Evidence-Based Management (EBMgt) is a topic growing in popularity in both the academic and professional worlds for a number of reasons. We’ve never had access to the volume of data that we do today coupled with the processing power available to make sense of it. In addition, we’ve learned that while hunches give us a gut feel we are comfortable with, we like to confirm it with data (which can be challenging when coping with Confirmation Bias).
One of the bigger lessons you learn when you continue past a Masters degree is that your opinion doesn’t matter. Everything you write about should be evidence based. Synthesis is great and a critical step in Bloom’s Taxonomy—meaning it is just fine to take multiple sources of information and combine them in ways that may not have been thought of before. But outright stating something as fact without backing it up is not acceptable.
The challenge for managers today is that there are right ways and wrong ways to apply techniques in EBMgt. Excel is a powerful tool that we all know and love, and tools like SPSS or R can do more programmatically. For example, taking a data set, cleaning it, analyzing it, and displaying the results in a meaningful medium is doable in R alone, but it requires a working knowledge of the ecosystem. The results can also provide valuable insight and due to the scriptable nature, repeated over time with no additional effort. As I am going through my own BI/EBMgt exercise, there are a few things I have learned that might be useful for you:
- No result is still a result. In some recent analysis I am doing, I found no correlation between different groups of data. That’s still a finding and can be useful to report.
- Choosing a tool you are comfortable with is important. Sometimes free tools are better!
- If you are not familiar with a particular statistical test, ask before you take action. Statistics is hard, yo! Even though something like a Pearson correlation is pretty common, be sure you know what the results indicate before you invest in some big project.
- Data collection methodologies are critically important. Be sure you are using well established methods to measure what you are looking to measure. For example, doing a quick survey of free-form questions may seem like a great idea, but if you are trying to find patterns in a more defensible way, go for Likert.
- When in doubt, prefer to use scientifically proven methods! If someone says they want to create data for EBMgt and says that they don’t want to be too scientific about it, warning bells should be going off.
There is a great article in the current issue of Academy of Management Learning & Education that discusses curricula for EBMgt. You can probably find many of the articles through your library or through some creative Googling.