Contact Us Log In

Habit Of Living

by Project Camus

Positive Impacts of Multi-Feature Mental Health App on User Mood: A Statistical Analysis

Positive Impacts of Project Camus on User Mood: A Statistical Analysis

: Syed Adel Ali
: 06/15/2023
: Project Camus

Correspondence to

This brief presents an analysis of changes in mood scores over time among users of Project Camus. Project Camus provides features such as journaling, AI-guided therapy sessions, visualization diaries (AI-generated art), issue and goal tracking, nutrition guidance, and habit-building tools.

Our goal was to investigate if there is a discernible improvement in mood scores correlated with the usage of our app over time.

We collected mood scores of a random selection of 100 users who had at least 20 days of usage data in 2023. An average happiness score was calculated for each day. The data was subjected to an ordinary least squares (OLS) regression analysis to estimate the effect of the passage of time (day) on the average happiness score.

Our regression model provided compelling results. The number of days using the app emerged as a significant predictor of the average happiness score. Each additional day of app usage corresponds to an approximate increase of 0.28 units in the average happiness score. The low p-value associated with the day variable (p < 0.001) strongly suggests this result is not due to random chance, thus confirming statistical significance.

The model's adjusted R-squared value stood at 52.7%, implying that over half of the variation in happiness scores can be attributed to the number of days using the app. The remaining variance suggests the potential impact of other influencing factors.

The Durbin-Watson statistic of 1.874 indicated minimal autocorrelation in our data, confirming that mood scores from different days are mostly independent. However, the skewness and kurtosis values hint at potential deviations from normality in our residuals, indicating the linear model might not perfectly capture the underlying relationship between days and mood scores.

It is crucial to consider the context in which users may decide to download and use our app. Often, individuals seek mental health resources, such as our app, during particularly challenging times when they may be experiencing lower mood states. This may provide some explanation for the initial average happiness score being lower compared to subsequent scores.

Our analysis shows an uptick in the average mood scores after the first day, which may be indicative of the immediate benefits users gain from accessing the variety of resources provided by our app. However, it's essential to note that this initial low point can skew the overall trend, creating a seemingly steeper increase in mood scores over time.

The trend line graph displays the positive correlation between number of mood scores documented usage and average mood scores.

While the increase of 0.28 units in average happiness score per day may seem numerically modest, it's important to put this into context. When dealing with psychological states such as depression or anxiety, every increment in mood can have significant real-life implications for individuals. Each step toward improved mood can mean a better day, improved interactions, increased productivity, or even a brighter outlook on life. Therefore, we interpret this incremental change as a positive sign, illustrating the potential benefits that consistent use of our app can bring.

This result, however, does not stand alone but contributes to a broader picture. An important takeaway is that about 52.7% of mood improvement can be attributed to app usage over time, suggesting that our app can make a substantial contribution to users' mood management. However, the remaining variance indicates that other factors not captured in this analysis also play a significant role. This could include personal circumstances, external stressors, or the specific ways users engage with the app's different features.

The potential non-normality in our residuals, suggested by the skewness and kurtosis values, is a limitation of the present analysis. It implies that a linear model might not capture the entirety of the relationship between app usage and mood scores, suggesting the need for further investigation, possibly with more sophisticated models.

While this study provides valuable insights into the positive influence of our app on mood over time, there are several potential avenues for future research. Further exploration of how different features of our app contribute to mood changes need to be explored. User experience is multifaceted, and understanding the specific aspects that most contribute to improved mood can guide feature development and user experience enhancement. Additionally, a closer examination of the initial mood scores at sign-up could help us customize our onboarding process, delivering more personalized support right from the start.

In conclusion, this analysis provides promising evidence that Project Camus can contribute to mood improvements over time. Nonetheless, the pursuit of delivering the best possible support to our users continues, with these findings illuminating some of the ways forward.

The statistical analysis reveals a positive association between the duration of Project Camus usage and improvements in mood scores.

Although the number of days using the app has a significant influence, the variation in happiness scores also suggests the role of other factors. Further research could be beneficial in exploring the effects of individual app features like AI-guided therapy or visualization diaries.

Introduction Methods Statistical Results Initial Mood Score Visualizations Discussion Conclusion