dataviz, program evaluation, performance management

Charting imputed, real, and projected data

January 30, 2020 | Pieta Blakely

shutterstock_47743384

In a previous post, I discussed the importance of keeping a data agenda: a list of data that will be helpful to your program but is not currently being measured. Once you’ve completed your agenda, you might be thinking, “Now what?” Getting started on charting your data may seem like a daunting task, but there are a few simple steps to get started.

1. Choose a measure to chart.
Often there is an obvious need to begin charting a specific performance measure. But if you’re unsure where to start, don’t get stuck in indecision — simply choose one.

Start by asking: What is the trend? Is it going in the right direction? It’s possible you haven’t collected this data in the past, or your data is incomplete - which means you don’t have a perfectly accurate trend line. Even with incomplete data, you can ask these questions and get started on creating your chart. How do you do this?

2. Chart your best guess.
Chances are you have some information that will help you get started. Institutional memory can come in handy, here: What can you remember from past years? Go through old records to fill in the gaps. And if you have some years but not all, simply chart what you have. You can also use benchmarks or comparable programs as a guide. What is average for your type of program? Is your program doing better or worse than the average?

All of these questions can help you create your chart. When charting, show that this data might not be accurate by shading it grey, as illustrated below:

Slide1

3. Begin collecting current data.
Once you have your measure properly defined and are collecting data consistently and accurately, you can start charting in a different color to show the actual data in contrast to the guesstimated data, as illustrated below:

Slide2

4. Project future data.
It’s difficult to know how your data trend line will change in response to program changes, industry trends, or factors outside of your control. In a future post, I’ll go into more detail on how to project future data. For now, you can estimate data projections by following your current trend line and using dashed lines to indicate projected data.

With all of these steps in place, you now have a complete chart that shows past, current, and future data:

Slide3

Remember: you can work on a problem even if you don’t have great measures. Don’t get stuck in a cycle of “I can’t work on this metric because I don’t have historic data.” There’s no wrong time to start, so get started now!

Tags: dataviz, program evaluation, performance management