What Did the Report Say!? Sampling 101

AS4308Most people have taken a statistics class at some point in their school career, but the reality is that after school the statistical learning often stops. When drug and alcohol testing companies like CMM Technology mention random testing, there is a good chance many workers have forgotten what the statistical method entails.

Employers, on the other hand, understand what random testing means. However, the ones who try to stay current on the latest research on drugs and alcohol are bound to run into terms that have little meaning. Understanding the basics of sampling gives employers the tools needed to intelligently respond to staff questions, whilst recognising commonly used research terms makes the information in written materials easier to assimilate.

No Head Shaking

First things first: When a company says it will randomly test employees for drugs and alcohols there are certain principles assumed to be at work.

  • There is a definable method for selection of those to be testing
  • The selection method is objective, eliminating personal biases
  • There an equal probability that each person in the defined group of people identified for testing will be selected
  • There is an independent chance a person will be selected
  • The employer will not interfere in the selection process so that it remains random

The statistical population is the group targeted for random testing. It may be a department, the entire organisation, job titles designated as high-risk positions, and so on. Stratified random sampling is a term used to divide the entire population of employees into smaller groups based on shared attributes, like high risk job duties. The smaller groups are called strata. The random sample then draws a number of people for testing that is proportional to the stratum’s size when compared to the population.

Terms, Terms, and More Terms

Random sampling can actually be applied in a variety of ways. There is random sampling, simple random sampling, cluster sampling, spatial sampling, quota sampling, and so on. However, when employers are reading research reports they will also run into terms like those used in this sentence:”Respondent driven sampling, traditional snowball sampling and street outreach methods were used to recruit heroin and amphetamine injectors from one outer-urban and two inner-urban regions of Melbourne, Australia.”1

This is a classic example of how critical information can get lost in a sea of words. Though the term “street outreach methods” is fairly obvious as to what it means, the term “snowball sampling” is unfamiliar to most. Snowball sampling is also called referral sampling or chain sampling because it is a method for identifying program participants by asking research participants for referrals. It is a non-probability sampling technique and no relationship to random sampling.

Respondent-driven sampling is a statistical method for combining snowball sampling with a mathematical model that compensates for the non-random sampling method. These terms and more are regularly used in articles discussing research on drug and alcohol addiction. Understanding the terms makes the information provided intelligible.

Catching Terms

There is general agreement that workplace data collections concerning drugs and alcohol are limited in terms of breadth and quality. It is impossible to understand the full extent of the problem until there is adequate data to assess.2 However, all the data that has been collected to date by various agencies and organisations all bolster the belief that workplace substance use is common, thus supporting the need for random alcohol and drug testing.

Terms get tossed about whilst people shake their heads in agreement as if they understand what is being discussed. They do not know what the terms mean but are unwilling to acknowledge they do not know, or they give up trying to stay informed. Mediscreen uses drug screening methodologies and procedures that reflect the best information obtained from toxicologists and industry experts. Coupled with a truly random selection of screening subjects, employers have the assurance that the right steps are being taken to minimise workplace drug and alcohol use.


  1. Danielle Horyniak, et. al. “Establishing the Melbourne injecting drug user cohort study (MIX): rationale, methods, and baseline and twelve-month follow-up results,” Harm Reduction Journal v10:11, (2013), Retrieved at
  2. Lindsay Breugem, et. al., “The Impact of Alcohol & Other Drugs in the Workplace,” Safework SA and Drug and Alcohol Services South Australia, (2006), Retrieved at
This entry was posted in Articles and tagged , , , . Bookmark the permalink.
  • Categories

  • Archives

  • Recent Posts

  • error

    Enjoy this blog? Please spread the word :)