Does this answer you query:
b) Sampling instructions
The CSO will require a sample of employees (whether salaried or waged) who are on the payroll and receive wages or salaries during a particular period, called the reference month (e.g. for NES 2007, the reference month is October 2007). The size of the sample (i.e. the sampling fraction), e.g one in one (i.e. all), one in two, one in ten, one in twenty etc, will be determined by the CSO and notified to the company.
Employees are defined as all people who were on the payroll and received wages or salaries during the reference month Therefore, unpaid employees such as voluntary workers or family members, employment agency staff (where these staff remain employees of the employment agency) etc. should be excluded.
It is important that the sample is representative of all relevant employees, without exception. If, for instance, separate payrolls are maintained for persons of different grades or in different locations, then all of these payrolls should, prior to the sample selection, be combined into one overall master list and then a sample selected from this list. However, if this is not possible, then a sample should be selected from each of the different groups.
Selection procedure
The easiest way to select the sample is to:
• Assign or generate a random number for all relevant employees (i.e. all employees who are on the payroll and received a wage or salary during the reference month)
• Select those employees above a cut-off point sufficient to fulfill the sample size requirements.
Examples:
(1) Company A has 99 relevant employees in October 2007 and is requested to select a one in ten sample. This equates to 9.9 employees. Following the usual rounding policy this implies a sample of 10 employees. The payroll software assigns a random number to each of the 99 employees and chooses the top ten (by value of the random number) employees.
(2) Company B has 19 employees in October 2007 and is requested to select a one in one (or 100%) sample. This implies that all employees are included in the sample.
The key points for the sample selection is that all relevant employees are included in the sampling frame (i.e. the population from which the sample is drawn) and the selection is at random (i.e. each employee has an equal chance of being selected for the sample). For this to be possible then the generation and assignment of the random numbers must be done at random.
