test
Grrr, something fucked up, at least from my POV. Anyway, here’s my question:
Lately I’ve been looking at http://www.colorgenics.com. It’s a very interesting site that determines your mood by giving you 8 cubes of different colors (red, blue, green, grey, black, purple, yellow, and orange), and gives you a profile of your mood based on what order you pick the cubes in. I figured out that there are thousands of different ways you could choose the cubes (I did 8 x 7 x 6 x5… Someone tell me if this is wrong, I suspect it might be, and I haven’t done anything close to math in months), and I tried testing it by choosing random cubes and seeing if it applied to me, to see if the profile generator just tossed the same crap at you no matter what you picked.
It didn’t. So far, the site has been fully and completely accurate in determining what I’m feeling.
So what is the site, and is it BS or not?
I’m not sure. It looks interesting, and, as it happens, I’m in a position to give it a test. Tonight I’m tired and a bit down because a fellow I go to church with died on Sunday. I took the test and it seemed to be a fairly accurate description. Since the grief should pass in a day or two, I’ll retake the test when I know my mood has changed and compare results.
Let you know,
CJ
Did it say your mood was “sketpical”?
I don’t know enough psychology to debunk this, but it it has the dull wet splat of Bullshit to my ears.
I tried it. It’s a load of bunk. All it did was tell me I’m lovable and looking for my perfect mate. I do NOT want a mate.
C
Well, you seem pretty lovable, mate.
JB
The general advice is on the level of astrology – pretty vague and easily applicable to your situation if you want it to. If someone really believes in it, they’re more likely to forget whatever it got wrong and only remember what was right.
And I find it difficult to believe that someone has written 40,000 separate entries for the colors. My guess is that it must give the colors a value, then do some sort of averaging so that only a few 100 answers need be generated. Or possibly a few paragraphs get shuffled around like so many tarot cards.
I got this in my response :
So, you mean I don’t trust … this colorgenics thing? Hey, I think they got something right! (Although I do have to give them credit for telling me I’m a borderline paranoid delusional. Not everybody can handle that sort of praise. )
I’ve taken the Colorgenics test a few times, and was simply amazed that it did accurately describe my inner feelings.
Just to be sure, I took the test 3 days in a row to see if there was any “trick” to it. Sure enough, it accurately described my mood for each of those 3 days.
I have no idea if there’s a scientific basis or not, but there IS something there…
Well, I won’t say that I’m convinced, but mine was pretty accurate, more so than the typical horoscope.
Well, I tried to try it, but I didn’t know what it meant to be “in harmony with a color”. Maybe I can first find a website where I can learn to speak “new age”…
I tried it, and it was pretty accurate. I was skeptical too, and tried different combinations, none of which seemed to suit me.
When my twelve-year-old daughter tried it, it was almost frighteningly accurate, and somewhat depressing…and not at all vague.
I was so amazed with it I sent it to all my friends, who all reported back with wide-eyed amazement as well!
i say it’s a load of bull.
first off, none of the colours were shades i found attractive, they were rather hideous. secondly, i chose black first, since it was the least ugly (and i do like black, but if they had had a rich plum, or rosey pink, it probably would’ve been picked first), and the first thing it says is “you are deeply depressed” blah blah blah. if i chose black last and kept the other colours in the same orders, that depression thing magically went away. :rolleyes:. i’m not depressed because i picked black first! and lastly, besides that inaccuracy, the description i got was obviously made with contrasting statements, so i could say “oh, i can kind of see that one” right after, “that’s a load of horse hockey”.
It’s crap. Pure, unadulterated, crap.
Choose cubes by some method, such as flipping coins or rolling dice. Better yet, have someone else, or even a computer program generate the decisions. Get your report.
Then, do it again. And again. Each time, note how the report finally generated is pretty good at describing some issues you are dealing with nowadays. Because they’re issues that almost everyone deals with all the time. Basically, the designer is a pretty good at inventing statements that most people would say, “Yeah, that pretty much applies to me.”
Otherwise, you will have to explain how Larry Wall designed the programing language Perl such that it correctly generates a list of cube selections that correctly describes my state of mind:
#!/usr/bin/perl
# removes a randomly chosen element from an array, and returns that element.
sub prand(\@)
{
splice @{$_[0]}, int(rand($#{$_[0]} + 1)), 1, ();
}
my @colors = qw/red green blue yellow orange purple grey black/;
while ($#colors >= 0)
{
print prand(@colors), "
";
}
I have heard of this before. In Soviet Russia they used this test on people being put through various stress tests for the Russian space program. They would do things like have a cosmonaut stay in the same constantly lighted room with no human contact for months at a time, or force him to go without sleep for extended periods, then periodically have them organize the colored squares to try and figure out how the person was feeling.
I believe testing was halted when 90% of the results came back “tired and cranky”. 
Well, it was more accurate than horoscopes usually are for me. The last time a horoscope was even remotely accurate was when I went outside in my summer pajamas to get the newspaper and got looked outside of the house. While I waited for someone to wake up and let me back in I read the paper. My horoscope for that day was “Something you want is just beyond a locked door.”
Bumpity bump bump
Okay, so for a way to test how accurate this thing is, I have an idea for a double-blind test. To perform it, it requires a friend and some pieces of paper. First, go to the site and perform the selections, but leave the computer before the mood explanation pops up.
At this point, your friend should get on the computer and, either by using the program I posted above, or by some other random method, choose cubes repeatedly to get another (different!)explanation to pop out. Your friend should print out each explanation, and write down on a separate slip of paper which of the explanations was the one that you chose. They should also randomly choose the order (say, by flipping a coin) of the explanations, and then hand them to you, and then leave you alone to rank them from 100% accuracy to 0% accuracy. Mark down your decision on each one, put them to the side for later referencing.
You should do this across a broad swath of time, maybe once a day or once every other day, until you have a good number of samples. You probably want to have at least ten separate tests done. More is better; say, twenty. You should not check your results against your friend’s records until you are done doing all the tests you plan on doing.
Once you are done, it’s time to do a statistical analysis on your data. Separate your accuracy ratings into two groups: the explanations that you had chosen, and the explanations that had been generated randomly. The mean of each group is easy to calculate, just add all the ratings together of that group, and divide by the number in that group. Calculating the standard deviation is more difficult. In each group, calculate the average of the square of each rating, and subtract from that, the square of the mean, that you calculated above. Then, take the square root of that value.
So, for instance, if a set had the following accuracy ratings: 90, 75, 85, 80, 80, 75, 95, 100, 100, 85, you would get a mean of 865/10 = 86.5, and a standard deviation of sqrt(7562.5 - 7482.25) = 8.96.
Now, for what those values mean: for Colorgenics to have indications of validity, you would need:
[ol]
[li]The mean rating of the explanations you chose should be high.[/li][li]The standard deviation of those explanations should be small.[/li][li]The mean rating of the explanations your friend chose should be average, or slightly below average.[/li][li]The standard deviation of those explanations should be large.[/li][/ol]
Specifically, for Colorgenics to have indications of validity, you would need to see that the mean rating of your friend’s choices plus its standard deviation, should be less than the mean rating of your choses, minus its standard deviation. For stronger proof, the difference between the two mean ratings should be more than twice, or better yet, three times the sum of the two standard deviations.
On the other hand, if the results are that the mean of the friend’s choices’ mean plus its standard deviation are greater than the mean of your choices minus its standard deviation, then there is enough overlap in the two samples that lends doubt to the belief that Colorgenics is a valid way of measuring mood. Worse, if the friend’s chocies’ mean plus your choices’ standard deviation is greater than your choices’ mean, then your experiments have found no statistical significance for the Colorgenics results.
There are some dangers of doing this experiment that could skew the results. If you are, intentionally or unintentionally, picking the same order of colors or close to it, you might notice some strong repetitions in one of the explanations you see. If you key into this, and give stronger weight to these explanations on later tests, even unintentionally, it would invalidate most of the data. A way to partially prevent this from happening would be for your friend to add more random explanations per trial, making it less easy for you to key in on those similarities.
I think its just a rehash of The Luscher Color Test. A well known test along similar lines, that has been out for years. It’s a book.& it profiles yourself by your Color choices & the order you do them in.
It’s BS, Du Hast, but then, you didn’t need me to tell you that. 
I put in the colors I truly preferred, and got something that didn’t describe me at all.
Um, nope, there’s nothing like that going on in my life at all.
Then I put in the colors in exact reverse order, and also got something that didn’t describe me at all.
Um, no, actually, I’m not. I have an extremely easy life as a stay-home mom, and as for being physically active–well, I do have a couple of pink fuzzy weights that I lift thoughtfully, while I’m sitting in bed reading.
It’s all randomly generated paragraphs, the same as with daily horoscopes.
Bushwah.
I did this test a few days ago and posted to my journal. The results were pretty close. There was only one thing I really disagreed with.
I still prefer the Myers-Briggs for accuracy.