Authors: Ruth Bijl, Royal New Zealand Foundation of the Blind, Mary Schnackenberg, Royal New Zealand Foundation of the Blind, Clive Lansink, Independent IT contractor.

Introduction

Serious vision loss affects as many as 94,000 New Zealanders according to the 2001 New Zealand Disability Survey.[i] In addition to difficulties associated with living and working in the community, vision loss affects people's ability to access the written word. It is estimated that, globally, less than 5 percent of printed material is made available in formats which are accessible to print-disabled readers.[ii] The Royal New Zealand Foundation of the Blind (RNZFB) provides accessible information to approximately 5,500 library users. The major accessible format produced and distributed by the Foundation to its library members is audio. This has been produced by the Foundation for 70 years. Audio is the preferred format of the RNZFB Library's, predominantly older, patrons.

The Foundation has identified that it needs to introduce new technology to support the distribution of its audio books and magazines. To support this change we undertook research into providing audio books and magazines to patrons via an internet delivery system. Trials conducted demonstrated that this delivery mechanism was achievable and acceptable to older members who do not use computers. Results from this trial are provided in this report.

Background

Age-related vision loss results in thousands of older people experiencing difficulty reading standard print and accessing print-based information. Difficulties in accessing written material are compounded by reduced mobility and confidence outside the home, and transport problems. Rehabilitation agencies and public libraries encourage reading for personal enjoyment and learning. They recognise that access to written material enables participation in society. They also appreciate the importance of access to society's written information as the basis of active citizenship.

The Royal New Zealand Foundation of the Blind (RNZFB) is the primary blindness rehabilitation agency in New Zealand. The RNZFB provides services ranging from Children's Orientation and Mobility and Adaptive Daily Living, Employment Services and Adaptive Technology to Guide Dog Services and Library and Accessible Format Production Services. The RNZFB has an operating budget of approximately $NZ22 million of which only about one third is government funded. The RNZFB has a membership of approximately 11,600 people, two thirds of whom are aged over 65 years.

The RNZFB's vision is that blind and vision-impaired people have the same opportunities and choices as other citizens to participate fully in society. Participation is enhanced by access to literature in formats which are readily accessible to blind and vision-impaired people. One such format is audio. Audio books and magazines are produced and distributed by the RNZFB for the Library's approximately 5,500 library borrowers. The RNZFB Library does not charge members any fees or service charges even for return postage costs.

Public libraries in New Zealand generally have poor collections of unabridged audio literature and do not provide magazines in audio formats. The RNZFB fills this gap through the production and distribution of audio 'talking' books and magazines. RNZFB production focuses on New Zealand literature and includes production of a range of popular magazines such as The Reader's Digest, New Zealand Gardener and the Woman's Weekly.

Currently the RNZFB's Library records talking books and magazines using the Digital Accessible Information SYstem (DAISY) 2.02 standard and distributes titles through postal delivery on Library of Congress 4-track cassettes. Transitioning away from the current cassette-based system is not optional. Cassette-based technology will not be able to be supported beyond around 2010 as the various components, including tape cassettes and machines, are obsolete and are no longer being supported or manufactured.

The internet provides an alternative method of delivery to the more usual postal method. Internet distribution is being considered as an option for the RNZFB's future library service. The RNZFB recently undertook two interconnected research projects to support the development of its audio library services to members. The research included an independent survey of members and a pilot to test an internet-connected digital talking book player. This report sets out the findings from the pilot of the internet-connected digital talking book player.

The research set out to establish:

• whether an operationally satisfactory internet capable system could be implemented;
• whether users with lower technical proficiency could access the prototype system effectively.

A survey commissioned from Phoenix research considered other questions including:

• what barriers and enablers were relevant to library users and potential users, and;
• if a new system were implemented, what would the likely up-take of that service be?

Essentially the RNZFB had three simple questions:

1. Could the 'physical' postal system be replaced with a 'virtual' postal system?
2. Could a senior friendly player be obtained?
3. Could an alternative reading experience meet readers' expectations and needs?

In essence these points can be considered against the terms achievability and accessibility.

Methodology

A two-pronged research approach was undertaken. Firstly, the RNZFB developed an internet-capable audio player, commissioned production of ten prototype machines, installed these in the homes of 40 members in the Auckland, New Zealand region, and analysed the trialists' responses to the player experience. The set-up and installer/instructor experience with trialists were also explored. Secondly, the RNZFB commissioned an independent market research company to survey 400 members - both library users and non-library users regarding library services and hypothetical changes to services. Two specific concepts were tested, one a CD-postal based service, the other, an internet based service. This paper discusses the results of the trial. The survey results have been reported separately.

The internet-capable player trial included:

1. development of a prototype internet-capable player,
2. a screening questionnaire,
3. installation of prototype players in trialists homes,
4. in-home trialling by 40 RNZFB Library patrons,
5. pre- and post-test questionnaires designed to illicit trialists' perceptions and experiences,
6. an observational approach to check practical skill acquisition,
7. a focus group with installer/instructor staff.

The prototype player trial ran in members' homes from 15 August 2006 until 22 February 2007.

The New Zealand setting and player conditions

Geographically, New Zealand is long and narrow. The total population is just over four million people. The distance from the RNZFB Library in Auckland to the furthest borrowers in Invercargill is 1186 kilometres (737 miles). It can take from one to six days for a book or magazine to travel from Auckland to the borrower. The postal service does not collect mail from the borrower. When the borrower has finished reading the book, they have to get the postal containers back to the nearest postbox which may be on their street or some kilometres away.

Not all RNZFB members are confident about moving about beyond their own home. This is particularly so with older members. Consequently, returning a book can create a significant barrier to accessing books and magazines in accessible formats. To overcome this barrier the option of an 'automated postie' was conceived by Clive Lansink, a member of the RNZFB and an IT professional. The concept conceived was not a robotic postal system but a virtual postal system. This could be achieved technically through the extension of the digital technology that the RNZFB had already applied to its production facilities.

An additional advantage of digital distribution was that issues of supply and demand would be minimised. Unlike a physical stock of books, the RNZFB would no longer have to worry about having sufficient copies to satisfy demand or removing excess stock.

A good proportion of the country can access broadband. Well over 90% of New Zealand households have access to ADSL broadband. Furthermore, downloading digital talking books off the internet is commonplace. However, not all RNZFB members had access to the internet and access to a PC to download talking books and the confidence to use it are less common amongst members. Less than 10% of the RNZFB library membership has access to the internet with the knowledge of screen readers to use an online public access catalogue (OPAC), let alone the confidence to download books.

As an idea, digital distribution seemed to offer benefits for RNZFB members and the library service. However, it was necessary to establish whether digital distribution was achievable for the RNZFB and acceptable to RNZFB library members. Establishing whether the concept was both achievable and acceptable required the development of a functional player and testing with a target population.

Development of a prototype player

A range of delivery systems were considered for the trial. Some were already in use internationally although not been tested in the New Zealand context. Most commonly, postal compact disc (CD) -based alternatives are in use in the United Kingdom by the Royal National Institute of the Blind (RNIB) and in Canada by the Canadian National Institute for the Blind (CNIB). The RNZFB considered that sufficient information on postal CD-based systems could be gathered from international sources and results extrapolated for the New Zealand context. The RNZFB considered that significant additional benefits could be obtained through an automated postal system and sought to trial an internet-based delivery system - the 'automated postie'.

The RNZFB commissioned HumanWare to build an internet capable talking book player. HumanWare built the player inside a standard Victor Reader Classic. The Victor Classic used for this trial worked as normal. However, the CD-ROM drive was replaced with a pocket personal computer (iPAQ). The iPAQ is wireless capable and was set up to connect wirelessly to a broadband modem supplied and connected by the RNZFB. Wireless connectivity allowed borrowers to move the players around their homes without the need to plug them into anything other than the electric power.

The players were equipped with one gigabyte of memory. The amount of memory was deliberately limited so that borrowers would be forced to test the 'automated postie' aspect of the internet service. To further ensure the 'automated postie' was trialled, borrowers were limited to just three books at a time.

The CD eject key on the player was given a new function. Essentially it functioned as a delete function but it also sent a signal to the RNZFB server. However, the language used around the process was carefully chosen to replicate current borrowing practice. When the eject key was pressed the message prompt said:

"Do you want to return this book to the Library? Press the play key to confirm or any other key to cancel."

If the borrower confirmed the return, the book was erased from the player's memory and the Library server told the book was 'being returned'. The server then signalled the Library Management System and initiated selection of the next book for the borrower. A message was sent to the player with the Uniform Resource Locator (URL) of the book to be downloaded. The player then downloaded the book automatically without the user's intervention.

Readers could also use a talking bibliography on the RNZFB's Telephone Information Service (TIS) to select and order their digital talking books independently. Titles could be automatically downloaded over the internet to a player which was designed to be easily used by blind and vision-impaired individuals who are unfamiliar with computers. Surprisingly, the TIS ordering system option was not used by many of the trialists. It tended to used by the more computer literate trialists, however, the option may have been used by more users over time. This could also be explained by some readers preferring more passive selection methods such as selecting preferred genres and getting a 'lucky dip' in their leisure reading experiences.

To support the player, the RNZFB installed broadband Asymmetric Digital Subscriber Line (ADSL) connections in the trialists' homes. The costs for broadband connection were paid for by the RNZFB so that no cost barrier was imposed through the trial. This was consistent with RNZFB member charging practices.

RNZFB library staff learned how to use the trial players. They visited trialists homes and installed the machines which involved setting up the wireless connection. Once the machine was functional in the trialists home, the library staff instructed the trialist in the player's use.

Ten players were developed and trialled with 40 library borrowers. Each trialist had a player for three to four weeks. All trialists downloaded, listened to and returned several books and/or magazines during the trial period.

The technical communications protocol

A simple protocol was developed to fully automate the interaction between a given player and the library computer server. The protocol made no assumptions about the internet connection. A standard broadband, 3G mobile or other connection could be used. We considered that wireless connectively would be more user-friendly as the reader could listen to their favourite book in bed, the sitting room or even the garden shed!

The protocol used File Transfer Protocol (FTP) because it was simple to implement. However a more secure protocol may be required to conform to Copyright legislation and protect users.

The RNZFB developed a 'book card' concept which was an Extensible Mark-up Language (XML) file that described a given book request. It could be sent in both directions between the server and the player and from the player to the server. This XML file contained information about the book and gave the player all the information it needed to download the book. The book could be on any given server although in the trial was limited to the RNZFB server. Attributes in the book card indicated the host name, directory and other relevant information. All related files were listed so the player could get each file in turn until it downloaded an entire book.

There was also an attribute related to the status of the request, and as the book card could go in both directions, meant that the server could get an up-to-date status as the request was processed. It was possible to view whether a request was waiting for download, currently downloading or fully downloaded and available on the player.

When the borrower returned a book, the book was deleted from the player and the associated book card sent to the server with a 'returned' status. The server could then determine the next content to be sent to that player (based on previous borrower requests). The server knew which books and magazines were on a player, and how much memory was available. So the server did not download more content than the player could receive.

Technically, it was also possible to send multiple copies of a book simultaneously. In effect the notion of queuing for the next available copy which happens with physical books was eliminated.

Demographics of the trialists

Trialists were screened to obtain a sample skewed towards the older age group. Forty trialists were selected to participate, 39 were from Auckland and one was from Wellington. All trialists were current RNZFB library users.

Potential trialists were also screened to obtain people who were prepared to listen to two books a week. This was considered an indicator of being prepared to take the time to explore the machine and provide detailed feedback on the usability of the machine. However, trialists were not in fact required to listen to two books a week for the month of the trial. The number of books read over the trial was ultimately at the member's discretion. Members who did not feel able to participate in the full trial and provide feedback due to ill-health or caring responsibilities were excluded.

The selected trialists ranged in age from 60 to 93 years, with one exception who was in his 40s. Most trialists were aged in their 70s and 80s. In terms of gender, 27 were women and 13 were men. This is reasonably consistent with the demographics of the older population.

In terms of vision loss, nine trialists were totally blind. The other 31 trialists had varying degrees of vision loss across a range of eye conditions but most commonly age-related macular degeneration. All trialists met the Foundation's criteria for membership which is defined in terms of visual acuity. By definition a Foundation member has visual acuity that does not exceed 6/24 in the better eye with corrective lenses, or has serious limitations in the field of vision generally not greater than 20 degrees in the widest diameter. A number of trialists also had hearing difficulties.

Technological skill level of the trialists

Trialists had to have a specific telco phone connection for practical implementation reasons associated with the manner in which the trial was set up. (Most New Zealanders have a phone connection.) The sample was deliberately skewed towards members with little or no computer experience. Initially library users who did not have internet connections were sought although this criterion was widened somewhat so that there was a range of people with access to various levels of computing and the internet or not.

The trialists had a range of abilities and confidence levels when it came to use of technology. Over half of the trialists (23) did not have a computer in their home and six of these had never used a computer because they thought computers were too complicated for them.

Other indicators of comfort with technology were also used such as use of an electronic fund transfer point-of-sale (EFT-POS) card. Seventeen trialists did not use EFT-POS cards. The questionnaire also asked about the use of their old 4-track talking book machine. Five trialists had had some difficulty learning how to use their old 4-track talking book machine.

Based on these indicators, the trial included people with a range of technological ability with a good number being very limited technology users with about one in seven considering computers too complicated for them to use.

Achievability and acceptability

The RNZFB had twin goals in relation to the digital player - it had to work within a normal New Zealand household and it had to be acceptable to the reader. More specifically, the digital player needed to be acceptable to an older person who may not be familiar with computer technology.

The 40 trialists were asked to try out the new Digital Talking Book player and received one-to-one instruction in their home. The internet connection was set up for them and they were asked to read two books and magazines each week. They had the player for three or four weeks.

We also asked them a series of questions. The questions related to their experiences with the old talking book machine and their experiences with the new player. The first set of questions was asked before they were shown the digital player. After they had received instruction and had the opportunity to experience the player for three to four weeks an RNZFB staff member visited them in their homes. The trialist then demonstrated their proficiency across each of the functions of the player including basic and advanced functions. Basic functions were generally within the range of getting a book, reading it from cover to cover and returning the book. Advanced functions included aspects designed to enhance the reading experience such as navigating from chapter to chapter, setting up a sleep function or marking a section of interest (book-marking).

Following this demonstration, the player was collected by the staff member and returned to the RNZFB. About a week after collection of the digital player from the trialist we telephoned the trialist to ask about their experience with the digital player.

Finally, we ran a focus group with the staff installer/instructors to get their impressions regarding what had worked well and what not.

Results

The main intention of the trial was to establish whether content could be delivered automatically and reliably to players over the internet. If so, whether, in particular, elderly borrowers with no computer experience would enjoy receiving their reading materials in this way and find it acceptable.

The results were very positive - not one person disliked the new player, and only one person would not recommend the player to their peers. As was expected, some adjustments were suggested and some technical problems were experienced but the trial demonstrated that the digital player was both achievable and acceptable within the trial context.

Achievability

A number of technical problems were experienced which hampered members engagement with the new players. The RNZFB recognises that efforts to minimise these problems would need to be managed before a national implementation programme was set in action.

Staff installing machines walked into an unknown technical environment that included other phone-linked services including modem connections, alarms and Life Link. Understanding the technical environment and how to deal with complicating factors in advance would improve installation efficiency and better communicate 'ease-of-use' to members.

The technical problems experienced included:

• Interruptions to connectivity with the RNZFB server.
• No interface between the telephone information services (TIS) and the library database. This resulted in the requirement for manual data entry.
• Issues in the relationship with the telco. Specifically, slow and inaccurate responses for broadband connections and poor communication. Billing problems were also experienced.
• Faults with nine out of forty trialists' machines. These faults affected the member's confidence.
• Help-desk support issues. Help-desk resourcing was inadequate. Specifically, staff would have valued having a player on hand so they could more easily visualise problems described over the phone. Remote access to computer diagnostics to improve the help-desk's ability to trouble shoot over the phone would also have helped. Staff recommended that this include automated information such as when players were off-line.

All of the technical issues were ultimately resolved so that the player could be trialled in the home of the trialist. However, technical issues impacted on acceptability for the trialist.

Acceptability

Making the new player work was easier for some trialists than others. Many more trialists found the new player about the same or easier to use than the old player, than those who found it harder. This was an important finding as it provided a direct comparison with the technology the RNZFB currently provides to members and considers acceptable for members' use. In terms of ease of use then, the digital player may be regarded as an improvement on the status quo.

The player itself needed some enhancements as identified through the trial. These included:

• The transformer became very hot and trialists were concerned that it may create a fire hazard.
• The transformer needed to be changed - it was difficult to plug into the power supply due to its large size. The power supply plug going into the machine was the same as, and could be confused with, the audio output plug.
• The buttons needed to be more definite and responsive. There was a delay between the time the button was pressed and the being sound heard. This made the user think they had used the wrong button. There should be no perceptible time delay.
• The buttons needed to be better contrasted in both colour and feel.
• The book return button needed to be improved. Books were sometimes returned unintentionally because the confirm button was also the play-stop key. It would have been better to have a physical difference in its location and 'feel' such as using a slide rather than a button. If a button is used to confirm the return action the button should not be the play key.
• Volume settings were not loud enough. The volume range needed to be increased as members commonly had hearing difficulties.
• Diagnostic information needed to be available. Ideally diagnostic information should display on the machine and be able to be accessed remotely (from the library helpdesk). Information needs to cover simple matters such as whether the machine is online as well as more detailed information.
• Setting the User Guide as the default proved frustrating. A separate control key to access the User Guide would have been preferred. The relatively high level of proficiency required to get in and out of the User Guide belied the purpose of the Guide.
• Keys that had more than one function caused confusion.

Players were designed to make use of DAISY navigation features (moving between chapters or magazine articles). It should be noted that DAISY navigation was unfamiliar to the vast majority of trialists. There was also a sleep timer. These were considered more advanced functions. While the navigation features were appreciated by some, not all trialists wanted or used the navigation features. Some trialists found it difficult to grasp the navigation concepts. However, these trialists were satisfied with the basic functionality of the player.

The additional navigation buttons added a layer of complexity that was beyond some trialists. Instructors suggested that consideration be given to providing a 'basic' player option. This could be achieved either through the use of a cover overlaying the more complex buttons or by making two versions of the player available. A cover could also be useful for teaching with the cover being removed once the user is familiar with the fundamental concepts.

By the end of the trial, one trialist was observed still having difficulty performing the more fundamental functions associated with player use. This trialist felt that they still needed more time to become familiar with the player. However, this trialist would still recommend the player to others. In total, three trialists reported that the new player was 'somewhat difficult' to use by the end of the trial. No one reported that the player was 'very difficult' or 'impossible' to use.

Key attributes - moving from acceptable to desirable

Acceptability can be considered a baseline measure. However, it is worth considering whether the digital player improved on the status quo. Responses from trialists demonstrated that the digital player was indeed an improvement and offered significant additional benefits to trialists.

The key factor to improving the service was driven by the internet delivery system. Getting books from the library quickly and without effort was very important to trialists and they would recommend the player mostly because of the 'internet factors' such as not having to post cassette tapes back to the RNZFB Library.

The other factors that were most important to trialists and which they experienced through the trial were improved sound quality and a wider range of reading materials. However, it is important to note that the sound quality is a factor that is not dependent on internet delivery and is likely to be achieved through any 'new' player. A wider range of reading materials is of relevance only if more material can be accessed. Potentially the barriers to accessing a wider range of materials can be reduced in a digital environment. The 'automated postie' though was a significant and valued enhancement.

The trialists were specifically asked in the post-questionnaire: "How did you feel about getting books directly to the player instead of through the post?" The responses from a selection of trialists speak for themselves:

"Oh wonderful, that's a huge help."

"Oh, that's grand."

"Excellent. That's the beauty of it."

"It's a godsend."

"Oh, amazing, excellent."

"Wonderful. I don't have to rely on someone else."

In practice a magazine can be recorded in the RNZFB studios, copied to the server and within half an hour be available for reading on multiple players. The potential for borrowers to obtain books on weekends and holidays such as on Christmas day also exists. Certainly, the trial demonstrated the potential to provide a full 24-hour, seven days a week service. It was this potential that gives an internet-based service the edge over any postal service.

"The best thing [with the 'automated postie'] is that I don't have to worry if the postman has put the parcel out of the rain."

"Even with everyone's best intentions, you can run out [of books]. And a day or night without a talking book can be a very long time. With digital talking books, they are there in a minute."

"As soon as I finish a book, another book arrives without any trouble at all - if I push the right button!"

Conclusion

The RNZFB had two key questions, could the RNZFB deliver books effectively over the internet and would the service be acceptable to our predominantly older membership? To answer these questions a trial was set up which developed a player and the technical support required to send books and magazines to people's homes digitally via the internet. Unsurprisingly, some technical difficulties were experienced. These had the potential to impact on acceptability. However, they were overcome and the objective achieved. Books and magazines were delivered to homes in a range of areas around Auckland using a virtual 'postie' - the internet.

But the end users were not just anyone. The end users of interest for this trial were elderly people with little or no vision. A number of these people considered computers too complicated for them to use. The RNZFB wants to send members books over the internet to avoid the limitations imposed by traditional postal services. As a consequence, it was necessary to consider not just the delivery mechanism but the receiver. This trial included the development of a simple receiver that was clever enough to function as a computer but be worked by the listener like a basic tape-cassette player.

This trial demonstrated that an internet-capable digital talking book player could be made to function. More than this, the trial demonstrated that the technology was acceptable to the end user and resulted in additional benefits: less time to obtain a book, no queues, enhanced sound quality and the potential to increase access to reading material.

[i]. New Zealand Disability Survey Snapshot 6: Sensory Disabilities

[ii] The Royal National Institute of the Blind (U.K.) bases this figure on Lockyer, S., Creaser, C. & Davies, J. E., 2004, Availability of Accessible Publications, Library and Information Statistics Unit, Loughborough University, Loughborough.