Talk:Forgetting curve

Ebbinghaus vs the Internet

It took me two full days working on my literature review to figure out that this article is talking about two very different "forgetting curves". The first/main "graph" "plots" "Memory" as a function of "Time remembered" (what even is "time remembered"? Why not just "time"?). I put "graph" and "plots" between quotes because it's clearly an illustrative drawing not based on specific data. The third graph additionally has extra curves with flatter slopes supposedly representing the effect of spaced repetition. This kind of visualization is in every online discussion of the Ebbinghaus forgetting curve, but clearly is not supported by his work.

From the article:
> The forgetting curve hypothesizes the decline of memory retention in time
> A typical graph of the forgetting curve purports to show

Somehow, I had missed these very important caveats. Clearly, we understand that these representations of the "forgetting curve" are unsubstantiated illustrations. Yet, in the next section, we make little distinction when discussing Ebbinghaus' work, which is completely unrelated and routinely conflated with the aforementioned curves in online discussions of learning, even by seemingly authoritative sources.

Ebbinghaus focused on "savings", or how easy it is to re-memorize something with repeated practice. Ebbinghaus' curve says "studying something once is not nearly enough to memorize it effectively, but the 5th time you study it you can expect to fully refresh your memory in a third of the time of your first study session" (actually, this applies to nonsense words; meaningful information is memorized much more effectively, but you see the point). This has nothing to do with "% information retained over time with/without study", which is what the first and third graphs are about.

I think this distinction needs to be made much clearer. The caption of the second plot (the actual Ebbinghaus forgetting curve) needs to change: "The forgetting curve [...]" does not accurately inform readers of the fact that it is not what is discussed in the first section of the article. In fact, I disagree with presenting the unsubstantiated, popular understanding of the forgetting curve as the main point of the article and discussing the actual scientifically supported forgetting curve as "History". I would personally make the main article about the Ebbinghaus curve and mention its misinterpretation in a subsection. Perhaps it would suffice to explicitly call out the popular interpretation as unsubstantiated in the blurb and specifying in the History section that Ebbinghaus' work is the source that was likely misinterpreted ?

The third figure also should not be in proximity of the discussion of Ebbinghaus' work in the "Increasing rate of learning" section. Speaking of which, do we even want that figure there in the first place? Is there any data supporting it? The oldest source I can find for it is the udayton.edu page linked on the Wikimedia commons page of the figure, which is now defunct but can be accessed at the internet archive (2004). Again, it looks like it's just an illustration with no data backing it and doesn't mention Ebbinghaus. I e-mailed the author of the web page about it 2 days ago and haven't heard back.

The second plot, which is taken from Murre & Dros (2015) is an accurate depiction of Ebbinghaus' data but the plot on page 10 with the power function fit would have been better than the plot on page 11 with the MCM fit. If we want a non-logarithmic plot, The Recovery of Unconscious Memories: Hypermnesia and Reminiscence. (Erdelyi, 1998) has a very good one on page two.

— Guy
2A05:6E02:1113:D510:4CE3:C6BD:A57F:F118 (talk) 10:14, 3 September 2025 (UTC)