How to Conduct a Statistical Analysis of Job Performance
- 1). Design a quantitative-based performance review. Design a survey that ranks employees on a scale of one to five for a given set of traits related to the job. Include bench markers such as productivity, quality of output, communication skills and time management.
- 2). Assign the management team the task of ranking each worker using the same performance evaluation. Ideally, one manager reviews all of the workers, which helps avoid extensive variation of the data arising from a difference in viewpoints and standards.
- 3). Compile the results. Transfer the results of the performance survey into a statistical database or spreadsheet such as Excel or Citrix. Ensure the data can be accessed pertaining to an individual's performance and how the person's data compares to the other worker: In Excel, for instance, this is achieved with the use of filters.
- 4). Use statistical measurements to assess the evaluation. Calculate the mean, mode, median and standard deviation of the data. Notate which employees exceed the average and in what categories. Perform the same for employees that fall below the average. Group employees into quartiles, such as top 25 percent and bottom 25 percent. Create a bell curve of the data and for each report, notate where in the bell curve the worker falls.
- 5). Provide a written report of the findings for the management team to evaluate. Formulate the report with an executive summary that details the department's overall strengths and weaknesses based on the findings. Then, discuss each category. Highlight the top performers and the weakest ones within the grouping. At the end of the report, provide recommendations. For instance, if you notice one worker excels in communication and working with others but has poor accuracy, a recommendation to transfer the worker may be appropriate.
Source...