How to compare the results of a survey from two time points and how to compare two groups within a time point?

Question

Looking for some advice on running simple statistics on some survey data I have.

Part 1. I have two surveys, one from 2009 and one from 2013, with the exact same question but different respondents. How might I compare this question between years to see if they are significantly different or not? The survey was conducted anonymously from a pool of about 4000 people. Several hundred responses were collected for each year (~500 for 2009 & 700 for 2013).

Part 2. What is the best way to test for significance among groups of people based on their characteristics within one survey? T-Test?

I unfortunately only have access to Excel to do my stats. Any help would be appreciated!

Answer 1

A few Suggestions:

1) Run Cluster iterations on both sets of data and compare the mean centers. 2) If there are too many questions and sets of questions belong to the same category, run a PCA to reduce the number of factors. 3) If your questions and responses are scalar - "rate on a scale of 1 to 5" then an attitudinal segmentation of people can be done.

It would be helpful if you can post the purpose of your survey, what you are hoping to achieve with it and a few more details about the nature of the data.

Answer 2

Is there only one question in the survey? If that is the case, just perform a T-test and check for significance in the difference.

If there are more than one single question that interest you, you can have two different appraoches:

1.- Check for difference on individual questions: this is a "more complete" study but it is also more likely to give you false positives at random. So, you will have to be more strict with p-values. 2.- Check for "overall difference" (see vagabond's answer)