Question

Statistical Error: Regression to the mean

Definition: In any event where luck or chance is involved, extreme outcomes are followed by more moderate ones.

1. Provide a Human Resource decision that has to be made as an example to regression to the mean.

2. Errors typically occur because the data used to make the decision is flawed in some way. What flawed data could lead to the error for this decision?

3. Think about what data could be used instead (to avoid the error: regression to the mean)?

4. What parameter or statistic will you use to represent the dataset?

5. How could this help avoid the error?

Answer 1

1) The human resource decision that has to be which is perfect example of regression to mean is the hiring process wherein the company hires an employee based on the recommendation of an existing employee. If for instance, the first time the existing employee was able to suggest a capable employee, however, in the subsequent suggestions the quality of man power suggested by him/ her deteriorated. This is a perfect example of regression to the mean.

2) The entire hiring process in the decision mentioned above depends on the subjective opinion of the existing employee. Maybe in the first time he/she suggested a quality person but later poor man power due to personal gains or even lack of judgement.

3) The data that could otherwise be used to avoid the error could represent the qualifications of the candidate. Further in house exams can be conducted for the suggested candidate to see how fit he/she is for the role.

4) The parameter that will be used to represent the dataset will be educational qualification and recommendation.

5) This dataset will incorporate the faith in good suggestions from the existing employees and the educational qualifications that will be best fit for the current job. Consequently, with a highly educated recommended employee the company would hit a jackpot in recruitment.