Got an applied project? You can build the capacity for data-driven decision making.

During graduate school, students are usually offered applied opportunities. What I love about applied psychology (e.g., evaluation, I-O psychology…) is that graduate students have the chance to bring their knowledge to a variety of industries — and built the value of data driven decision making. To me, that is priceless. Exposing the field of applied psych is great…and so is making others aware of all the great things can be done when something other than anecdotes are the decision making tool of choice.

So what’s my experience with this? I also had these great applied opportunities, and I started to realize that I was an advocate for my field. I had a new perspective about the entire experience — if my client walked away feeling like they wasted their time, I didn’t do a very good job.

My second to last semester I completed a needs assessment and process evaluation for a company in another state. This company is phenomenal — great idea that’s meeting a need, lean bottom line, and an office full of great people. Where’s the but? Well, it’s that data wasn’t driving their strategic planning. Needs assessment? What’s that? There I had it — an opportunity to BUILD the capacity of evaluation in this organization.

In short, the project went well — everyone learned a lot, services were revised, future planning was focused on data. But that’s all while you’re still there, right? In the back of your obsessive applied psych mind…you know this was one project, and the long lived method of luck and “educated” guessing (oxymoron’s make for good blog topics) could be revived and become the preferred decision making tool — again.

Data driven decision making. What does the data tell us? How do the statistics relate to what we are seeing financially? Your clients said they wanted this service…but was it a representative sample?? I felt like a broken record…because I hoped that between demonstration, dialogue, and bringing my client along for the experience would lead to an appreciation and PREFERENCE for data to inform their decisions.

Created by Chris Lysy

Well, two months later I was in a follow up meeting…sitting in on an unrelated project…and I heard it:

“We need to make decisions based on what the data tells us. It can’t be what I like, or what makes sense to me. Let’s use the tools we have to make changes using data.”

…you know that moment when someone says their idea of an awesome day is binge watching Frasier and eating pizza rolls, and you’re like “…that’s hot.” Bam. There it was. The sexy side of being an advocate for our field.

You see, change is hard. Pushing for a better method (that isn’t always easier) can be a challenge. And hey, being a grad student is a special level of hell at times. Sometimes you want to drop the results and peace. You don’t always want to screw with Excel for hours to get something other than a canned report (but seriously, talk to me if it’s taking you hours to craft good viz). But at the end of it, you have an opportunity to see the results put into action. Your very presence is a disruption — a potential catalyst for change. The credibility of our field? It’s on ALL of our shoulders. So, the next time you’re burnt out, remember the potential to impact decision makers and the responsibility to your colleagues (oh, and call on them when you need help!).

Created by Chris Lysy

Perfect datasets don’t exist but cool tools do. Here are two I use.

One component of my day-to-day is retrieving reports and turning them into something useful. Simple, right? Ha! I bet you laughed, too. Data comes in all shapes and sizes…in fact, it usually comes in every form besides the one we need. As data scientists, our job isn’t only to interpret the data and provide someone with the results to make an informed decisions – although that’s great. Visualizations? Those are great, too. But, if you don’t start with the right data, you might as well not analyze or visualize it. …so what?

Well, canned reports suck (unless they’re super customized by your IT team, but who has time for that?!?). When I ask for a data pull, I hate being told “no” or “yes, I’ll get that to you in a few weeks.” Usually when I need a report, it’s because someone important needs to make a decision. So instead of waiting around for someone to do it for me, I put on my data manipulation cap and get ready to finesse some Excel files.

Enough with the back story, right? You need tools! Admittedly, my most common problem is having all the data I need…spread across a bunch of out-of-the-box reports. Easy you say? Sure, if I was only dealing with a few rows of data. We’re talking thousands of rows. So, here are a few tools I’ve come across that have turned my data nightmare into something usable:

  1. Consolidate
  2. Merge Table Wizard for Microsoft Excel (the app with the world’s longest name) – Mac
    1. If you’re a PC user, use ActiveData

Screen Shot 2017-04-17 at 8.35.21 PM

So let’s talk about these tools…

Consolidate is great when you have several entries for one company (or participant) that you need summed, averaged, or a number of other functions. It’s as easy as five steps:

  1. Select your data.
  2. Go to the Data tab.
  3. Choose the Consolidate function.Screen Shot 2017-04-17 at 9.07.41 PM.png
  4. Select your data (be sure to note where the label that you’re consolidating is at).Screen Shot 2017-04-17 at 8.37.35 PM
  5. Choose your function.

Screen Shot 2017-04-17 at 8.38.22 PM

Done! Now you’ve got your data in one place. Screen Shot 2017-04-17 at 9.14.02 PM.pngIn my case, I needed each company’s entry to be consolidated into one entry, and I wanted the sum of all of their orders.
Next, I wanted to tie e-mail addresses to the company. That’s where my fancy application came into play. Here’s a quick walk-through on merging two reports to make one useful report.

  1. Open both reports.
  2. Open Merge Tables Wizard (Mac)
    1. ActiveData (PC)Screen Shot 2017-04-17 at 8.47.35 PM
  3. Select the data you want to merge on (pick the dataset with the majority of the data). This is your primary dataset.
  4. From the other worksheet (or, in my case, a workbook), select the data you want to merge. Screen Shot 2017-04-17 at 8.48.36 PM
  5. Your tool will likely verify that it identified the variable you want to merge on (i.e. company).Screen Shot 2017-04-17 at 8.48.46 PM
  6. The default is to add a new column for any data that does not exist in the primary.
  7. After that, the remaining steps are fluff. Select what you want and run it!

Boom. Screen Shot 2017-04-17 at 8.49.34 PM.pngIn several minutes, you just created a data file with all the data you needed. Of course, this is an oversimplification. But, you can apply it to anything – especially those nasty datasets.

So, what about you? What are your favorite tools for forcing Excel to comply?Screen Shot 2017-04-17 at 9.14.02 PM.png

I stopped blaming Excel for my basic data visualizations. Here’s when and why.

Early last year I realized my research findings only meant something if I involved the right people. And, to do so required me to be a creative and adaptive evaluator. Around the same time, I realized that I could have all the answers, but without a good way to communicate them, no one would actually notice.

Screen Shot 2017-04-02 at 8.32.46 PM.png
Look how easy it is to be basic!
I started searching for solutions. I found blogs by industry celebs (Ann K. Emery, and Stephanie Evergreen, to name a few) and began to pour over their archives. I found a lot of cool stuff. My world was quickly shaken when I realized that meant APA formatting might not be the best way for my voice to be heard (my condolences to the die hards, there is a time and a place). But, the big takeaway was that I figured out that, “Yes, it sucks because Excel made me do it,” was a crappy excuse for not visualizing my data in a better way. I learned to stop accepting the norm. You see, the purpose of Excel is to be a tool – not to do all my work for me because guess what? I’m the expert! I know those gridlines are garbage (and unnecessary), and that I should really figure out the RGB numbers for my brand.

So when did I stop blaming Excel? When I started valuing my data visualizations as more than entering data, selecting insert, choosing between a column or bar chart, and dragging it into my equally as boring report. When I realized that sweet visualizations in Excel just meant more thought and a few extra steps.

Screen Shot 2017-04-02 at 7.13.36 PM
I stopped accepting default settings!
Need an example? Check out Basic to Bomb Column Charts! Now go…find your inspiration for better visuals!

Screen Shot 2017-04-02 at 8.53.49 PM
Not too bad for a few extra clicks…

Basic to bomb column charts.

Screen Shot 2017-04-02 at 8.54.25 PM
The basic Excel column chart.
Let’s take a quick look at how a basic column chart in Excel can be turned into something you might want to look at.

  1. Enter your data into Excel
  2. Insert a column chart (if you’re basic, save and close here…)

Oh, good! You’re ready for more. Okay, here it goes:

  1. Get rid of that white space by decreasing the gap width.
  2. Remove the gridlines. They add unnecessary clutter for readers
  3. Trash the x-axis.
  4. …the y-axis, too!
  5. Add data labels (and for the love of all things, don’t leave them hanging in space – move them onto their respective bar).
  6. The title. I’m a fan of Stephanie Evergreen’s piece on this. Instead of a simple word or two, write a sentence that describes the dataset and LEFT justify it.
  7. Format! 10pt labels, 12pt title, non-default font and colors.

Bonus: Make everything a table. Sort them — highest to lowest (i.e. descending first), so readers don’t have to do the ordering themselves. 

Questions? Shoot me a message or tweet. 

Screen Shot 2017-04-02 at 8.53.49 PM
Our column chart gone bomb!

Can we talk about logic models?

Over a year ago I had the pleasure of co-presenting on logic models to a group of individuals from non-profit organizations. The presentation included time to discuss the pros and cons of the logic model. This really gave my team and I the ability to work through the problems with the attendees. The most voiced problem?

Continue reading “Can we talk about logic models?”