This is perhaps the most extensive and time-consuming experiment I have ever undertaken in my life. Moreover, it involves a small population sample consisting of just one individual, making it highly subjective as it relies on my memory and perception of time.
4 Data Science Experiments on Life Statistics to Consider
- Initial data analysis and trendline examination.
- Correlation assessment and significance testing using Pearson and Spearman methods.
- Application of ARIMA model for data fitting.
- Decomposition using Fast Fourier Transform.
But why engage in such an endeavor? Establishing routines, like any form of self-assessment, serves various purposes for me. This endeavor commenced during a challenging phase in my life, aiming to explore how different habits could influence my mood and mental well-being. The ultimate goal was to “hack” my own brain by identifying statistically significant factors contributing to my happiness and overall health. This knowledge could not only enhance my own life but also potentially offer insights or assistance to individuals facing similar challenges.
This experiment serves as a prime example of the versatility of data science applications. While my focus was on personal tracking and journaling, the scope of data science extends far beyond. One can delve into analyzing diverse aspects of life, whether it involves monitoring a pet’s behavior, tracking local weather patterns, or assessing public transportation delays. The opportunities for personal analysis are abundant, given that data is omnipresent—it’s merely a matter of identifying and monitoring the relevant information.
Strategies for Life Statistics Tracking in Data Science
I dedicated a few minutes each day to jot down personal notes detailing my activities and the time allocated to each category.
10 Key Life Statistics to Monitor
- Sleep
- Writing
- Studying
- Sports
- Music
- Hygiene
- Languages
- Reading
- Socializing
- Mood
The variables I monitored evolved over time, with some new entries emerging, others fading away, and a few merging together. The final set of variables, for which I consistently recorded data, includes Sleep, Writing, Studying, Sports, Music, Hygiene, Languages, Reading, Socializing, and Mood—comprising ten essential aspects of my life.
Exploring Life Statistics Data
Initially, I examined individual time series data for four variables: sleep, studying, socializing, and mood. Using Microsoft Excel, I created plots illustrating the daily hours spent (in blue) alongside the five-day moving average (in red), which I deemed a reliable metric for my situation. The mood variable was rated on a scale from 0 (awful) to 10 (excellent).
The statistical data provided in the footnotes of each plot—total hours, mean, standard deviation, and relative deviation—offered valuable insights into each variable’s characteristics and trends.
In summary, my sleep patterns remained relatively stable, with occasional fluctuations. While my academic dedication fluctuated significantly, I managed to strike a balance between work and study commitments. Socializing, despite its high variability, exceeded my expectations in terms of total hours invested. Maintaining a stable mood, with minimal variability, was a positive trend that I aimed to sustain.
Correlational Analysis of Life Statistics
To delve deeper into the data and uncover potential relationships, I conducted Pearson and Spearman correlation analyses. These metrics allowed me to identify patterns such as the impact of studying on sleep duration or the association between language learning and musical pursuits.
The significance of correlation coefficients was assessed to filter out noise and identify meaningful relationships. By applying statistical tests and visualizing correlation matrices, I gained valuable insights into the interplay between different variables in my life.
Time Series Analysis of Life Statistics
Treating the data as time series enabled me to explore temporal relationships and autocorrelations among variables. Leveraging ARIMA modeling, I identified autoregressive patterns in mood and studying habits, shedding light on the dynamics of these factors over time.
Fast Fourier Transform analysis provided further insights into periodic trends and seasonality within the data. While some variables exhibited distinct frequency components, the presence of noise complicated the interpretation of results. Filtering techniques, such as moving averages, were employed to mitigate noise effects and enhance the clarity of frequency analyses.
Key Takeaways from Life Statistics Study
Despite the complexity of human behavior, systematic data collection and analysis offer valuable perspectives on personal habits and their interconnections. Through a comprehensive examination of life statistics, I uncovered nuanced relationships and trends that influenced my well-being and daily routines.
The experiment not only provided insights into my own behavior but also underscored the significance of introspection and self-awareness. By leveraging data science techniques, I gained a deeper understanding of the factors shaping my life and mood, paving the way for informed decision-making and personal growth.
In conclusion, this experiment exemplifies the power of data-driven insights in enhancing self-awareness and fostering positive lifestyle changes. By embracing data science principles and methodologies, individuals can embark on a transformative journey of self-discovery and continuous improvement.
For a visual representation of the cumulative data trends across various variables, refer to the logarithmic chart presented below:
Cumulative sum of each series, logarithmic Y axis. | Image: Pau Blasco Roca