A number of you might remember me harassing you to do the online ‘Complexity Test’ some time back last year, and promised to get back on how deviant you are compared to the general population.
For those who aren’t too familiar with it, I’m interested in how people perceive complexity. What is complexity? How do we define it? How do we quantify it? For example, most people might see a star more complex than say, a circle. So I got a whole bunch of people to arrange and rank four sets of shapes; shapes like these:
And when I have enough data, I’ll try to compare our human algorithmic prowess with more standardized numerical methods. It’ll also be interesting to see how people perform relative to the general population mean.
So as promised, here are some preliminary results. The graph below shows how males and females (from a popualtion size of 90 people) perform in the test relative to the overall population mean, with the leftmost columns being the ‘best’ predictors and the rightmost the ‘worst’ predictors. See below for a more detailed explanation for the X-axis legend.
Overall, girls seem to be better predictors of complexity but have a wider spread; the best and worst predictors are both ladies, and the worst ones differed by a substantial lot.
Guys on the other hand seem to have a tighter curve, but neither shone nor failed as much as the ladies.
Legend:
- Regular Joe/Jane – You got around 50-75% of the shapes right. Yes, you’re pretty much a run-of-the-mill human, with human error rates.
- Deviant – If this was a graded test, you failed. You got only 25-49% of the shapes right. You choose a square when everyone else chooses a circle. A chimp might have better success than you at those IQ tests with shapes.
- Bloody Aberration – You got pretty much everything wrong. You should be locked up; who knows what goes on in your head. As Goya would say, your sense of reasoning is asleep, and it breeds monsters.
Those who have already done the test left their email with me can look forward an even more detailed and complex analysis of their behavior, future and personality mailed to them to see how deviant they are. It’ll be as accurate as tasseography. All I can say now is that Tommy is so far the best male predictor!
As said, these current results are preliminary, and yes, I do need more test subjects to improve my statistical power, so please, if and when you have 15 minutes or so, click the link below to access the page and do the ranking.
COMPLEXITY TEST LINK
Deadline for result compilation is a 2 weeks from now (end March) so do it fast!
A few pointers for the test:
Whats important is that each shape be given a rank. There are ten ranks, with 1 being the most simple, 10 being the most complex. Important: both values 1 and 10 MUST be used. This is to allow for standardization of the data. You can however skip rank numbers inbetween, i.e. a series of shapes can be ranked 1,2,3,3,6,6,7,9,10,10, where i have skipped values 4 and 5 because the jump in complexity between the fourth and fifth shape is just too high to have only one or two jumps in rank.
Please make sure you complete each set properly; sets with unranked shapes will have to be discarded. you can however complete sets separately and submit them individually (as it can get a little tiring after a while), but please remember to leave your particulars in the field provided so that i can track and concatenate the data.
So what is complexity? Its up to your own innate senses to decide. You shouldnt think too much about whether this shape is more complex than the other – just go with the gut flow.
So thats it, and thanks so much for your valued help! Do spread this link to people you know, but please make sure that they are reliable as the test can get a little tiring, and it should be taken seriously, since this is going to be part of published work.
As a reward, I’ll send a report to everyone who has completed the test to show you how ‘deviant’ you are from the rest of humanfolk in terms of perceiving complexity!
YC
Read Full Post »
