Data cleaning is the process of identifying and removing poor data from a set. This is a standard practice in research and let's face it, you're basically one of the pros!
How do I identify poor data?
Always follow the golden rule 🥇 - Assess the quality of each participant's individual data based on the elements outlined below. If more than 30% of your data qualifies as poor, it's time to say goodbye!
Flatlining
If your graph has flat, horizontal lines that do not follow sine-style waves with peaks and valleys (∿), then it is possible the participating individual experienced data connectivity problems or device placement issues.
Missing Data
If your graph has sections of data missing, it is likely there was some sort of connectivity issue. This issue could have come from a poor wifi connection, an extended distance between the smartwatch and paired phone, or accidental closing of the app while the experience was in session.
How do I remove poor data?
Partial or poor participant data sets can be removed by first navigating to the “Second-by-Second” tab. Toggle off “All Participants,” and toggle on the participant’s data you need to remove. Once their data is displayed, click on the three dots to the right of their graphed data. Once you click this, select “Remove”. The data from this participant will be removed from the aggregate score for the overall experience.
There you go, it's that easy! You have successfully cleaned your data!
