Types of Feedback

How To Find Sampling Error In Our Surveys

A sampling error represents a deviation in sampled audience versus the overall population. These errors are common - but there are ways we can avoid them.

July 8, 2021

How To Find Sampling Error In Our Surveys

A sampling error represents a deviation in sampled audience versus the overall population. These errors are common - but there are ways we can avoid them.


There is no such thing as a perfect, flawless survey. Sampling error are always going to represent a threat to the integrity and the validity of our research surveys; as such, it is vital that we remain aware of the different kinds of sampling error, and how they affect our data.

In this article, we will take a look at the most common types of sampling error, the ways we control our sampling error, and what reducing sampling error mean for the results of our market research surveys.

What Is A Sampling Error?

A sampling error occurs whenever there is a deviation in our sampled audience compared to the overall population being tested. In layman's terms, this means we’ve made a mistake somewhere during our surveying process - or that we’ve either over-represented or under-represented some variable in our test.

If we wanted to survey every person in America to find out what the most popular flavor of bubble-gum is, it could not be done with 100% accuracy in the results. Even if we hired pollsters in every city to count the results and assumed everyone answered honestly, there will be errors in the data we’re being given.

Although we can never perfectly extrapolate the results of our audience sampling to the greater population, there are ways that we can reduce the impact these sampling error have on the validity of our data.

Before we take a look at how we can reduce sampling error in our surveys, we must first familiarize ourselves with the most common forms of sampling error. Learning how sampling error can affect our data in different ways will allow us to gain a better understanding of how we can diminish their impact.

Types of Sampling Error

Whether you’re sampling random off people off the street - or you’re pulling from a targeted demographic segment over the internet - these errors are going to occur often in your surveys.

The reason we cannot avoid these errors is because we can almost never sample the entirety of a population; because this is true, we are always extrapolating our findings about our sampled audiences onto the greater population.

After we look at the most common forms of sampling error, we'll take a look at how we can use confidence levels and the standard deviation of our sample size to determine how closely our audience matches the true population.

Here are the most common types of sampling error that occur in research surveys, and how they occur throughout the research process:


Sampling Error

Sampling error in its most basic form occur when there is a variation between the number of samples we measure in an audience vs. what the actual ratios look like in an overall population.

For an example of what a sampling error might look like, let’s assume we survey 100 baseball players about their batting stance. We want to know what the population of left-handed batters looks like compared to right-handed batters. We pick a sample size of 100 batters, and find out that 47 of them are left-handed; an impressive number, given that only about 10% of people on average are left-handed.

If we were to change our sample size to 1000 baseball players, we would find that our percentage of left-handed batters is much closer to 10%; this is because left-handed batters were over-represented in our initial sampling of batters.

Controlling for Sampling Errors

We can control for sampling error in a number of ways.

  • We can pay careful consideration to the design of our sampling; taking special care to ensure that our final sampled audience accurately represents the population we’re testing
  • We can increase the size of our sample audiences, so that we may reduce the chances that certain elements are over-represented in our data
  • We can divide our audiences up into further segments, and run surveys that test specific demographics as opposed to a larger, less focused group of people

Population Specification Error

A population specific error occurs when the researcher chooses the wrong audience to survey. Demographics are extremely important when it comes to surveys, as researchers are always looking to sample audiences that are as representative of the overall target population as they can be.

If a wine retailer decides to survey 100 teenagers about their average wine consumption habits, they would be creating a population specific error. More innocently, a sampling error might occur if the researcher has only a partially complete understanding of their audiences.

For example, we might see a manufacturer of cotton swabs interview audiences about their cotton swab consumption as it relates to hobby, and arts & crafts. This survey audience would be leaving out a key demographic of people who - against the manufacturer’s wishes (and we all do this) - use the swabs for ear-care and personal hygiene.

Controlling for Population Specification Errors

We control for population specification errors by conducting preliminary research of our audiences, and identifying our target demographics. Before we spend our marketing budgets researching untested audiences, we must first have a clear understanding of who our target markets are and what they value.


Selection Error

A selection error happens when pollsters opt-in to surveys themselves; the issue being that only respondents who are interested will participate. Because of this, our sampling will have left out what is likely the vast majority of potential respondents.

The US Census goes to great lengths to avoid selection error by motivating citizens to respond and by making the process extremely accessible. Not only can the census be completed online - the department employees citizens to go door to door throughout neighborhoods and communities.

Controlling for Selection Errors

To control for selection errors, we want to be prepping our audiences before the survey and following up with them after the survey is finished.

By initiating contact with our audiences early on in the surveying process, we can make more of an effort to drive their engagement with our survey. Offering survey incentives like raffles or coupons is one way we can promote a greater response rate in our surveys. Initiating surveys at the appropriate time is also crucial; for topics like user-experience surveys, retailers make sure to introduce shoppers to the survey immediately after a purchase has been made.



If your goal is to survey as wide and diverse a sample group as you can get, then you’ll need to make accommodations. Always give your audiences ample time to reply to surveys.

Sample Frame Error

A sample frame error happens when researchers target an element of a population incorrectly during their sampling.

An example of error in sample framing occurs when researchers look at the number of guns in America. There are more guns in the United States than there are citizens; but this does not mean that every citizen owns one, let alone multiple guns. There are private collectors that own dozens or perhaps hundreds of firearms. There are also police forces and military bases where firearms are used for professional purposes, and are out of the hands of private citizens.

Controlling for Sample Frame Errors

We can control for sample frame errors by conducting thorough research of our audiences beforehand. We will need to explore the uncontrolled variables that exist in our study, and attempt to frame our survey accordingly.

We can look at existing surveys and studies done by other researchers, and try to mimic their framing of the audiences we’re looking to sample.

Non-Response Error

A non-response error happens when the respondents that do participate in our survey differ greatly from those who do not.

In some cases, we may fail to consider a certain segment of our audience when disseminating our surveys. Other times, respondents refuse to participate for one reason or another. We cannot make our surveys compulsory, so it is up to us to try to be as effective as we can when capturing audience interest in responding to our survey.

Controlling for Non-Response Errors

We can apply the same practices we use to control for selection errors to manage the impact of non-response errors on our surveys.

Offering incentives and removing barriers that make it harder for participants from responding to your survey is the best way to motivate our audiences to engage with our surveys thoroughly and to bolster our survey sample size.

Key Takeaways

Sampling error represent an ever present threat to the validity of our survey results. By remaining aware of how different types of sampling error can occur in our work, we can design our surveys in such a way that promotes better data for our analysis.

If we have an understanding of what our sample size is and the standard deviation of our population, then we can calculate our sampling error.

Sampling Error Formula

Z x (σ /√n)

Z = Z Score (Based off of the desired confidence interval)
σ = Population standard deviation
n = Sample size

The smaller the value of our sampling error, the closer our sample is to being truly representative of the actual total population. Researchers typically use a Z-score of 1.96 - a score that correlates to %95 confidence in the accuracy of our results.

Having a low sampling error means we are operating within a lower margin of error; thereby generating the results we need to gain a clearer understanding of the wants, needs, and desires of the entire population we're targeting.

What Helpfull Does Best

Helpfull is the premiere surveying software for anyone looking to get the most out of their marketing efforts. With a vast variety of different question types to choose from, Helpfull surveys can be generated and delivered to thousands of panelists in minutes.


These are just some of the features that make Helpfull a must-have tool for those looking for quality consumer feedback:

  • No Locked-In Price Plan
  • Affordable Surveying Options
  • Create Simple Surveys with Pre-Set Questions, or Design Your Own
  • No Extra Cost For Multiple Questions in a Single Survey
  • Wide Variety of Polling Demographics to Choose From
  • Get Responses & Feedback in Real Time
  • Save Your Favorite Answers for Easy Reference
  • Aesthetically Pleasing UI and Easy-To-Use Interface
  • Easy to Access Prior Surveys



With Helpfull’s open-ended survey design, users can design visually-striking, multi-question surveys for any occasion!

A simple survey could be all that stands between your company and the renewed success of it’s online marketing efforts.

An intuitive user-interface, coupled with the ability to gather hundreds of consumer responses in just minutes, are just a few of the features that make Helpfull the ultimate tool for any artist, designer, marketer, or inquisitive spirit.

Refine your marketing - sign up with Helpfull today, and get surveying within minutes!



Types of Feedback

Survey Data Collection Methods For Calculating Online Survey Audience Size

Choosing our target audience demographics is an integral component of the surveying process. Learn how to calculate perfect size for your survey audience.

Read story
Poll vs Survey: Definitions, Differences, and Where They Fit In Your Marketing Research

A poll consists of only one question; surveys are made up of multiple questions. Learn the differences and how each tool is used for information gathering.

Read story
10 Tips To Create The Best Survey Email

Whenever you’re looking to send out a survey email, it’s important to get it right. Here are the 10 best tips for creating effective survey emails.

Read story
Likert Scale Survey Questions - How To Use Them, And What They Can Do For Your Business

Likert scale surveys are a reliable way to collect valuable customer data. Learn how we can use the scale to generate feedback for our business insights.

Read story