Data Visualisation – A Great Topic For Building Your Reputation

Transcript

Data Visualisation – A Great Topic For Building Your Reputation 

Subscribe to our Newsletter!

Do you want to boost your career as a statistician in the health sector? Our podcast helps you to achieve this by teaching you relevant knowledge about all the different aspects of becoming a more effective statistician.

SUBSCRIBE

[00:00:00] Alexander: Welcome to another episode of the Effective Statistician. This is another short Friday episode and today I want to [00:00:10] talk a topic that is close to my heart because I have used it again and again and again to build my reputation. [00:00:20] And that topic is data visualization. Now, what is data visualization? For all the different things that we do as statisticians, [00:00:30] what makes data visualization?

[00:00:33] Alexander: So different and unique compared to all the other things, [00:00:40] you know, we can be experts in real world evidence and how we make bias adjustments and [00:00:50] how we can handle confounders that happen over time and change over time. How we can make robustness analysis, [00:01:00] all that type of thing. We can be super clever in terms of adaptive design and all kind of different [00:01:10] group sequential design things.

[00:01:12] Alexander: We can be super clever in terms of multiplicity analysis and how we [00:01:20] can best get most out of the alphas that we have. Now, the problem and the challenge with all these kind [00:01:30] of different things is the same. They need a lot of explanation. Very often, you first need to even [00:01:40] explain the problem. Multiplicity, for example, well, most people even struggle with understanding what a p value is in the first place. [00:01:50] And let alone what multiplicity is, let alone all the different techniques that exist there to get most of out of [00:02:00] it.

[00:02:00] Alexander: Now, that’s very, very different with data visualization. Data visualization is easy. [00:02:10] You can just show people what is possible. And you can talk them through examples, you can show [00:02:20] them how what they currently get, and what they could get, just as a contrast. Well, at the moment, we always, for example, produce [00:02:30] these plots using SARS, and we just put them in the appendix, and then we have these line and bar charts, and that’s it.[00:02:40] 

[00:02:41] Alexander: What we could do It is something like this, something that is animated, that, for [00:02:50] example, shows how the path of response increase over time for different groups or how [00:03:00] individual patients behave over time how different groups of patients behave over time, how different endpoints. Work together, and [00:03:10] that, you know, over time, quality of life improves as symptoms get better.

[00:03:17] Alexander: Now, all of these kind of [00:03:20] things you can show very, very nicely using data visualizations. And that’s why data visualizations are [00:03:30] a great way to build your reputation.

[00:03:34] Alexander: Now, if you want to become better at data visualization, of [00:03:40] course you need to spend some time learning about it. Most statisticians have no formal [00:03:50] training about data visualization. We all learned, kind of, by experience. working on our jobs, how to do these [00:04:00] kind of different things. We learned about, okay, yeah, a line graph, a bar chart, a couple of other charts.

[00:04:06] Alexander: We learned about these and we learned, [00:04:10] yeah, they should be more clean. And maybe we picked up some rules about how to use colors and a couple of other [00:04:20] things. But we never learned the theory behind that. What are Principles, design principles, [00:04:30] what are data visualization principles that exists and that are tested and tested and confirmed [00:04:40] again and again.

[00:04:42] Alexander: We don’t need to reinvent the wheels here. There are a lot of theories and models [00:04:50] available that we can learn from. We through this can. Make much more conscious decisions about what will [00:05:00] be a good data visualization, what won’t be a good data visualization, how we need to adapt data visualizations to different scenarios.[00:05:10] 

[00:05:11] Alexander: So invest in your skills to become a better data visualization expert. Someone said. [00:05:20] can use these kind of different techniques to make sure that the communication is better. And when you do that, then you [00:05:30] will build a bigger and better reputation for yourself. You will be able to showcase what you can [00:05:40] do much better, even with all the other things that I have just talked about.

[00:05:44] Alexander: Because Very often, you will need to run kind of various [00:05:50] scenarios, for example, to, to, to, to, to, Showcase, okay, in which scenario, under which assumptions will, which multiplicity [00:06:00] testing yields the best results, or what are the different design features that we can have to design a study in terms of making it [00:06:10] adaptive, and what are the consequences under different scenarios.

[00:06:15] Alexander: For many of these different statistical [00:06:20] approaches, you will need data visualizations as well. So invest in them to be seen as someone that [00:06:30] can communicate clearly. It’s probably not news to you that our community, let’s put [00:06:40] it that way, It’s a little bit limited in terms of communication skills, or at least we are perceived to be limited in terms of communication skills.[00:06:50] 

[00:06:50] Alexander: And yes, there are some outliers. Generally, we were not trained at university to be great communicators, unfortunately. [00:07:00] And we are also very often not the, let’s say, most extroverted people. to build it nicely. And that is not a bad [00:07:10] thing. It’s just the things that we need to take into account. So improve your communication skills using data visualization skills.

[00:07:19] Alexander: And that [00:07:20] will make us as a function, you as a person, better known, and you’ll get a better personal branding through that. Or [00:07:30] if you do that as a group, as a department, you will get a better Branding for your department, for your function, and hopefully we overall as a [00:07:40] community. So please do invest in data visualization skills.

[00:07:45] Alexander: I will put some links to various free [00:07:50] and paid things into the show notes. Check out these show notes and you will find a lot of different resources there. Have a [00:08:00] nice weekend, if you’re listening to this on a Friday, and otherwise, have a nice day. Bye.