home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.math.stat
- Path: sparky!uunet!wupost!gumby!destroyer!cs.ubc.ca!newsserver.sfu.ca!rivas
- From: rivas@fraser.sfu.ca (Jorge David Rivas)
- Subject: Re: Test for Linearity ??
- Message-ID: <1992Dec11.220220.10224@sfu.ca>
- Sender: news@sfu.ca
- Organization: Simon Fraser University, Burnaby, B.C., Canada
- References: <1992Dec10.234906.61926@cc.usu.edu> <2975@tymix.Tymnet.COM>
- Date: Fri, 11 Dec 1992 22:02:20 GMT
- Lines: 49
-
- In article <2975@tymix.Tymnet.COM> rick@daffodil.tymnet.com (Rick McFarland) writes:
- >Why don't you try calculating the correlation coefficient: the closer it
- >is to 1 or -1, the more linear your data.
- >
- >
- >
- >------------------------------------------------
- >"ad astra per aspera"
-
- I do not think that the corr. coef. would tell much at all right away.
- If he is asking about linearity I'm sure he has in mind fitting a straight line.
- He says:
-
- " I have a question regarding, linearity
- of data points. Does anybody know of any statistical
- test for linearity. I would appreciate any kind of
- help, references etc."
-
- I would think that PLOTTING the data first would be necessary in order to asses
- the rest of the test:
-
- ie: if there is apattern other than somekind of random scattering along some
- invisible straight line (ie: curvilinear like a parabola) then he should fit
- a polynomial model (possibly of second order), then the reduced model with only
- the linear term and THEN asses linearity by means of the corresponding tests.
-
- If not consider looking at : "Problem solving: a statistician's guide"
- Christopher Chatfield,
- Chapman and Hall, pages 117-118
-
- In it there are 4 sets of data constructed by Anscombe (1973). All have r=.82,
- however:
-
- set 1 is your typical "around the straight line" cloud
-
- set 2 is 3/4 of a seemingly "branches down" parabola on which all the point are
- perfectly lined on it.
-
- set 3 is a set of 11 points in which 10 are perfectly lined on a stright line
- and the 11th is an obvious outlier
-
- set 4 is a setof 11 points, but this time the first 10 make a perfectly vertical straight line. The 11th one is again a far away outlier.
-
- REMEMBER: ALL them have r = 0.82
-
- Jorge David Rivas (undergrad) Dept. of Math & Stats
- Simon Fraser University
- Burnaby, B.C. CANADA
-
-