SuperMemo user survey

February 1999

New article by Dr. Piotr Wozniak: The true history of spaced repetition

Most of the presented results have been compiled from data submitted by users of SuperMemo via questionnaires available from this website as well as questionnaires in-built in the newest releases of SuperMemo that allow of automatic submissions via-e-mail

Our 1999 user survey shows remarkable similarities with results obtained in 1994. Seemingly nothing has changed: our average user is about 30 years old, spends 30 minutes per day on repetitions and has about 5000 items memorized in his collection.
Similarly, 50% of users have a university education, only 3% are women (!), students and engineers prevail.
One in two users, when asked what he (or she) likes most in SuperMemo says: the effectiveness of the learning method.

If there has been any major change in the world of SuperMemo since 1994 it could be summarized in one word: the Internet!
In 1994, 88% of users surveyed quoted press as the source of information from which they have learned about SuperMemo. Currently, the press is down to 3% with 79% of users quoting the Internet as the place where they first seen the word: SuperMemo.

Secondly, there has been a remarkable increase in customer satisfaction. Only 16% of users surveyed in 1994 called SuperMemo an excellent piece of software. Now this figure stands at 50%. The satisfaction with the SuperMemo method of learning also increased slightly. 74% of users consider it an excellent aid in learning, up from 69% in 1994.

Contents:

  1. The user
  2. Age
  3. Country
  4. Occupation
  5. Education
  6. Using SuperMemo
  7. Speed of learning (!)
  8. Opinions about the SuperMemo method
  9. Opinions about the SuperMemo program
  10. Opinions about the SuperMemo website
  11. How did you learn about SuperMemo?
  12. What do you like most in SuperMemo?
  13. What do you dislike most in SuperMemo?
  14. Proposed improvements to SuperMemo
  15. SuperMemo Website improvements
  16. SuperMemo Help improvements
  17. How to increase the popularity of SuperMemo?
  18. Versions
  19. Most requested library collections
  20. Winners

The user

Only 2.3% of questionnaires were submitted by women (see Survey 1994)! The proportion of female users is actually much higher but women clearly have been less inclined to send us their feedback.

Of those who responded, nobody was able to point to a competitive product in the area of spaced repetition.
100% stated that articles presented on this website are interesting and understandable.
Also, 100% of users say that the price for SuperMemo is OK (although a significant proportion proposed reducing prices in SuperMemo Library; see Library Survey coming soon).

Age

The number of users below the age of 20 dropped from 27% to under 1%. Still the average age was about 33 with the following age distribution (see Survey 1994):

  • 1% - below 20
  • 18% - 21-25
  • 38% - 26-35
  • 27% - 36-45
  • 14% - above 45

Country

Country statistics were similar to sales statistics except for a higher proportion of respondents from Eastern Europe (esp. our native Poland, Czech Republic, Slovakia, Hungary, Romania and Bulgaria):

  • 27% - US
  • 11% - Poland
  • 8% - UK
  • 6% - Netherlands
  • 5% - Canada
  • 4% - Germany
  • 3% - Czech Republic

Occupation

Traditionally students and computer scientists predominate. Still very few teachers are users of SuperMemo!!! (see Survey 1994)

  • 36% - student
  • 12% - programmer
  • 10% - engineer
  • 7% - physician
  • 4% - scientist

Education (see Survey 1994)

  • 6% - Ph.D.
  • 32% - MSc, MD
  • 19% - BS
  • 2% - BA
  • 3% - MBA
  • 30% - student
  • 0.5% - secondary education

Using SuperMemo

Average use time was 2.14 years, with 31 min/day, and 5375 items memorized (see Survey 1994).

Speed of learning

Not all users provided all components needed to compute the speed of learning: total learning time and the number of items memorized. Additionally, we have eliminated results for less than 3 months of learning and for less than 300 items memorized. Consequently, the obtained data is much less representative than the overall sample and differs significantly from the results obtained in 1994 (see Survey 1994).

The average speed of learning was 163 items/year/minute (243 in 1994). 65% of users reported speeds between 50 to 250 items/year/minute. The lowest speed reported was 25 items/year/minute and referred to a difficult job-oriented material packed with difficult to digest formulas in physics.
The highest speed was 1000 items/year/minute and was reported by Peter Cool from the Netherlands. Peter has memorized 2300 words in over a half a year working only 3 minutes/day; however, he did not count the time when he massively memorized new words (up to one hour per day). He also noted that French vocabulary is relatively easy to learn for Dutch people, esp. that he has attended a course of French 25 years earlier.

The greatest number of items memorized: Gabor S. (USA, Advanced English), 10,000 in one year, 20 min/day (speed: 500 items/year/minute). Gabor's record must be taken with a correction because of the fact that he was familiar with a great deal of vocabulary included in Advanced English. We are waiting for news from record holders from earlier years if they have improved upon their results (NB: at SuperMemo World we have four employees with over 10,000 items memorized; among them, the host of SuperMemo Library, Michal Hejwosz has memorized 12,000 in the last four months!).

There has been no correlation between the age and the speed of learning (0.07), as well as the university degree and the speed of learning (0.18).

A slight negative correlation with the speed of learning was found for (1) the time spent for learning per day (correlation: -0.36; users with speeds below 100 spent 2 more minutes per day) and (2) years of use (correlation: -0.19; users with speeds below 100 spent 4 more months in the learning process). There has also been nearly no correlation between speed of learning and the opinion about the method and the program (-0.2 and -0.1 respectively).

Opinions about the SuperMemo method (see Survey 1994)

  • 74% - Excellent aid in learning
  • 26% - Useful idea
  • 0% - Nothing special
  • 0% - Artificial approach to learning
  • 0% - Waste of time

Opinions about the SuperMemo program (see Survey 1994)

  • 50% - Excellent piece of software
  • 41% - Reliable product
  • 9% - Good software that is not user friendly
  • 0% - Poor quality software
  • 0% - Bag of bugs

Opinions about the SuperMemo website

It was a very pleasant surprise to find out that you like SuperMemo website even more than SuperMemo itself! Here are the opinions:

  • 83% - Enlightening
  • 17% - Interesting
  • 0% - Nothing special
  • 0% - Boring
  • 0% - Exaggerated claims
  • 0% - False information

How did you learn about SuperMemo? (see Survey 1994)

We have investigated the ways you find out about SuperMemo in most detail. Not only user questionnaires and registrations were used but we have also added a special field that can optionally be used while ordering SuperMemo. All these data point to the fact that the Internet is the greatest news for SuperMemo since the first efforts to market the program were undertaken in 1988. The Internet has replaced the press as the main source of information about the program.
We are glad to see that a word-of-mouth still plays an important role with 11% of users discovering SuperMemo via recommendation.

  • 79% - Internet -- of which (if specified):
    • 55% search engines (of which 68% is Yahoo!). Most often used keywords: spaced repetition, memory software, mnemonic, memory, teaching aids, creativity, accelerated learning, superlearning
    • 45% links -- of which:
  • 11% - Friend's recommendation
  • 5% - SuperMemo freeware
  • 3% - Newspapers -- of which
    • Poland 30%
    • Czech Republic 20%
    • Netherlands 20%
    • Slovenia 10%
    • Australia 10%
  • 1% - Software fairs -- of which:
    • Comdex 66%
    • CeBIT 33%
  • 1% - Course in learning techniques

What do you like most in SuperMemo? (see Survey 1994)

  • 55% - the method of learning
  • 14% - possibility to add one's own material (esp. images, sound, video, etc.)
  • 10% - interface
  • 4% - user-friendliness
  • 16% - other (improved concentration, programmability, potential, Advanced English, possibility to use Japanese fonts, possibility to integrate mind-maps from Mind Manager, etc.)

Other positive opinions and quotes:

it keeps me coming back to learning all the time,
without it I wouldn't spend so much time learning,
ability to save graphics with a question,
ability to import Q&A from a text file,
ease of use (with a little time investment you can achieve a lot),
very user friendly software,
the fact that you can also learn about how our memory works,
most useful software in my collection!,
the excellent technology of learning,
ease of creating items in basic mode,
Awesome, just what I've been looking for!!!!,
Thank you very much for your great work
With it's new features SuperMemo 98 is a wonderful piece of learning software
This program is exceptional!
This is the time that I finally graduate...

What do you dislike most in SuperMemo? (see Survey 1994)

  • 26% - not intuitive in use (mostly in reference to SuperMemo 8)
  • 9% - interface
  • 8% - bugs (mostly in reference to SuperMemo 8.2 and 8.4)
  • 6% - slow (exclusively in reference to SuperMemo 8)
  • 4% - no possibility to export the material to Access or Excel
  • 34% - other complaints
  • 10% - full satisfaction

Other dislikes and quotes:

Too few collections
Too big files
I don't like the clock and there is no easy way of hiding it
Too many choices when grading the answer (3 would be enough: good, pass, fail)
Inaccuracies in English Grammar and Advanced English
Too many windows
Not sticking to MS standards (e.g. resizing components, property sheet with double-click, etc.)
I do not like help in SuperMemo 8
Too much change from SuperMemo 7 to SuperMemo 8
Too many statistics
Self-grading
Limitation on the size of text
(this has been eliminated in SuperMemo 98)
No statistics for a single day of learning

Trouble with importing QA text to the correct node
Lack of a good backup method
Poor printing
Lack of multi-user options
ABC of SuperMemo is too large for lower resolutions
The slow rendering of gif images

Proposed improvements to SuperMemo (see Survey 1994)

  • 9% better printing options
  • 8% better user friendliness (hopefully improved with introducing difficulty levels in SuperMemo 98)
  • 5% remove the limit on the size of texts (implemented in SuperMemo 98)
  • 3% automatic backup
  • 2% verbal report on the learning progress

Other propositions:

Adding mathematical formulas (implemented in SuperMemo 98)
Spell checker
Undo
Better search facilities: incremental search, boolean search, regula expressions
I would like to be able to operate SuperMemo without using a mouse (
this is possible)
Record WAV files (
this is possible with the sound component's Recorder option)
More standard way to edit components (like in Microsoft software)
Editing translations (
it is possible with Shift+click)
Changing fonts throughout the collection
(it is possible with Edit font or with global templates)
Combine the SuperMemo algorithm with a browser as GUI
Implement a Linux version
(we plan to port SuperMemo 2000 to Linux once Borland's Kylix project is completed in late 2000)(information: luis@supermemo.com)
Easier input of Q&A pairs
Integrating with to-do lists
(implemented in SuperMemo 99)
Interfaces with databases created using other software, as Access, Excel
I would like to be able to create the exposition material in HTML
(this is possible)
I would like to create learning material via the Intranet
Improve import features
Speed up the gif drawing, it was faster in SuperMemo 8.4
(it is not an issue in SuperMemo 2000)
Possibility of importing text file containing Q&A into template consisting of text as question and sound-component as answer
(it is possible)
Supporting further audio formats
(e.g. RealAudio) (RealAudio and MP3 are supported in SuperMemo 2000)
Randomly selecting one of several customized sound feedback (currently there are only sounds for success and failure)

SuperMemo Website

SuperMemo website is liked for richness of information (38%), frequent updates (25%), easy navigation, scarcity of images, screenshots, fast loading, freeware and software upgrades. It has been criticized for an excessive number of hyperlinks and information scattered among many pages.

Most requested enhancements:

  • 36% - more library collections (30% of which stressed freeware collections) (library is currently growing at the rate of two new collections per week)
  • 32% - more articles on self-improvement, health, memory, speed-reading, etc.
  • 20% - chat-room, discussion forum, mailing lists, etc.
  • 12% - tips on how to effectively formulate knowledge (already on the site!)
  • 8% - tips&tricks on how to use SuperMemo (see: Hints&Tips and Guide)
  • 7% - more links
  • 5% - Site Map (see: Site Map)

Other opinions and propositions

Provide collections for earlier versions of SuperMemo (library stores a number of collections compatible with SuperMemo 6 and SuperMemo 7)
More collection samples when ordering
Knowledge network of SuperMemo users
Using ScreenCam for demos
Comparison with a program that doesn't use SuperMemo method but just random repetitions
Database of books on memory and related areas
Theoretical and practical papers on fast reading techniques
I really enjoyed Dr. Wozniak's commentary on memory, which sold me onto the program
I would like to see some case studies of people who have accomplished goals using SuperMemo
SuperMemo Press with screenshots
The Dutch flag "hanging" below the bottom of my screen
More links to school/education/(US)K12 sites...
Metatags for learning
Sometimes I can't find what I'm looking for and saw before (e.g. SuperMemo 7 freeware collections)

Most requested improvements to SuperMemo Help

We will successively cover the questioned area according to your priorities:

How to increase the popularity of SuperMemo?

  • 14% - Increase the presence at universities
  • 7% - Better marketing (that's quite obvious but ... not specific enough!)
  • 7% - Banner campaign on the net (until now we scored very poor click-through rates!)
  • 6% - Make it simpler
  • 6% - SuperMemo for PalmPilot, Windows CE or Psion with easy exchange of data with a PC
  • 5% - Testimonials and referrals (e.g. from Tony Buzan, other users, etc.)
  • 4% - Advertise in my country
  • 3% - Ask for links (esp. on creativity sites)
  • 2% - Shareware/Freeware (e.g. release SuperMemo 7 as freeware or even ... make all SuperMemos free and earn on selling collections)
  • 1% - Make SuperMemo Library freeware

Organize easy collection exchange between users
Partnership with other companies (SuperLearning, Accelerated learning, etc.)
Hire a Public Relations consultant
Create a mnemonic collection with a peg system
Get in contact with publishing houses of schoolbooks
More public presentations
Write a book
Don't make SuperMemo a program for all people because you're making simplifications which destroy the idea!
Correct bugs in Advanced English
Try advertising in Microsoft Certified Professional Magazine (www.mcpmag.com).
Easier input of data (e.g. with voice)
Version for PalmPilot might be a breakthrough!
Links on educational sites, and shareware sites
Merge with US encyclopedias (like it has been done with a Polish encyclopedia)
Target collections on specific audiences (e.g. medical students)
Attack universities. You got a goldmine there!

Versions

85% of users did not fill out the version field and we believe these were mostly new users of SuperMemo 8 or SuperMemo 98. Among the remaining 15% the percentage of those who used earlier versions was significant (see Survey 1994):

  • 9% - SuperMemo 2
  • 13% - SuperMemo 5
  • 22% - SuperMemo 6
  • 26% - SuperMemo 7
  • 82% - SuperMemo 8
  • 69% - SuperMemo 98

Most requested library collections: (see Survey 1994)

  • 33% - Computers (Internet, JavaScript, HTML, Windows NT, Programming, Perl. C++, Unix, Latex, OpenGL, etc.)
  • 25% - Languages (English 45%, French 10%, German 9%, Spanish 6%, Japanese, Italian, Russian, Swedish, Polish, Chinese, Greek, etc.)
  • 13% - Medicine
  • 8% - Mnemonic techniques
  • 7% - Career-specific material
  • 3% - Touch typing
  • 11% - Other

Winners of the questionnaire competition

Jorge R. Garcia de Alba, USA
Pawe³ Bura, Poland
Tomas Cermak, Czech Republic
Othello D. Cobal, Phillipines
Leonard Erwin, South Africa
Giuseppe Esposito, Italy
Marius Felecan, Romania
Ed Freeman, UK
Mary Garland, Ireland
Adam Gourley, USA
Edward GK, Hong Kong
Bjørn Hallberg, Norway
Ronald Y. Hanagami, USA
Mark Hooker, Sweden
Miro Hrzasky, Slovakia
Willfrid Janzen, Canada
Jan Krabec, Czech Republic
Peter Lada, Hungary
Fernando Ladeira, Brazil
Sean McAree, UK
Ian Moulster, UK
Mark Nanut, Slovenia
Jens Nielsen, Denmark
Antonio Paterno, Italy
Lu, Ming-Pin, Taiwan
Victor Hitiel Rodriguez Sardina, Japan
Philip Stoyanov, Bulgaria
Jorge Villalobos, Colombia
Cezary Wagner, Poland
Armin von Werner, Germany


See also: Survey 1994 and Survey 1993

1.2.27