Data Stacking
Data Stacking
Data stacking also commonly referred to as data rotation, is a data transformation technique whereby data for a single person is stacked in multiple rows. This technique is useful in instances where a comparative analysis is required. It’s especially useful in projects where results need to be reviewed side by side for various ads, media, and/or concepts to be shown to responders.
Data Stacking
Technique
stack
verb
1. arrange (a number of things) in a pile, typically a neat one. “the books had been stacked up in three piles”
2. shuffle or arrange (a pack of cards) dishonestly so as to gain an unfair advantage. ” I know the cards are stacked”
Data stacking also commonly referred to as data rotation, is a data transformation technique whereby data for a single person is stacked in multiple rows. This technique is useful in instances where a comparative analysis is required. It’s especially useful in projects where results need to be reviewed side by side for various ads, media, and/or concepts to be shown to responders.
In survey research, when a responder is shown multiple media and subsequent questions is asked about the exposure, data is generally in flat format, one responder per row.

In above example, 3 ads were shown and after each ad, 3 questions were asked, in such a scenario, the data in its
natural format would not allow for a comparative analysis.
In such cases, the data needs to be stacked. Stacking here would create three lines of data per responder with ads shown appearing in one column and all subsequent questions in their respective columns.
Example below

Once the data has been reformatted (stacked) in the above format it can easily be plugged into any analytical/ tabulations software for performing comparative analysis.
Use In tabulations
Stacked data can be very easily fed into any tabulations software for running a comparative analysis, also referred as grid tables. Example below

How it’s done
• Stat testing of the
ads/media/segments/concepts can be
done automatically via
tabulations software,
no manual stat testing needed
• Reduces manual work in
formatting tabulated output to
allow for a side by side view
thereby improving turnaround time
• Helps in reducing the reporting
turnaround time
Disadvantages of stacked data files
Data files that have been stacked cannot be used to run analysis for questions that are outside the ad/media/concept evaluation section. As the data file will have multiple rows per responder, it would artificially inflate numbers for non-media evaluation questions. E.g. in a survey of 100 responders, where each person was shown three concepts, the stacked data file will have 300 lines of data, compared to 100 in unstacked/flat/standard file.
Contact
Toronto – Corporate Headquarters
100 Sheppard Ave East. Suite 503
Toronto, Ontario M2N 6N5
+1 (416) 549-8000
+1 (888) 224-6198
info@b3intelligence.com
Copyright © 2023 b3Intelligence, All Rights Reserved | Disclaimer | Privacy Policy