web sites. These aggregator sites are legally separate from both employers and travel providers, and provide no actual services related to travel or employment. No case law or statutory laws specifically address aggregator web sites, although in 2004, the New York State Attorney General required

Priceline.com to become accessible. The advanced notice of proposed rulemaking from the US Department of Justice, titled “Accessibility of Web Information and Services Provided by Entities Covered by the ADA,” ( would enact regulations that would clearly cover these types of aggregator sites.

Background on Employment and Employment Web Sites

In 2008, the Society for Human Resource Management in the USA found that job postings on company web sites and national online job boards (such as CareerBuilder.com) were two of the top three methods for employee recruiting (the third method was employee referrals) [14]. However, such aggregator sites are only a part of the problem. Individual employers often have their own online application processes. A previous study of 31 corporate e-recruiting web sites in 2002 found that accessibility of these web sites was lacking [1]. Added to this is the fact that all of the companies in the Fortune 500 currently have online application processes [22].

Technology has dramatically changed the hiring process for many employers and prospective employees. Electronic processes have the capability to level the playing field for individuals with disabilities, in that the many misconceptions regarding disabilities can be removed from the equation for at least the initial part of the job search process. This is important, because research has shown that even seemingly small details such as the visual formatting of an electronic resume can impact the employment opportunities for individuals with disabilities [20]. The challenge is that there are often accessibility problems with web sites that prevent them from being fully functional for individuals with disabilities.

Background on Travel and Travel Web Sites

Travel web sites have become extremely popular as people worldwide rely on them to obtain travel prices from a variety of sources through a single web site. The benefits of this are travel flexibility, the ability to comparison shop, and sometimes lower prices than can be found on the primary web site of an airline or hotel. Due to the dominance of online travel booking, calling an airline or hotel for a reservation over the phone can result in higher prices as well as booking fees. A recent study comparing the accessibility of major airlines in the US noted that many of the airline web sites suffer from accessibility problems. With an inaccessible web site, an airline is legally obligated to offer the same price over the phone when people identify themselves as having a disability, but some of the airlines are not in full compliance to the regulations in this respect [8]. Since airlines and hotels commonly partner with aggregator web sites (such as Priceline or Travelocity), the question is whether those web sites must also be accessible, since they are essentially being used to “outsource” the public accommodation that the parent company would normally provide.

In the US, Department of Transportation regulations mandate that people with disabilities cannot be discriminated against in the pricing of airline travel [17]. In the UK, it is also illegal for an airline to discriminate with its flight reservation and booking web site [2]. While compliance to these regulations is still lacking [8], the legal, non-technical question is whether aggregator sites are covered under current laws.

Research Methodology

There are three potential methods for accessibility evaluation: automated evaluation, expert inspection, and having users with disabilities perform usability testing. Many accessibility evaluation efforts utilize automated software, such as A-Prompt, Deque WorldSpace, InFocus, RAMP and WebXACT. However, automated evaluations are not as accurate as expert evaluations or user-based testing, because many of the accessibility problems (such as link titles or alternative text that do not relate to content) cannot be ascertained by automated means. Expert inspections using a screen reader are considered to be the most effective form of accessibility evaluation [9], and the accuracy improves when multiple evaluators inspect the same interfaces [5]. Inspections by experts without visual impairments can effectively assess compliance with technical and legal standards, such as the Section 508 standards for web sites.

There are major differences between expert inspections and usability testing. Expert inspections tend to focus on technical accessibility, rather than true usability. Expert inspections often serve as a comparison between a set of accessibility guidelines, and a set of interfaces. That means that expert inspections are most effective in ascertaining whether a web page or web site is in compliance with the law. On the other hand, usability testing, involving people with disabilities, provides a higher understanding of ease of use. As an example, interfaces, such as audio CAPTCHAs, may technically be accessible for blind users; however, blind users still cannot complete the tasks at a 50% success rate, so they are impossible to use [7]. Usability testing is task-focused, where users with disabilities attempt to complete a series of tasks, however it is not inspection focused. As such, usability testing may not ascertain compliance with the law, since the users may not wind up testing all of the various components of the regulations, in attempting to complete their tasks.

Due to the various strengths and weaknesses, the combination of expert inspections and usability testing with people with disabilities is ideal, however, is it rarely performed, due to the time and cost involved. In short, expert inspections give a more thorough view, and are more effective at evaluating legal compliance, but usability testing is more effective at determining task success and ease of use. For this evaluation study, funding was not available to pay participants with disabilities to perform usability testing, and we were interested in investigating legal compliance, so expert evaluations were the most appropriate. The expert evaluations that we conducted utilize the web site accessibility standards of Section 508 (1194.22) of the US Rehabilitation Act, which involve 16 guidelines, identified as paragraphs “a” through paragraph “p” [18].

Expert inspection consists of an expert with vision inspecting a web site using a screen reader and a checking for compliance with each specific guideline from Section 508. One of the benefits of having an expert with vision inspect the web site while listening to a screen reader is that is it possible to ascertain if there was any content available on the screen to a visual user that is not available via audio. Subsequently, the expert inspected the underlying code to apply guidelines where code inspections are necessary. The process was guided by the “Absolute Minimum Accessibility Inspection” which was developed by the author and used effectively in other accessibility evaluations [5]. Each paragraph of the guidelines was weighted equally. We did not measure the number of times that a specific paragraph was violated, only the existence of a paragraph violation, because the number of paragraphs violated is considered to be a more accurate measurement of accessibility [6]. Table 1 presents a list of the guidelines, along with a short description of each of the guidelines (note that the descriptions are those of the authors, and not a part of the regulations).

Table 1: Short Description of Each of the 16 Paragraphs of the Section 508 Web Guidelines

(a) Text Equivalent (have a text equivalent for any graphical elements)
(b) Synchronized Equivalent Alternatives (have captioned video, transcripts of any audio, or other alternatives for multimedia)
(c) Use of Color (color should not be used as the only method for identifying elements of the web page or any data)
(d) Organization (style sheets are encouraged, but users should still be able to utilize a web page when style sheets are turned off)
(e) Redundant Text Links on Server-Side Image Map and (f) Client-Side Image Maps (redundant clickable links for server-side image maps, and accessible client-side image maps are preferred)
(g) and (h) Row and Column Headers (use appropriate headers and markup to allow easy navigation of a table)
(i) Frames (title all frames and label all frames for easy identification and navigation, e.g., use “navigation” “main content” and “search” rather than “top” or “bottom”)
(j) Screen Flicker Frequency (limit or eliminate the use of flickering, which can provoke seizures)
(k) Text-Only Page Default (if a web page cannot be made accessible, provide an equivalent text-only page, and make sure it is kept up to date)
(l) Scripting Languages (make sure that equivalents for any non-accessible scripting are included, e.g., for those who are not using pointing devices)
(m) Linked Plug-In or Applet (if any plug-ins are required, make sure to provide a link to an accessible version of the plug-in)
(n) Online Electronic Forms (all forms must be properly labeled and accessible)
(o) Method to Skip Repetitive Navigation Links (all web pages should have a link which allows a user to skip directly to the main content, bypassing any site navigation information)
(p) Alerts on Timed Responses (if any page responses are timed, the user should be given the opportunity to indicate that more time is needed)

We selected eight travel aggregator web sites, and eight job employment aggregator web sites, which are popular in the United States (refer to Table 2).

Table 2: Employment Aggregator and Travel Aggregator Web Sites Selected for Evaluation

Employment Aggregator Sites: / Travel Aggregator Sites:
Monster / Travelocity
Craigslist / Priceline
CareerBuilder.com / Hotwire
Dice / Hotelbook.com
TheLadders.com / Expedia
CollegeRecruiter.com / Orbitz
Yahoo! HotJobs / Kayak
Jobfox / Hotels.com

For each web site, four individuals did an individual expert inspection of the home page (but not the rest of the site) using a screen reader, followed-up by a code inspection, and guided by the “Absolute Minimum Accessibility Inspection” approach which has been used in previous expert accessibility inspections [5]. Note that due to technical problems with one data file, there were only three individual inspections of the Craigslist home page. After doing an individual inspection, the four individuals met, discussed the differences between their respective evaluations, interpreted, re-inspected the homepage, and then agreed upon one common evaluation for each web site. This is a common approach from the general usability inspection literature, where multiple expert reviewers will perform individual reviews, then note where each review was lacking, and come up with a combined review which has a higher level of validity than a single individual review [12].

Results

Employment Sites

The employment sites that violated the fewest paragraphs of Section 508 were Craigslist, with no violations, and TheLadders.com, with only one paragraph violated. Monster and Dice violated the most paragraphs with four and five paragraphs violated, respectively. The most common violation was the lack of a method to skip repetitive navigational links (paragraph o). Table 3 shows the results of which employment aggregator sites violated which paragraphs of Section 508.

Table 3: Results of Which Employment Sites Violated Which Paragraphs of Section 508

Para. / Career
Builder / College
Recruiter / Craigslist / Dice / Jobfox / Monster / The
Ladders / Yahoo!
HotJobs
a / X / X / X / X
b
c
d / X / X
e, f
g, h
i / X / X
j / X
k
l / X / X
m
n / X / X / X / X
o / X / X / X / X / X / X / X
p

CareerBuilder.com had violations of three paragraphs of Section 508 (l, n, and o). These included inaccessible drop-down menus that can only be navigated with a mouse, a poorly labelled form (“s_jobtypes” would be read to the user rather than a meaningful label), and no “skip navigation” link. CollegeRecruiter.com also had violations of the same three paragraphs of Section 508 (l, n, and o). These included all the drop-down lists using JavaScript without equivalent alternative on CollegeRecruiter.com, which are inaccessible to keyboard navigation, and also, there are no descriptive labels for the “sign up” form, and no “skip navigation” link.