Ballard Integrated Managed Services, Inc. (BIMS), is experiencing a significant problem with morale and employee turnover at their Douglas Medical Center operation. At Douglas Medical Center, the employee turnover rate had risen to over 64%, which is high even for the low-skill nature of BIMS’ general work. To understand this phenomenon, Debbie Horner, the HR manager for BIMS on-site, designed a survey of 14 questions. This survey was included in the paychecks of all employees to encourage participation.
The purpose of the survey is to identify the problem driving the unusually large number of employees to stop working for BIMS. Using the survey as the study instrument, the questions were designed to first identify which employees were at risk and may already exhibit this low morale. The next set of questions attempts to categorize the employees in terms of their satisfaction with a variety of aspects of their job including pay, training, treatment and security. Finally, the remaining four questions give relevant information about the employees to help provide summary stats about the respondents.
A variety of hypotheses can be used to reframe the research question: what is driving high turnover among Douglas Medical Center Ballard Integrated Managed Services, Inc. employees? Judging by the survey questions, Ms. Horner is making some basic assumptions as to the outcome. The four possible hypotheses that can be investigated through the survey data are as follows.
• Employees are leaving because they are under-trained.
• Employees are leaving because they are poorly paid.
• Employees are leaving because they are treated unfairly personally or their division is treated unfairly.
• Employees are leaving because the company does not communicate effectively.
• Employees are leaving because they do not have job security.
A basic statistical analysis of the data should show some relative trends that may influence an individual to prefer one hypothesis to another, but the small sample size and inherent respondent bias prevent this from giving statistically significant conclusions.