Coffee Case Essay

774 Words Nov 22nd, 2012 4 Pages
Business Report
Starbucks coffee would like to understand the factors that determine the amounts of money a customer deposits into the prepaid Starbucks debit account. In doing so a sample of 25 Starbucks customers are examined in our research. The demographics that are being cross-examined for contribution are age, days per month, cups per day, and income. The statistical tool we use to better understand the predicting power these factors play in the contribution of money deposited into the prepaid cards is a multiple variable regression. The regression, which is ran using statistical analysis on excel takes a random set of data with uncertain characteristics and finds correlations in the data and is able to measure the significance of
…show more content…
The F-test had a value of 15.3, which is meaningful when comparing to the alpha of .05%. After establishing the fact that the overall model is significant, the T-test is used to measure the significance of each individual variable.
For the first regression was the overall regression, the T-test results implied that the only variable with significant predicting power was income. Although the T-test concluded that days per month had a slight significance level above the rest, it was removed from the re-regression because it was concluded that the days per month at Starbucks did not affect how much people were putting on their debit cards. After analyzing the outcome of the model, it is conclusive that age, days per month at Starbucks, and cups per day do not play a major role in the predicting the amount a customer will prepay in to their debit card.
Once all other variable were eliminated we decided to regress our model again to see the strength of the predictive power of income alone. The simple regression performed stated that the overall predicting power of R Squared was 72.2%, the adjusted R Squared then measured at 71%. When comparing the first and second regression it is noticed that during the second regression the percentage of the adjusted R Squared slightly improved.
Starbucks is also trying to figure out the forecasting power of the number of days per month a customer

Related Documents