I am currently using EpiData manager for my data entry. Few questions are relevant for Married, so i want to jump those questions which are not applicable for unmarried. i could do this successfully with Jump option. But there are few questions which comes in the later part of my questionnaire which is applicable only for married, which i want to be skipped/jumped automatically based on the previous entry of marital status. How can this be done???
2. How do i have the value coded options visible in the output csv/stata file?? should i create the variables as interger/string variable???
I can address the second question: For Stata, you can code this any way you want because you can later change it in Stata as you need. In many cases, you can leave the value as a string and Stata can still perform the requested statistics. In cases where you need the value to be numeric, you can use destring to have Stata change the value from string to numeric. If you have several nominal categories of a variable, you can use encode to provide a numeric value for that category, and the label will be that of the category name. So basically, as you can see, Stata will allow you convert your data as you'd like. Incidentally, if you have all your data saved as a .csv file, you can change the string values to numeric while importing the file into Stata. This saves the steps I mentioned earlier.
I hope this helps
Ariel
Assuming you are using EpiData version 3.1 (see http://www.epidata.dk/index.htm - be sure to go to the correct download page for EpiData 3.1!), here is example code to conditionally skip the two variables "exvar06" and "exvar07" if the variable "married" = 1 (meaning "yes"). In the example, in the check block of "exvar05" that is to be entered immediately before the variable "exvar06" I build in an "after entry" check that conditional on the value of variable "married" being 1 (= yes) enters the code for "not applicable" (say: 8) into "exvar06" and "exvar07" and jumps to "exvar08". I included a message to the coder to alert him/her to jump to "exvar08" in order to prevent coding errors and/or to alert him if inconsistently "exvar06" or "exvar07" have been answered. Of course, you may skip the message to speed up data entry:
* ------ Begin EpiData code: ---------
exvar05
MISSINGVALUE 9
COMMENT LEGAL USE exvar05
MUSTENTER
TYPE COMMENT
AFTER ENTRY
IF (married = 1) THEN
LET exvar06 = 8
LET exvar07 = 8
HELP "Continue entering data with question 'exvar08'!"
GOTO exvar08
ENDIF
END
END
exvar06
MISSINGVALUE 9 8
COMMENT LEGAL USE exvar06
MUSTENTER
TYPE COMMENT
END
exvar07
MISSINGVALUE 9 8
COMMENT LEGAL USE exvar07
MUSTENTER
TYPE COMMENT
END
exvar08
MISSINGVALUE 9
COMMENT LEGAL USE exvar08
MUSTENTER
TYPE COMMENT
END
* ------ End EpiData Code. ----------
Thanks Mr. Ariel Linden and Mr. Dirk Enzmann ..
Mr. Dirk Enzmann, i am using Epidata Manager not the Epi Data 3.1, i am unable to find the code typing option in the Epidata Manager....
I am not absolutely sure, but I believe that more complicated programmed checks as I showed you in the example are only possible with EpiData 3.1 and not with the EpiData Manager. On the download page of EpiData it says: "Relational databases and complicated programmed checks can only be made with EpiData (Classic) Entry v3.1."
EpiData Manager is still in the developing stage and there are plans to implement programmed checks in the future. Meanwhile I would recommend to use EpiData Entry 3.1, instead.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis