Graph Crank

Well, I've had a pinch of time so here's my latest in Mathematica:

I received a bunch of data from a Google Survey my teacher put up for Korean moms to look at (all users of MissyUSA who are mostly Korean women immigrants in the US.) At first, I was attempting to find very unrelated variables and graph them by hand (as in typing the actual labels of the graph, and the place to find the data, etc.) However, that took a long time and was ineffective so I made a module to crank out graphs in 2D ListPlot form for numerical data points. My algorithm is as such:

Before the Module:
Since I'm using Google Surveys, I extract the data from the website published to import the "FullData".
After that's imported, we have to remove the question titles (but I kept these handy since I needed them for the next part of my code), Flatten the data, and Partition the flattened data into whatever length each person's response to all the questions were added to 2 (in this case, there were 37 questions + 1 timestamp + 1 bullet = 39 boxes). Then, take the last 37 boxes or whatever number of boxes (by using -37)by mapping it. Afterwards, for this module, you have to Select the Integers using, well, Integers, as the filter.

In the Module:
This module is solely to make ONE graph, not several, yet. We will utilize this module to make hundreds of graphs by using a for loop.
Anyways, in the module, we have five parameters: the x axis label, the y axis label, the x-axis variable number (in the survey, what question are you seeking to put on the x-axis), the y-axis variable number, and the data set to analyze (in this case, from Google Surveys). We use a for loop to make a point with one number from the x var. number question and one from the y var. number question.
We Append that point to an empty list we declare in the beginning of the Module.
After the for loop, fit a line/curve, with obviously, Fit, to the "empty list" that had all the points appended in it in the previous for loop.

Then, Plot the line and ListPlot the data points. Use the x var. number and the y var. number with a "vs." in between as the title to show what you are comparing.

For Looping:
We use a double nested loop to traverse the questions with integer responses (which we had filtered beforehand). Set one variable to the be the counter of the x var. and the other nested to be the y var. This way, we can get ALL possible combinations. Afterwards, inside the nested for loop but before using the listPlotMaker module, put in an if statement to ensure that the x and y var. numbers are NOT the same (that way, we don't have unnecessary repeats of shockingly perfect correlations). Then, just use the listPlotMaker module inside the loop to create each graph.

ALGORITHM FIN

Anyways, that was a really long-winded explanation, but it cranked out a huge number of graphs. I can't seem to find any good correlation though...

School Life Survey 1 Wrap Up

School life survey 1:

Question: Are the claims of parents that social networking is harming today's generation true? How about the assumption that more "friends" really means more friends?

Hypotheses:
1. If more time is spent on social networking sites, then there will be more emphasis on social standing. (This hypothesis was just Null--no proof for or against since there was not enough data and I was attempting to use a Boolean-style approach, where Social Standing was 1 and Academic Standing 0. Came out like 2 straight lines.)
2. If more time is spent on social networking sites, then there will be less trustworthy friends.
3. If there are few friends on social networking sites, then there will be more hours spent on social networking sites.
4. If more time is spent on social networking sites, then there will be a lower weighted GPA.

These were hypotheses added on later (I made the above hypotheses on the spot as I posted my School Life Survey 1):
5. If there are more friends then there will be a lower GPA.
6. If there are more trustworthy friends then there will be less FB friends.

Data:

Hypothesis 1: Null.

Hypothesis 2:

Hypothesis 3:

Hypothesis 4:

Hypothesis 5:
(FB Friends to GPA)

(Trustworthy Friends vs. GPA)

Hypothesis 6:


Conclusion:
The data given suggests a strong negative correlation between the time spent on social networking sites and GPA. Also, there is a very slight negative correlation between the number of trustworthy friends/facebook friends and GPA. Otherwise, the data is too sparse to draw anything even semi-definitive.

Limitations/Uncertainties:
This data was taken largely from students from Mission San Jose High School in Fremont, CA. One or two points each come from Illinois and the East Coast (I do not know which points, obviously, as this was an anonymous survey, but my peers from those parts said they took the survey). Mission has a very high emphasis on academia, which leads to the relatively high amount of people who said they appreciated academic standing over social standing. Data is here: http://spreadsheets.google.com/pub?key=trtorCi2YDWnRugPcM3tkdg&single=true&gid=0&output=html

School Life Survey 1

This is a survey I made in order to see if there is any validity in what many adults claim to be correlations in teenagers' social lives. I don't want to show my hypothesis here, but I'll save a page right now of my hypothesis. By the way, there are several questions but I'm going to use them in pairs to find correlations. Also, this is a completely anonymous survey so please respond with all honesty.



Thanks.