| T O P I C R E V I E W |
| annal82 |
Posted - 10/30/2008 : 07:27:46 AM
Hello,
I just bought Origin 8 and I used it to fit my data to a sigmoidal function (Boltzmann function). In the results sheet are reported, in a specific block, the results of the ANOVA on the regression coefficient for each subject (I had previously selected the "indipendent fit-consolidated report" option for multi-data set). My problem is that the p-value cell (p < F) shows only a zero instead of the exact p-value. How can I know the exact level of significance of F? I tried to see if I could change the cell setting (I need to know at least two decimal values of p) but I'm not allowed to. Also copying the value on an excel file and increasing decimals, did not solve the problem.
Is there something I do not understand or I do wrong? This problem is present for each ANOVA I run testing a regression coefficient.
Thanks for your help
Anna Sforza |
| 3 L A T E S T R E P L I E S (Newest First) |
| rdremov |
Posted - 11/04/2008 : 4:58:49 PM
Alternatively, you could right-click on the table (e.g. ANOVA), select "Edit Format", navigate to NamedStyles->ogData, right click on Style branch to add new node "Format" and type in value ".7" (without qoutes) to force Origin to display 7 decimal digits for that particular table. |
| easwar |
Posted - 10/31/2008 : 1:04:50 PM
Hi Anna,
To change the numeric display in report sheets, such as number of digits etc, you can go to Tools->Options dialog and on Numeric Format tab, make your changes. You will then be prompted if the setting should be saved on start up. If you say yes, this will apply to all such sheets and other places. If you say No, it will work for the current session in all places and you can print out the sheet etc and then close Origin and on start-up it will go back to the default.
Easwar
OriginLab
|
| Echo_Chu |
Posted - 10/31/2008 : 06:09:37 AM
Hi, Anna,
Usually we won't round the result. It looks like 0 is exact level of your P value.
Prob in ANOVA is computed by this equation
1 - fcdf(FVal, 1, DF)
where is FVal is value of F Value, DF is value of DFof Residual. fcdf is a Labtalk funtion to compute F cumulative distribution function at x
Could try and compute the P value to see whether it is equal to 0?
Echo
OriginLab Corp |
|
|
