How to handle data from multi-country surveys
If you are handling a market research project that is being conducted in more than one country, you are not only likely to encounter language issues and dealing with many stakeholders, but the issue of managing data is probably an important consideration.
The ideal solution, but….
Ideally, from a data perspective, you would like to have the same questionnaire in every country with data entered in exactly the same format. If that’s possible, then you should have no problems, but there are several potential issues.
Common problems with data from multi-country surveys
The most common problems encountered with data from international projects are:
- There are additional or different questions in some or various countries
- There are different brand lists/responses applicable in different countries
- The data is being entered locally using different software platforms
- Data has been entered in different formats or layouts in each country
All of these can result in data being different for each country, which can make the production of analysis and reporting complex or laborious.
How to resolve data differences – Solution #1: Manage Data Entry
One way to ensure that analysis and reporting are straightforward is to ensure that data is stored in a uniform format. This means that questions that only appear in some countries are stored in reserved data locations and that any different brand/response lists are unified using a common code list across all countries.
These types of task can be handled easily in QPSMR and Snap. Both of these products can manage country differences and ensure that your data is stored using one data map. QPSMR has a freely distributable data entry module so that if you have different suppliers in different countries, you can manage data mapping centrally. Snap works in a similar for online projects using a database to save the different country variations.
How to resolve data differences – solution #2: Post Data Entry Management
Sometimes, you are not able to control how data is stored for each country. In such a situation, you could, for example, have ten sets of data from ten countries which are all in a different layout. MRDCL is a good tool to manage such data. You can enter data locations in a spreadsheet so that MRDCL reads data from where it happens to be coded for each country. Similarly, brand/response lists can be stored in spreadsheets, so that Brand 5 in Country 1 can be matched with Brand 8 in Country 2, for example. Data could also be in different data layouts and be controlled from the same spreadsheet. MRDC can assist with such problems by providing either software or a service to manage the task.
The key is to manage the project
Whatever solution you adopt, being conscious of data management issues in advance is always good practice. We are often asked to consult on such projects and planning is the key to success, particularly if the project is part of a tracking programme where questionnaires may vary from wave to wave. If you need any help or advice, just ask.