Mar 7, 2017

Using Excel to Perform Survey Analysis

Excel is a great way to take responses from a survey and to complex statistical analysis. What I have shown in the video and in this excel template is the count formulas that are required to report on various sub-groups for a given question or multiple questions. There are a wide variety of ways to go about survey analysis, but this captures the fundamental mechanics of understanding your respondents answers.


Often times within a survey you will want to know multiple things about a given question. For example, if you have received their age and or gender, you may want to know how many people answered a different question a certain way and were below the age of 35 or how many people had a given condition and also answered a certain way to a given question.

There are also a lot of ways to visualize the responses and dynamic statistical analysis insights that are made with sub-group reporting style.

Another reason why I built this video is so survey creators can better see the best ways to structure questions so they can be queried easily. For example, the sub-groups you plan on isolating against each question has to be binary in nature. If it is not, further formulas and data manipulation would have to be used to make it binary in nature. That is one limitation of excel, but it also forces a cleaner analysis.

Coming Soon...I am going to add to this template a correlation analysis. This will give the correlation coefficient to two data points. For example you will be able to see how strongly or weekly the physical activity level relates to age or amount of hours watching TV relates to number of health conditions or what have you.

No comments:

Post a Comment