WolframAlpha Facebook Report

This is a delightful exercise that everyone seems to love. WolframAlpha will provide you with an extremely detailed analysis of your own Facebook data including visualizations, world clouds, graphs, and more.

Graph of Facebook Activity over time

 

 

 

Here’s how:

  1. Go to WolframAlpha.com.
  2. Type “Facebook Report” and execute the search.
  3. Allow WolframAlpha to have access to your Facebook account by clicking on “Analyze my Facebook Data” and following the directions.
  4. Wait while the data is analyzed.

Note: Sometimes the report seems to stall after 100% of the data is analyzed. If this happens, simply repeat steps 1-3. The second time, the report seems to load just fine.

Enjoy!

Possibly Related Posts:


Share

Battling Bad Science (and Statistics)

If you ever needed a REASON to calculate the highest point of a parabola that opens downward, here’s one.

Possibly Related Posts:


Share

Moving Math from Analog to Digital

Arthur Benjamin has been on TED in the past (see Mathemagics) and has done a really phenomenal job.

Here’s his latest 3-minute appearance, called “A Formula for Changing Math Education.”

The problem is that the very short talk does not present a “formula” for changing education, just Benjamin’s idea that the pinnacle at the top of the math pyramid should be statistics instead of calculus. There is nothing in the the short talk that suggests any kind of coherent plan for how it could be done, or even a suggestion that he has a plan. That’s what I would want to know about. Of course, it’s only a 3-minute talk and it’s certainly possible that he had nothing to do with the name of the talk.

I did agree with these two statements, but want to add my own two cents:

1. “very few people actually use calculus in a conscious meaningful way in their day to day lives” … but I’m not sure we teach people how to use calculus in a “conscious meaningful way” nor are many of us required to use calculus for the simple reason that our superiors don’t understand it at all. Calculus could be used in a “conscious meaningful way” but our society chooses not to engage. As a matter of fact, very few people actually use statistics in a conscious meaningful way in their day to day lives. Enough said.

2. “it’s time for our mathematics to change from analog to digital” … here I agree, kind of. It’s time for our mathematics to include both analog and digital, and it’s definitely time for our mathematics teaching to change from analog to digital. What happens in most math classrooms is based on a factory-model of education that developed before computers even existed. Even though the world has changed, the instruction (for the most part) has not.

I found it more interesting to read through the comments that followed the short TED talk. There is an interesting conversation taking place there. One wise commenter pointed out that it’s possible that there should not be just one pinnacle on the math pyramid. Both Calculus and Statistics could be considered penultimate goals of a mathematics education. I think that’s dead-on.

If there’s anything I’ve learned during the process of writing my dissertation, it’s that the system of collegiate mathematics education is extremely complex.  There will be no “easy” fix to the system, even if someone is able to convince a majority of the stakeholders that their change is the correct one.

Possibly Related Posts:


Share

Mathematics of Coercion

predicting-the-future

Bruce Bueno de Mesquita, a consultant to the CIA and DOD, uses mathematical analysis to predict the outcome of “messy” human events in this 2009 TED Talk: Three predictions on the future of Iran, and the math to back it up.  He claims that we can use mathematics to predict the outcomes of complex negotiations or situations involving coercion (everything that has to do with politics and business).

His modeling is based in Game Theory, which (he says) is based on three assumptions that (1) people are rationally self-interested, (2) that people have values and beliefs, and (3) people face limitations.  The CIA verifies the predictive ability of the model, claiming it is correct 90% of the time even when the experts are wrong.

To build a model of the outcomes, he says he need to know (1) Who has a stake in the decision? (2) What do they say they want? (3) How focused are they on one issue compared to other issues? (4) How much persuasive influence could they exert?  Using this, we can predict behavior by assuming that everybody cares about two things: the outcome (effect on their career) and the credit (ego).  In the model, you must be able to estimate people’s choices, chances they are willing to take, values, and beliefs about other people.  Believe it or not, history is not necessary for the model.

Other than the mention of mathematics and a really general look at game theory, there was not a lot of mathematics in this talk.  There was one concrete mathematical example that you might be able to utilize in one of your classes (especially if you teach a little combinatorics as part of Probability and Statistics or Liberal Arts Mathematics):

To build a model that predicts the outcome of complicated social events, we need to look at the interactions between all of the people who have input in the decision-making (the influencers).  The number of interactions between n influencers is n!  If we double the number of influencers in the interaction, does that double the number of interactions?  (to use this example, play from 4:24 to about 7:15)

Possibly Related Posts:


Share

Teaching Statistics with Clickers

I haven’t talked a lot about clickers on this blog, mostly because there’s no easy way for me to try using them for a semester and because I’ve focused a lot of my free time on learning to teach math online.

My college has a set of clickers that can be checked out by faculty, but it’s always a guessing game (when you “check out” equipment) whether the previous user will remember to turn in the equipment before your class meeting time.

If it was affordable, I’d probably just purchase a full class set myself and then not have to worry about the costs to students (but that seems a little extreme at roughly $2000 for a class set).

In the meantime, Derek Bruff has two great posts on teaching Statistics using classroom response systems (or clickers), so I’m sending you there!

Check it out here (Part I and Part II).

Possibly Related Posts:


Share