Week 3 of 540 - Survey

This week we are starting to plan a survey where we can analyze the results. As mentioned last week, the hardest part will be in designing questions that do not bias the answers. We will need to select a question that will be interesting to the people that we can get to participate. Once we have that topic, we will need to design neutral questions that do not cause the participants to select one answer over another. Once we have the topic and questions, we then will need to administer the survey and analyze the results. This is where the real math begins, and where we can start using the science of statistics to interpret the answers.

WE also did more work this week to look at the differences between mean, median, and mode. Our homework really drove home the point that two data sets might have the same mean (average) but may in fact be completely different. One might have a very closely grouped set of data, and the other may have a huge range that happens to balance out to the same average. Usually you want more consistent results, although of course that depends on the questions you are trying to answer.

CEdO540 - Week 2, Samples

This week we really started looking at the ways to evaluate a sample of numbers using statistics. WE evaluated several samples using the common ways of measuring central tendencies, Mead, Median, and Mode. It was interesting to learn that the Mean is not always the best way to evaluate a sample, as it could be thrown off by a single outlying value. You should also look at the median and the Mode, as they might help you find out is the data being affected by one or more values on the ends, or if the data is closely grouped around the center.

Comparing the central tendencies to each other can also give you a lot of information, for example if the mean is to the right or left of the median, then that can tell you something about the data. It is obvious that we will need to do a lot more evaluation and comparison of the central tendencies, as well as look at the data sample as a whole, in order to get the best picture of the results.

In the future we will be looking at how to collect data for evaluation. I can see that designing a survey that will not bias the results is going to be a challenge.