In an effort to better understand the dynamics of police shootings, we conducted a comprehensive statistical and clustering analysis. Here’s what we found:
Kolmogorov-Smirnov Test on Age Distribution
The Kolmogorov-Smirnov test was performed to assess whether the age distribution of individuals involved in police shooting incidents follows a normal distribution.
Given the extremely low p-value, we reject the null hypothesis, indicating that the age distribution is not normal.
Cumulative Distribution Function (CDF) of Age
The Empirical CDF of age shows the cumulative probability of encountering an individual of a certain age or younger in these incidents. This visualization further confirms that the age distribution is skewed and not normally distributed.
Clustering Analysis
We employed the K-Means clustering algorithm to segment the data based on age, race, and signs of mental illness. Four clusters were identified:
- Cluster 0: Middle-aged individuals (avg age ~39) from a specific racial group, all showing signs of mental illness.
- Cluster 1: Younger individuals (avg age ~31.6) from a different racial group, none showing signs of mental illness.
- Cluster 2: Young individuals (avg age ~31.3) from another racial category, none showing signs of mental illness.
- Cluster 3: Older individuals (avg age ~54.9) from yet another racial category, very few showing signs of mental illness.
Conclusions:
- Mental illness appears to be a significant factor for incidents involving middle-aged individuals from a specific racial group (Cluster 0).
- Young individuals, irrespective of mental illness, form a significant portion of the incidents but belong to different racial groups (Clusters 1 and 2).
- Older individuals from a specific racial group are also involved but rarely show signs of mental illness (Cluster 3).
Be First to Comment