Important Factors To Consider When Interpreting Your Survey's DataImportant Factors To Consider When Interpreting Your Survey's Data

Your survey data is only as valuable as your ability to interpret it accurately. The most insightful responses can be rendered useless (or worse, misleading) if your analysis is poor. It's worth noting that poor analysis is not limited to a lack of analytical skills on the part of the surveyor or her team. It can be due to unclear objectives, poor design and question choice, and uneven representation within the population. Each of these can damage the integrity of the responses.

This article will provide a short list of key factors to consider when interpreting and analyzing your survey results. I'll describe a few traps that ensnare novice surveyors. I'll also explain why your results need to be cleaned prior to interpreting them.

Valid Responses Cannot Save A Bad Design

One of the most common mistakes made by beginning surveyors involves using poorly-designed questions. Even if respondents' answers are entirely valid, the results might be useless given your needs. For example, suppose your company wants to find out the median annual income of your website visitors. However, rather than letting respondents fill in the amount of their annual income, you provide a scale of one of five options: $25,000, $35,000, $50,000, $75,000, and $100,000.

Even if every respondent within your population offers a valid answer, it is practically impossible to derive an accurate median figure. The problem here is not with the data. Nor is the problem with the surveyor's analytical skills. The problem is in the design; the data cannot be interpreted according to your objective (identifying the median income).

Limiting Potential Misrepresentation

Another problem that will prevent you from interpreting your survey results accurately involves a skewed representation within the population. Making matters worse, the problem often goes unnoticed, which causes the responses to be misleading. For example, suppose you are surveying your entire customer base. Customers are evenly distributed across multiple product classes, so you expect an even representation. But, what happens if response rates within those product classes are wildly different? That will skew the results. The larger the variance in response rates, the bigger the problem.

The issue is exacerbated if you and your team fail to identify the skewed representation in your survey responses. Any action taken on those skewed responses can have a negative impact on your business.

Cleaning Your Results

You will need to clean your pool of responses before interpreting and analyzing it. In some cases, respondents will fail to complete enough of a questionnaire to make any of their answers meaningful. You should discard all of their answers. In other cases, there will be duplicate answers. Those too, should be discarded.

If you're including open-ended questions in your survey, you should create a formula for categorizing responses based on thematic elements. Answers to these types of questions often provide the most insight, but they can be challenging to interpret. Categorizing them makes interpreting them easier.

Digging Beneath The Surface

Once you have designed your survey with the right questions, eliminated skewed representation, and cleaned your results, it's finally time to analyze and interpret the data. Depending on the type of analysis you need to perform, there are software applications that can help. Keep in mind those applications will be all but useless when interpreting answers to open-ended questions.

The key is to have a clearly-defined goal with regard to the information you're trying to derive from your population's responses. Otherwise, you might choose an inappropriate method for analysis. Mistakes made at this point can lead to disastrous results in terms of cost, time, and misguided effort. Whether you use frequency tables, cross tabulation, or regression analysis, your ability to accurately interpret your survey responses hinges upon the method you choose.
by Frank Lucer
References and Bibliography
SurveyGizmo is a leading provider of customer survey software, for more great ways to use surveys to enhance your business check them out online at
Rated:NR/0 Votes
Add To My Article Reading List
Add To My Article Reading List
Print Article
More Article By Frank Lucer
More Article by Frank Lucer
More Articles From Ezine Marketing
More Articles From Ezine Marketing
Related Articles and Readings
Paid Surveys - Paid Surveys - Paid Surveys - Paid Surveys By: Benoit FANTIN
Paid Surveys If you are looking for jobs you can do inside the comfort of your homes then you should search for internet jobs One of the most popular and in demand work from home is working on paid ...
Who Can Take Online Surveys? How to Get Many Surveys? By: Anand Giri
Online surveys are easy, simple, interesting and rewarding You do not need any qualification or technical skills to ...
What Kind of Person Do British Market Researchers Want? By: Don Long
Sometimes, it might seem like Americans have got paid surveys completely covered Free paid surveys are usually commissioned by companies interested in feedback from their ...
Tips For Getting Paid To Do Surveys By: Lauren Dzuris
Have you ever heard of getting paid to take surveys If not, this is one of the newest ways to make money ...
Learn the Way to Improve Your Chances of Getting Paid For Surveys By: Jason Bacot
Are you a person who doesn't fit in a 9-to-5 mold If this is the case, you still need a job to be able to make enough cash to live your life ...
The information provided in this article and/or the comments is the sole responsibility of their respective authors and does not necessarily reflect the opinion of  does not endorse any article and/or comments published by our web users unless otherwise noted. 

Member Panel

login to submit articles and more


  • » Active Categories: 419
  • » Active Articles:251893
  • » Active Authors:32244
  • » Active Members: 31843
  • » Statistics Updated:
    - Sun Apr 1st, 2018 06:13PM EST