On Sat, Apr 25, 2009 at 2:06 AM, Peter Bienstman <[email protected]> wrote: > > Really wonderful, thanks! > > Once the final version of your thesis is published, do you mind us putting a > link to it on our website? > > Peter > > On Saturday 25 April 2009 03:41:09 am Bill Price (formerly Notyourbroom) > wrote: >> Hello, >> >> As my first attempt at an experiment, I examined the use of Mnemosyne >> in the context of a Mandarin Chinese course. I had two groups, one of >> which used the off-the-shelf version of Mnemosyne and one of which >> used modified version that had no scheduling algorithm. (In other >> words, the second program always scheduled cards for review the next >> day, regardless of user input, so those users would activate just one >> or two sub-decks to study each day, depending on what they felt the >> greatest need to study.) >> >> Because it was conducted in the context of an academic course and >> lasted for three full weeks, the experiment was difficult to control >> to any acceptable level of rigor, and the results are essentially >> confounded because so many factors (number of cards studied per day, >> number of cards studied in total, specific material studied, etc) >> varied from person to person and between the two groups. The >> differences between the two groups are still quite striking, though. >> >> Check out this box plot: >> >> http://tinyurl.com/cydnu4 >> >> It looks to me as though some folks were predisposed to score at or >> near ceiling level regardless of the intervention (as would be >> expected from a pool of subjects recruited from a class essentially to >> do extra coursework), and that this kept the means of the groups from >> separating. However, the large discrepancy of variance between the >> groups suggests that the spaced repetition intervention is boosting >> the scores of low performers—for whatever reason, be it the spacing >> effect or else a simple cultivation of good daily studying habits. >> >> To quote from my introduction, >> >> "This thesis argues that the spaced repetition intervention resulted >> in a significant increase in evaluation scores and that the intuitive >> repetition intervention did not; that the spaced repetition >> intervention was in particular of better help to struggling >> individuals than the intuitive repetition intervention was; and that >> the spaced repetition intervention appeared to promote a distribution >> of scores in which subjects clustered closely in the higher score >> range, whereas the intuitive repetition intervention resulted in a >> wide distribution of scores with high intra-group variance. While the >> differences between the two intervention groups are clear, the present >> study was unable to determine which specific factor or factors >> contributed the most to the success of the spaced repetition group." >> >> I finished the first draft of my full manuscript today: >> >> http://tinyurl.com/cqfpjz >> >> It will need further revisions and copy-editing (see if you can find >> the error in the abstract!), but the content is essentially complete. >> >> —Bill
"This thesis argues that the spaced repetition intervention resulted in a significant increase in evaluation scores and that the intuitive repetition intervention did not;" Is this really warranted? Seems to me that equally consistent with the evidence is the suggestion that both groups' scores were increased - just the spaced group increased *more*. At least, I'm not seeing any score comparison with students who did neither intuitive nor spaced repetition. As you yourself write later: "The research hypothesis was formulated to test the claim that a computer-scheduled spaced repetition rehearsal regimen would yield better long term memory retrieval results than a self-scheduled studying regimen" "While the differences between the two intervention groups are clear, the present study was unable to determine which specific factor or factors contributed the most to the success of the spaced repetition group." A good point. I tend to think that what's going on here is less procrastination, or rather, perhaps the lower-scoring members of the intuitive group are choosing easy topics. That intuitives are wimping out certainly seems to be supported by 5.2.4 where you write about how the spaced repetition students wound up doing quite a few more (32.6%) repetitions than the intuitives. One could check for the former by seeing how many students in either group were punctual & always reviewed; and perhaps one could check for the latter by plugging their data into the spacing algorithm and seeing whether the intuitive's reviews were disproportionately favoring cards which would be ranked a 4 or 5 (that is, they were reviewing stuff they already knew instead of the ones they should've been - the ones they didn't already know). Note the first suggestion is not entirely obviated by your attrition statistics, since as far as I can tell, one could be irregular in review and still complete the study, and the learning data mentioned in 4.2 is never used to tell us how compliant everyone was; all we get is "Additionally, although most subjects exhibited very close adherence to the research protocol, some subjects skipped one or more days of rehearsal sessions, leading to further gaps in the learning record as well as possible extraneous learning effects. In short, this study suffered from a rough beginning and occasional subject noncompliance." (The second suggested analysis is not the same as checking whether the intuitives favor more recent questions.) But a good study for what it is. Certainly I'll be reading through the bibliography in the future, and I look forward to the final version (be worth linking to from the Wikipedia article on Mnemosyne at least). -- gwern --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "mnemosyne-proj-users" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/mnemosyne-proj-users?hl=en -~----------~----~----~----~------~----~------~--~---
