Easy Way to Turn Survey Information Into a Statistical Table

Surveys are a popular data collection method for researchers in a wide range of fields, including marketing, sociology and public opinion. Summarizing hundreds or even thousands of responses into a statistical table of results can appear challenging. However, a few basic steps can make the task manageable, enabling you to present survey results in a clear and concise manner that your client or audience will understand.
  1. Code Responses Numerically

    • The first step in preparing survey data for presentation in a statistical table is to convert the data into quantitative form. One of the most popular ways of doing this is to assign a numeric value to responses. Many surveys, for example, use a Likert scale of responses, in which a respondent answers each question with one of a series of responses. A typical response scale consists of the following: strongly agree, agree, disagree, strongly disagree and neutral/no opinion. A researcher could code these responses by assigning a value of 4 to strongly agree, 3 to agree, 2 for disagree, 1 for strongly disagree and zero for neutral or no opinion.

    Frequency Distribution

    • An easy and convenient way to display survey results in a single table is to construct a frequency distribution table. A typical approach to such a table is for each question in the survey to represent a row. The columns, meanwhile, would consist of the number and percentage of respondents who indicated agreement or disagreement, depending on the number of response categories.

    Summary Measure

    • Another approach to summarizing survey results is to use a measure of central tendency, a statistic that reveals the center of a set of data. Measures of central tendency include the mean (arithmetic average), the median and the mode. A common mistake among novice researchers, however, is to take numerically coded survey responses and compute a mean value of the responses to each question. This is not a desirable approach to Likert scale data. Adding a “strongly agree” value of 4 to three “disagrees,” valued at 2 each, and dividing by 4 gives us a mean value of 2.5 with no real indication of its significance. Instead, construct a table in which each question comprises a row, and in the columns, the mode, or the most frequent response, of each question. For example, if 50 people answered a survey, and 33 of them responded “agree” (coded 3 in our example) to a single question, then the value of the mode for that question is 3, which tells us that the largest number of respondents agreed with that question.

    Creating Tables

    • Statistical tables can be easily created using a spreadsheet program, such as Excel, or using the table and chart wizard in Microsoft Word.

Learnify Hub © www.0685.com All Rights Reserved