How Does Exercise Impact Productivity?

Introduction

I love data. I love data so much that I was gifted a pair of socks that say ‘I love spreadsheets’ that I am currently wearing as I write this.

The reason I love data is that it gives clarity into areas of life that might otherwise be confusing or involve a lot of guesswork. 

A byproduct of my love for data is that I track almost everything I can, and I’ll usually look at this data in an attempt to make improvements to my life. 

One such improvement is being more productive with my free time. 

I have a lot of hobby projects like this website and different coding projects like my AI chatbot and buzzword detecting Chrome extension. Working on these things gives me a sense of happiness and fulfillment, so it is in my best interest to find ways that I can continue to pursue these hobbies.

A while ago, I examined the impact of sleep on my productivity with some moderate success, but I anecdotally noticed that the more I was running, the more productive I seemed to be, so I decided to apply the same scientific method that I did for sleep to my running habits.

Methodology

I downloaded my Strava running data from 2024 as well as the data from Clockify for 2024, which I use to measure my productivity on coding and writing tasks. I then stripped out any hiking data from the Strava data (as this adversely impacts the ‘Seconds of exercising’ metric, due to hikes being inordinately longer than running).

I then compared the correlation betweens seconds run and seconds of productivity with a Pearson's correlation coefficient, and a Point-Biserial Correlation test to examine the relationship between the Boolean value of ‘whether I exercised that day’ and seconds of productivity.

Results

In terms of seconds of exercise vs seconds of productivity, results of the Pearson correlation indicated that there is a significant small positive relationship between Seconds worked and Seconds exercised, (r(357) = .127, p = .016).

This implies that longer runs are slightly better for boosting productivity than shorter runs.

In terms of whether the binary value of ‘did I exercise’ impacts productivity, overall there was a statistically significant relationship between exercising on a given day and productivity (r(357) = .134, p = .01).

A positive relationship between exercise and productivity was seen both on weekends (r(102) = .151 p = .129) and weekdays (r(257) = .180, p = .003), however only the weekday correlation was statistically significant at p < .05.

Learnings

It appears that there is a significant positive correlation between exercise and productivity. To maintain and increase productivity, exercise must be taken. Longer runs are better, but any form of running will boost productivity.

Interestingly, this relationship is more statistically significant during the week, highlighting the need for rigidity and discipline in maintaining a midweek training schedule for the sake of boosting productivity.

Limitations

The first limitation is that this test only measures ‘exercise’ as running. There are multiple different forms of exercise, and their impact on productivity should also be measured.

Another limitation is that what this test might be picking up is an overall level of mental wellbeing and organisation that is affecting both variables, rather than exercise directly causing the increase in productivity. This is the classic “correlation does not imply causation” fallacy. However, it could be argued that exercise is a core building block (along with sleep) of a healthy schedule so is indicative of a healthy mindset. I will look into a way I can measure overall healthiness and organisation and it’s impact on productivity for a future post.