Identify discrete, differentiable categories for the responses to the survey question or questions. In a survey that asked consumers what they like about a brand of cereal, for example, response categories might include taste, nutrition and cost. Create a numerical code for each category, such as Taste -- 1, Nutrition -- 2 and Cost -- 3.
Mark the survey responses with the applicable category numerical code. Every response must fall into one of the categories. If a response does not fall into a category or could fit into more than one category, then add a category with a new numerical code for that response.
Record decisions made while coding the survey data, including how categories were applied. Survey responses may be confusing or misleading. In order to note potential bias or inaccuracies introduced through interpretation, record the logic used to categorize unclear responses.
Count the number of times a category was noted in the coded survey responses. Afterward, the coded categories and number of responses in each category are quantitative data because they can be expressed numerically. The data can be measured, analyzed or presented as needed.