4 Comments

What an awesome weekend read ) (I may be a taxonomy and conceptual model freak, so this is just what I needed. Not coffee. This.)

You've managed to take rather common concepts and organize them in a way that helps me think about projects I work on or see around. It makes me look for answers – what models have bigger or smaller focus in the project and how that influences the project, its understanding or just branding.

I would like to ask you about the IPCC's risk visualization. I agree that when it comes to easy readability / understandability of the chart, it's so custom that one needs to learn how to read it and we can't build to much on our existing models. What I wasn't sure was if by "an egregious example" you mean just the level of our existing mental models of how to read visualization or whether you find the chart really bad overall? Because I actually find it really helpful (of course after learning how to read it, hh) and I was wondering if you would see a better way to display such information given the complexity of the topic / data and uncertainties that there are.

Expand full comment

Thanks for your nice comment. I am so happy to hear it helps you think about your work in a a new and maybe even better way. That is exactly what I was hoping to achieve.

Regarding the IPCC figure. For me it's an egregious example because in my mind the mental model of how bar works in traditional charts and how they are used here clash. There are many other problems. For instance, the continuous nature of the shift in risk clashing with the discrete nature portrayed in the legend. Every time I see it I have to re-read the description to figure out how to read it.

Part of the problem is to figure out what I am supposed to derive from this graph. What am I supposed to learn? Is the goal to see at a given temperature value which systems become critical? If this is the case I think there are better representations. A possible way to go is to use a table with temperature in the columns, the various systems that could be affected in the rows and the risk level in the cells. I can't tell if this is the best possible solution. I am pretty sure there could be some very clever ones. It would be nice to test!

Expand full comment

Thanks for the insightful post.

Do you think we can achieve a better visualisation model by utilising semantic icons and visual metaphors for laypeople? How do we measure that? Using performance metrics with non-experts who look at visualisations for non-work related tasks is not the best option. How do we balance between the two and measure their understanding without interference with their interaction?

Expand full comment

There is a lot to do in the space of metaphors and affordances. I think we do not have a systematic enough understanding of how people make sense of visualizations and how they derive meaning from them.

Expand full comment