# Case Study of Store24 (a): Managing Employee Retention

Topics: Regression analysis, Statistics, Prediction interval Pages: 6 (1636 words) Published: October 4, 2011
Case Study of Store24 (A): Managing Employee Retention
Summary: The top executives of a chain of convenience stores, Store24, are attempting to come up with ways to increase employee tenure at their stores. We need to determine the relationship of employee tenure to store profits before they commit to this. They collected profits, management and crew tenure, and site location factors of 75 stores. Based on the data, we recommend that Store24 researches ways to increase employee tenure, more specifically manager tenure. Holding store location factors constant, manager tenure has the greatest impact on profits of Store24 stores. Problem and Matching Objectives: In defining the problem, we need to determine whether employee drives performance and how well employee tenure explains financial performance compared to site location factors. If we can find proof that employee tenure drives performance, we need to determine how to distribute the budget in extending retention of managers and crew. Our project is to set up a relatively reliable model to predict the corresponding change in the store performance, specifically the store profitability, with an increase in the employee tenure. Data Analysis: Based on the data we have, we plan to set up a multi-regression model with the expected profit as the depend variable Y and all other factors (Manager tenure, Crew tenure, Competitor number, Population, Visibility, Pedestrian foot traffic volume, 24 hour open or not, and located in residential or industrial area) as independent variables. From the scatter plots below, we can see that management and crew tenures drive performance.

We consider Visibility, PedCount, Res, and Hours24 factors as dummy variables, and we calculate correlation coefficient to understand the correlation between profit and the independent variables. According to the result below, we know that Mtenure, Ctenure, Population, Visibility=4, Visibility=5,Pedestrian=4, and Pedestrian=5 have positive relationship with profit, and the other factors have negative relationship. Moreover, Mtenure, Population, and Competitors have stronger relationship with profit.  | Profit| MTenure| CTenure| Pop| Comp| Visibility_3| Visibility_4| Profit| 1.00000| 0.43887| 0.25768| 0.43063| -0.33454| -0.15679| 0.07725|  | Visibility_5| PedCount_2| PedCount_3| PedCount_4| PedCount_5| Res| Hours24| Profit| 0.23444| -0.25220| -0.03351| 0.24688| 0.30023| -0.15948| -0.02569| Based on the adjusted r squared statistic from the regression statistics below, we can see that we can better predict financial performance when all factors are included rather than when only site location factors are included. Regression Model 1 (All Factors)|

Multiple R| 0.806609692|
R Square| 0.650619195|
Adjusted R Square| 0.576160990|
Standard Error| 58204.66364|
Observations| 75|
| Regression Model 2 (Site Location Factors) |
Multiple R| 0.622830552|
R Square| 0.387917896|
Adjusted R Square| 0.281046417|
Standard Error| 75806.72712|
Observations| 75|
|
Run a model with all the significant independent variables from Model 1, the result of regression Model 3 shows below: Model 3 (All Significant Factors)|
Multiple R| 0.800747891|
R Square| 0.641197186|
Adjusted R Square| 0.585134246|
Standard Error| 57585.2311|
Observations| 75|
| ANOVA| | | | | |
| df| SS| MS| F| Significance F|
Regression| 10| 3.79261E+11| 37926081027| 11.43709531| 6.07311E-11| Residual| 64| 2.12228E+11| 3316058841| | |
Total| 74| 5.91489E+11|  |  |  |
|
Compared with Model 3, Model 4 and Model 5 below respectively have large difference of adjusted r square. The incremental adjusted r-squared attributable to manager tenure and crew tenure is 0.25739 and 0.02792, respectively. Therefore, the manager tenure is more important in predicting profit at Store24. Model 4 Regression Statistics...

Please join StudyMode to read the full document