| Author |
 Topic  |
|
|
kmgraju
Malaysia
18 Posts |
 Posted - 01/06/2015 : 10:30:41 AM
|
Dear all
What type of analysis is most suitable to check the significance.
case 1. One independent variable and one dependent variable
Case2. Two independent variables and one dependent variable.
Thank you
Govind |
|
|
Hideo Fujii
USA
1582 Posts |
Posted - 01/06/2015 : 6:16:15 PM
|
Hi Govind,
If you have only one independent variable, you can run the linear fit (regression) tool, and if two, the multiple linear regression tool. In either case, it outputs the ANOVA table, and tells the Prob>F , i.e., P-value. In the linear case, it even reports the judgement at 5% level in the English text.
Hope I didn't misunderstand what you wanted to know.
--Hideo Fujii
OriginLab
|
 |
|
|
kmgraju
Malaysia
18 Posts |
Posted - 01/06/2015 : 8:41:02 PM
|
Dear Mr Hideo Fujii
I also expected the same thing.
Thank you for your immediate response.
One more question
Case 3 : three different groups, two independent variables
and one or two or more dependent variables.which test is more suitable?
One way ANOVA or two way Anova or regression
thank you
Govind
|
|
|
|
Hideo Fujii
USA
1582 Posts |
Posted - 01/07/2015 : 10:06:43 AM
|
Hi Govind,
In statistics, there are methods like MANOVA or MANCOVA (both are currently not a part of Origin) for multiple dependent/independent variables, but I lack the knowledge nor experiences of these. Not sure, but Origin's non-linear curve fitter, which allows multiple dependent/independent variables, may serve you as it also produces ANOVA table, if you can provide the model.
Sorry for not being able to help you much as I'm not a statistician. Maybe someone in this forum can tell further?
--Hideo Fujii
OriginLab |
|
|
|
Sam Fang
293 Posts |
Posted - 01/08/2015 : 02:48:43 AM
|
Hideo is right. When independent variables include both continuous and categorical variables and there are multiple dependent variables, MANCOVA is needed.
Sam
OriginLab Technical Services |
|
|
| |
Topic |
|
statistical significance