4
$\begingroup$

I've recently come across the following article: http://www.super-memory.com/help/il_full.htm The author makes the claim that "Incremental learning" is the best way of learning and goes as far as to claim that the problem of "forgetting" is nearly completely avoided. For me the whole thing just sounds like advertising for the program "SuperMemo" but apparently the author Piotr Woźniak holds a PhD and is in fact a researcher. https://en.wikipedia.org/wiki/Piotr_Wo%C5%BAniak_(researcher) Is he a known fraud or is the information presented a mainstream opinion shared by psychologists? Thanks in advance

$\endgroup$
  • $\begingroup$ I think the question should focus on this claim "Once the art of incremental learning is mastered, the advantages go far beyond the advantages of the interruption or spaced repetition." Because spaced repetition is fairly well studied. $\endgroup$ – Fizz Apr 8 at 19:52
  • 1
    $\begingroup$ There is a long list of publications in relation to SuperMemo here, but a few I've tried finding on scholar do not show up, even though they are fairly recent (90s) publications. Google search doesn't show up anything either. From my own experience having looked up various research, including stuff in the 80s, it has always been easier than this. Call me a skeptic. $\endgroup$ – Steven Jeuris Apr 8 at 21:54
  • 1
    $\begingroup$ @StevenJeuris Good catch; most of those publications are in Polish venues I am not familiar with, but there was one at least in a journal I've heard of...the content seems like it is not what that journal typically publishes, and I went to the archives of the journal and the issue they cite doesn't even exist (there are no supplemental issues in that journal). Seems very very fishy. $\endgroup$ – Bryan Krause Apr 9 at 16:45
  • 1
    $\begingroup$ Ah, now I think I have it figured out. Eur J Neurol publishes their annual meeting abstracts in a supplement, but the older ones aren't online. So these "publications" were presented at a conference but never peer reviewed or submitted to a peer reviewed journal. That's a kind of laughable way to try to make your work seem legit when it is not. Red flags galore. $\endgroup$ – Bryan Krause Apr 9 at 17:03
  • 1
    $\begingroup$ @StevenJeuris Peer-reviewed conferences are not the norm in this field. I would argue that this is a proper journal, but their choice to not make sure their conference abstracts are archival is because in this field poster abstracts are just that, poster abstracts, and no one expects those to be archival (in this field). They are typically where people present in-progress work that later becomes part of a full publication. $\endgroup$ – Bryan Krause Apr 9 at 17:28
4
$\begingroup$

The merits of any given body of work is best assessed by looking at it, regardless of whether you want to label it as 'pseudoscientific' or not based on that judgement.

Unfortunately, this proved to be quite troublesome for the SuperMemo application, which makes some pretty strong claims that require justification to be considered academic. For example:

Incremental learning is the fastest and the most comprehensive way of learning available to students at the moment of writing (2013).

Incremental learning easily ensures 95% recall of top-priority learning material for lifetime

To be considered scientific, there would have to be publicly accessible publications backing up those claims. I found the following list of 50 publications in reference to SuperMemo. But, I could only find the following 4 publications out of the 50 listed on Google Scholar:

  • A hypothetical model of processes involved in the release of norepinephrine in the adrenergic synapse. (Gorzelańczyk & Woźniak, 1991): cited by 2; the abstract seems irrelevant to the claims made
  • The educational technology of repetition spacing in information systems development. (Wozniak & Abramowicz, 1997): 4 references, all self-citations; no tracked citations on Scholar; furthermore, listed twice in the list of 50 publications
  • SuperMemo (Wharton, 1994): a 5-page software review, representing a subjective case study at best ("As a new German student, I have been using SuperMemo ..."); no tracked citations
  • Optimization of repetition spacing in the practice of learning. (Wozniak & Gorzelanczyk, 1994): 4-pager, 7 references, cited by 79

The fact that the majority of these publications are not available indicates these were not made 'archival' by the conferences/journals they were presented at. This gives a first poor assessment of the adequacy of this presented work. Rigorous, peer-reviewed, work would be accessible and cited by subsequent research.

The last paper (4 page 'short communication') with 79 citations presents:

A universal formula for computing inter-repetition intervals in paired-associate learning has been determined for the knowledge retention level of 95%. It is claimed that the formula could be used in the practice of learning for a wide range of subjects, regardless individual learner's capacity.

The method they employed to evaluate their algorithm:

The subjects taking part in the experiment were 7 unpaid volunteers, students of computer science at the Technical University of Poznan. In the period of 18 months, they memorized and repeated altogether over 35000 items of their choice (the items had the form of Polish-English word pairs).

I'll leave it up to you to decide whether or not you agree with their conclusion:

The obtained results unequivocally indicate the superiority of progressive repetition spacing over massed or distributed spacing and may be used in practice of learning in order to minimize time necessary for memorization and retention of the learned material.

Or, alternatively, follow-up on the 79 citations to see how experts in the field formulate the work. One example (Lindsey et al., 2014), emphasis mine:

Outside the academic literature, two traditional adaptive-scheduling techniques have attracted a degree of popular interest: the Leitner (1972) system and SuperMemo (Wozniak & Gorzelanczyk, 1994).

Interestingly, Wozniak seemed to have been a reviewer on that paper!

Conclusion

While the SuperMemo application does seem to have made a noticeable commercial impact, of which the Wired article "Want to remember everything you'll ever learn? Surrender to this algorithm" and citations to the one paper is a good indication, it does not seem to be embedded in rigorous published scientific research.

This does not necessarily make the application 'pseudoscientific' as it might still base itself or correspond to 'proper science', but, the way the website represents itself as scientific is grossly overstated. I would not place much trust in anything written there.

Wozniak, P. A., & Abramowicz, W. (1997). The educational technology of repetition spacing in information systems development. In Systems Development Methods for the Next Century (pp. 341-344). Springer, Boston, MA.
Gorzelańczyk, E. J., & Woźniak, P. (1991). A hypothetical model of processes involved in the release of norepinephrine in the adrenergic synapse. BioSystems, 25(4), 275-281.
Wharton C. (1994). SuperMemo. CALICO Journal Vol. 12, No. 1 (Fall 1994), pp. 66-70
Wozniak, P. A., & Gorzelanczyk, E. J. (1994). Optimization of repetition spacing in the practice of learning. Acta neurobiologiae experimentalis, 54, 59-59.
Lindsey, R. V., Shroyer, J. D., Pashler, H., & Mozer, M. C. (2014). Improving students’ long-term knowledge retention through personalized review. Psychological science, 25(3), 639-647.

$\endgroup$
  • $\begingroup$ How do you find the reviewers for the Lindsey paper? And I thought that reviewers should be anonymous? (and if yes, does that mean that if the reviewers have conflict of interest, they should honestly report to the editors?) $\endgroup$ – Ooker Apr 10 at 1:21
  • $\begingroup$ @Ooker Reviewers are not always anonymous. I inferred he might have been given the acknowledgements at the bottom of the paper in question, but this could also have been informal feedback. Yes, usually they need to mention conflicts of interest. To double check, you could verify the reviewing process of the journal. This is typically documented. $\endgroup$ – Steven Jeuris Apr 10 at 6:56
0
$\begingroup$

WIRED, has an article just about him and his lifestyle: Want to Remember Everything You'll Ever Learn? Surrender to This Algorithm

An anecdote: I built a flashcard deck for Anki program. From my observation Anki is one of the leading software in the market for spaced repetition software. The program's algorithm is based on SuperMemo 2 algorithm..

So we have these names: WIRED, Anki, spaced repetition. They seem to converge to one conclusion.

$\endgroup$
  • 1
    $\begingroup$ The Wired read is a fascinating piece, but I'm not sure how it answers the question. The SuperMemo doc makes the claim that "incremental learning" (whatever that is) is better than spaced repetition. $\endgroup$ – Fizz Apr 9 at 14:30
  • 1
    $\begingroup$ "SuperMemo shuffles all his potential knowledge into a queue and presents it to him on a study screen when the time is right. Wozniak can look at a graph of what he's got lined up to learn and adjust the priority rankings if his goals change. These techniques are designed to overcome steep learning curves through automated steps, like stairs on a hill. He calls it incremental reading, and it has come to dominate his intellectual life." So possibly "incremental learning" means just en.wikipedia.org/wiki/Incremental_reading but I'm not sure. $\endgroup$ – Fizz Apr 9 at 14:33
  • $\begingroup$ @Fizz thanks for the information. I think the question has two actual questions: Is incremental learning/reading legitimate, and is Wozniak fraud? I guess I have answered the latter; maybe you want to answer the former, as you have read it more than me $\endgroup$ – Ooker Apr 9 at 15:46
  • $\begingroup$ from the Wikipedia link: "Incremental reading was the first of a series of related concepts invented by Piotr Wozniak: incremental image learning, incremental video, incremental audio, incremental mail processing, incremental problem solving, and incremental writing. "Incremental learning" is the term Wozniak uses to refer to those concepts as a whole." Checking the reference of that bit, it just points to the SuperMemo link OP gives. So well, no third party confirms this $\endgroup$ – Ooker Apr 9 at 15:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.