What is a Crosstab, and do I need one?

The questionnaire has been written, responses collected, and now it’s time to analyze the results of your research project. While it can be tempting to look at your data among all your respondents, looking at distinct subgroups of respondents (for example Millennials, or Females) can provide a more complete picture and add depth to the narrative your data is trying to tell.

It is possible to look at subgroups manually using raw data, but this is a tedious process and doesn’t provide any statistical testing to confirm the differences you may see. The most effective way to compare groups within your data is by using crosstabs. 

Each crosstab allows you to see your study’s data among up to 15 sub groups (for example Male and Female would be 2 of the 15 groups). The subgroups used will vary per study and should be chosen strategically to reflect actionable groups within the data. Often demographic filters such as gender, age, or geographic region are used. If sample size allows these can also be layered with other questions to create more specific groups like your target customer, or purchasers of a specific brand or customer segments based on a segmentation you may have developed.

In addition to seeing how each group has answered the questions in your study, the crosstab will provide statistical testing between these groups. This is especially helpful when analyzing the differences between groups. When a significant difference is noted, you can confidently draw conclusions about the disparity between these groups vs a simple directional difference that may be due to chance. 

Significant differences are especially important when conducting certain studies where you are comparing different visual stimuli to determine the most effective i.e like an Ad test, Concept test, Package test, Logo test etc. Using crosstabs we can compare the different stimuli to each other and determine significant differences to help us evaluate which one is the most effective.

There are very few cases when a crosstab would not be recommended to be included in the scope of a research project. Studies with small sample sizes or simple analysis only requiring the total dataset do not necesitate crosstabs. But more often than not, crosstabs are the best way to get a well rounded view of your data.  Crosstabs allow you to examine your data among different respondent subgroups and detect significant differences between these groups. This extra step in your analysis can give you the confidence that you are making the best decisions for your business. 

Have more questions about crosstabs or other research topics? Contact one of our researchers at hello@lab42.com.