00. Sign Up for Thursday: http://doodle.com/poll/6e4r52ipyitkpkgn
0. The Survey
Data: Today in class we are going to be assessing the information from the survey that you took. I've got copies of them all, but you are going to want them in this form as well:
The answers we care about as a .csv are below.
Perspective: We have been asked to analyze the data that came out of the survey as well as make suggestions if we were to give this survey in the future.
Discuss the best ways to interpret the data
What kind of visuals are going to be best for us in this instance?
What possible outcomes would we derive from different aspect?
Are there any questions we should consider as bivariate as opposed to univariate?
Are there any questions where we should consider removing data?
Goal: Our goal is two-fold.
1) Analyze the data we have, creating any possible changes or observations of the data
2) Create follow up of clarified questions and create a new google form to ask those questions.
Resources:
I came across this today: http://passyworldofmathematics.com/online-graph-makers/
Please be careful, and continue to use chart common sense.
---
These questions are very open ended, and are designed to elicit discussion, creation, and a feeling of 'bigger' which was alluded to in our discussion of how they should feel. The first two questions are built with a goal of sorting through data, making a conclusion, or otherwise creating something from the data.
1. VT Youth Risk Behavior Survey
Data: http://healthvermont.gov/research/yrbs.aspx <-- more specific links below.
Perspective: You have been asked to analyze the data the state has put out about the YRBS (Youth Risk Behavior Survey). Specifically, You have been placed in charge of figuring out where funding should go for your county (Bennington College is in Bennington County). Should the focus be on
drinking and driving?
Minimizing self-harm?
Heroin awareness?
Something else?
Goal: Each year, the county creates a focus for minimization of a specific Risk Behavior. The school districts have a minimal amount of time and money to work on education and minimization of these behaviors. Create a SHORT report for your county outlining possible focuses, and give a recommendation on what you think should be the focus for the coming year.
(note: it should be clear that personal bias will play a role in this--but make sure you are backing up your reasons with specific numbers, reasoning, etc.)
Resources: There are more of them out there, but here's a few useful links. All were found by poking around on the main site linked above:
Statewide Overview -- highlights Statewide
Please note any other resources you use while investigating this question.
2. Data on Amazon Purchases. (note -- if you are interested in analyzing your own data for this part, let me know. I'll talk more about this in class)
Data: The data on all Amazon Purchases I've made since the start of 2008 is below (amazon.scrubbed.csv).
I downloaded this data this morning and did not delete any of the purchases.
I did delete the names of the different products.
I left the category of each purchase, according to Amazon's categories.
There's two price categories -- list price and price paid.
Perspective: I wanted a large chunk of data -- but not too large -- for us to run an actual analysis on. I assume my habits have changed over the years, but am not sure how to justify that. I'd like to run some tests and see if habits have changed in any appreciable way. I'd like for you to work through the questions below and see what you can make of the shopping habits.
Goal: I'd like you to check the following things:
Can I say that my shopping habits on Amazon have changed in terms of price over the course of the years?
Can I say that my shopping habits have changed in terms of category over the years?
I've been thinking about whether or not Amazon tends to give bigger discounts on more expensive items or cheaper items. I'd like for us to check that.
Assume that the data here is a random sample of items.
Set a threshold where you think the cheap/expensive.
Decide which is a better determinator for a discount -- amount of money off (probably better for bigger items) or percentage of the price off (probably better for smaller items)
Run an analysis of your data, using any appropriate methods and thoughts.
Create a graph that details one aspect of this data in an interesting or meaningful way.
Resources: If there any object that you HAVE to know what they are, please do not hesitate to ask. Simply email the date, cost, and category.
3. Build a new survey.
Data: None Yet.
Perspective: Y'all did great finding the flaws and issues that are prevalent in surveys. As such, now it's your turn to make one. You will be searching for the same information that we were on the first survey: information about this course.
Goal: Your finished product should be a survey you would want the class to take.
It should be made on some survey site
I used google forms; if you have a google account you can make on one google drive.
www.surveymonkey.com is another option. I think they require a user name, but it's free
there's others out there.
It DOES NOT have to focus on every aspect of the course. Rather, design a series of questions that focus on a specific aspect of the course.
It should be through, but concise. There is no question limit, but 5 - 10 would be a good range. Ask as many questions as you need, but not more.
Consider what you would do with the information once you receive it. In other words, what graphs, ideas, or thoughts you are trying to make and connect.