I'm doing a project about different phases of democratic transitions and their possible effects on economic growth. Because of the large amounts of missing data in econometric country data, I have decided to use multiple imputation to fill in the blanks.
Unfortunately, I can't get stata to perform the mi estimations. I have used the following commands:
- mi query -
- mi set flong [or any other style] -
- mi register imputed [variables to be imputed]
- mi register regular [other non-passive variables]
- mi impute [tried a host of different methods like - regress - , - mvn - , - logit - etc.], add([usually 5, but other numbers have been tried])
All of the above works fine. Then I attempt the actual estimation...
- mi estimate: regress [or some other function] variable1 variable2 -
... and get this error message:
- file not estimates -
- r(610); -
I have encountered the same error with many different datasets; both original, simulated and hand-typed. As the generic error description says that the file is not of a stata-format, I have tried to use csv-datasets and datasets from different versions of stata. All to no avail.
The original plan was to use the Amelia II-application for R to do the imputations and then have stata do the estimates (since I'm pretty non-fluent in R), so I have also tried using imported Amelia II-imputations with the - mi import - command. They too work fine up untill the estimation.
I have combed the web and help-files for a solution to this problem, so I really hope you guys can help.
Thank you for your consideration!
- Bertel Teilfeldt Hansen
PS. I'm completely new to statalist, so please excuse any breaks of protocol.
*I have read and understood the statalist FAQ*
STATA: Stata 11.0 (in administrator-mode)
OS: Windows 7 (64 bit)
Memory: 4 GB ram
Processor: AMD Athlon II Dual-Core M300 2.00 ghz
INTRODUKTIONSPRIS: 250 foldede visitkort, 99 kr. - klik her og se