We have used CAR as a dependent factor and rest factor as in dependent. The results shown by Regression analysis on Excel/ SPSS software are vague. I don't know the reason for getting it. Values are correct if I am taking 4 factors at a time but have wrong values if I use more factors.
Multiple R 1
R Square 1
Adjusted R Square 65535
Standard Error 0
Observations 8
Standard Error 0
t Stat 65535
P-value #NUM!
What can be the PROBABLE ERROR in the said data. Please look and let me know.
As Michael says:
The number of predictors should be less than the number of obesrvations.
A very lenient rule of thumb: the number of cases (observations) should be at least TEN times the number of predictors (including intercept).
In fact: if you have N observations and N RaNdOm predictors; the r-square will be 1.
The remarks on multicollinearity are true but do not really explain the unexpected results.
If predictors are highly correlated the results can not be generalized; they are only valid for the sample.
Greetings
I think you may need more data points. Perhaps you can breakdown the periods into months or weeks. You would need sufficient data points to reach possible statistical significance under the normal distribution...
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS