Big impact at the margin which is not legible in the average

From Childhood Roundup #3 by Zvi Mowshowitz.  

A section on homeschooling but with a crucial observation individual effort at the margin versus averages.  Read at the link for all the links within the article.  Emphasis added.

Time Spent on Children is Historically Anomalous

Reminder that working mothers today spend more time on childcare than did housewives of the 1960s. Quality time is great, quantity time has gotten out of hand.

Bryan Caplan points out in his version of my old post More Dakka the alternative hypothesis that also explains all the ‘parents don’t matter’ measurements.

Take parenting. Most readers summarize my Selfish Reasons to Have More Kids as “Parenting doesn’t matter.” But that is only one possible interpretation of the twin and adoption data. The data is also consistent, however, with the theory that most parents are barely trying to get results - at least on many relevant margins. I pondered this in depth before I started homeschooling my kids. I’m always stunned by all the economists who fail to teach their kids about supply-and-demand. All my kids know about these holy diagrams. What’s the difference between their kids and mine? I did ten times as much.

Yep. I interpret the ‘parents don’t matter’ statistics in a similar fashion. If you fail in one or more of many ways one can fail, that does a lot of damage.

You can then do vastly better than baseline, but that requires you to put in unusually more work or work unusually smart, or both. That doesn’t happen often enough to show in the stats. There is still a lot of room to make a big impact on the margin, but also not in ways systematic enough to show up in the statistics.