This is my first case study using spreadsheet and R as a graduate project from the Google Data Analytics Certificate course. This is also my first notebook at Kaggle, I had a lot of fun exploring the platform. At first I didn't know where to start. I went through a few notebooks created by other users to get inspiration and direction, and started to work on the data with spreadsheet, and then visualize the cleaned data with R. This exercise encourages me to do some further study in data analytics. I would love to do more case studies on my own, the next question would be how to find the perfect dataset to work on.
Business problems
- How people are already using their smart device?
- What are some trends in smart device usage?
- How could these trends apply to Bellabeat customers and help influence Bellabeat marketing strategy?
Cleaning process and tools
- Using spreadsheet SUM function to add up VeryActiveMinutes, FairlyActiveMinutes, LightlyActiveMinutes, SedentaryMinutes columns all together to create a new column named "TotalMinutes".
- Using spreadsheet COUNTUNIQUE function to check how many users participated in the data collection, which is 33.
- Using spreadsheet pivot table to group by Id, in order to get average number of TotalSteps, VeryActiveMinutes, FairlyActiveMinutes, LightlyActiveMinutes, SedentaryMinutes, TotalMinutes, Calories for each user. Decrease decimal points of each column.
- Export pivot table to Posit Cloud, using R to generate plots to better understand the relationship between the values.
Analysis and visualizations using R
Key findings
- From the first histogram we can find out that not all users use their Fitbit 24/7.
- The scatterplot "Relation Between Calories Burned And Total Time Spent With Fitbit" shows that there is no positive relationship between the average time spent with Fitbit per day per person with the average calories burned per day per person. This indicates that users mostly use Fitbit to track their everyday health data.
- The scatterplot "Relation Between Sedentary Minutes And Total Minutes" shows positive relationship between the average sedentary time per day per person with the average time spent with Fitbit per day per person. This indicates that most users spend a majority of time sitting while using Fitbit.
Recommendations
- This analysis was based on a dataset of 33 Fitbit users, and we do not know the users' gender or age, which is not sufficient for the study. Since women are the target clients of Bellabeat, it is best to collect more data from women users of any health tracker in order to conduct further analysis.
- Bellabeat offers stylish health trackers designed for women, using heart rate data and users' input data to track period and provide lifestyle suggestions based on users' menstrual cycle could be a good way to separate Bellabeat from other health trackers on the market.
- Design a system to encourage users to move or stand up once in a while to increase users' physical activity and improve health.
Comments
Post a Comment