The document describes how to select a stratified sample from a population of 800 full-time employees, with 25% managers and 75% non-managers. The sample of 32 is drawn proportionally with 25% (8 employees) from the manager stratum of 200 and 75% (24 employees) from the non-manager stratum of 600. Random samples are selected separately from each stratum using a random number table to ensure the overall sample accurately represents the composition of the full employee population.
The document describes how to select a stratified sample from a population of 800 full-time employees, with 25% managers and 75% non-managers. The sample of 32 is drawn proportionally with 25% (8 employees) from the manager stratum of 200 and 75% (24 employees) from the non-manager stratum of 600. Random samples are selected separately from each stratum using a random number table to ensure the overall sample accurately represents the composition of the full employee population.
The document describes how to select a stratified sample from a population of 800 full-time employees, with 25% managers and 75% non-managers. The sample of 32 is drawn proportionally with 25% (8 employees) from the manager stratum of 200 and 75% (24 employees) from the non-manager stratum of 600. Random samples are selected separately from each stratum using a random number table to ensure the overall sample accurately represents the composition of the full employee population.
The following example illustrates how to select a stratified sample. Example A company wants to select a sample of 32 full-time workers from a population of 800 full-time employees in order to estimate expenditures from a company-sponsored dental plan. Of the full-time employees, 25% are managers and 75% are nonmanagerial workers. How do you select the stratified sample in order for the sample to represent the correct percentage of managers and nonmanagerial workers? Solution If you assume an 80% response rate, you need to send 40 surveys to get the nec- essary 32 responses. The frame consists of a listing of the names and e-mail addresses of all N = 800 full-time employees included in the company personnel files. Because 25% of the full-time employees are managers, you first separate the frame into two strata: a subpopulation listing of all 200 managerial-level personnel and a separate subpopulation listing of all 600 full-time nonmanagerial workers. Because the first stratum consists of a listing of 200 manag- ers, you assign three-digit code numbers from 001 to 200. Because the second stratum contains a listing of 600 nonmanagerial workers, you assign three-digit code numbers from 001 to 600. To collect a stratified sample proportional to the sizes of the strata, you select 25% of the overall sample from the first stratum and 75% of the overall sample from the second stratum. You take two separate simple random samples, each of which is based on a distinct random starting point from a table of random numbers (Table E.1). In the first sample, you select 10 managers from the listing of 200 in the first stratum, and in the second sample, you select 30 nonmanagerial workers from the listing of 600 in the second stratum. You then combine the results to reflect the composition of the entire company.