Learning Goal: I’m working on a statistics project and need the explanation and answer to help me learn.
You are free to design whatever survey interests you. You can do a survey on people, things, animals (no harming animals!), vegetation, whatever you’d like. Here are the key components I will be looking for:
- Is it a survey of information that isn’t easily available by a Google search (it should NOT be easy to get answer)?
- Is it something that would interest a group of people (doesn’t have to interest everyone, but some group out there should be interested)
- Is your population well defined, and is it large enough to warrant a sample (It can’t be a population that would be easy to get a census from)?
- Is your sample size appropriate to the survey (does it satisfy the central limit theorem for that parameter and test)?
- Do I, the reader, understand what you are trying to find out?
You also need to describe in detail your sampling method and process. I will grade harshly if your sample is not adequately randomized.
The sample that you collect needs to be used to make a prediction about the population. To do this, you will be performing a Hypothesis Test. The hypothesis tests we will be covering are the following:
- Population Proportion (single population)
- Population Mean (single population)
- Population Correlation
You are free to use any of the hypothesis tests we cover in the class that best fits your survey. You are free to discuss with me for suggestions on which test is best for your data. If you’d prefer to do a hypothesis test not covered in class (chi-square, ANOVA, Difference of Two Means or Proportions), just make sure you ask me first so I know that you are using the correct test for your study.
As with all hypothesis test you need to define your hypothesis well and test it. As you will learn in the coming chapters, your goal is NEVER to prove a null hypothesis. You sole goal is to see if your data disproves the null hypothesis. Whether you reject or fail to reject the null hypothesis doesn’t affect your grade, so don’t “p-hack” your way to a conclusion. Let the data stand on it’s own merits.
Attached is an outline of a template Hypothesis test that can be applied to any of the tests listed above. This sample is just to help you organize your ideas. You must write your own study in your own words. Ask me questions if you don’t understand how to use it.
Here are the things I expect in your submission:
- Description of Survey
- Description of Sampling Method
- Full list of Data Collected (Your raw data)
- Hypothesis Test
- Final Conclusions and observations, as well as how you might adjust your survey in the future to answer any further questions.
Attempting to do it all with only a couple days to go will be obvious and graded harshly. Give yourself plenty of time to work on it.
Most importantly, do something that interests you! This will go easier if you are doing something you are truly interested in. Feel free to ask me questions on how best to conduct this assignment. I am here to help.
If you are going outside in a public area, you are responsible to make sure you are being safe. Don’t go anywhere that makes you feel uncomfortable. Bring a friend to help you. Follow the laws and rules of the area you are doing the study. If you are asking questions in public, it MUST be in a public area and not an area of business (like in front of a store that forbids people from bothering its customers). Also if asking a question, please don’t ask anything that would embarrass or make the person upset with you. Make the question as neutral as possible.
Your paper should be a minimum of 600 words (not counting the raw data). A lot of explaining and descriptions of how and why you did everything is important.
You must submit it either as a Word document or PDF. All other formats will result in an automatic zero. Your submission must also be college level work. If you are in college, expect yourself to work at a college level. If I find your work not at that level, I will just give you a zero.
Requirements:
