Experiment Setup: Study Drugs

One of the experiments I am currently carrying out is on the effects of different “brain boosters” or “study drugs” on learning.  My intention was to create something close to a “one man blind clinical trial” on the different drugs, so I had a friend remove the labels from each of the bottles.  I chose the drugs carefully for a number of reasons which I’ll cover when I publish the full results, but here I’ll explain my basic experiment setup.


I’ll be taking each drug for 1 week and recording the results – but the difficult part was figuring out how to quantify my results.  Researchers have struggled with this problem for years.  Hermann Ebbinghaus famously invented the idea of a ‘nonsense syllable’ in the late 1800s.  The advantage to his system was that there was no information crossover – everything he was learning was completely new.  I have no interest in “wasting space” in my brain on nonsense, so I am learning French instead.  Admittedly, this means that certain words will naturally be easier to learn because they are similar to English (a foreign language is never completely foreign, after all).


So I am using Anki, a wonderful SRS flashcard application, in order to memorize French words.  Anki natively supports time-boxing, so I have it set to use a 20 minute session length for each study session (here’s why).  I’m keeping track of how many flashcards I study each day in those 20 minutes, as well as the average ease.  Because SRS systems require that you rate the “ease” of each word (in Anki’s case, a number 1-4), I can simply look at the average ease for any study session to have a sense of how hard the material was.  It is not a perfect system, but it does a reasonable job of letting me know how well I’ve mastered the material.  I also use a tiny free application for Mac OSX called Think, which simply blacks out the rest of the screen and forces me to focus entirely on my studies.


I wanted more ways to quantify my results, though, so I took a look at a company called NeuroSky.  One of their better-known products is the Mindflex Game, which is essentially a board game played using your mind.  By wearing a small headband you can actual control the path of a ball through the air with your mind (thanks to a clever system of fans).  The product I was interested in, though, is called the MindSet.  It looks like a pair of slightly bulky headphones (and it can act as headphones), but it can transmit data about your brainwaves to a PC, Mac, mobile device, etc.


Unfortunately, there was no software that existed to do exactly what I wanted.  Developers have created games for the MindSet and there are expensive ($200-$500) research-grade pieces of software that exist, but neither was right for me.  So I decided to put together my own solution.



In about 4 hours on a Saturday morning I made the application you see above, which simply records the primary data coming in from the MindSet.  The two most interesting things are the attention and meditation levels, though it also records all available brainwaves (alpha, theta, etc.).  These “attention” (focus) and “meditation” (relaxation) values are simply another way for me to look at how my brain is functioning during my study session.  Will certain drugs increase my focus, and will my focus be linked to better recall?  Only time will tell!


(If you’d like a copy of this simple application for OSX, feel free to drop me a line.  You’ll need a NeuroSky MindSet to be able to use it, but its fairly easy to set up and use.)

Finally, to put it all together, I am using Google Docs to keep a daily entry for each of these values.  When I finish the experiment in another few weeks I’ll analyze the results and make the data a bit more clear.  But for now…

Click here to see the Google Docs spreadsheet as I update it daily with my results!



Potential Problems


No experiment is perfect.  Here are a few things which I have already identified as problems with my setup.  If you can identify any more please let me know!


  • Not all French words are created equal.  Some will naturally be easier or harder to memorize; statistically this should hopefully even out over each phase of the trial (each week).  If I had more time I would allot a month for each drug.
  • By the very nature of how learning a language works, the words that I add later may be characteristically different than those at the beginning.  For example, at the beginning I will be learning colors and directions and basic verbs, where by the end I should be learning complex adjectives and nouns.
  • I have allotted one day between each drug to clear my system (6 days of pills, 1 day off).  This may be insufficient – I have not found enough information about each drug to be 100% positive (will some of the drugs be stored in body fat, for example, and released over time and therefore effect the next phase?  Do some of the drugs take longer to take effect?  I honestly don’t know).



  1. Avatar
    Jon Hughes -  April 11, 2011 - 4:35 pm 6

    I’m really looking forward to this experiment — how did you go about choosing the drugs you would include?

    • Avatar
      Zane Claes -  April 11, 2011 - 8:37 pm 8

      A lot of googling 🙂 The first is a studied Alzheimer’ drug, the second was put together by a well-known neuroscientist and is marketed as a focus & memory booster, and the third a cocktail of vitamins and oils… a veritable cross-section of the market. I’ll do a follow-up this week with more details.

  2. Avatar
    Ron Murphy -  April 27, 2011 - 10:33 pm 42

    Interesting. Do you think there will be a progressive learning effect, so that as you acquire familiarity with the word forms of the language they become easier to learn? How would you determine the impact on this? Wouldn’t you need to repeat the trials over longer periods, maybe with other languages, rotating the drugs so that they are used at different stages of the experiment to compensate for this effect.

    • Avatar
      Zane Claes -  April 27, 2011 - 11:15 pm 44

      Wonderful point, Ron, and one I’ve been thinking about more. To start, I think that you’re definitely right that doing this experiment several times in different orders and with different languages would have a nice balancing effect. I think I will do exactly that (though I will also publish the first results by themselves), but if I can I will try to collect data from several months and languages.

      The use of Anki as my study tool should have a mitigating effect on this problem. Anki chooses only information which has a high probability of having been recently forgotten meaning that the difficulty curve should be rather flat. If you have not already I highly recommend that you check out the research on SRS.

      Still, you may be right. One thing I will definitely do is take a look at the overall trend in the data. I would expect, if you are right, that there will be a linear trend in terms of “ease.” In this case I can compensate for the trend by applying a trendline and considering the offset from the predicted value instead of the literal value.

Leave A Comment

You must be logged in to post a comment.

Back to Top