Sample size

Everything You Needed to Know About How to Calculate Sample Size

by Ivana Taylor  

 

 

Online survey software is a vital tool for businesses that are interested in receiving helpful feedback from their customers, potential clients and other individuals, and creating a great web survey can be critical to the success of any business.You can create the best survey in the world with high-quality survey software, however, but if your survey sample size is too large or too small, or is directed toward the wrong audience, the results may be inaccurate. Maximize the accuracy of your next survey with a thorough understanding of how to calculate sample size.

How to Calculate Sample Size

Before you decide on the exact number of individuals to include in your survey, there are a few factors that you should take into consideration. Here is an article that explains these factors in depth and how to determine the best sample size for your survey, but you will find a basic rundown of the process below.

Factors to consider include:

• Objective of the Survey:  If you are interested in using the results of your survey to make a big decision or you plan to project the results on a larger population, the sample size is critical. If, however, you are merely using the results to get a feel for a specific group’s preferences the size of your sample may not be as important.
• Diversity: How diverse is the population of individuals or groups you wish to survey? While similar respondents can typically be surveyed using a smaller sample size, surveys that require a more diverse population will receive more accurate results with a larger sample size.
• Accuracy: How critical is the accuracy of your survey results? Sample size and targeted audience can play a significant role in the accuracy of your survey. For example, if you are planning to use the results of your survey to make a large investment or make a decision that will affect a large population, the accuracy of your survey is likely extremely important and a larger sample size is essential.
• Rate of Response:  In order to target the most effective sample size for your survey, you should first consider the number of respondents you will need as well as the willingness of your audience to participate in your survey. If you estimate that only 50% of your audience will respond, and you need at least 100 respondents for example, you will likely need to target a minimum of 200 individuals.
• Audience: Another factor that should be considered when it comes to how to calculate sample size is the audience itself. You wouldn’t want to target homeowners when surveying about the quality of local apartment life, for example. If you send out a survey that is targeted toward a specific audience that is qualified to provide accurate answers, your sample size may not need to be as large as it would be if you decided to survey random individuals in the area.

Once you have reviewed the above factors, it’s time to calculate your sample size. This sample size calculator is an excellent free tool that is designed to help users determine the most effective sample size based on the population to be studied and the confidence level you wish to achieve. When combined with the data from the factors above, the calculator can help you pinpoint a much more accurate estimate of the size of your sample.

Online Surveys

Simple Steps to Successful Surveys

by Zontziry Johnson 

 

 

When it comes to creating a new survey, where do you like to start? Do you start with a laundry list of questions and then filter from there to the end result? Do you start with a potential list of respondents, and then write the survey questions based on the respondents? Now…where should you start?

Start with purpose

I’ve been in many meetings where the survey project started with the end result (we want a bunch of slides that show us pie charts and slices of the data based on these particular demographics), the audience (how about we start with our customers, then open it up from there), or even the laundry list of questions that the customer wanted to have answered from the survey (we’d like to get a net promoter score to start a feedback loop, but we also want to get satisfaction ratings for customer service, the overall customer experience – how about we ask what color of the product they’ve liked best?). I’ve also sat in meetings where the approach was simply, “We don’t really know what we want to ask, we just want to do a survey.” 

The best place to start when approaching any kind of survey project is with the purpose of the survey itself. Are you (or your client) wanting to measure how your brand rates against other brands? Are you wanting to get a baseline customer satisfaction score? How about trying to decide what features are most important to your potential customers?

It’s all right to be thinking about your respondent audience, your list of questions, and even what you hope the responses will show. Just be certain you come back to the actual intent of your survey, or, as I like to put it, the one burning question you want to have answered from your survey. 

Follow with relevant questions

To me, creating the questions for a survey once you’ve defined your purpose can still be difficult to do because there is always more information that you want to gather from your respondent audience. It doesn’t matter how well a research report has been written, I find myself always walking away with more questions that I wish had been asked in the survey, or hoping there is going to be a follow-up survey to dive deeper into a particular aspect of the report that I found compelling, but just want more data. 

The trick is not to follow that train of thought in the early stages of writing the survey.

How can you avoid that pitfall? Go back to your original purpose for the survey. This can be difficult to do. As much as I love to teach that you come up with that primary purpose and stick to it, I know it’s difficult. For example, you could state that your one burning question is: “Do our customers like us?” But there are so many questions you could ask, from asking about the product offering (is it enough, is it too much, is it too little, is it too difficult to navigate, is it too…) to the entire customer experience (were you greeted when you entered the store, were you asked if you needed help, could you find someone to answer your questions when you needed someone), answering the seemingly simple question, “Do our customers like us?” can end up being a 45-minute survey experience! 

The remedy? Identify things like key performance indicators, and possible narrow down the original question you came up with in step one. For the example, “Do our customers like us?” you could easily narrow that down to something like, “What do our customers like most about us?” or even, “How do our customers like us compared to our competition?”

Personalize the experience

Once you have decided on the survey questions, it’s time for you to plug it in, add any needed logic, and talk about personalization. If all you have is a first name of a potential respondent, it still is worth your while to include that in the survey invitation at a minimum. Everyone expects personalization in any kind of customer interaction, from a newsletter addressed to them to a survey invitation with their name on it. 

Top it off with customization

Customizing your survey should be the very last thing you do. Make sure that everything else is working before you start to customize the survey experience. Customizing should be the icing on the cake, rather than a primary focus. Content first, then you can spend time making it look pretty. This reduces the risk of having a really nice-looking, but poorly-functioning survey. It also reduces the risk of finding that the error in your survey was from the programming, and not the CSS formatting you’d just applied. This same rule applies to many other types of content, but is especially true for a survey. In my own opinion, respondents are far more interested in a survey that is straightforward and doesn’t feel like it’s wasting their time than they are in a survey that looks cool but doesn’t make sense or is terribly long and boring.

Zontziry (Z) Johnson is the Community Manager for QuestionPro. With 9+ years experience in the marketing research industry, she is continually enthralled with the ever-changing possibilities behind how to ask people what they think.