This dashboard analyzes Fitbit quantified self data to explore relationships between physical activity, calorie expenditure, sedentary behavior, and sleep patterns. The dataset was obtained from a publicly available Fitbit export containing daily activity and sleep tracking data.
The data was cleaned in R using the tidyverse and lubridate packages. This included converting date variables, extracting day-of-week labels, and aggregating sleep data to the daily level. The activity and sleep datasets were then merged by user ID and date to allow direct comparison across behaviors.
Each question is paired with a visualization suited to the type of relationship being explored: - Scatterplots show relationships between continuous variables - A boxplot compares activity across days of the week - A time-series line chart evaluates activity consistency over time
This dashboard follows course principles of clarity and simplicity. Visual clutter is minimized through a clean theme and limited gridlines. Color is used consistently to highlight patterns without distraction. Separate tabs are used to organize each question, improving usability and reducing cognitive overload.
The dashboard was built in R using flexdashboard for layout, ggplot2 for visualization, plotly for interactivity, and tidyverse for data preparation.
There is a clear positive relationship between steps and calories burned, indicating that higher physical activity leads to increased energy expenditure.
Activity varies across the week, with some days showing higher median step counts than others.
The trend suggests that as sedentary time increases, calorie burn tends to decrease slightly.
Step counts fluctuate over time, showing that average activity levels vary from day to day rather than remaining stable.
The relationship between sleep and activity appears weak, suggesting that more daily activity does not necessarily lead to longer sleep.