Tilted Forum Project Discussion Community  

Go Back   Tilted Forum Project Discussion Community > The Academy > Tilted Knowledge and How-To


 
 
LinkBack Thread Tools
Old 12-15-2004, 08:44 PM   #1 (permalink)
Upright
 
Location: Houston
Statistics question

Quick Question.
True of False: An r of -1.0 proves a strong cause and effect relationship between x and y.
Thanks.
BlitzkriegKommt is offline  
Old 12-15-2004, 09:06 PM   #2 (permalink)
Psycho
 
Location: Atlanta, GA
I think it's false, simply because it's a correlation thing not a cause and effect thing. Just because things correlate does not mean one causes the other.
__________________
"Great spirits have always encountered violent opposition from mediocre minds" -- Albert Einstein

"A clear indication of women's superiority over man is their refusal to play air guitar." --Frank Zappa
shred_head is offline  
Old 12-15-2004, 09:17 PM   #3 (permalink)
Drifting
 
amonkie's Avatar
 
Administrator
Location: Windy City
Shred's got it - r values only show the the correlations or relationships between two variables, and never, at ANY time, do they say that X actually causes Y.
__________________
Calling from deep in the heart, from where the eyes can't see and the ears can't hear, from where the mountain trails end and only love can go... ~~~ Three Rivers Hare Krishna
amonkie is offline  
Old 12-15-2004, 09:21 PM   #4 (permalink)
Upright
 
Location: Houston
Oh I see, there could be a lurking variable or something to that effect, thanks for clearing that up.
BlitzkriegKommt is offline  
Old 12-15-2004, 09:22 PM   #5 (permalink)
Crazy
 
Location: U.S.A
Definately false,
Correlations are measures of association and do not indicate cause and effect. I think this sort of question is on most early stat quizes/tests.
A correct answer would be that it indicates a perfect inverse correlation between x and y.
lpj8 is offline  
Old 12-18-2004, 02:56 AM   #6 (permalink)
Insane
 
ok, since we have that question answered...

then how DO you prove a causational effect between two variables?
Amano is offline  
Old 12-21-2004, 04:32 PM   #7 (permalink)
Tilted
 
Location: Salt Lake City
You cannot say whether it is true or false.

If r-squared equals 1 than obviously you have perfect correlation. -1 or 1 could both give an r-squared value of 1.

Not enough information.
belkins is offline  
Old 12-22-2004, 07:47 PM   #8 (permalink)
Psycho
 
1010011010's Avatar
 
Location: Virginia Beach, VA
False. "Correlation is not causation."

High correlation (even perfect correlation) proves nothing about causation. Both X and Y could be caused by an uncontrolled variable, yet have no direct connection between them.

Even so, I think Bayesian statistics would give that statement a "mostly true".
__________________
Simple Machines in Higher Dimensions
1010011010 is offline  
Old 12-23-2004, 11:20 PM   #9 (permalink)
Psycho
 
Location: Ouuuterrrr Spaaaaacccceeee
Quote:
Originally Posted by Amano
ok, since we have that question answered...

then how DO you prove a causational effect between two variables?
You can't. But you can make a very very educated guess by doing an experiment with a control group. In other words, you try to get two nearly identical groups and administer the independent variable (which will mean the IV will be the only difference between the two groups) and see if the variable makes an effect. Quasi Experiments will work too, but we won't go into that...
RoboBlaster is offline  
Old 12-23-2004, 11:45 PM   #10 (permalink)
Upright
 
Location: Chapel Hill, NC
Quote:
then how DO you prove a causational effect between two variables?
My guess us that in the kind of research you would be doing where you are measuring simple correlations you can't prove a causal relationship.

But, if you've got a theory: F=ma.

Take a mass, apply a bunch of different F's to it. Measure the a's. Plot the results. Find that yes, indeed, 'F' is proportional to 'a' by a constant 'm' which turns out to be a property of the mass which we can also measure with a scale.

The data supports the theory, and insofar as the theory is correct, F caused a.

Strictly speaking, you can never prove a theory to be true beyond a shadow of a doubt, but you can demonstrate its validity to such a high degree that it would be silly to doubt the accuracy of the theory.
Drayab is offline  
Old 12-26-2004, 12:36 AM   #11 (permalink)
Crazy
 
Location: U.S.A
Statistical tests are not set up to prove that anything is true, rather they are set up to test the null hypothesis that some phenomena or event happened by chance, or that there are no differences in groups of subjects/events. They intend to disprove the null hypothesis.
A statistically significant result indicates evidence that the phenomena could not have happened by chance thus, we fail to reject the null hypothesis.
A statistically significant pearson product moment correlation is merely saying that it is highly unlikely that the obtained correlation happened by chance. However, as many people have commented, correlational results do not imply causation because of the potential for spurious variables.
For example, if one were to look at the correlation for per pupil expenditure money for each state in the US, and the academic achievement (SAT scores) of the students within each state, there is a inverse correlation between the amount per pupil expenditure and academic achievement (SAT scores) (More money = Less achievement). However, this correlation fails to take into account the percentage of students taking the SAT for each state. The states that spend the least per pupil have the fewest percentages of their students taking the exam and the states that spend the most per pupil have the highest percentages of students taking the SAT.
Maybe in some states they only want the smart kids to take the SAT, maybe they don't have the expectation for all kids to go to college, so they don't push the SAT on students with lower academic achievement. All of which are valid questions.
Therefore, after you statistically control for the percentage of students taking the SAT in each state, there is a positive correlation (as expected) between per pupil expenditure and SAT achievement.
lpj8 is offline  
Old 01-23-2005, 02:22 PM   #12 (permalink)
Born Against
 
raveneye's Avatar
 
There's a very strong correlation between being born and dying. Therefore, birth causes death.
raveneye is offline  
Old 01-23-2005, 04:16 PM   #13 (permalink)
Cracking the Whip
 
Lebell's Avatar
 
Location: Sexymama's arms...
ah, r values.

That brings back some memories.

Little did I know when I took sadistics that I would actually use them.
__________________
"Of all tyrannies, a tyranny exercised for the good of its victims may be the most oppressive. It may be better to live under robber barons than under omnipotent moral busybodies. The robber baron's cruelty may sometimes sleep, his cupidity may at some point be satiated; but those who torment us for our own good will torment us without end, for they do so with the approval of their own conscience." – C. S. Lewis

The ONLY sponsors we have are YOU!

Please Donate!
Lebell is offline  
Old 01-23-2005, 06:15 PM   #14 (permalink)
Born Against
 
raveneye's Avatar
 
I teach university stats. It's possible to teach it in a stimulating, fun, interesting way. But "sadistics" is a good descriptor for what uninspired teaching does to the subject.
raveneye is offline  
Old 01-26-2005, 08:09 AM   #15 (permalink)
Crazy
 
Location: U.S.A
Quote:
Originally Posted by raveneye
I teach university stats. It's possible to teach it in a stimulating, fun, interesting way. But "sadistics" is a good descriptor for what uninspired teaching does to the subject.
I also teach a university stat course as a GA. We have a good time with it, and I think the students enjoy it. I try to keep my classes relatively small, as it makes things easier for me and the students.
lpj8 is offline  
 

Tags
question, statistics

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are On
Pingbacks are On
Refbacks are On



All times are GMT -8. The time now is 11:03 PM.

Tilted Forum Project

Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Search Engine Optimization by vBSEO 3.6.0 PL2
© 2002-2012 Tilted Forum Project

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360