In DS23's last week of training, Collin of DS17 gave us a training session on Accessibility. In this session we discussed reasons that may cause the user to have difficulty interpreting a dashboard.
In this blog post I will be going over these reasons and give a quick note on how you can improve upon this aspect. Let's get started.
Too many colours
In this chart, there are too many colours. If I gave you 2 seconds to find where the Storage Sub-Category -- could you realistically do it?
In order to lessen cognitive load, reduce the number of colours on your charts. Of course, it depends on the scenario but the maximum number of colours should be around 4 to 6.
One way of reducing the colours in the chart above would be to focus on a specific Sub-Category (e.g. Storage) by using an IF statement in a calculated field:
If we place this calculated field into the Colour marks card and Sub-Category into the Detail marks card we have the following chart:
Much easier to find, isn't it?
Too much text or not enough description
It might be tempting to supplement your dashboard with lots of text. If I give lots of detail it means that the user can get a lot more out of the dashboard, am I right?
Well, not quite. Having a wall of text can be hard to read and may be counterproductive. Have a go at reading the paragraph below:
Have you ever read a wall of text? Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Especially when there is no paragraphs. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Perhaps the text-to-chart ratio is just way off. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Sed ut perspiciatis unde omnis iste natus error sit voluptatem accusantium doloremque laudantium, totam rem aperiam, eaque ipsa quae ab illo inventore veritatis et quasi architecto beatae vitae dicta sunt explicabo. But, oh my goodness, look at all this insight I'm providing to this chart! Nemo enim ipsam voluptatem quia voluptas sit aspernatur aut odit aut fugit, sed quia consequuntur magni dolores eos qui ratione voluptatem sequi nesciunt. Neque porro quisquam est, qui dolorem ipsum quia dolor sit amet, consectetur, adipisci velit, sed quia non numquam eius modi tempora incidunt ut labore et dolore magnam aliquam quaerat voluptatem. Wow, so much text. Ut enim ad minima veniam, quis nostrum exercitationem ullam corporis suscipit laboriosam, nisi ut aliquid ex ea commodi consequatur? A bit overwhelming isn't it? Quis autem vel eum iure reprehenderit qui in ea voluptate velit esse quam nihil molestiae consequatur, vel illum qui dolorem eum fugiat quo voluptas nulla pariatur? Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum. Not great to read, right?
Are you confused? Did you skim over sentences? Sounds about right.
Personally, I am guilty of adding too much text to my dashboard. Looking back at my final interview dashboard for The Data School. I could have cut some of it down and therefore had more space for charts. A reason behind this was because the main purpose of the dashboard was to be able present it live to an audience, which the long text was not suitable for.
With this being said, it's about striking a balance.
Not adding enough text, for whatever reason, might lead to a bad experience for the end user. Have a look at this incomplete and unpublished Makeover Monday about vaccines that's sitting on my Tableau Public.
How is the end user meant to easily interpret this chart? I have no title, no description. Is this second dose? First dose? Both doses? Moreover, how am I meant to use the chart/dashboard?
Text in a dashboard should be used to both explain the chart and describe how to use it.
Colour confusion e.g. green/red for good and bad
It's natural instinct to associate with red as bad and green as good, especially when these are paired together. In the case above, we have the percentage change on sales compared to the previous year. Would you want to confuse your stakeholder and make them instinctively think that 2019 was a good year? I would assume not.
One way to avoid the issue of bias/learned association would be to use a different colour palette. For example, Tableau's default blue and orange.
Another way to avoid colour confusion is to also include a legend. In the example below, the colour legend is incorporated into the sub-title.
Link to this dashboard: What's the point of Valentine's Day? | #MakeoverMonday 2021 Week 7 | Tableau Public
Finally, relying only on colour to describe whether something is good or bad is a bad choice for someone who has visual impairments or blindness. In the example below, the table incorporates shapes in order to distinguish between a good and bad month, compared to the previous year.
Link to dashboard: Simple Sales Table By Month (Current Year vs. Last Year) | Tableau Public
Inconsistent formatting
A keen eye for detail is key here. Fonts. Sizing. Colours. Make sure they are consistent. If they are all over the place, then it is really jarring for the user.
Here's an bad example:
- Inconsistent font usage (this includes the charts and legends)
- Legend isn't in the same place for both charts
- Map is not in the same place for both maps
- Orange means bad for the map on the left. But on the right the more orange the better
Not enough whitespace
Whitespace is an area of your dashboard which has nothing in it. Without whitespace your dashboard becomes crowded and will lack something called "breathing room". If this is the case, it will be harder for your end user to digest the information. This is not what you want. You want to make things easier, not harder.
Create white space by using inner and outer padding. On top of that, utilise blanks if you need to and in general don't just throw every single worksheet onto a dashboard. Of course, it depends on the specification but don't be afraid to make the dimensions of your dashboard larger.
Have a look at this dashboard below:
In this dashboard, the charts push up quite close to the edges of each individual 'square'. On top of this, the grey gaps between each of these squares is quite minimal.
Now have a look at the same dashboard, but this time with more whitespace via inner and outer padding:
I haven't added much padding but it has made a big difference in how it looks and how it's processed by the end user.
Confusing order/flow
In western culture, we read from left to right and from top to bottom. Be aware of this when creating dashboards that this is the 'journey' of the end user's eyes.
An example of confusing order/flow would be to have filter or title at the bottom or bottom right. Moreover, you want to avoid making your user have to dart their eyes back and forth e.g. if your colour legend is far away from the relevant chart.
Another example would be to have big numbers e.g. KPIs in an unintuitive place, like near the bottom or on the right hand side.
This aspect also applies to creating universally-designed dashboard for those who have no pointing device/mouse and are limited to using the keyboard. Those who only use a keyboard navigate the dashboard by using Tab. In this case, even the order that you put worksheets into your dashboard can alter the order/flow for the reader.
For more reading have a look at this blog post, where Tableau ran an eye-tracking study to better understand how common design elements draw visual attention.
Difficult interactivity
This can mean putting in impossibly small button(s) which requires lots of precision from the user. Or it could mean overcrowding the chart so individual marks are hard to select.
Right now I'm watching a TV series, on my laptop, where the video player has the playback history thinner than a grain of sand. It is incredibly frustrating and takes a long time when I have to rewind to a specific part. Imagine how hard it would be for users who perhaps have an impairment!
Small buttons also affect keyboard users and/or those with visual impairment. Small buttons are hard to see whether they are highlighted and in focus, when the user is navigating the dashboard.
If you're going to make your dashboard interactive, make it easy for all users.
Overly complex charts
I've seen it. You've seen it. Those charts which are mind-blowingly complex. I understand where it comes from, especially from a data art perspective where their use-case is slightly different.
But you have to consider when you're making your own dashboards: Is it necessary? They say that all data can be simplified to a bar chart.
I must add, that if you have approximately over 1000 marks on your worksheet, then these marks get rendered by the server. If this is the case then it could mean that those who use text-to-speech may not be able to use your dashboard.