This guide will walk through data preparation tips for survey data.
Dedoose has a “Survey Importer” tool that can instantaneously link quantitative (e.g., scores, continuous variables, demographics, Likert-scale and multiple-choice questions, etc.) and qualitative data (open-ended responses) in your survey. The importer does this by checking each column title, examining the data in each column, and then using this information to organize and link the data.
For a thorough overview of the Survey Importer, please view the Institute for Mixed Methods Survey Data Webinar below.
Formatting Your Survey Data
Excel or CSV is the best format for uploading survey data. When preparing your spreadsheet there are several things to consider:
1. Titles in the column headers of the quantitative data will become Descriptors and data in the columns will become the field values.
2. The qualitative data will be compiled into a single document for each participant and each response excerpted and coded according to the corresponding column header. In other words, the title column for each qualitative question in your survey will become a code and applied to the data so you can quickly view and code specific questions.
For this to work well, the Excel file needs to be carefully prepared. Below are several factors to consider:
Quantitative or Categorical Data
- Each column containing quantitative (or closed-ended, descriptor) data needs a column header that refers to the category the data belongs to. In Dedoose this is called a Descriptor field. For example, for a survey question like 'Please indicate your age group,' it may be best to edit to 'Age Group' in the column header.
- If data are represented by numeric proxies, e.g., 1, 2, and 3 for 1 = 10-25 years, 2 = 26-40 years, 3 = 41-50 years, you'll be happiest in Dedoose if you use the actual values and change 1s to '10-25 years' the 2s to '26-40 years and so on. This is a fairly straightforward process with Excel’s search-and-replace feature.
- For columns with continuous numeric data, you convert these data to a categorical form. For example, imagine collected data from a depression scale where scores can range from 0 to 100. From clinical practice or other guidelines, we may know that scoring between 0 and 50 is considered 'not depressed,' 51-70 = 'minimally depressed,' 71-85 = 'moderately depressed,' and 86+ = 'severely depressed.' Converting these data prior to import will allow you to explore your excerpting and tagging activity as a function of these groupings.
- For columns containing qualitative (narrative or open-ended) data you will want to change the header to whatever will be most informative as a code. For example, a header like 'Please describe your experiences with X,' might be shortened to 'X experiences'.
Type _ddqual_ as a prefix to the column title of qualitative data to ensure Dedoose reads the data as qualitative.
- Revise your column title to be concise and descriptive. Dedoose will pre-code your qualitative survey data using the column title as the code.
Other Resources can be found on the Dedoose website.