The word accessibiity is a fairly general term that can mean lots of different things. In software development though, accessibility means something quite specific; it means that software can be used by everyone, irrespective of any physical or technical difficulties they may have. To ensure it meets accessibility standards, any piece of software should be designed so it can be used by those with vision difficulties, motor impairments, and vestibular disorders. Statistics software is no different.

Accomodating those with accessibility needs is a very clear ethical issue, but it will soon become a legal one too. New legalisation across Europe and the US will effectively make it unlawful for institutions to recommend web apps or software to individuals that do not meet accessibility standards. Therefore, it is critically important to ensure that the stats package you recommend to students is accessible. If not then, frankly, it shouldn't be adopted by your institution.

Accessibility of current stats software

Fortunately, all the serious players in statistics have already made their software accessible and have accessibility statements to support them. This is the case for well-known proprietary software, including SPSS, SAS, MATLAB, so these are all safe to use in your teaching. If you're not able to pay for these, then free stats software is also accessible: R (through the R console) is thought to be the most accessible stats package available, and R Studio has recently been updated to ensure it meets accessibility standards too.

These software packages are fantastic tools for professional statisicians but, while they all meet technical accesibility requirements, they might not be accessible academically. This software is generally designed for professional statistians, and not those who are just learning statistics for the first time (e.g. first-year psychology undergraduates). Because of this, I had an additional motivation for providing a new stats tool for students; one that was acessible both technically and academically.

Designing statscloud to be accessibile

Before I wrote a single line of code for statscloud, I spent almost two years designing the app on paper. As you might imagine, designing a statistics app that is accessible to everybody and works across all devices (phones, tablets, desktops and laptops) takes an awful lot of thought and planning, and isn't really something you can make up as you go along. Similarly, accessibility isn't something you can retro-fit to your app afterwards or put on a 'to-do' list of things to add in later; accessibility is central to the whole app and something you need to think about right from the start. Because of this, I made some decisions about the design of statscloud very early on.

Supporting those with motor impairments

The first obvious design choice I made was to not have dozens of tickboxes when users are running analyses. Not only do these overwhelm users who are learning statistics for the first time, they look awful and aren't touch friendly. Tabbing through dozens of them, or memorising keyboard shortcuts for them all isn't much fun either, so those had to go.

Instead, the statscloud interface was designed such that it can be used without a mouse. This is obviously the case for touch-device users, but also those who are unable to use a traditional mouse on a desktop. To support these users, statscloud has a whole bunch of keyboard shortcuts you can use, and every input in the app can be accessed with the 'tab' key on your keyboard too. There are no boxes that require ticking when running analyses either; statscloud runs every test for you so, essentially, every 'box' is already ticked for you by default.

Supporting those with vestibular disorders

Another consideration was how results are presented on-screen. Generally, software should avoid rendering big chunks of content on-screen unpredictably; content should only appear when you expect it to, and ideally without any spinny wheels disrupting the process. Having your results appear in one long continuous output isn't ideal, and auto-scrolling to new results as they arrive would have been an unthinkable design choice as this would be incredibly disorientating for those with vestibular disorders.

StatsCloud gets around both of these issues by ensuring that the results of an analysis are presented to you the second you push the 'Run Analysis' button. (In fact, I haven't actually designed a spinny wheel animation yet as we haven't needed one!) By presenting all analyses in the same format and allowing you to toggle the analysis selected in a sidebar, you can browse through each analysis easily and without large chunks of content changing on-screen. To support users with vestibular disorders further, any animations (e.g. when editing charts) in the app are turned off automatically if you have requested to turn off animations in your device's accessibility settings.

Supporting those with limited / no internet access

Another accessibility feature software should have is that it should, where possible, work for those who have limited or no access to the internet. I said in a previous blog that, despite the name, statscloud doesn't actually use the cloud for your statistical computations, and there are a couple of of reasons for this:

The first is user-experience. Sending data backwards and forwards across a network constantly can lead to a lot of lag, a screen that constantly fills with spinny wheels, and a pretty choppy user-experience overall. The second is cost. Running a server can be expensive; to make use of a stats server, they'll either be an indirect cost for you (you'll have to set up, run and maintain it yourself) or a very real one (you'll have to pay a subscription to access one). Neither of these things are conjucive to accessibility.

By keeping the stats analyses on your device, neither of these things are an issue. As a result, statscloud is the only statistics app that allows you to run statistical analyses on any device, free, and without an internet connection.

Summary

These are just some of the ways statscloud was designed to be accessible, and you can find out more in our 'accessibility statement' here. In it, you'll find more details on how statscloud supports those with motor impairments, visual impairments and considers those with vestibular disorders too.

We believe accessibility is incredibly important, and we will continue to work to ensure statscloud is a statistics package that can be used by any student, irrespective of their technical of physical ability.