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.
2. 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 <. 1, 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 method, regarding correlations is the size of the samples which changes the correlation significance. The reason for the difference in the sample size, of the proposed method, is it was used as a trial base, at only selected stores and carefully administered whereas the traditional method information was gathered informally.
Content validity is used to estimate how the methods are actually measuring the desired characteristics. Tanglewood has an interest in improving citizenship and absence. According to the observation above, the most effective measure of these characteristics biodata and conscientiousness. Content validation is most appropriate when the sample size is few and criterion measures are not available.
3. In order for Tanglewood to find good employment candidate it should look to the predictors and choose the ones that predict the characteristic most important to the company. Accordingly, they should continue to rely on the application blank to obtain information on education and work experience which accurately predicts performance and promotion potential.
The managers have already expressed an interest in retaining the interview as a predictor this is possible however the interview should be standardized to a structured interview with open ended questions to invite the applicant to provide the information needed. This standardization will allow for applicants answers to be compared. Another good predictor for Tanglewood to administer is the conscientiousness part of the personality test which has proven to be an accurate predictor of citizenship and absence.
These characteristic are very important to maintain the desired culture at Tanglewood. They could also consider using biodata which is a very good predictor of all identified characteristics however the applicant may find this a difficult test due to the expectation of the applicant describing life experiences which they may not have experienced due to their age or lack of life experiences.
4. Tanglewood should have success with the proposed validation procedures. Use of the application blank is very common and widely excepted practice. The changes to the interview, standardization and open ended questions, will allow for comparisons among the applicants. The conscientious test should, also, be easily accepted. The biodata test may possibly prove difficult to administer in the more rural areas due to the expectation of applicant to describe life experiences and situations the do not apply or have not been experienced. However, these proposed validation procedures should yield more qualified applicants better than the traditional predictors and allow Tanglewood to obtain more employees with the characteristic sought by the company.