Communication and Visualization
From visualization to presenting, communication is key for data science. Use this space to discuss ideas and challenges
- 13 Topics
- 34 Replies
Welcome to the Communication part of the Community 👋Some ideas for discussion for this section:How to effectively explain technical concepts to non-technical stakeholders Best practice approaches for visualisation How to make great presentations How to write well And anything else you can think of!
This might be a bit old school, but what is your preference for visualisation and communication between Microsoft PowerPoint and Google Slides? I’ve used both, I like the amount of tools Google Slides provides, however I think PowerPoint has an overall better user experience. What do you think?
I’m looking for tips on how to best format shiny webapps and their linked components! Beyond a css file (Cascading Style Sheets), which describes how HTML elements are to be displayed, I can’t find a good way of applying formatting to shiny webapp elements. For instance, formatting a valuebox element by changing font sizes/styles, switching the output value with the descriptive text/title and moving anys symbols from the bottom left to the top right hand corner.Does anyone have any tips on easily applying formatting, or some good go to resources on css files?
In my first year as a PhD student, I made the massive mistake to choose the wrong type of plot to present some results to my supervisor. Obviously, that constructive feedback is stuck into my mind and I’ll never forget it now:Having great results is not enough, I also need to choose the right plot to present them. In the meantime, I found this chart that I wish I had back then. I think it helps quite a lot when deciding which plot to use depending on your data.Have you ever chosen the wrong type of plot to present your results? 😅
Many times, at the end of some piece of analysis or meeting a milestone in a project, I found myself wondering how to best present the results so that others see what I see, i.e., why I think my results are interesting and helpful to others.I’m keen to know if people have other tricks/secrets to make a presentation interesting/engaging/useful/helpful. I’ll share below my current approach for building a presentation: Understand my primary audience. Not everyone in my audience may be relevant or may be interested to take any actions on the back of my presentation so I’ll try to narrow it down to a group of people and I’ll focus to get my message across to them specifically. This doesn’t mean I ignore everyone else, I’ll try to make the presentation as inclusive as possible but without spending too much time explaining concepts that are quite common knowledge to my primary audience. Plan your presentation: how to build the story + how to present the story depending on the audience (heav
I want to brand the scale of count so it reflects my customers brand colours. Their brand colour is hex colour #FF3C82. Ideally it would be a scale of different colours so they could see which referrer source has the highest count, and which has the lowest. Can anyone help? ggplot(refer_source_cust_orders, aes(x = prop, y = reorder(referrer_source, prop))) + geom_point(aes(colour = count)) + coord_flip()
Tell me about a technical concept you've had to explain to a non-technical stakeholder, and what did you learn?
In my experience, I found explaining the output of basket analysis was quite difficult!When you are dealing with measures such as lift, confidence and support, it’s easy to show the difference between these with a formula. However, explaining this verbally was slightly more complicated and I had to be very careful with my choice of words, as at times, the high-level difference between these can be quite marginal.I had to spend a bit of time thinking about how to visualise this to make it accessible and impactful for the stakeholder. Showing them just two of the measures (lift and count) helped with this and also provided less opportunity for me to waffle and get my definitions mixed up. I also learnt that having a practical explanation was useful! Including a slide before all the analysis on ‘how to de-bunk’ a basket analysis visualisation seemed to help with its interpretability. What technical concepts has everyone else had to explain and how did you go about it?
An interesting article/podcast section on the interpretability of bar chartshttps://www.sciencefriday.com/segments/bar-graph/Research here: https://jov.arvojournals.org/article.aspx?articleid=2778118#247308956TLDR: According to the research 20% of people misinterpret bar charts when they are used to display averages of things - consistent across education levels, ages, and genders.Suggestion: plot all points instead - human minds are great at finding the centre of clusters.
Already have an account? Login
Login to the community
No account yet? Create an account
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.