Statistics paper - Essay Example

Commercial sample showed a strong negative correlation (r=0.98) while residential sample showed a strong positive correlation (r=0.97) between the bill size and the delay. The date of a payment can be predicted by the regression equations derived separately for commercial and residential samples. Late bill payments always trouble the service providers by interrupting their future planning and investments. Delay can be varied from a day to few months or sometimes a year. It is really helpful if a service provider can predict the date of the payment.

The size of the bill may have a relationship with the number of days the payment delays. It may take a long time for a customer to pay a large bill than a small bill or it could be vice versa. However if such a relationship is found the service provider can use it to predict payment dates and work accordingly. The mean values of the total and sub samples (commercial and residential) were calculated. Regression analysis (linear) was performed using Excel and SPSS statistical packages to find out possible correlations between the size of the bill and number of days the payment delayed.

Size of the bill was considered as the independent variable in this analysis. Regression analysis for total sample (figure 1) showed no correlation between bill size and delay. However regression analysis for commercial sample (figure 2) showed strong negative correlation. This explains their behavioral trend of paying large bill more quickly than that of the small bills. The residential customers expressed a totally different behavior than the commercial customers. They delay the payment as the bill size increases.

Amongst the commercial customers the bill size negatively correlated with the delay while it was positively correlated with the delay amongst the residential customers. The date of a payment can be predicted by the following regression