Re: chi square validity?
Hi On Tue, 18 Dec 2001, Benjamin Kenward wrote: > Let's say you have a repeatable experiment and each time the result can be > classed into a number of discrete categories (in this real case, seven). > If a treatment has no effect, it is known what the expected by chance > distribution of results between these categories would be. I know that a > good test to see if a distribution of results from a particular treatment > is different to the expected by chance distribution is to use a > chi-squared test. What I want to know is, is it valid to compare just one > category? In other words, for both the obtained and expected > distributions, summarise them to two categories, one of which is the > category you are interested in, and the other containing all the other > categories. If the chi-square result of the comparison of these categories > is significant, can you say that your treatment produces significantly > more results in particularly that category, or can you only think of the > whole distribution? Yes, this is equivalent to planned contrasts (assuming it was planned) in ANOVA. In ANOVA with the critical condition as the last one, the contrast would be -1 -1 -1 -1 -1 -1 +6 (or some variation on that, e.g., normalized coefficients). I remember long ago in an epidemiology class learning how to partition chi^2, but I do not remember off hand whether the contrast ends up being equivalent to collapsing groups 1-6 and doing the 2-group chi^2, or whether there was a way to partition a SS for the numerator and use a common denominator from the 7-group chi^2 for the test of contrasts. The following link suggests that the two are not the same. http://www.sas.com/service/techsup/faq/stat_proc/freqproc919.html Best wishes Jim James M. Clark (204) 786-9757 Department of Psychology(204) 774-4134 Fax University of Winnipeg 4L05D Winnipeg, Manitoba R3B 2E9 [EMAIL PROTECTED] CANADA http://www.uwinnipeg.ca/~clark = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: chi square validity?
On Tue, 18 Dec 2001 14:19:34 + (UTC), [EMAIL PROTECTED] (Benjamin Kenward) wrote: > > Let's say you have a repeatable experiment and each time the result can be > classed into a number of discrete categories (in this real case, seven). > If a treatment has no effect, it is known what the expected by chance > distribution of results between these categories would be. I know that a > good test to see if a distribution of results from a particular treatment > is different to the expected by chance distribution is to use a > chi-squared test. What I want to know is, is it valid to compare just one > category? In other words, for both the obtained and expected > distributions, summarise them to two categories, one of which is the > category you are interested in, and the other containing all the other > categories. If the chi-square result of the comparison of these categories > is significant, can you say that your treatment produces significantly > more results in particularly that category, or can you only think of the > whole distribution? Mathematically, the statistical test is okay: There is no problem if you decided at the outset that 7 categories should be scored as D-and-not-D, so you would do a 2x2 contingency table test. Other inferences, of course, are more problematic. "Multiple-tests." Deciding on a test based on the outcomes is a form a cheating in the hypothesis testing, if you don't take that into account in the reporting of it. If your overall test is significant -- with 6 d.f., I think -- then it is somewhat conventional to look at the separate contributions by cell, without being too shy. If the overall test is *not* that happy, then you ought to state that, and offer further guesses as purely exploratory or suggestive numbers. Then you can describe one cell's contribution "versus the others." -- Rich Ulrich, [EMAIL PROTECTED] http://www.pitt.edu/~wpilib/index.html = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: chi square validity?
Ben, The other posters give some good advice, but you could also set simultaneous confidence intervals for all 7 categories. If you are trying to establish "equivalence" then you could define some range that would be appropriate (say, within 3%). The CI's are easy enough to compute and I think a Bonferroni approach would work okay here. There are various methods of computing the intervals, but I would suggest a Wilson-type interval. (For a discussion, see the 2000 issue of the Journal of Statistical Software that discusses setting CI's for the multinomial) Warren May [EMAIL PROTECTED] (Benjamin Kenward) wrote in message news:<9vnj9m$s2c$[EMAIL PROTECTED]>... > Hi folks, > > Let's say you have a repeatable experiment and each time the result can be > classed into a number of discrete categories (in this real case, seven). > If a treatment has no effect, it is known what the expected by chance > distribution of results between these categories would be. I know that a > good test to see if a distribution of results from a particular treatment > is different to the expected by chance distribution is to use a > chi-squared test. What I want to know is, is it valid to compare just one > category? In other words, for both the obtained and expected > distributions, summarise them to two categories, one of which is the > category you are interested in, and the other containing all the other > categories. If the chi-square result of the comparison of these categories > is significant, can you say that your treatment produces significantly > more results in particularly that category, or can you only think of the > whole distribution? > > Thanks, > > Ben = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: chi square validity?
[EMAIL PROTECTED] (Benjamin Kenward) wrote in message news:<9vnj9m$s2c$[EMAIL PROTECTED]>... > Hi folks, > > Let's say you have a repeatable experiment and each time the result can be > classed into a number of discrete categories (in this real case, seven). > If a treatment has no effect, it is known what the expected by chance > distribution of results between these categories would be. I know that a > good test to see if a distribution of results from a particular treatment > is different to the expected by chance distribution is to use a > chi-squared test. What I want to know is, is it valid to compare just one > category? In other words, for both the obtained and expected > distributions, summarise them to two categories, one of which is the > category you are interested in, and the other containing all the other > categories. If the chi-square result of the comparison of these categories > is significant, can you say that your treatment produces significantly > more results in particularly that category, or can you only think of the > whole distribution? Yes, as long as the choice of which category to do it for is not based on the data... no fair just testing the most extreme one. Glen Glen = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: chi square validity?
Glen wrote: > [EMAIL PROTECTED] (Benjamin Kenward) wrote in message >news:<9vnj9m$s2c$[EMAIL PROTECTED]>... > > Hi folks, > > > > Let's say you have a repeatable experiment and each time the result can be > > classed into a number of discrete categories (in this real case, seven). > > If a treatment has no effect, it is known what the expected by chance > > distribution of results between these categories would be. I know that a > > good test to see if a distribution of results from a particular treatment > > is different to the expected by chance distribution is to use a > > chi-squared test. What I want to know is, is it valid to compare just one > > category? In other words, for both the obtained and expected > > distributions, summarise them to two categories, one of which is the > > category you are interested in, and the other containing all the other > > categories. If the chi-square result of the comparison of these categories > > is significant, can you say that your treatment produces significantly > > more results in particularly that category, or can you only think of the > > whole distribution? > > Yes, as long as the choice of which category to do it for is not based > on the data... no fair just testing the most extreme one. good advice in every case. Jay > > > Glen -- Jay Warner Principal Scientist Warner Consulting, Inc. North Green Bay Road Racine, WI 53404-1216 USA Ph: (262) 634-9100 FAX: (262) 681-1133 email: [EMAIL PROTECTED] web: http://www.a2q.com The A2Q Method (tm) -- What do you want to improve today? = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =