Baseline in My Profile
How is the baseline or reference line in My Profiles calculated? Is it the average of all the current users, all the past users, or some standard objective reference?
A FAQ would enhance this website.
A FAQ would enhance this website.
1stone | 4 years ago
Reply
cognitivefun | 4 years ago
Reply
"How is the baseline or reference line in My Profiles calculated? Is it the average of all the current users, all the past users, or some standard objective reference?"
It's based on the averages of the first three measurements for every existing user at every time of calculation.
This does mean the baseline is expected to fluctuate from time to time, but fluctuations should also stabilize after the sample size gets larger. For example, the auditory digit span average is already fairly stable, and confirms expectations based on literature (~7).
It also means that it is very hard to go under 100%, which isn't a bad thing because the idea is to look at performance relative to the untrained average.
This does mean the baseline is expected to fluctuate from time to time, but fluctuations should also stabilize after the sample size gets larger. For example, the auditory digit span average is already fairly stable, and confirms expectations based on literature (~7).
It also means that it is very hard to go under 100%, which isn't a bad thing because the idea is to look at performance relative to the untrained average.
cognitivefun | 4 years ago
Reply
Yes, once you go into building a successful site, it rapidly becomes a full-time job :-)
But keep it up... it's addictive.
I'm not sure the first three measurements are necessarily representative of an untrained person. I spent a couple of weeks 'playing' anonymously before I registered. And judging from other high scores by anons, this may be widespread. Also, I would venture to say that the average user of this website is a bit off the average.
How about adding a second red line in the performance chart, showing the current average of the (trained) users?
As a second suggestion, it may be a good idea to take the median rather than the average in each individual score; I often get a run of good timings, only to be distracted and get one outlier which destroys the arithmetic mean average.
But keep it up... it's addictive.
I'm not sure the first three measurements are necessarily representative of an untrained person. I spent a couple of weeks 'playing' anonymously before I registered. And judging from other high scores by anons, this may be widespread. Also, I would venture to say that the average user of this website is a bit off the average.
How about adding a second red line in the performance chart, showing the current average of the (trained) users?
As a second suggestion, it may be a good idea to take the median rather than the average in each individual score; I often get a run of good timings, only to be distracted and get one outlier which destroys the arithmetic mean average.
1stone | 4 years ago
Reply
"As a second suggestion, it may be a good idea to take the median rather than the average in each individual score; I often get a run of good timings, only to be distracted and get one outlier which destroys the arithmetic mean average."
i.e. an issue of outliers. The stats are swayed by outliers now since they are just simple means. This is a valid and good suggestion, and I'll keep it in mind. It will require a site-wide recalculation though (but I suspect most people will end up with better scores, due to things like that you mentioned).
cognitivefun | 4 years ago
Reply
"How about adding a second red line in the performance chart, showing the current average of the (trained) users?"
I have thought about this before, but think it might be harder. Some users have gotten very good at some tasks and can really pull down the average response time. While this may be interesting as a comparison, at some point the numbers do reflect individual, physiological limits, that may be less useful, or demotivating.
A theoretical example is reaction time. Assuming equal processing speeds in the brain, a motor task executes slower in a tall person than a short person, simply because signals need to travel further. While the difference is pretty trivial in this example, it shows that interpersonal comparisons may not be the best impetus for self improvement.
I'm leaning towards letting people publicize their stats if they want; then you can track other users on your own, and would be able to find who is an "expert" at some task, and follow them. Otherwise, it's usually not fun to find yourself starting to run, but realize that everyone else is flying!
A theoretical example is reaction time. Assuming equal processing speeds in the brain, a motor task executes slower in a tall person than a short person, simply because signals need to travel further. While the difference is pretty trivial in this example, it shows that interpersonal comparisons may not be the best impetus for self improvement.
I'm leaning towards letting people publicize their stats if they want; then you can track other users on your own, and would be able to find who is an "expert" at some task, and follow them. Otherwise, it's usually not fun to find yourself starting to run, but realize that everyone else is flying!
cognitivefun | 4 years ago
Reply
