Statistics: Change of Lecture Approach (Case Study Method)

 

1. Strategy:
1. correlate customer retention length (in months) to churn (do they leave or not)
2. correlate churn-rate with female, churn rate with account length
1. hypothesis testing also possible
3. customer value with churn rate
1. joint probability
4. correlate age with churn-rate
1. teenagers and churn (Use Anova)
2. elderly and churn
3. compare correlation coefficients (R value)
5. correlate free voice mail and churn, and high intl volume with no intl plan (conditional probability)
6. correlate total charge with churn, account length
7. Sort data by region, do hypothesis test (Anova approach: compare mean between two parameters, is the difference significance)