Ask your survey data a question: Pivot Tables

A Google Form is a great way to conduct a survey. It even produces nifty graphs of the results. But, it does not give you a breakdown of the data. More specifically, it does not answer Independent/Dependent variable questions, eg "Are Males or Females more likely to..." That's where a Pivot Table comes in handy.

Create a survey, share it to gather responses, then analyse your data using a Pivot Table as a way of asking your data questions about the interaction of Independent vs Dependent variables.

Step 1: Create & Share a Survey

  • Use a Google Form to set up a survey.
  • Click Share. You can use bit.ly or Rebrandly.com 
  • TIP: If you want clean, no-hassle data, avoid using multiple selection format questions (checkboxes).

 Step 2: View Survey Responses

  • You can get a quick summary of the data by clicking on Responses. But it is a one-dimensional analysis. Eg 50 said Yes, 20 said No, but we can't see quickly who said what (eg Male vs Female).

Step 3: Create a Pivot Table

Here is some fake sample data to work with. Make a copy of the file so you can play with the data.

Note:

  • There are 3 Independent variables (IV): Gender, Age Group and Location.
  • There are 3 Dependent variables (DV): Twitter, Facebook and Computer Test Score
  • There are already some pivots (see the tabs at the bottom) that you can explore.

To create your own:

  • Insert > Pivot Table > New Sheet > Create 
  • NB - you must first click on any cell in the data, and there can be no empty columns 

Step 4: Arrange the Variables

Basically, you use a Pivot Table to play with variables. Independent Variable (IV) is a relatively fixed characteristic; Dependent Variable (DV) is where you hypothesise you may see an effect.

So, ask questions: eg "What is the effect of Gender on Twitter usage?"

  • Independent Variable as Row
  • Dependent Variable as Column
  • COUNTA for any variable for Values

Your setup could look like this

 Pivot2.jpg

If you choose a numeric field (eg Computer Test Score), you will use AVERAGE as the summary.

Step 5: Chart

Insert a Chart to visualise your findings.

  • Select the data in the Pivot Table (do not include the Total columns, but do include the row and column labels)
  • Insert > Chart
  • For non-numeric data (eg Yes/No) select 100% Stacked Column as the Chart Type. This is useful as it converts the summary data to a percentage, so you can compare your IVs even if you have different numbers of people in each category.
  • For numeric data use a Column Chart

Related Articles