Remain Detached from the Outcome

I’ve been working on my Black Belt certification for several months. The past few have been devoted to a project related to understanding how improvements over the years have impacted business performance. Many of the improvements were my own initiatives.

Be Objective

As I sifted through data, I was really hoping to prove my improvements were, in fact, improvements. Many of the changes that had been made felt like they really helped improve the operation and contributed to reaching business goals. I just wanted the data to say so.

I caught myself trying to confirm the results I wanted with the data. Yes, confirmation bias is real and it’s easy to wrapped up in. Once I recognized what I was doing, I had to tell myself to do the analysis and let the chips fall where they may.

Data Analysis is an Experiment

In the spirit of scientific thinking, the experimenter can’t get emotionally tied to an expected result. We simply form a hypothesis, run the experiment, check the results, and adjust the approach until the goal is reached.

It’s easy to get misdirected with scientific thinking in that we can keep analyzing data to get us the results we want, which, of course, is errant thinking. Edison tried over and over to make a light bulb that wouldn’t burn out right away. Each of the thousands of times he didn’t get the results he wanted, he changed something that might work. Eventually it did.

Know the Result You’re Seeking

Proving one’s self to be right isn’t scientific and really isn’t beneficial. The result to seek with data analysis, or any experiment, is simply the truth. Take the emotion out of process. The point, after all, is to learn. And, sometimes, learning can be humbling.

Think about it this way–the expectation to set before the analysis is to learn what the data tell us–nothing more.

Now that I’ve competed a fair bit of data analysis for my Black Belt project, I’ve learned that many of my continuous improvement efforts haven’t produced the long-term results I had planned. While this is disappointing, at least I can be satisfied with the fact that I now have data to help me take the next step.

---