Fact Sheet 8

Analysing Basic Survey

Download PDF Factsheet 8 (677kb) - Download PDF Factsheet 8

1

Clean Data: After importing (or manually entering) data into a program such as Excel, take a look over the information for incomplete responses, multiple responses from the one participant, or any data entry errors / inconsistencies. Be consistent in your cleaning method; if every participant has responded twice, decide which response to keep and apply that decision to all.

2

Tally the Frequency of Responses: Generally one of the most important parts of survey analysis is determining the frequency of each response.  For more information on this, view the University of Wisconsin Guide to Using Excel to Analyse Data.

Tally the Frequency of Responses

3

Create Percentages from Frequency Data: Calculate the percentage for each response by dividing the frequency of responses by the total number of responses for that question, multiply by 100 to obtain a percentage (eg. 165 / 350 * 100 = 47%).

Create Percentages from Frequency Data

4

Finding the Average Answer: Finding the average response to a particular question is a useful way of summarising data for reporting purposes.  To find the average: code responses to their corresponding numbers (eg. Strongly Agree = 5, Agree = 4, Neutral = 3, Disagree = 2, Strongly Disagree = 1), calculate the total number for the question (eg. Participant 1 selected Agree (4), Participant 2 selected Disagree (2), and Participant 3 selected Neutral (3).  The total would be 4+2+3 = 9). Divide the total by the number of responses (eg. 9/3 = 3. In this case, the average response was a Neutral response).
If using Excel to analyse data, you can use the AVERAGE function (detailed instructions contained on page 19 of the University of Wisconsin Guide to Using Excel to Analyse Data.
Note that it is only appropriate to calculate an average response when you have 10 or more responses to a question.

5

Dealing with Neutral Responses: A neutral response generally indicates that the participant is unsure of their view on the question.   There are two generally accepted ways of analysing these responses: report them alone (eg. 100 people agree, 20 were neutral, 100 people disagree), or add the responses to both ends of the scale (eg. 120 people agree, 120 disagree).

6

Text Responses: Analysing text responses (or qualitative data) can be time consuming, but usually will provide the most useful responses.  There are 4 key steps involved in effective analysis:

  1. Organise the data in a way that is user friendly.  If responses have been collected using different methods, be sure to collate all data into the one excel sheet, or document.
  2. Read through the responses and note any themes or patterns that stand out.
  3. Tally the number of responses that correspond to each theme (eg. 6 people may have mentioned that they do not oppose the installation of additional EGMs, but they are unhappy with the location – this would be an interesting theme to report upon).
  4. Create percentages from the frequency data.  Be sure to report on all key themes found, even if a particular theme is only represented by 2% of the participants.
© 2017 Guides to Application Processes For Electronic Gaming Machines All Rights Reserved.