Asked 17th Jan, 2018
What should I do if R-square is 1.000 or 100%?
My research topic is relationship between soil fertility and plant growth.
The dependent variable is 1 (THT) and independent variable is 18 (soil chemists components and soil nutrients components).
And my sample is 5 plots for soil A and 13 plots for soil B.
When I use the multivariate linear regression on SPSS, the result of R-square and R is 1.000, but there is no appear a number of Adj R-square and SEE.
On the ANOVA table, still no appear a number of F. and Sig.
And the coefficients table, still no appear a number of t. and Sig.
So, what must I do?
I must add the sample or increase the independent variables?
Or, I formerly use 2 procedure :
1. t test -> Sig of variables -> regression -> pearson correlation
2. pearson correlation -> Sig variables -> regression -> t test
If I use another software, Minitab, the result on output is different between SPSS.
So, what can I do?