This week, a visualisation of the climate change caught my attention.
Last month, January, was an abnormally warm one. It broke the record of being the warmest month of January since the recording started. It was 1,13°C warmer than the average month of January and further continues the trend on the record that had been set by December 2015.
The picture above illustrates the difference between January 2016 and an average one. It’s colored according to the scale underneath. The more the color tends to be red, the warmer than usual the temperature was on that location and the more blue, the colder it was. It’s a nice illustration that is quite intuitive. Red is commonly associated with warmth and blue with cold. However if people wanted to know the exact difference between two places, it is not very easy to deduct. The problem comes in 2 stages: (1) getting the exact location of cities on the illustration and (2) getting the exact temperature on that location.
The location can only be dirived from the rough world map that’s drawn in the background, so for a country like Germany it’s hard to see where the red/orange line splits the country and for other inland countries it’s hard to locate them.
The temperature is only given in a color scale which despite being very intuitive, is not accurate (exact). People can’t derive numbers from it, only estimates. And places within the same color can still differ a lot from each other, since red covers 2 points on the scale of Celcius.
The second problem this illustration has, is the scale/legend used. The categories represented by colors are of log-scale size, which to a human is not very intuitive – at least not when it comes to temperatures. Also, the two outer categories are joined, probably because of the few outliers but still the darkest red can be both 4.1°C or 12.9°C, which makes a hugh difference. Also as mentioned above, it does not offer an exact representation of the temperature.
So the conclusion: it’s a nice illustration to give a rough idea, but it lacks detail and when not read properly, it can be misinforming too.
This week, I came across a couple of visualisations that have to do with my hobby – sewing.
This visualisation gives an overview of all the different styles of necklines a garment can have. A stylised picture of every type is shown for visual understanding, accompanied by the correct term and a short explanation. This kind of visualisation also exist for dresses, skirts and trousers, as well as for undergarments – any kind of garment that has a specific name to indicate the style.
Visualisations like this are very powerful in the sense that it is much harder to explain such a style in just words. If someone refers to a certain type in an article, it is easy to know what they mean by taking the visualisation in front of you and looking it up. Combining the different visualisations (trousers, tops, skirts,…) can also be very helpful in the designing process of a new garment.
In this week’s InfoVis I look into how not to visualize information, and more concretely how not to make powerpoints.
Life after Death by Powerpoint is a well-known video by Don McMillan. He teaches us how not to present information and use powerpoint.
The video shows him making a presentation supported by powerpoint. He points out all the wrong ways of presenting information and using powerpoint by actually doing them, sometimes in an over the top way to make his point even clearer. Some very valuable lessons can be taken from it.
He starts off with an overly crowded slide with a graph which is more like a maze than a graph. It is too complex and interwoven, and also too small to be readable by the audience.
Secondly, he points out how two graphs can be wrongly connected by comparing their shapes by showing the graph of “Number of Powerpoints by Day” and “Home Foreclosures by Day” and saying PowerPoint caused the mortgage meltdown. The lesson to learn from this one is to always make sure that you know what kind of information is being compared or used and take valid conclusions not just based on mere things like the shape of graphs.
Thirdly, he points out people tend to put everything they will say on PowerPoint slides. This makes your slides crowded, wordy and boring. The audience will loose attention rapidly. It is thus very important to always make a good selection of which information to represent. Bombarding your audience with information will not give good results.
McMillan also points out that font size and font type matter. The audience needs to be able to read your text to be able to interpret the information correctly.
Furthermore, he depicts that text should stay stationary to not distract the audience from the message, nor should there be too many animations. This can be extended to the lesson to limit unnecessary decorations, stick to the information that you have and that you want to represent.
Another common mistake is to make too many bulletpoints. Once again this shows to not give too much information in one view to not bombard the audience with information.
It is also important not to use too many abbreviations as those may hinder easy information interpretation.
Lastly, it is important to not use too many charts. Not all information is fit to put into charts nor should all information that is available be provided to the audience. Once again, make sure to only select relevant information and present this in a meaningful way.