This is an Eval Central archive copy, find the original at freshspectrum.com.
In today’s vlog I walk through 5 of my favorite quantitative data dashboard design tips.
Okay. So last week I fielded a question about data dashboards. And in that particular video, I talked about why they failed and why you might want to consider using a website instead of a regular kind of dashboard. Uh, quantitative dashboard. But I also know that there are plenty of you that have tons of quantitative data and just want to create. The regular kind of dashboard that you see all over the place.
So let’s get into that today. You know, I, I think that issue. The thing that we should talk about first. Are the reasons why we have this data dashboard in the first place. So this is a person. Okay. I don’t even know what exists. And that’s true. A lot of the time.
There are a lot of data sources within organizations that. Are not necessarily accessible to the decision-makers stakeholders who could really get a lot of value out of that data. So that’s the reason why a lot of dashboards are created or at least the theory behind creating dashboards.
And I’ve talked about this a little last week, but.
I see dashboards as being an interface between the data and a user, some user. So. It can be hard to. Have a person just go, okay, well, I want to access some data. I’ll just go into SQL and pull some data and then I’ll analyze it and pull it up over here. That just doesn’t happen. I mean, some people it does, but most people, it does not happen. You have to reformat data most of the time to make it useful.
It’s a resource. You have to process it. In some way. And when you process it, our goal really is to reduce the amount of overwhelm. Because it can really quickly, we can get overwhelmed by the amount of data. And depending on who you are. I mean, there are some organizations like places like a retail establishment.
Are analyzing tons and tons and tons of data. And that data is you cannot access it. Most research and evaluation data is a little different though, because it’s accessible enough. You could probably pull it up in Excel and just put together some pivot tables and do some analysis.
Even so it’s enough data that having some kind of interface makes sense. So what are my tips? Well, let’s go through five. I think. Yeah, about five. Alright. Tip number one. It does not have to look like a car dashboard. Like it doesn’t have to have the little gauges or look like it’s speed limit thing.
Everything doesn’t even have to fit on one page. You can have it. So it’s up and down. Scrolling, scrolling, scrolling, actually that’s a much better idea most of the time than having anybody click on anything. ’cause. I don’t know if you know this about people, but they’re pretty lazy. They don’t like clicking on things.
So something has to be really valuable for somebody to click. That’s why so many social media sites. Now you just have the endless scroll, scroll, scroll, scroll, scroll, scroll. Well, you can develop dashboards in the same way. All right. Tip number two. It really comes down to familiarity and frequency.
Um, in terms of how you would design your dashboard. Alright. If you are really familiar with the data. Where if the, the people, the user that you’re trying to reach is really familiar with the data. It doesn’t really matter how it’s formatted. I mean, you just give them tables, you give them numbers. They’re going to understand it.
So they’re very familiar. And you have a lot of data. You create a dashboard. If they’re very familiar and you have a little bit of data, you create some kind of report. It doesn’t really matter what it looks like. They’re going to get it. If they’re not so familiar or they need to be introduced to the data.
Then you need an engaging report if it’s less frequent and if it’s more frequent, then you need some kind of annotated engaging dashboard. The idea is that we have to not Jess. Deliver the data and the information, but we also have to deliver the context behind that. So we have to walk our user through it.
All right. Tip number three. Um, and this is particularly useful for researchers and evaluators.
If something’s useful. Um, and insightful. That’s what you want to put into a dashboard. You want something useful?
I know a lot of researchers and evaluators who will look at a dataset and they will systematically start analyzing it and pulling it apart. And then pull it into pieces in a dashboard. Like you have to share all that stuff. But a good dashboard. Isn’t just a set of comprehensive documentation. It’s not just replacing a big table with a bunch of charts.
You know, it is something that is supposed to be useful. So that means picking and choosing, deciding what’s valuable. What’s not valuable. And you can use some different UX. A user experience design approach is user interface. Design approach is card sorts. Stuff. I’ll talk about in future episodes.
All right. The interactive design mantra. Um, I learned this pretty early on by Shneiderman. Um, I think that’s the name. I hope I didn’t just mess it up, but if I did, I’ll go back and change it. All right. So wait. There are three things. Overview first. Zoom and filter details on demand. The idea is.
When we’re developing dashboards that are interactive. There’s a different level, a different element. And that’s that you can explore, you can dive into it. So it means approaching the data in a little bit different way. And yet, sometimes that means starting from a. Bigger picture kind of point of view.
And then allowing people to zoom in and filter on the things that they really want to know. And then when they want more details, they click and they can get details. Right? So you have an overview then you can filter. This is like, if you see a dashboard in shifts, like a full country, And then you have, like, let’s say it’s the us.
And you have a dropdown filter and all the states are in the filter and you can click from the U S as a whole to individual states. So that’s your filtering. And it lets you zoom into different states. Maybe you can even zoom on the screen on a map. And then when you want more information about a particular area, you click and it gives you details. So that’s kind of how you design it.
Um, it’s a little bit different than designing a regular report. But. That’s pretty much it. All right. And the other thing is stay consistent. It’s great. If you know how to create lots of different chart types. But you don’t have to create lots of different chart types. Actually, it probably takes away. Anytime you change your chart, you kind of put that in somebody’s head that they have to like understand a new type of chart.
It’s better. If all the data is similar or kind of on similar dreams to, you know, just use bar charts, line graphs, scatter plots, these are all through things that are clickable filterable. Zoomable so they’re really nice charts to use in these kinds of situations. I’d avoid pie charts most of the time, unless you only have a couple of numbers.
I’d avoid, um, anything too complicated. I would. Try to avoid using, uh, multiple axes and anything too, like crazy. It’s better to do something simple. And repeated over and over and over again, then to jam everything into one chart. Which is usually what people try to do. When they try to turn it into a car dashboard.
So full circle, don’t do a car dashboard. Let it flow. Um, Yeah. If you have any questions, any requests for future? Episodes of this vlog. Let me know. I did get a request about doing something on qualitative dashboards. And we’ll talk about that next time, or I don’t know, in a couple of vlogs, we’ll see how things go.
But if you have any requests, please let me know. Again, just visit my site, down below, leave a comment here on YouTube. For how send me an email, do something like that. Either way. I hope you have a great day and we’ll talk soon. Alright. Bye.