Data analysis is the process of assessing the data you’ve gathered to extract useful statistics and draw conclusions. It allows you to make sense of the information gathered and answer your key research questions.
But when the data comes rolling in, you may feel a little overwhelmed. It’s often hard to know where to start. Especially because there are a few different options. We recommend developing a plan and thinking carefully about how you’re going to organise and analyse your survey data.
First things first, what questions did you set out to answer? What did you want your survey to tell you? These are your research questions.
For instance, let’s say you held a conference for people working in education, and you wanted to know what the attendees thought of your event. Normally, you will have come up with several key questions you wanted answers to, including:
Attribute each survey question to the relevant research question. We’ve provided an example in the following table. Doing this will help you know which survey questions to refer to for specific research topics. For example, to find out which parts of the conference attendees liked the best, look at the answers to questions 3–6.
|Research question||Survey question(s)|
|How did attendees rate the event overall?||1. Overall, how satisfied were you with the conference?|
2. How useful was this conference compared to other conferences you have attended?
|Which parts/aspects of the conference did attendees like the best?|
Which parts/aspects of the conference need to be improved?
|3. How would you rate the difficulty of the workshop?|
4. Overall, do you think the conference provided too much, too little or about the right amount of networking?
5. In general, how would you rate the food at the conference?
6. Do you feel the temperature in the conference building was too hot, too cold or just right?
|Who are the attendees and what are their specific needs?||7. Are you a teacher, student or administrator?|
8. How large is your school?
9. How old are you?
There are different ways of analysing data, depending on whether it’s qualitative or quantitative. Because most surveys include a mixture, you’ll probably use both types of data analysis techniques.
Quantitative research focuses on facts and figures. Answers are often presented numerically, but can also be expressed as words, such as yes/no questions or multichoice.
When it comes to analysis techniques for quantitative data, some basic statistics can help you understand your results and identify patterns. This includes descriptive statistics such as minimum and maximum, mean (or average), median and standard deviation (looking at the distance from the mean). And the good news is, there’s no need for other data analysis tools or number crunching—all these statistics are automatically calculated in the Analyze Results [A1] section of your SurveyMonkey survey.
Qualitative research often investigates opinions and perceptions. While it’s a little harder to analyse than quantitative data, it provides some valuable insights. For example, it often looks at the motivations behind decisions, meaning it’s a useful way of adding context to your quantitative data.
Two key types of qualitative data analysis are content analysis and grounded theory. Content analysis actually refers to a range of methods for analysing text-based data. One example of this is Sentiment Analysis, which is a way of identifying the emotion behind people’s comments. When this functionality is enabled in SurveyMonkey, your survey answers will be categorised as Positive, Neutral, Negative or Undetected. Meanwhile, grounded theory involves looking at the qualitative data to explain a pattern. You can analyse individual comments manually and then keep track of them by creating and adding your own tags in SurveyMonkey. You can also filter responses by tag.
As well as looking at overall survey results or at results for each question, it can be incredibly insightful to segment your data and compare results across different segments. You could choose to segment your results by demographic characteristics such as gender, age, location or occupation and compare between them. For instance, you might discover that those aged 30–35 enjoyed your event more than those of other age groups. Or perhaps that, overall, teachers enjoyed the networking sessions more than students.
You can also use some nifty filtering methods for data analysis and comparison survey-wide. For example, SurveyMonkey allows you to filter your entire survey data based on sentiment analysis, on specific demographics or on tags, so you can compare results across these groups.
Humans are visual creatures. This means we find information easier to understand and remember when it’s displayed as an icon, graph or image, rather than as screeds of text or rows of numbers. Given this, it’s a good idea to display your findings visually before sharing them with your team or adding them to your report. Some common visualisation options for quantitative data are pie charts, bar graphs and line graphs. With SurveyMonkey, you can change between these different charts and graphs, choosing the most compelling option for each result and even customising them. And with your qualitative data, you can create a word cloud, which is a visual representation of the most common words and phrases from your open-ended responses.