Re: can multicollinearity force a correlation?
Wuzzy wrote: http://www.accessv.com/~joemende/insulin2.gif Appologies, i also forgot to divide the KCAL in food by the 31 as this represents kcal. It seems to me logical to advise decreasing food intake and increasing physical activity to improve insulin sensitivity. I would probably avoid reporting the R^2, or try a different model (non-linear) You may also look at http://noppa5.pc.helsinki.fi/uudet/corr/avp/avp20.html -- Juha Puranen Department of Statistics P.O.Box 54 (Unioninkatu 37), 00014 University of Helsinki, Finland http://noppa5.pc.helsinki.fi = Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at http://jse.stat.ncsu.edu/ =
Re: Who said Correlation does not imply causation.
Stu wrote: Silvert, Henry wrote: Might I go one step further and point out the correlation does not establish a causal relationship primarily because it does not point to directionality, at least not without a working hypothesis and some background support. Absolutely. Without both a working hypothesis and a literature search one could say that cancer causes cigarette smoking. Stu Garfield High School Los Angeles If you have two variables (numerical) you can allways calculate the correlation coefficient. Usually I show this to my students http://noppa5.pc.helsinki.fi/koe/corr/cor7a/cor7a.html and ask if it is sensible to calculate or use correlation in this case. -- Juha Puranen Department of Statistics P.O.Box 54 (Unioninkatu 37), 00014 University of Helsinki, Finland http://noppa5.pc.helsinki.fi = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
making measurements
Hi I have put my making measurement-page on http://noppa5.pc.helsinki.fi/koe/dens/densitometer.html On this page you can - measure gray-values using densitometer, - calculate reliability - plot scatterplot (true- vs. measured values) - change accuracy - visualize, what happens, if different measurers have different scales. Works with ns4.7, ie5 opera5 Please comments. I am also looking someone, who is expert in making measurements and is ready to write some text on this problem. regards Juha -- Juha Puranen Department of Statistics P.O.Box 54 (Unioninkatu 37), 00014 University of Helsinki, Finland http://noppa5.pc.helsinki.fi = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: 2x2 tables in epi. Why Fisher test?
Ronald Bloom wrote: In sci.stat.consult Elliot Cramer [EMAIL PROTECTED] wrote: In sci.stat.consult Ronald Bloom [EMAIL PROTECTED] wrote: Herman as usual is absolutely correct; the validity of the Fisher test is analagous to the validity of regression tests which are derived conditional on x but, since the distribution does not involve x, are valid unconditionally even if the x's are random. If I take your analogy in the direction that leads back to the Fisher test, I should be able to paraphrase the above as the validity of the [Fisher test] which [is] derived conditional on [the fixed marginals] but, since the distribution does not involve [the fixed marginals], [is] valid unconditionally even if the [marginals] are random. Please clarify what is meant by the distribution does not involve [the fixed marginals]. I am not clear on this: the Fisher test statistic (hypergeometric upper tail probability) certainly *does* depend on the fixed marginals in this case -- they appear in every term in that tail sum. Usual the assumptions for Fishers exact test are not true. What you can fix are the row margins, or column margins or grand total or Element of row i and column j. In these cases the exact Fisher test is biased. At least in Survo (may be in some other programs too) it is possible make the test also in these cases. Look at http://www.helsinki.fi/survo/q/qu1_03.html regards Juha -- Juha Puranen Department of Statistics P.O.Box 54 (Unioninkatu 37), 00014 University of Helsinki, Finland http://noppa5.pc.helsinki.fi = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: Histogram for discrete probability distribution
"Jose Ramon G. Albert" wrote: Try having the points enumerated be the centers of your rectangles with each rectangle having an AREA of 1/6. Thus the first rectangle should have its corners at 1.25 and 1.75 (and have 1.5 as its midpoint). Now since the width of your rectange is 0.5, let the length of your rectangle be 1/3. This will have your rectangle have an area of 1/6. And do the same for the rest. Cheers. Jose Ramon Albert Statistical Research and Training Center J S Bldg., 104 Kalayaan Ave. Diliman, Quezon City PHILIPPINES www.srtc.gov.ph I use classification 1(.2)5 or 1.05(.1)5.05 so that everyone can see: "It is discrete " Juha -- Juha Puranen Department of Statistics P.O.Box 54 (Unioninkatu 37), 00014 University of Helsinki, Finland http://noppa5.pc.helsinki.fi = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: hyp test:better def
Milo Schield wrote: I agree with Dennis that students need to be exposed to the use of Bayesian priors within the process of teaching classical hypothesis testing. Using Bayesian priors can be very difficult for some students. (Why do we take the uniform prior ??? ) For to teach decission making in the class, I have made some WWW-pages . If you have Netscape 4 or better (this is not for IE-users), pleace look my DHTML-pages http://noppa5.pc.helsinki.fi/koe/dhtml.html and there Making decissions 1 (Critical value) Making decissions 2 (Probability) Some of my students said "it was usefull". Regards Juha Puranen -- Juha Puranen Department of Statistics P.O.Box 54 (Unioninkatu 37), 00014 University of Helsinki, Finland http://noppa5.pc.helsinki.fi === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
