Margin of error can predict how survey findings will be distributed relative to the true population being studied. A margin of error of plus or minus 4 percentage points, at a 95% confidence level, means that if the same survey was fielded 100 times, the results will be within 4 percentage points of the true population value, 95 of those times. Survey data may not perfectly match the population being studied, but margin of error allows researchers to feel confident about a certain percentage of the time, within a range above or below the ideal estimate, represented by a margin that is believed to be least in error.

Computing Margin of Error

It is important to remember that surveying is a method of balancing and the margin of error is a way of evaluating how successful your survey is. If the margin of error presents as a smaller percentage, the more certainty you place in the results. The greater the margin of error, thehigher the probability for the data to inaccurately embody the opinions of the total population.Picture a scenario in which you are trying to decide between Concept A and Concept B for a new product package. Now, let’s assume your target market consists of 400,000 budding consumers. This number will represent your total population.You decide to distribute a survey to 400 of your prospective customers. This number is defined as your sample size.When the data returns, you learn that 65% of respondents say they prefer Concept A. The following step requires inputting a confidence level into a margin of error calculator. Confidence level (not to be confused with confidence interval, which is a range of values that contains the true mean of a population) determines how certain a researcher can be regarding their sample accurately revealing theattitudes of the total population being tested. It is common for researchers to set confidence levels at, 90%, 95% or 99%.To properly demonstrate, let's enter the quantities illustrated above into the margin of error calculator.

• Sample Size (indicates how many respondents will participate in the survey) = 400
• Population Size (indicates the size of the total population from which your sample will be drawn) = 400,000
• Confidence Level (the percent at which you want to have confidence in your results) = 95%
• Proportion (this number indicates the percentage that you want to calculate your margin of error; the number should be between 0-100. In this specific example, we will use 65% as our proportion value. If you require one figure for the whole survey it is standard practice to input 50% for proportion as it gives the higher ME value, and use it for every figure. Keep in mind that smaller proportions will result in slightly smaller errors since we are answering the question in what range may the real population proportion fall). = 65%

Once the numbers are calculated a margin of error of 4.6% (4.9% rounded up) should have been reported. Remember 65% of your survey participants selected Concept A? This margin of error means that there is a 95% probability that, approximately, 60% to 70% of the targeted total population preferConcept A for your new product image. The figures 60 and 70 were calculated by adding and subtracting the margin of error from the Concept A sample response of 65%.

The Takeaway

As we know, margin of error supports the understanding of whether or not a sample size is an appropriate representation of a total population. If a margin of error reveals as too largeof a number, it is important to then increase the size of the sample. Doing this will ensure that the attitudes of the population surveyed match those of the total population more closely. Knowing your margin of error, and associated concepts like sample size and confidencelevel, is an important part of the balancing act that is designing a survey. Being able to properly calculate these concepts will allow you to proceed on your research journey with conviction.