posted
So, I warned you all that I was a datahead.
I did a statistical breakdown of the entrants in thesefourbeliefthreads. I only took those entrants who had specified percentages for all 27 religions. I'd like to add in all the others, and maybe I will after I read up a little bit more on optimal data reconstruction.
Anyway, here are some of the features. I'm pretty close to plotting a two-dimensional graph projection that would capture much of the information distribution, but I'm not sure when I'll get around to it (and, honestly, I want to see if anyone's interested in these results before I go to the added effort. Or maybe I'll just do it anyway. This stuff is fun for me).
Statistical Results ------------------------ Most Similar Beliefs (not counting double entrants): UofULawGuy, Occasional
posted
Here is the 2-D projection, even though nobody is showing any interest so far.
<edit>BTW, I should mention that Jay is in the extreme bottom right corner. He's easy to miss, since he's in the very corner, and some browsers might not display the full image</edit>
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
posted
That's fascinating. Seeing the projection in particular was really interesting. I've done these quizzes twice before -- if I take it again tonight and supply my results, can you add me to the 2D projection? Posts: 10886 | Registered: Feb 2000
| IP: Logged |
posted
Katie (<edit> and twinky and anyone else </edit>-
Thanks for being interested. I feel validated. If you want to post your full 27 (here or wherever), I'd be more than happy to include you.
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
posted
That is really fascinating. It's interesting to see how everyone stands compared to each other. I like the 2D representation a lot, because it shows that I'm actually close to some people. Based on the list of data, I assumed that I held different beliefs from almost everyone at Hatrack, and that would be a little daunting.
How are the X and Y axis labled? I'm not a big data person, but what does it mean to be higher or lower compared to farther left or right?
Posts: 2596 | Registered: Jan 2006
| IP: Logged |
posted
I know I've taken this quiz at least twice; have I failed to post the results? ... Maybe I just posted the top ones.
Posts: 1522 | Registered: Nov 2005
| IP: Logged |
posted
Yeah...the axes. I had hoped they'd have a clear definition, but they don't really.
The best I can say about the X axis is that it roughly corresponds to general belief versus specific belief. People who had a pretty uniform spread across all 27 belief systems move to the left, while people with more belief variation (very high in some, very low in others) move to the right.
People who score high on the Y axis generally have high scores for 'Secular Humanism' and 'Unitarian Universalism', while people who score low have high scores for 'Roman Catholic', 'Eastern Orthodox', 'Mainline to Conservative Christian/Protestant', and 'Seventh Day Adventist.' It's sort of a traditional 'Liberal' vs. 'Conservative' divide, although I might interpret it more as preference for 'unstructured' vs. 'structured' belief systems.
2-D Projection Methodology: I performed an SVD decomposition on the data vectors, then kept the two strongest singular values and zeroed out the rest. The vectors corresponding to the two strongest singular values became the two principle axes of the graph. I estimate that the projection maintains about 65% of the relational information from the original 27-Dimensional space.
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
I'll add you to the graph tomorrow; it takes a little while to update it and if more people post their results I'll be able to do it all at once. Here are your stats, though:
I wonder what would happen if I removed Jay and Dag from the data set. There's a large group of users (31/61 or 52%) for whom Jay, Dagonee(05) and Dagonee are the furthest neighbors (although sometimes they reverse Dagonee(05) and Dagonee). If I removed Dag and Jay from the data set, would that large group end up with a more diverse set of furthest neighbors, or would they simply move to a new invariant group? Hmm...
<edit>Well, I tried it, and I got a significantly more diverse set of furthest neighbors. I'm not sure what that means, though</edit>
Why wouldn't Aria be a closer neighbor than TMedina? Of course, I can't see my own dot yet, but I can't see having TStorm and Alcon as closest neighbors and simultaneously being closer to TMedina than Aria on the 2D chart.
Posts: 6394 | Registered: Dec 1999
| IP: Logged |
The problem is that the 2D visulization necessarily compresses things from the 27D original (where I'm taking the distance measurements). The dimensions where Aria is far from you contain less general relationship information than others (if that makes sense). Since I could only keep the two most information-rich dimensions, 25 (slightly to significantly) less information-rich dimensions had to be left out.
Which is all to say, the 2D plot is an approximation of the actual data, since we can't visualize in 27 dimensions.
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
posted
I don't know how hard it is, but would it be possible to make charts, probably using different dimensions, to show the differences amongst the clusters on the current graph?
For example, if you drew a line with a slope of 1 from the origin (assuming we're in the (+,+) quadrant of the Cartesian plane), you would divide the group roughly in half. Then a chart of each one would let us see how different those who are similar on the two main dimensions are from each other.
Posts: 26071 | Registered: Oct 2003
| IP: Logged |
After responding to Karl's question, I got to wondering what the 3rd dimension would be if I had included it. I can't make a good 3-dimensional image to show you all, but the results were really interesting.
The next most significant dimension separated people who ranked highly on "Jainism" and "Hinduism" on one side, and people who ranked highly on "Secular Humanism" and "Nontheist" on the other. In this dimension, Jay and Dagonee move in with Alcon, Vera and others, while from Cythera, vonk and Luet13 move out. It's pretty cool to see, as I rotate around in the 3D; I'll see if I can't find a good angle to demonstrate in "2 1/2-D."
Anyway, to further answer your question Karl, in that dimension (call it the number of gods dimension) you correlate very strongly with TMedina, and much less strongly with Aria.
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
posted
That makes sense Senoj, I think. Aria might be, for instance, twice as far from me as TMedina, but to show that you'd need at least a 3rd axis. She only looks closer on the 2d projection. Right?
Posts: 6394 | Registered: Dec 1999
| IP: Logged |
Thanks for doing this. I'm intrigued by your results, and I'm impressed by your dedication in pursuing the quiz out to a logical conclusion.
Posts: 1813 | Registered: Apr 2001
| IP: Logged |
Thanks for doing this. I'm intrigued by your results, and I'm impressed by your dedication in pursuing the quiz out to a logical conclusion.
+1...you just made that quiz a whole lot more fun. Throwing out the outliers (no offense Dag and Jay) might yield some interesting comparisons as well.
Posts: 1412 | Registered: Oct 2005
| IP: Logged |
I'm guessing you're laughing about me including Jay in Karl's list. I guess it looks a little funny, the way I put it, but my intent was just to give a relative feel for the total distances we're talking about.
As a note, I hope no one feels this is a divisive activity. The point isn't to divide Hatrack, but just to help see some of the interesting results from these quizes we seem so fond of taking.
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
I didn't figure you were trying to be mean. I just wanted to make sure that everyone realized that this is just to help us get to know each other, not so we can laugh at people who are different (not that I thought you were doing that, or that anyone here would do that, for that matter). That's one reason I've been using the term "neighbors" for everyone, regardless of how near or far they may be from each other.
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
posted
. Mainline to Conservative Christian/Protestant (100%) 2. Church of Jesus Christ of Latter-Day Saints (Mormons) (99%) 3. Jehovah's Witness (89%) 4. Mainline to Liberal Christian Protestants (82%) 5. Eastern Orthodox (80%) 6. Roman Catholic (80%) 7. Orthodox Judaism (78%) 8. Bahá'í Faith (74%) 9. Seventh Day Adventist (74%) 10. Orthodox Quaker (71%) 11. Islam (68%) 12. Sikhism (60%) 13. Christian Science (Church of Christ, Scientist) (54%) 14. Reform Judaism (53%) 15. Hinduism (47%) 16. Liberal Quakers (46%) 17. Jainism (39%) 18. Unitarian Universalism (37%) 19. Mahayana Buddhism (31%) 20. New Thought (30%) 21. Theravada Buddhism (30%) 22. Scientology (29%) 23. Neo-Pagan (27%) 24. New Age (22%) 25. Nontheist (16%) 26. Secular Humanism (14%) 27. Taoism (10%)
Is it to late to get in on the compairison? Oh, by the way, remind U of U Law Guy he is supposed to be studing, not crusing the net.
Posts: 1167 | Registered: Oct 2005
| IP: Logged |
Except, do you mean that you just take the two dimensions that are the most meaningful and just completely ignore the rest? Which two dimensions are those then? I would hope there would be a way to compress the rest of the dimensions into the two that are present. Are the "distance" statistics you give based on all 20-some dimensions or just the two?
I took this quiz a few days ago and got different results now from when I took it then. I think it's more because a few of the questions are worded in such a way that my answer does not correspond directly to any of the choices, but comes somewhere in between some of them. For instance, in the what happens to people after death, both of the first two choices have elements of what I believe: I believe people are immediately judged, as in the first answer, but then after the second coming the elect will be given a paradisical Earth to live in, like the second answer states. So I might have chosen two different answers in the course of the few days between when I took the test not because my belief changed but because I probably randomly picked one over the other because of my mood.
Anyway, here are my results:
quote: 1. Seventh Day Adventist (100%) 2. Orthodox Quaker (96%) 3. Mainline to Conservative Christian/Protestant (93%) 4. Eastern Orthodox (86%) 5. Roman Catholic (86%) 6. Mainline to Liberal Christian Protestants (67%) 7. Orthodox Judaism (56%) 8. Islam (54%) 9. Liberal Quakers (53%) 10. Hinduism (50%) 11. Bahá'í Faith (47%) 12. Unitarian Universalism (43%) 13. Jainism (41%) 14. Jehovah's Witness (40%) 15. Church of Jesus Christ of Latter-Day Saints (Mormons) (38%) 16. Sikhism (31%) 17. Reform Judaism (30%) 18. Mahayana Buddhism (30%) 19. Theravada Buddhism (30%) 20. Nontheist (24%) 21. Neo-Pagan (21%) 22. New Age (21%) 23. Secular Humanism (21%) 24. Taoism (20%) 25. Christian Science (Church of Christ, Scientist) (16%) 26. Scientology (11%) 27. New Thought (10%)
posted
Hmm, I retook the test again and I got Seventh Day Adventist at 100% again, but this time with Conservative Protestent, Catholic, and Orthodox all higher, and my bottom 10 shifted around like crazy.
I'm rather puzzled at how that works. For one thing I'm not 7th Day Adventist, I'm part of a very liberal Protestant denomination (Episcopal church) although my own views are more conservative. I am rather glad that Scientology showed up so low, although I'm also puzzled at how Mormomism showed up pretty low as well despite the fact that it seemed to show up pretty close to the mainline Christian denominations for other people (e.g. if they were Mormon, the other Christian denominations would be high, if they were non-Mormon but Christian, Mormonism would show up high). Not that I'm surprised that it ranks low for me, I'm just surprised at the disparity between my results and those of others.
Posts: 142 | Registered: Apr 2005
| IP: Logged |
posted
In my own case at least, I don't think the questions are worded in such a way as to accurately peg my beliefs. Contrary to popular belief, I think there is room for some interesting variety of "beliefs" among those who would ultimately be forced to choose "There is no God; Don't know; Not Important" or whatever the "default non-theist" answer was for each question.
I wonder what I answered differently from Twinky, because his results seem more accurately to reflect what I actually believe than my own do. I think I might have misread an option or something.
Posts: 6394 | Registered: Dec 1999
| IP: Logged |
posted
I've remade the original projection to include new posters. Here it is. I'm working on some other interesting plots, too.
Posts: 2926 | Registered: Sep 2005
| IP: Logged |
posted
Poor Twinky is all alone out there in the ether.
At least Dagonee and Jay have themselves to keep them company.
I also agree with KarlEd, in that I found myself dissapointed with some of my choices where I was continually being lumped in with a broad non-theist group. Because the choice ended up being the "best fit" even if not completely accurate.
Posts: 8741 | Registered: Apr 2001
| IP: Logged |
posted
I think "There is no God" is a hugely different assertion than "I don't know" and "It doesn't matter", so they shouldn't be lumped together like that.
Personally, I don't "know". However, I'd go so far as to state my belief as "There is no God" for certain definitions of "God".
For instance, for some people God=Creator. What if it turns out that all life on Earth really was created by an intelligent designer (ID). Is this ID "God"? What if it turns out that this ID was a very advanced, atheistic alien race? Are they now collectively "God", or were the people who believed "God" created the Earth mistaken?
At any rate, I do know I'm a lot more nontheist than I am "Liberal Quaker", whatever that is, despite the results of the survey. (And "Mainline to Liberal Christian Protestant" is so far from what I am as to be floor-rollingly hillarious.)
Posts: 6394 | Registered: Dec 1999
| IP: Logged |
posted
Okay, I took it again to get on the graph and just see how close I am to KetchupQueen and the rest. (Although I would agree, Karl, that "I don't know" is VASTLY different from "It doesn't matter." I think it goes to your level of religiousity - how religious you are determines whether you think your doubts are important or not.)
My results:
1. Church of Jesus Christ of Latter-Day Saints (Mormons) (100%) 2. Mainline to Conservative Christian/Protestant (97%) 3. Jehovah's Witness (87%) 4. Mainline to Liberal Christian Protestants (84%) 5. Eastern Orthodox (72%) 6. Roman Catholic (72%) 7. Baha'i Faith (70%) 8. Orthodox Quaker (68%) 9. Orthodox Judaism (65%) 10. Seventh Day Adventist (64%) 11. Islam (55%) 12. Liberal Quakers (51%) 13. Christian Science (Church of Christ, Scientist) (48%) 14. Sikhism (48%) 15. Hinduism (41%) 16. Unitarian Universalism (40%) 17. Reform Judaism (39%) 18. Jainism (36%) 19. Mahayana Buddhism (29%) 20. Theravada Buddhism (28%) 21. Neo-Pagan (28%) 22. New Thought (25%) 23. Scientology (19%) 24. New Age (18%) 25. Secular Humanism (15%) 26. Nontheist (13%) 27. Taoism (11%)
Posts: 1522 | Registered: Nov 2005
| IP: Logged |
Printer-friendly view of this topic
![[Frown]](frown.gif)
![[Smile]](smile.gif)
![[ROFL]](graemlins/rofl.gif)
![[Wink]](wink.gif)
(And "Mainline to Liberal Christian Protestant" is so far from what I am as to be floor-rollingly hillarious.)
