I have a conjoint survey in several languages. How can I copy and paste the data together for one export/utility run/simulator?
Version 7 and prior used .dat files, which are just text files that can be opened with a simple text editor such as Notepad and combined. For version 8, the easiest route is to log into your Admin module directly and download a CSV file and work in a program such as Excel.
In general, there are two components to consider when merging data files: general data and conjoint/MaxDiff data
For general data, if the survey has the same question names and number of answers, a simple copy and paste of the data files together should be sufficient. If one survey has questions the other does not, then during import, version 8 will ask you a little about the new variables and append them to your data record. Version 7 and prior will require variables to already exists in order to import data for them (as a question, pass-in field, etc.). Missing data is fine, as it's treated just like a respondent skipped that question.
For conjoint data, adaptive methodologies contain the necessary design information in the data file, so copying and pasting data together shouldn't cause any issues as long as the conjoint exercises contained the same settings, attributes, levels, etc.
For "fixed-design" methods, such as CBC and MaxDiff, a design is generated before the survey is fielded. These design files sit in the study folder and have a unique design ID. When the respondent takes the survey, the design ID is written to their record. If you try to combine data files with two different design IDs, you won't be able to analyze respondent data. The design ID is created when you generate the design, so if you save a copy of the study for another language, if you delete levels and paste the levels back in in a different language, you'll have a different design ID. If the design is fundamentally the same, you just hit the Generate Design button again in the software, then you can probably do a find and replace in the data files to make it look like everyone saw the same design ID, but you might want to contact support and go over the situation with them first.
The last issue to consider is respondent numbering. If you field two surveys, you will have multiple respondent 1's, 2's, etc. In version 8, one option would be to take the CSV file and add a constant, such as 10,000, to the respondent numbers of a second survey. In all versions, however, you can change the settings of exports and utility estimation to use sequential respondent numbering, which uses the respondent's position in the data file, rather than their assigned respondent number when they took the survey, to provide a unique identifier for each respondent.