At the end of 2021, I observed around 400 days' worth of recorded data. Various analyses could be performed with such data and wanted to test the relationships among the variables. For example, I wanted to know how meditation affected my lifestyle and other variables, such as screen time or the total percentage of the day. By finding various cause-and-effect relationships among variables, I was able to derive meaningful insights about my lifestyle.

At the end of every month, an evaluation is done to analyze my lifestyle. These are all of the past assessments beginning from February to November of 2021. As shown above, one can observe improvements in the wake-up time and total percentage as time progresses.

learn more about the graph Below visualizations represent meaningful relationships within variables. By utilizing **Pearson's correlation coefficient**, various statistical relationships can be observed. When the test between two variables results in a p-value less than 0.05 (meaning there is only a 5% chance that they are statistically unrelated), it is considered statistically significant. Thus, when the p-value gets closer to 0, it demonstrates a stronger relationship between tested variables. As shown below, a red line represents a p-value of 0.05, so if a colored line is above the red line, it is most likely statistically significant.

Note that since there are numerous variables, it may seem a little messy; however, the purpose of these visualizations is to grasp a general depiction of the meaningful relationships of the data. Thus, one can merely consider them as **a map for further analysis**.

**Observations found in November**

- In the Meditation plot (row:2, column:2), it seems the meditation variable was statistically correlated with (1) Screen time, (2) Reading, (3) Rise time, (4) Total, and (5) Work done %

These results can lead to a hypothesis. For example,**since (1) Screen time correlates with meditation, it may have a negative relationship where the higher values in meditation result in lower values in Screen time.**(Correlation does not always have to have a positive relationship, where if one goes up, the other goes up as well)

- Notice that every graph has an empty correlation with the variable "Run." It is vacant because I did not run a single day in November, which resulted in no comparable data in the variable.

**Observations found in All Data (400+)**

- Similar to the plot in November, the Meditation variable (row:2, column:2) seems to correlate with (1) Multiple (numeric representation of the day's overall satisfaction), (2) Phone pickups, (3) Screen time, (4) Reading, (5) Rise time, (6) Total, (7) Tot To-do, and (8) Work done % variables.

The correlated variables are similar to November plot, but they are not the same. This change in correlation can be meaningful and can depict a lifestyle change.

- Notice there are more graphs in the All data correlation plot than the November plot. It is because of the addition of the new variables in December, which include (1) Mentality, (2) Productivity, (3) Social, (4) Tech Consumption, and (5) Overall Satisfaction.

**Proving the hypothesis: Relationship between Meditation & Screen time**

Previously in 2.1, observations found in the November correlation plot led to a hypothesis that the __meditation variable__ **negatively correlates** with the __Screen time variable__. Thus, the above code can help to evaluate such an assumption further.

As can be seen above, the first graph demonstrate the relationship between corresponding variables. It can be observed that the p-value is 0.03077 and the stat value is -0.451.

The stat value is the correlation coefficient between two given variables, which illustrates a positive correlation as the value approaches 1, and a negative correlation as it nears -1. In this case, since the calculated stat value is -0.451, Meditation and Screen time variables are negatively related as predicted.

Every month, my lifestyle is evaluated to modify or improve my daily habits. Through the correlation tests of the numerous variables and data observation, various insights emerge that help me understand my weaknesses and strengths to always strive for a better version of myself. In the future, these evaluation processes will be automated and will also provide me answers to the question of "how" I can push myself for a healthier lifestyle.

Go to Notion Automation Further Statistical Analysiscreated with

Website Builder Software .