Means with the same letter are not significantly different

samer_badri75 — Wed, 06 Feb 2019 03:32:43 GMT

* day_day.sas;
options pageno=1 linesize=80;
goptions reset=all;
title "Two-way ANOVA for day";
title2 "Data from day and Reynolds (2007)";
data day;
	input location $ exp $ day;
	* Apply transformations here;
	y = log10(day);
	datalines;
Cool	EC	22
Cool	EC	20
Cool	EC	19
Cool	EC	19
Cool	AC	65
Cool	AC	51
Cool	AC	70
Cool	AC	78
FB	    EC	9
FB	    EC	10
FB	    EC	10
FB	    EC	10
FB	    AC	11
FB	    AC	10
FB  	AC	10
FB	    AC	11
RES	    EC	15
RES  	EC	12
RES	    EC	15
RES	    EC	11
RES	    AC	13
RES	    AC	14
RES	    AC	14
RES	    AC	13

;
run;
* Print data set;
proc print data=day;
run;
* Plot means, standard errors, and observations;
proc gplot data=day;
	plot y*location = exp  / vaxis=axis1 haxis=axis1 legend=legend1;
	symbol1 i=std1mjt v=star height=2 width=3;
	axis1 label=(height=2) value=(height=2) width=3 major=(width=2) minor=none;
	legend1 label=(height=2) value=(height=2);
run;
* Two-way ANOVA with all fixed effects;
proc glm data=day;
	class location  exp ;
	model y = location  exp  location *exp ;
    lsmeans location  exp  / adjust=tukey cl lines;
	output out=resids p=pred r=resid;
run;
* Two-way ANOVA with interaction;
title3 "MODEL WITH INTERACTION - USE THIS OUTPUT IF INTERACTION SIGNIFICANT";
proc glm data=day;
	class location exp;
	model y = location exp location*exp / ss2;
	lsmeans location*exp / slice=exp exp=location;
	output out=resids p=pred r=resid;
run;
goptions reset=all;
title "Diagnostic plots to check anova assumptions";
* Plot residuals vs. predicted values;
proc gplot data=resids;
	plot resid*pred=1 / vaxis=axis1 haxis=axis1;
	symbol1 v=star height=2 width=3;
	axis1 label=(height=2) value=(height=2) width=3 major=(width=2) minor=none;run;
* Normal quantile plot of residuals;
proc univariate noprint data=resids;
	qqplot resid / normal waxis=3 height=4;
run;
quit;

Greeting All

I run one way-ANOVA and I got the results as showed in the pictures attached, can get a similar result in Two-way ANOVA.
Thanks in advance

Re: Means with the same letter are not significantly different

PaigeMiller — Wed, 06 Feb 2019 13:17:05 GMT

I run one way-ANOVA and I got the results as showed in the pictures attached, can get a similar result in Two-way ANOVA.

Yes.

Re: Means with the same letter are not significantly different

samer_badri75 — Wed, 06 Feb 2019 15:09:30 GMT

Ok , what’s the code that I need to add?

Thanks

Re: Means with the same letter are not significantly different

PaigeMiller — Wed, 06 Feb 2019 15:14:05 GMT

lsmeans location*exp  / adjust=tukey cl lines;

Re: Means with the same letter are not significantly different

samer_badri75 — Wed, 06 Feb 2019 17:19:43 GMT

Many thanks

I add it and it works perfect

topic Re: Means with the same letter are not significantly different in SAS Programming

Means with the same letter are not significantly different

Re: Means with the same letter are not significantly different

Re: Means with the same letter are not significantly different

Re: Means with the same letter are not significantly different

Re: Means with the same letter are not significantly different