Talk:Forgetting curve

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
WikiProject Cognitive science  (Inactive)
WikiProject iconThis article is within the scope of WikiProject Cognitive science, a project which is currently considered to be inactive.
 

Could we have some meaningful detail here? 'halve' over 'days or weeks'? It sounds like the explanatory use of the "Factor of 10" Not helpful.


I beg to disagree with "not helpful". Most people have no idea what is the true forgetting rate. 50% in days or weeks is roughly true and the best short expression of the rate I can come up with. I added more data to the entry but this verges on theorizing. The best argument against "not helpful" is that, following some not-so-honest mnemonists, people tend to belive that we can learn once and remember for ever. In other words, they belive in forgetting rate of 0. Some reputable scientists also embarked on an effort of proving that true, but the only "zero rate" possible comes from implicit repetition


Well. In distinction from 0 (zero) I suppose it's helpful. Otherwise you have to admit it's rather broad.


"Gedaechtnis" should be spelt "Gedachtnis" with an umlaut (two dots) above the 'a'. I don't know how to to this in Wiki - can anyone help out and make the change?


Dear author I am very interested to know how you define S in your equation! how would you, numerically define the information strenght? or is that open to discussion!

  • Yes, I would like to know this as well. I would also like to know where the entire equation came from. I cannot find it in any of Ebbinghaus' writings. It does not seem to agree with his data.
  • Is this section dead? There's doesn't appear to be any activity in 2 years. This was posted on 7/13/11. — Preceding unsigned comment added by Griswold62 (talkcontribs) 11:37, 13 July 2011 (UTC)
  • NOT the author: the equation seems do be from the Wozniak paper and I'm guessing it's just a guess at the simplest possible equation to model exponential decay of memory. The equation is similar to a half-life equation, but instead of a half-life, it's a 1/e life... e being about 2.7 something. I take it e was chosen to allow for using the natural log (ln). S then, is essentially the (1/e)-life of the memory; the amount of time it takes before the recall strength (R) is 1/e. — Preceding unsigned comment added by 68.7.175.67 (talk) 20:26, 14 February 2019 (UTC)

Of course it should be taken with a pinch of salt. It is excelent example of "math envy". That guy might have just tried to be interesting doing math. Sure it looks cool and "sophisticated" and more than 78,3% of people won't understand it :). —Preceding unsigned comment added by 86.61.232.26 (talk) 11:24, 16 April 2009 (UTC)

graph[edit]

This graph isn't a very good depiction of the forgetting curve, which is presented in the textbooks as much steeper in the early period and then leveling off fairly fast, not unlike an "L" lying on its side. Also, Ebbinghaus found that the most rapid drop could be measured in hours, not days or weeks, at least in the case of nonsense syllables. --Jcbutler (talk) 20:23, 9 April 2008 (UTC)

Yes, AND the graph presents multiple forgetting curves, which sort of implies that the forgetting curve itself somehow predicts the change in decay of a memory after review.... something it certainly does NOT predict. I vote to replace the image with one taken from actual data such as Ebbinghaus's own data. Perhaps there's an image we can take from his own work... — Preceding unsigned comment added by 68.7.175.67 (talk) 20:30, 14 February 2019 (UTC)
Agree, I believe this may not be a representative graph of the forgetting curve and rather confusing. The green curves, I believe, are the forgetting curve after reviewing the material with spaced learning. I will look into this and substitute it with a better image. Danieltaocf (talk) 17:40, 5 April 2020 (UTC)

I think the graph that was added by Danieltaocf greatly improves the strength of this article. It is much easier to understand than the one that was there prior and is overall a more accurate depiction. I know it is supposed to be a simple rendering, but I think the lack of units on the axis make it somewhat misleading. The graph looks like it is very close to approaching zero, but this is not necessarily the case for the curve. The plateau should be more pronounced, in order to effectively understand the concept. Schachterjo19 (talk) 19:30, 24 April 2020 (UTC)

Personal Conversation retention ability question[edit]

During a conversation where you are told what to do one time and then the conversation continues about other things. What is the lokely hood that you will remember the task, or that sttement? —Preceding unsigned comment added by 71.156.59.120 (talk) 16:25, 20 March 2009 (UTC)

Description section[edit]

Section refers to what sounds like someone's recollection of a study or commentary, but doesn't say where from. If I don't get to this, someone else might like into it. WWB (talk) 03:56, 29 May 2009 (UTC)

Description could be improved, I think the stregth of memory as a related concept can be explained later, after more information on the forgetting curve is introduced. "hypothesizes decline of memory retention in time" is a rather short and broad description for an important concept. Ecehanyurukoglu (talk) 11:58, 24 April 2020 (UTC)

Units[edit]

What are the units for R & S in the forgetting curve formula?

  R = e^(-t/S)  

This formula is a variation of the standard exponential half-life decay function

  y = y0 * e^(-(ln(2)/h)t)  

where

  S = h/ln(2)  

In that formula, h is in the same units as t (time), so I assume that S is also in the same units as t (time).

But what is R?

In the half-life formula, y and y0 are usually mass units. Since there is no R0, I am assuming that it is "1".

  R = 1 * e^(-t/S)  

This means that R goes from 1 to 0, which suggests that it's a probability. Is it the probability of recall?

It would be helpful to have this added to the article. — Preceding unsigned comment added by Griswold62 (talkcontribs) 00:40, 11 July 2011 (UTC)

  • Yes, s would have the same units as t; s is essentially the amount of time before R = 1/e. Time units cancel out and R has no units; it's a probability. Yes, the probability of recall. — Preceding unsigned comment added by 68.7.175.67 (talk) 20:33, 14 February 2019 (UTC)

Article is wrong -- Ebbinghaus' decay process is logarithmic, not exponential[edit]

This article is incorrect about the Ebbinghaus data fitting an exponential decay function. Ebbinghaus never mentions the R = e^(-t/S) function. In sections 28-30, his data fits a logarithmic function, which Ebbinghaus mentions. It also fits a power function almost as well. I ran his data through a curve fitting program and got a this power function

  R=0.317778t^(-0.1268656)  

and this logarithmic function

  R=-0.0455882×ln(t)+0.333966  

There is no exponential function that come close to fitting this data. — Preceding unsigned comment added by Griswold62 (talkcontribs) 00:52, 20 July 2011 (UTC)

Update 4 December 2017: it's pretty disappointing to see this section entitled "Article is wrong" to be unactioned six years after the comments were made. These comments are correct: the exponential formula given here is not correct. It links to a paper by Wozniak and Gorzelanczyk (1995), but this formula is not a finding or conclusion of that paper. In fact the formula is preceded by the words "Assuming the negatively exponential decrease of retrievability". So basically I consider this completely spurious. In short, the reference given does not support the claim made, it just happens to mention the same formula. Kitjohnson9 (talk) 09:07, 4 December 2017 (UTC)

2019-02-13: I agree completely with both commenters and changed the article to include Ebbinghaus's 1885 equation. I moved the R=e^(-t/s) equation to an 'equations' section and removed the wording implying that 'today we use this equation'... as it not true; I also cite a source, a 1999 article that found simple equations such as this do **not** fit the data well. The article, as well as the existing link to the Murre and Dros paper provide a good discussion of some other equations if anyone is interested, but I haven't bothered to add any to the 'equations' section. — Preceding unsigned comment added by 68.7.175.67 (talk) 20:40, 14 February 2019 (UTC)

Source[edit]

This Wired article has a lot of good information that should be incorporated into the article. JKeck (talk) 04:06, 20 May 2012 (UTC)

More on history[edit]

As Hermann Ebbinghaus being one of the first to carry out a series of rigorous experiments on the shape of forgetting, the history section of the wiki page could be expanded with facts such as why he chose to use meaningless syllables.Danieltaocf (talk) 18:23, 5 April 2020 (UTC)


Daniel, I agree with your idea to improve upon the history section of this. I think your edits are good and informative. However, I think this must be continued. The last sentence in the history section mentions a "large number of experiments in experimental psychology... based on highly controlled artificial stimuli"; these experiments should be explained below your addition. Regardless, this is a great start! Gdg1500 (talk) 23:43, 23 April 2020 (UTC)

Spacing effect[edit]

I suggest that it is worthwhile to add a section about beating the forgetting curve, namely the spaced learning and spacing effect. Danieltaocf (talk) 19:08, 5 April 2020 (UTC)

Daniel, I like how you expanded on the forgetting curve by introducing other psychological concepts. I am somewhat concerned, however, that the references you listed at the bottom of this section (9/11, Kennedy murder) cannot necessarily be compared to other examples of the curve, just because of the amount of coverage these events get. Thus, this makes it harder for one to forget because they are so often reminded of them in the news and popular culture. Therefore, I think they are quite extraneous situations that are not too comparable. Schachterjo19 (talk) 19:27, 24 April 2020 (UTC)

The Spacing Effect already has its own entry... no need to clutter this article with that IMHO. A link to the wikipedia entry and short explanation that the shape of the curve seems to be influenced by repeated study bouts is all that would be necessary, as the spacing effect is very much a separate issue -- Ed, 16:19, 2020-08-09

Peer Review[edit]

Daniel, I think you have a great start with edits to be made in this article. I agree that there is potential for a more effective image to be used to represent the forgetting curve, as it can be an easily understood concept if portrayed as such. I agree that added further information on how Ebbinghaus informed his study of the curve would help you to build out the background of the phenomenon. I think you could also add previous studies that showed application of the forgetting curve that have resulted from Ebbinghaus' findings. Finally, while I do not know necessarily if it is possible to "beat the forgetting curve" but you could potentially link articles to other mnemonics and strategies that help to improve long-term memory. Czaharris (talk) 22:10, 23 April 2020 (UTC)


Daniel, I too think that the updated picture was a good start in terms of your edits. It is critical to have data, such as pictures and graphs, that accurately correlate to one's argument. Therefore, you have really strengthened this Wikipedia page in more ways than one. It would be very interesting to see if this curve has evolved at all since one's lifestyle today is a lot different than what the curve was measured out to be whenever it was invented. Also, as Czaharris said, it would be interesting to see if the outliers of the curve experience a similar trend in terms of forgetting over time. User: johnaguilar007 —Preceding undated comment added 03:01, 24 April 2020 (UTC)

I agree with the above that the contributions made have been valuable to the page, in particular with regard to a more useful image and a description of what "savings" means in the context of the study that led to the initial equation. The citations also helped make a more complete scientific picture of how the forgetting curve has been studied. A couple minor suggestions, I think it would be useful to replace the equals sign in the first equation with an "≈" symbol to express that this is an approximation if someone is to lift that equation from this article without reading the context. Also, potentially changing the wording of "Battle the Curve" to something more academic in nature. But overall still great job! Ask.krier (talk) 19:05, 25 April 2020 (UTC)

Peer Review[edit]

I think you did a great job on additions and rewording to make it flow better. Brynneosh (talk) 00:48, 25 April 2020 (UTC)