MAY 7TH 2019: with just a few days to go before hand-in, today was spent making some adjustments to the report proposal now that I have done all of the research around the topic that I want to do and a draft proposal had been submitted for review by my tutor, Jamie.
Jamie gave some great feedback.
- The format of the report needs to be made clearer.
- There needs to be a ‘motive’ for doing this research and writing this report, for example a link to future practice.
- The length of the literature review can be reduced by using inline citations to reference texts and there is no need to describe who the authors are or what they do in this (this information can be found by looking at the attached bibliography.
- Stating that sticking to the 5,000 word limit is a ‘challenge’ either sounds like my project scope is too broad or I am unable to write concisely.
- ‘Code’ isn’t academic terminology.
I have now stated the format of the report at the beginning of the document and in the ‘aims and objectives’ section have said the following about how this relates to my practice:
Understanding these challenges will be beneficial for my future practice as a UX designer with an interest in creating cross-platform websites that are compatible with a variety of inputs, for example mouse and keyboard, voice and gestures, for the growing market for accessible websites.
As mentioned in previous posts, all of the professionals that I have spoken to about this project have been really fascinated by it and are keen to read about my findings and hear how the project progresses. They’ve all mentioned that this is a growing market and an exciting one to be at. I caught up with some people from Earthware yesterday and they said that a lot of their software doesn’t incorporate accessibility and they’re only starting to think about it. There’s more evidence that this is a growing market and not everybody is serving it yet.
I think this statement nicely links my interests in input methods to accessibility and my practice as a UX designer.
I’ve managed to get the literature review section down from around 500 words to around 360, not by reducing the number of sources but instead by using citations. It’s amazing how many words ‘this report was written by four researchers from Google and IBM in conjunction with the University of Nevada in 2007’ (or similar) adds to the overall word count when you’re doing this 7 times over.
I removed the comment about 5,000 words being challenging to stick to. Apart from anything else, it’s obvious that this is a challenge and it’s one that every other student at NUA faces. Removing it also removes 30 words from the report proposal word count.
The word ‘code’ was used in the context of ‘I will code a prototype’. I have since changed this to ‘I will design and develop a prototype,’ because there is design involved with this too. It’s not all coding.
- Is ‘mobile e-commerce website’ too broad?
This is something that I have reflected upon recently. The types of challenges that the blind will face on different e-commerce sites will be completely different. Below are some examples that I’ve come up myself:
Mobile clothes retail website
Applications that are very visual and the way the product looks and feels influences your purchasing decision, such as clothing, bring up some challenges. A few examples are below.
- Knowing exactly what the clothes look like – colours, designs and shapes especially.
- Knowing the size of the clothes.
- Visualising how they’d look wearing the clothes.
- Accessing reviews and looking at photos on reviews – how do you create screen-reader friendly images when users are uploading photos and developers don’t have any control of what’s being uploaded?
Mobile digital music website
Digital applications are theoretically less problematic, but the following issues may still arise:
- Searching for music through categories on the website.
- Using forms and menus and select formats to download the songs in.
- How to know when the music has successfully downloaded.
- How to know how to access the music once it has downloaded.
Choosing whether to focus on physical or digital products
As far as I can see, I have three options:
- Create a prototype ‘physical product’ website, such as a prototype retail clothes website.
- Create a prototype ‘digital product’ website, such as a prototype digital music website.
- Create a prototype ‘all-rounder’ website, such as something like Amazon, which sells both physical and digital products.
I personally feel that creating something like Amazon might result in a prototype that doesn’t have any real ‘focus’ or ‘identity’ and would probably need to include some form of search facility to mimic what shopping on Amazon is actually like (we tend to just browse aimlessly on Amazon since it sells just about everything) or I’d have to be clever about the types of things that my test candidates can purchase, i.e. they’d need to only be able to purchase things that are available in both physical and digital forms, such as music, films, games and software. Photographs might also be another option.
The choice of e-commerce website to make also depends on the answer to the following question from Jamie.
Methodology and testing improvements
Introducing iterations to the development
- Could your design be built on the findings of usability testing of an existing website?
- How about testing existing services and analysing problems?
If I were to make a prototype that featured both digital and physical products then Amazon would be a great site to test with the blind to see what usability problems they ran into. I haven’t personally tried using Amazon on an iPhone with VoiceOver yet, but I would suspect that with them being the market leaders in e-commerce websites, the fourth most visited website in the US and the tenth most visited in the world (Alexa, May 7th 2018) they have already considered accessibility. There may not be much to improve upon and I may just need up making a clone of the Amazon website.
Prototyping a specific type of e-commerce site could make for a more interesting initial usability testing session. If I were to produce a clothing website, I could use Next’s or Primark’s websites as good candidates. If I were to produce a digital services website I could use something like the Steam Web Store or iTunes or Spotify or similar (I know that Spotify’s business model is a little different, but the same kind of principles apply).
However, focusing on e-commerce sites that sell physical products, such as clothes, throws additional usability challenges to overcome, such as images and how the blind user knows what they’re about to buy. This could make for a more interesting project, so this is what I’m going to focus on.
At the moment I am considering the clothing website route because it’s something that people are buying more and more online and clothes are highly visual products that people only buy if they like the look of them. Somebody might buy a car because it’s cheap or economical, food because it tastes good, perfume because it smells good, a phone because it runs the software they want. People can often overlook the appearance of these products, but when it comes clothes I suspect the appearance and how the user thinks the clothing would look on them is the top buying factor.
How to design a good prototype for testing
- What usability problems will the prototype address? Should there be an explicit research phase before designing a prototype?
Creating a clothing website for the blind will be a very interesting challenge because clothes are so visual. There will need to be research into the specific problems that the blind face when using clothing retail websites on a mobile device before I can begin to develop my own prototype. Part of this research will be conducting usability testing with an existing retail clothing website to identify problems that blind users encounter when using these sites. However, in order to do this the following will need to happen:
- Use a clothing retail website on an iPhone with VoiceOver myself so I have an idea of what can be improved. Possibly the Next or Primark website. Or even both and do an A/B test with the blind user to see which they preferred to use.
- Research clothing retail websites and a bit about the fashion industry – two things I don’t know a lot about.
I enjoy research, this is isn’t a problem for me.
Additional improvements to the methodology
Jamie also asked:
- Who will be the testers?
- What specialist hardware or software may be required?
- What does success look like?
I’ve mentioned in blog posts but interestingly not in the report proposal (until today) about who will test my prototype. I think it’s fiarly clear that the way this project is going now is that this is discover what can be done to help improve web accessibility for the severely visually impaired (so either completely blind or near-blind). This means that near-blind or completely blind people need to test my prototype. As of yet I haven’t approached anybody and asked them to test for me, but I am considering contacting several local charities and organisations such as the Norfolk & Norwich Association For The Blind and Action For Blind. These are now mentioned in my report proposal.
I don’t expect it to be easy to find candidates and I don’t expect these organisations to neccessarily work with me, but sometimes these sorts of places are willing to get involved with projects like this if they can see the benefit to the wider community.
I don’t anticipate on needing anything too specialist besides an iPhone with VoiceOver activated, but what wasn’t mentioned in my report proposal until today was some of the things I wrote about in the previous post about my dissertation, such as taking additional lighting to help record in if the environment is dark and considering dual-audio outputs so that the tester can use headphones and you can still hear the audio through a pair of speakers.
As for success, that can be measured in two ways:
- From a usability perspective: the user is able to purchase a product and understand what it is that they’ve purchased. Basically they can succesfully complete the user flow.
- From a ‘what I want to get out this project’ perspective: I am able to design, build and test a mobile e-commerce website. Not only will I learn more about accessible web design, but I will also learn more about e-commerce websites in general and specifically retail clothing websites. With e-commerce being such a big part of the lives of consumers in 2019, there is no doubt that having knowledge about how e-commerce works and how to design successful e-commerce websites will be a very valued interest and skill to have in the UX industry. Especially e-commerce websites that work for the disabled.
Re-evaluating the question
Currently, the question/title is:
‘How can the accessibility of mobile e-commerce websites be improved for the visually impaired?’
Let’s see how it fits the requirements listed in Stella Cottrell’s book ‘Dissertations and project proposals: a step-by-step guide’:
- Is my title clear? Yes, it tells the research exactly what this report is going be able – it’s about question what can be done to improve e-c0mmerce websites for the visually impaired’.
- Is my title specific? Yes, but ‘e-commerce’ and ‘visualy impaired’ are fairly broad terms that can be made to be more specific.
- Does it incorporate a question? It is a question!
- Is it precise? It’s not a massively long title but it’s not the snappiest either. It’s as short as I can make it.
- Does it fit conventions? Yes, arts research titles tend to be questions. A lot of my friends who did this last year used questions for their research report titles.
The only area that the title could be improved is when it comes to specifics.
The focus of the report has shifted from ‘visually impaired’ to ‘severely visually impaired’, so the title could be changed to reflect this.
The focus of the report has also shifted from e-commerce websites in general to clothing e-commerce websites. I think that adding this detail into the title could help to make it more obvious that this isn’t an extended essay or a discussion piece – it is a technical report that documents an experiment.
So, the new title could be: ‘How can the accessibility of mobile clothing e-commerce websites be improved for the severely visually impaired?’
This is better because:
- It’s still clear, even if it is 2 words longer.
- It’s very specific now.
- Still incorporates a question
- Still fits conventions.
Using ‘severely visually impaired’ rather than ‘blind’ also doesn’t limit me to testing this with just the blind – there may be people who are near-blind that would also be very good candidates for testing something like this. My own research has also shown that there are different levels of blindness – some people cannot see anything or sense any light, whereas others can sense light (i.e. they’d know if they were in or outdoors). This opens me up to more testing candidates, making my method easier to execute.
State of the improved report proposal
With Jamie’s suggestions, this report proposal is now much more specific and thorough than before and even manages to be shorter – 1,075 words compared to the 1,079 that it was before.
As part of the technical report I will need to employ some degree of project planning.
At this stage, this is what I’m thinking of:
May-June 2019: research e-commerce and specifically clothing e-commerce websites – test the Next and Primark sites on an iPhone with VoiceOver to see if I personally think they’re usable on mobile devices for the visually impaired.
June-August 2019: begin to construct the usability tests and contact organisations about usability testing the Next and Primark sites. Try to get them onboard with the project and explain that there will be involvement from me on this for several months. Also begin to write the initial stages of the report, such as summarising all of my research.
August-October 2019: conduct the usability testing of the Next and Primark site to discover the specific usability problems that I need to solve, build my own prototype and then test this with the blind users.
October-December 2019: write up the rest of the report.
This is very basic and nowhere near detailed enough, but I feel that this is possibly achievable. I’ll have a better idea as I get working on my project as to how long each bit will take and some tasks might be done simultaneously, for example bits of the report can be written whilst testing and coding is going on.
Alexa.com. (2019). Amazon.com Traffic, Demographics and Competitors – Alexa. [online] Available at: https://www.alexa.com/siteinfo/amazon.com?toggle=true&utm_expid=.NFDkwnQTSf2ZNn_fyyCLoQ.1&utm_referrer=https%3A%2F%2Fwww.alexa.com%2Fsiteinfo [Accessed 7 May 2019].
Cottrell, Stella (2014) Dissertations and Project Reports: A Step by Step Guide. Palgrave Macmillan
Zangre, A. (2019). 51 Noteworthy Online Shopping Statistics in 2019. [online] Learn.g2crowd.com. Available at: https://learn.g2crowd.com/online-shopping-statistics [Accessed 7 May 2019].