6 minute read
By Guest Blogger
Posted in Customer Engagement
What is random sampling? It’s important to fully understand this concept since choosing the proper sampling method is one of the most important aspects of data science. Done properly, random sampling enables you to successfully generalize your survey results to the target group (e.g., customers or employees). To achieve this, random sampling needs two key requirements:
Randomness: an equal chance of selecting any member of the population ("probability sampling").
External selection: respondents are chosen to participate rather than deciding to take the survey themselves.
Drawing employee names from a hat is a simple example of random sampling. If every employee’s name was put in that hat and 25 names were drawn, you would have an accurate random sample. Every single employee had the same opportunity of being selected. Therefore, the 25 members of the final respondent pool should constitute an unbiased representation of the entire employee population.
Imagine a big jar of gold and green marbles. You do not need to count all the marbles to estimate what percent are gold. Grabbing a handful and counting just those will give you a reasonable estimate.
Now, to make it more difficult, these marbles are different sizes which means there is not an equal chance of selecting either type. If the green marbles are larger, you’re more likely to grab them than the gold. If that happens, then random sampling isn’t at work, and you would overestimate the percent that are green.
So, what do we mean by random sampling when dealing with people? Well, you need to make sure that there is an equal chance of selecting any individual within the population for your survey.
In the early 1990s, Bill Ablondi conducted a landmark market segmentation of mobile professionals. The survey warranted a random sampling of workers who weren't at their desks to be successful. Since we were using telephone surveys that called their office line, this was quite a challenge!
If we only called a phone number once, our survey respondents wouldn’t represent truly random sampling. Meaning we were more likely to get a worker who was less mobile by only placing one phone call. As a result, the sampling frame Bill designed required calling the same phone number on a dozen different dates to minimize the likelihood of missing the most mobile professionals.
In contrast, convenience sampling doesn’t require this type of extra effort to remain unbiased. In this method, surveyors select participants from the closest – or most “convenient” audience group.
However, random sampling doesn't mean that selection needs to be truly random. It is perfectly acceptable and common to get a list of email addresses or phone numbers, sort them, and then mark every Nth name (e.g., every other, every third, every tenth) as a candidate to be surveyed. Although there is a repeated pattern to the participant selection method, this would still constitute an unbiased representation of the overall list.
The second key part of random sampling is that the researcher chooses the potential participants. A quick comparison:
A formal restaurant places a postcard with every other bill. Diners choose whether or not to fill in and mail the postcard.
That same restaurant instead hires a pollster to stand outside the restaurant for a week and approaches every other exiting patron with a survey.
The first example suffers from self-selection bias, whereas the second offers true random sampling with external selection. Many of those people interviewed by the pollster would not have completed the postcard survey.
If you are going to be making important decisions based on your survey, you should use random sampling to ensure that the results represent your target population.
Now that we know what random sampling is, we can look at the concept in further detail. Organizations can further optimize their random sample by choosing the proper method for their needs. There are four main types of random sampling worth knowing.
Simple random sampling relies on a numbering system to choose participants from a particular list.
For example, let’s say you have a list of first and last names. You would assign each entry a random number using a database platform like Microsoft Access or Excel.
From there, you would choose the required number of entries for your survey at random. If you needed 50 survey participants, you would take the list entries with random numbers 1-50.
The first step to cluster random sampling is dividing the population into “clusters.” For cluster random sampling to be effective, each cluster needs to be representative of the overall population.
An example would be an elementary school survey. Let’s say you wanted to survey third-grade students. If there are five third-grade classes in an elementary school, choosing one of those classes to take a survey would constitute a valid cluster random sampling.
Similar to cluster random sampling, stratified random sampling requires splitting an overall population into smaller groups. Each of these smaller groups will consist of members with the same attributes. Then, random sampling will be done using a few members from each group.
Let’s look at another school survey example. Students in a particular school are already “stratified” by grade level or area of study. To achieve a stratified random sample for an overall satisfaction survey, school administrators may select a small handful of students from each department or grade level to make up the total survey sample.
Systematic random sampling is a method we referenced earlier in the article. We talked about how sampling can still be considered “random” even if it follows a pattern. Systematic random sampling does exactly that.
This method involves using a “system” of looking at a list of entries and comprising a sample group out of every Nth member. A real-life example might be conducting a survey at a mall or retail store. If you approach every 10th person with your survey, that would be a systematic random sampling of the overall mall-going population.
With all of the sampling methods available, many researchers and organizations wonder why they should go with random sampling. Here are some of the benefits using random sampling can offer you:
Lower margin of error
Easier to form sample groups
Eliminates researcher bias
Findings can be assumed of the entire population
Simple research and data collection
Random sampling aims to collect pertinent data without any unconscious bias. The simple and efficient nature of random sampling makes for easy knowledge management. Whether you’re a veteran researcher or your first day on the job, it’s hard to compromise random sampling data.
Of course, as with any data methodology, random sampling does have some drawbacks. Here’s a brief rundown of things to watch out for when implementing random sampling:
Random sampling occurs in a “vacuum.” No additional knowledge is considered.
It requires population grouping to be effective.
A large sample size is necessary.
Can be costlier than other data collection methods.
Can be time-consuming. Although data collection is simple, it can take a long time since it’s done on the individual level.
Now that you're familiar with what random sampling is, we hope you see some ways to use it in your organization. Sampling employees at random is a great way to get a handle on the overall sentiment of a company. Those random sampling results can then be used to foster first-class workforce optimization. The same sampling method can be used to gather customer-facing data. Demographic questions can be asked of your customer base to help create more-informed ideal customer avatars.
For more on this and other ways you can use random sampling to improve the way you do business, contact Verint Systems today.
Did you like this story?
Subscribe for more Customer Engagement insights