The Magical Number Seven, Plus or Minus Two

I've read the original paper before, and I think people are misusing its title and the number 7. If you read closely, the conclusion of 7 being a magic number is a bit of satire. In fact, just think about the title itself, a "magic number" with such a wide range?

The paper discusses various studies where the "magic number" (mean) varies between 5 to 12. And at the end Miller sort of throws his hands up and poetically concludes, "And finally, what about the magical number seven? What about the seven wonders of the world, the seven seas, the seven deadly sins, the seven daughters of Atlas in the Pleiades, the seven ages of man, the seven levels of hell, the seven primary colors, the seven notes of the musical scale, and the seven days of the week?"

So it's pretty clear "7 plus or minus 2" is not meant to be interpreted quantitatively, but rather it's that: depending on the task, there is some limit to our working memory. I think a more accurate title would be "humans seem to use chunking for various short-term memory tasks". And should be tempered by the standards of rigor from the past; compare Miller's paper to the precision in a recent paper like this one: https://www.zora.uzh.ch/id/eprint/151291/1/Thalmann.et.al.Ch...

Cue all the spammy "10 UX design principles that everyone must know" articles that put the number 7 on a pedestal...

There is an interesting conversation about this in https://www.arraycast.com/ (I think it was the second episode).

They talk about the cognitive load needed to understand code in different languages, and how thanks to the use of primitives and less levels of abstraction, they find array languages easier to work with (eg: with implicit loops, you do not need to keep track of a counter and an exit condition).

Even if you do not share this opinion, I recommend this podcast. There have been some threads in hn about it.

The original paper is from the 1950s and this has been studied a lot since then.

From doing a lit review on working memory capacity, it seems that our capacity is highly dependent on the context of the task/information and the individual's expertise/skills. One particular replication study I remember was not able to replicate the results. I tried applying it to more complex tasks (e.g., coding) and it did not work at all.

I recommend Radvansky's book "Human Memory" if you want to learn more.

I know not to obsess with 7, but I find it interesting my 2 year old is obsessed with 7.

Not even sure why, maybe the counting YouTube videos infected him.

I wrote an article a while back arguing that "7 plus or minus 2" implies objects in OO software development should be decomposed to have 9 or fewer members: https://tobeva.com/articles/brain-oriented-programming/

This is called digit span, it is one of the subtests for measuring IQ

I wonder if this is a hardware limitation, or a software limitation?

Could the average human functionally extend this limitation with a bit of training or adaptation?