Means, Medians and Inferential Data Analysis, Oh My!

woo.jpg

Something amazing just happened. I spent a couple of hours with Dr. Bryan Auday (neuroscience researcher and professor of psychology at Gordon College in Wenham, Ma) this week. He introduced me to a software package called SPSS. And I’m in love. With the software. And maybe with Dr. Auday.

As my data set approaches “completion” I began to experiment with it to see what kinds of data analysis could be performed on it. And I ran into some issues right away.

My old friend, Microsoft Excel, isn’t too shabby. With a few google searches and a lot of carefully crafted formulas involving SUMPRODUCTs and COUNTIFS and dollar signs and cell ranges and operations with strings, I was able to figure out not only how to get a descriptive analysis going, but I began to be able to compare some aspects of the data to other aspects of the data. For example, not only do I know how many churches collaborate vs. how many don’t collaborate, but I can also now see OF THOSE THAT DO collaborate, how many of them are led by lead pastors who are Boomers vs. those that are led by Millennials. Or I can look for correlation between denomination and collaboration. Or zip code. Or I can try to narrow down just what each respondent means when they say they collaborate (on a series, on a sermon, evaluation?) There’s some real power for making inferences from the data here.

BUT this process is super-clunky, and a misplaced comma or dollar sign, or a copied-and-pasted formula that “automagically” increases the cell column or row by one (trying to be so helpful, Microsoft) can foul up my best and most vigilant attempts to get accurate analysis. Too much room for human error.

ALSO this process doesn’t play nicely if I am going to add more data to the project as the months go on. I would have to go back to every single formula and update the cell ranges to reflect the new data being added.

LASTLY, it just takes a loooooong time to write the formulas and keep the results straight. There has to be a better way. And there is.

Enter a software package like SPSS, which is enterprise-level statistical analysis software available to large institutions like corporations… and maybe a few local area Christian colleges… It runs for over $2000 per seat. All I’m saying is that Dr. Auday may just be the man behind the curtain, pulling all the levers and switches, turning my data set into the great and powerful Oz. He showed me today how these tools work, and their beautiful point-and-click workflows to look for correlations within a dataset. This has changed my whole world.

Wait! I just went online and found a “gradpack” — a full version of the sofware, licensed for two full years if you’re a student, for $175. I know there will be a learning curve here. But that’s cheaper than Spotify! This is gonna be fun…