Data Analysis 101: The Z-Score is Your Friend

August 10, 2022      Kevin Schulman, Founder, DonorVoice and DVCanvass

This likely speaks volumes to my social network but I consider the z-score a friend.  The z-score is a way to compare apples to oranges.

First, a baseball example then a fundraising copy one.

Babe Ruth is an apple from 1919 with 29 home runs.  Barry Bonds is our orange from 2001 with 73.

Did Barry Bonds have the best single-season home run performance?  73 is a lot more than 29.  As even the non-baseball reader can imagine, the game has changed a lot from 1919 to 2001.   Here are but a few,

  • The ball was heavier and larger in 2019
  • The ballparks were, on average, much bigger
  • The season got much longer over time
  • And, oh yeah, Barry Bonds used steroids

How then to compare when so much is different?  The z-score is like the gold-standard.  It makes any two performances comparable by converting Barry Bond’s 73 runs to a relative measure of how much better that was compared to players of his era.  Same for Ruth; his 29 is converted to a number that shows his relative performance against others of his era.

So the comparison becomes how unusual was 29 for that era compared to how unusual the 73 was for 2001.  Most z-scores fall between a -2 and +2 with the average always being 0.

  • Barry’s z-score is 3.42, which is really high and says he was way above average for his era.
  • But, Babe Ruth’s z-score is 5.7, which is off the charts, unprecedented good.  (*z-score figures for Babe and Barry are courtesy of John Cottone and Jason Wirchin, academics and baseball nuts)

Using a lot of nouns in your fundraising copy is bad, it makes the copy feel dense and heavy and readers are less likely to read it.   But how many is too many?  What’s the average number of nouns for good copy vs. bad?

We know the noun usage rate for academic abstracts and press releases and scientific papers.   We also know the noun usage rate for personal letters.

You want your noun usage rate in your fundraising copy – the orange – to compare favorably to the personal letter, an apple.

Our Copy Optimizer (sign up here for your own account) tool to score your copy uses z-scores to see how your copy stacks up against our gold standard, the personal letter.   If your z-score is high (over a +1) we flag it by color coding all your nouns in your copy and suggesting you replace some with pronouns or remove them entirely.

No messy math that you can see but simple, useful math behind the scenes that lets us compare apples to oranges.

Now you too can have a z-score as your copywriting friend.

Kevin

P.S. Want to defend your copy with objective scores?  Consider signing up and creating your own, secure Copy Optimizer account.  It takes 2 minutes, is low cost and you can cancel anytime.

 

 

2 responses to “Data Analysis 101: The Z-Score is Your Friend”

  1. Tina Cincotti says:

    Hi Kevin, I realize this is not what your post is actually about but I’m curious about something related to how the z-score is calculated in the baseball example you mentioned…. Is the racial segregation of Major League Baseball one of the factors taken into consideration when determining something like the best single-season home run performance? Thanks in advance for sharing any additional info you have on this! –Tina

    • Kevin Schulman says:

      Hi Tina,

      No, the z-score is only creating apples to apples by comparing players to their era of players and saying how much better/worse were they than others playing at same time with same circumstances. It can’t account for the “non-players” – i.e. the black players who could have played but were prevented from doing so because of segregation.