Tanglewood Case 4

1. Tanglewood currently uses traditional predictors such as work experience, education and interview, to predict whether a candidate has the characteristics to excel in the areas of citizenship, absence, performance and promotion. These predictors are proven to predict a future employee’s promotion potential. However, only work experience can also predict performance, the rest of the selection tools do not statistically predict citizenship, absence or performance. The worst statistical predictor of performance, in the traditional method, has proven to be the interview score with a high p-value and a very low correlation.
Tanglewood would like to more accurately predict the characteristic of future employees who will excel in citizenship and absence. We have been experimenting with different predictors, with some success, in the Seattle area, for the last year. The best predictors of these characteristics have proven statistically to be the Marshfield customer service biodata questionnaire & essay, with a correlation of .22 and a p-value of <.01 for citizenship and a correlation of -.17 and a p-value of <.01, and the personality exam which measures extraversion and conscientiousness focusing on the conscientiousness with a correlation of .18 and p-value of <.01 for citizenship and a correlation of -.33 and a p-value of <.01 for absence. However, in my opinion the biodata questionnaire is not a practical tool due the cost and the fact potential employees may find the essay questions difficult to answer with possible lengthy answers and the questions responded to using the 5-point scale many applicants may feel the answers require further clarification than the 5-point scale. Another problem with biodata as a predictor is the questions are asking for life experiences which many of the applicants may not have yet experience due to their age and other limitations.
The biggest difference in the two samples, traditional and the proposed