Approximately 1900 University of North Florida students completed a survey during the 2006 spring semester that included 8 open-ended and 20 multiple choice items. Most of the open-ended items asked each participant to report “why” he or she chose a previous multiple choice response option.

Researchers initially used a traditional approach to analyzing open-ended or qualitative responses. This approach involved an identification of common themes that recurred across the responses of each item. These common themes then subsequently served as the basis for a coding scheme used by independent raters or ‘data coders” to categorize responses to the open-ended survey items.

Multiple researchers initially collaborated to create reliable categories for each qualitative item. These researchers then used these categories as a basis for the qualitative data to be coded. Each qualitative item was treated independently and assessed by two data coders (a total of 16 data coders). If there were discrepancies between raters, a third rater was asked to asses an item blindly (without knowledge of the previous raters’ coding). Outcomes from this traditional qualitative analysis approach were entered into SPSS for Windows™ 13.0 for further analysis.

Responses to the eight open-ended survey items are currently being analyzed by an automated approach (SPSS Text Analysis for Surveys™ 2.0). Outcomes from this automated approach will be compared and contrasted with the traditional approach.