calculating statistical difference

Hey everyone,

I am trying to figure out how to calculate if a survey statistic is different from a benchmark or goal. The statistic is a little different than normal, because the survey question has 3 options: positive/neutral/negatives. The statistic I am using is basically the "net positive" where I take the sum of all the positive responses minus the sum of all the negative responses, divided by the number of responses (Pos-Neg)/N

I'm not exactly sure how to go about trying to figure out if the statistic is different from the goal. For example, if I survey a population and I get back 90 positive responses and 10 negative responses for this survey, which gives me a "net positive" score of 80%, how would I say if this is statistically different from a goal of 70%?


TS Contributor
How did you come up with the benchmark / goal? Was it from a previous survey or surveys? If so, you can look at the distribution of scores from the prior surveys and then determine statistical significance via knowledge of the likelihood of various scores......
No, unfortunately the goals are somewhat arbitrary. I guess I was thinking of trying to calculate a confidence interval for the survey statistic, but I don't know how I would do that for this type of metric.