Analyze survey data with pivot tables

The analysis of survey results typically involves several steps, including data cleaning, summarizing the data using descriptive statistics, and using statistical techniques to test hypotheses and explore relationships:

• calculating measures such as mean, median, mode, and standard deviation to summarize the data (all these measures are supported in SeekTable).
• involves examining the relationship between two or more variables by grouping the data into tables that show the distribution of responses across different categories (pivot tables are perfect for that).

SeekTable is suitable for survey or questionnaire results analysis: it can summarize data by answers, calculate percentages, cross-tabulate and compare subgroups. SeekTable can visualize results with various charts.

You can check these demos to get some understanding about what survery reports may be created with SeekTable:

• American Time Use Survey (ATUS)
• Stack Overflow Developer Survey 2019

How to perform survey results analysis in SeekTable:

1. Export your survey data to CSV. No problems if you have multiple choice questions: SeekTable can handle them correctly.

2. Upload CSV file into your SeekTable account (up to 50Mb zipped).

3. Configure additional measures / remove excessive ones from the cube.

4. Create survey reports by categories, questions and answers.

If survey data is too large for upload you can import it into SQL database (say, PostgreSQL) and configure it as a live data source in SeekTable.

Configure median measure

In survey analysis it is often to enough to get just mean (average) values; another important measure is median (or 50% quantile) which reflects a middle value.

You can configure median (or any quantile) measure in the following way:

1. Go to cube configuration form ("Edit Configuration" on the cube details view)

2. Add new measure

3. Choose "Quantile" type

4. Add a 1st parameter with name of CSV number column

5. Add a 2nd parameter if you want to specify quantile value (between 0.0 and 1.0). If this parameter is not specified 0.5-quantile is calculated (which is median).

Unpivoting survey data

It might be possible that your survey tool exports questions as a separate columns in the CSV, and if you want to use the list of questions as a separate dimension you'll need to 'unpivot' your survey export result.

Fortunately you don't need to do that manually in Excel as SeekTable can perform CSV 'unpivot' operation on-the-fly (only when needed), and multiple choice questions will be processed correctly.