The project will focus on analyzing the benefit package options for the employees of a mid-sized San Diego retailer, to determine if a correlation can be made between benefit choices and retention. The top management of the organization has realized that it is very expensive to hire and train new employees every year. Therefore, they would like to find out if the benefit package offered is a significant factor in improving employee retention and thus overall organizational performance.

In the article, "The Balanced Scorecard--Measures that Drive Performance" the authors Kaplan and Norton (1992) talked about several financial and operational measurement systems that can clearly identify an organization’s future performance. An organization's measurement system may also affect the behavior of managers and employees. The authors discuss a balanced scorecard for senior management that helps them to gain critical information from all perspectives (i.e., customer, internal, innovation, learning and financial) without suboptimization or information overload. The success of these perspectives can be measured by the following four parameters:

Goals: What needs to be achieved to become successful?

Measures: What parameters should be used to become successful?

Targets: What quantitative value should be used to determine success of the measure?

Initiatives: What should be done to meet goals?

The goal of the retail franchisee is to become more financially and operationally successful through increased employee retention. The top management believes that by retaining the employees longer, they can save cost of new employee training, and improve efficiency and customer service.

The management has realized that it is very difficult to identify or pinpoint the measures that are directly related to employee retention, since outside influences such as the economy and the job market in a region will play a strong role in an employee’s job choices. However, the company seeks to control internal factors as much as reasonably possible, and would therefore like to determine whether increasing an employee’s flexibility of choice will affect their internal job satisfaction. Although it seems apparent that giving the employee’s greater flexibility and choice would increase job satisfaction, the company needs to determine how to tailor their options to maximize employee needs within budgetary constraints.

The general assumption is employees who can customize their benefits will be more satisfied and therefore stay longer in the company. However, the level of impact these choices create among employees needs to be determined. In order to do that it is important to identify the correlation of employees’ age, sex, education level, years of service, and position in the corporate level with their usage of, and choices for, employee benefits.

If it is possible to establish a strong positive correlation between the above factors with the preference for types of benefit plans then the management will have a good decision making tool that they can use to improve employee retention level. It may be decided to offer different types of employees different options depending on their needs, or to offer a flat dollar amount per employee (or a percentage rate based on salary) that can be used in any or all of several types of benefit plans.

Identifying Operational Variables (Independent & dependent)

The proposed independent operational variables are employees’ sex, age, education level, years of service, and position within the company. The dependent variable will be the choices made in the benefit options

Based on the research problem the management has decided to adopt a combination of Ex Post Facto research design (Davis & Utts, 2002, p. 116). In order to determine the exact variables that affect benefit preference, the management has decided to conduct an experience survey, on a sample of the employees. They have also decided to gather information from external sources like literature search on similar research.

If the experience survey and literature search both support for the proposed independent variables, the management would like to proceed with the next stage of the research, which is data collection. The management believes that the data collection process will be fairly simple. Much of the background data can be collected from secondary sources, such as human resource records. The remaining information can be gathered through a preference survey.

Before collecting a sample it is very important to carefully define the population. Since this company is relatively small (about 300 employees), it has been decided that the entire population will be surveyed. This will include employees that are not currently eligible for specific benefits, with the intention that gathering their information as well will give the company information as to the pool of employees that may desire to move to full-time permanent employment. However, the first question of the survey will include an exclusionary question such that if the employee does not wish to ever be full-time, or they are under 18 years of age, (such as a high-school student), their survey will be eliminated from the population for scoring purposes.

If the company was larger, then a Stratified Random Sampling (Davis & Utts, 2002, p. 139) would be appropriate for this study. The sampling would then include four homogenous groups of employees – Store-level Management, Store-level Clerks, Administrative Management and Administrative Clerks. Presuming the population was 30,000 employees rather than 300, the distribution would be handled as such:

Examples of the strata or groups are distributed as follows:

Store-level Management: 2,500

Store-level Clerk: 21,500

Administrative Management: 500

Administrative Clerk: 5,500

According to Arsham (2002), if there are k number of groups and N_i denote the size of group i then N denote the overall population size. Further, if n denote the sample size, then the stratified sample is:

n_i = n(N_i/N)

where, i = 1, 2, . . . . , k.

For this study, let Store-level Management = 1, Store-level Clerk = 2, Administrative Management = 3 and Administrative Clerk = 4. Therefore, N₁ = 2,500, N₂ = 21,500, N₃ = 500, N₄ = 5,500, moreover, N = 30,000. Let the sample size n = 100 then the sample that should be collected from each group will be distributed as follows:

n₁ = 100 * (2,500/30,000) = 8.33 (8)

n₂ = 100 * (21,500/30,000) = 71.67 (72)

n₁ = 100 * (500/30,000) = 1.67 (2)

n₁ = 100 * (5,500/30,000) = 18.33 (18)

After determining the sample size for each group, the final sampling should be done randomly, by using random number table generated by computer.

References

Arsham, H. (2002). Chapter 2: Descriptive Sampling Data Analysis. Statistical Thinking for Managerial Decision Making. Retrieved November 30, 2003, from http://ubmail.ubalt.edu/~harsham/Business-stat/opre504.htm#rwhyrssm

Davis, D., & Utts, J. (2002). Statistics and Research Methods for Managerial Decision [University of Phoenix Custom Edition]. Stamford: Thomson Learning Inc.

Kaplan, R. S., & Norton, D. P. (1992). Balanced Scorecard--Measures that Drive Performance. Harvard Business Review, 70(1), 71-80.

Copyright (c) 2003, Mahmud Zaman
www.zamanfam.com
All rights reserved