Re: [R] About stepwise regression problem

2011-10-07 Thread Gavin Simpson
On Thu, 2011-10-06 at 20:12 -0700, pigpigmeow wrote: > chris, > I'm not using lmer, i just use gam mixed with smoothing function and linear > function > I have the following question, > 1.if I reject the variable term which has greater the p-value no matter the > variable term is smoothing term or

Re: [R] About stepwise regression problem

2011-10-07 Thread Frank Harrell
Removing variables because of high P-values is not a valid procedure. Use of AIC or BIC is just a restatement of P-values. AIC can be quite useful if you have posited a very small number of fully pre-specified models (e.g., 2 or 3) and want to choose between them. Stepwise variable selection wit

Re: [R] About stepwise regression problem

2011-10-06 Thread pigpigmeow
chris, I'm not using lmer, i just use gam mixed with smoothing function and linear function and summary of the model, it shows Family: gaussian Link function: log Formula: newNO2 ~ pressure + s(maxtemp, bs = "cr") + s(avetemp, bs = "cr") + s(mintemp, bs = "cr") + RH + s(solar, bs = "cr") +

Re: [R] About stepwise regression problem

2011-10-06 Thread Chris Mcowen
I hadn't seen that page Dennis, that makes the case much more succinctly than my anti stepwise ramblings! Furthermore, "pigpigmeow" if you are using a random effects model i.e lmer - where are you getting your p-values from? And what do they mean in this context? I would strongly advise using i

Re: [R] About stepwise regression problem

2011-10-06 Thread Dennis Murphy
Hi: Please read this: http://www.stata.com/support/faqs/stat/stepwise.html Dennis On Thu, Oct 6, 2011 at 6:41 AM, pigpigmeow wrote: > using AIC/BIC, I'm not know too much about this. I just know using p-value to > perform stepwise regression > > if I used p-value to perform multimodel stepwise

Re: [R] About stepwise regression problem

2011-10-06 Thread pigpigmeow
using AIC/BIC, I'm not know too much about this. I just know using p-value to perform stepwise regression if I used p-value to perform multimodel stepwise regression, is it correct in the first message box? -- View this message in context: http://r.789695.n4.nabble.com/About-stepwise-regression-

Re: [R] About stepwise regression problem

2011-10-06 Thread Chris Mcowen
Why not use AIC / BIC based model selection which in this case makes more sense than artificial P-Values, and eliminates the problems of stepwise regression. For background read Burnham and anderson - multimodel selection. Chris. On 6 Oct 2011, at 08:14, pigpigmeow wrote: If I used mixed mode

Re: [R] About stepwise regression problem

2011-10-06 Thread pigpigmeow
If I used mixed model ( linear + smoothing function) How can I used p-value to perform stepwise regression??? Is it similar of Pr(<|t|) of linear term and P-value of smoothing term but T value of linear and F of smoothing term? please help! -- View this message in context: http://r.789695.n4.