Comments for Citizen-Statistician

Comment on Research Hack: Paper and Reference Management by samarthbhaskar

samarthbhaskar — Thu, 09 May 2013 00:03:11 +0000

I used Zotero a ton in the past, but this seems like a good academic paper archival and reference system too http://t.co/18RsE4mjPR

Comment on About Us by rob

rob — Mon, 22 Apr 2013 16:26:20 +0000

Great! Thanks for the tip.

Comment on About Us by Tal Galili

Tal Galili — Sat, 20 Apr 2013 07:40:57 +0000

Hi there,
Great post on the course.
I noticed that you have some other R related posts which you didn’t categorize as R, so they were not picked by r-bloggers,com.
I thought you’d might wish to notice that for the future (or even categorize them that way now).

With regards,
Tal

Comment on A Course in Data and Computing Fundamentals by grzewap

grzewap — Fri, 19 Apr 2013 19:48:11 +0000

A Course in Data and Computing Fundamentals http://t.co/mbYbTtOVJU

Comment on Datasets handpicked by students by Derek

Derek — Mon, 15 Apr 2013 13:20:22 +0000

You can find lots of interesting, complete country-level and international datasets at Quandl. Much of it is from international sources, like the UN, OECD, World Bank, WHO, WTO, etc (http://www.quandl.com/about/sources). Quandl has an R package (http://www.quandl.com/help/r) for those so inclined.

Comment on participatory sensing by Phillip Kent

Phillip Kent — Sat, 13 Apr 2013 08:22:14 +0000

You might find useful to look at the Census at School project
http://www.censusatschool.org.uk/
In this, a school gathers data by questionnaire and uploads their own results to an international database (not sure about the upload part, it may be managed by someone rather than self-uploading). Key point is that every school answers the same questionnaire hence increased likelihood of comparable data. Using the database students compare themselves with others and pose statistical queries. There are some online visual tools for doing that.
Phillip

Comment on participatory sensing by Rob

Rob — Fri, 12 Apr 2013 18:20:24 +0000

Corey, I think you’re right-on about TP. Too often, we think of the act of creating a graph as an intellectual activity. But “make a histogram” is no different then “compute the mean of these numbers” as far as real statistical thinking goes. And TP really makes students engage with the data and think about what they want to display. I’m seeking ways of integrating TP into Mobilize, but since the project is ensconced in a Computer Science curriculum, they are eager to learn and teach R (via Rstudio). What’s needed is a TP for R! Something like ggplot —which considers the building blocks in a mathematical sense–but for kids. Gisela, I think this is what you mean when you talk about going beyond the ‘cool’ factor. That’s a big problem. Sometimes, the WordPlot becomes the means itself, and not the portal to further discovery. (Has anyone ever discovered anything with a Wordplot?)

I think the challenge is to focus on questions that could—bear with me here–be put into a two-way table.

The rows represent the scope of the data: It’s about me; It’s about my community; It’s about my world. The columns are about the form of the question: What have I learned about x? Do the data generate questions/hypothses about x?

For instance, if I have data about hours of sleep of my classroom, the data say something about me: I’m typical, according to my classmates, in the number of hours of sleep last night. And they say something about my community: my classroom slept about 8 hours last night, on average, but there are some real insomniacs among us. And they suggest hypotheses about our world: are all classrooms like ours? I bet we sleep more than others. Etc.

Jerzy, I like the Feltron stuff very much, but I always struggle about where to go next. I think Feltron teaches us much about how to make pleasing displays. And Nathan Yau’s new book (Data Points) has gone some distance to convince me of the beauty and personality of these displays. But it all stays at the “me” level. (Although I may be wrong about that—I’m not that familiar with it.) The challenge in the classroom is to bring in more context, I think.

I wonder if the quality control people can guide us here? The goal isn’t just to record sleep but to improve sleep. (Sorry for harping on the sleep variable. I’m sure it speaks to my frame of mind.)

Comment on participatory sensing by Gisela Martin

Gisela Martin — Thu, 11 Apr 2013 16:49:48 +0000

I think the real challenge here, with respect to both teachers and students, is that while PS useful, how do you give perspective beyond “gee this is cool” and how would this integrate into what you are trying to do with the course at an intro level? Technical progress has made the learning enviroment much more relativist and while that’s liberating in one sense, it deprives students of structural understanding. This may not be a barrier for the top students (or teachers) who can self-motivate, but I am more concerned about the modal ones, and does this give them the foundation they need for effective careers?

Comment on participatory sensing by Corey Andreasen

Corey Andreasen — Thu, 11 Apr 2013 15:07:25 +0000

Some of the questions in the last paragraph are addressed by technology like Tinkerplots. TP is designed for middle school, but is really ideal for anyone’s first exploration of statistics. Displays are made naturally as the user drags cases on the screen. It’s very cool stuff, and really gets students to make sense of the data!

Comment on participatory sensing by Jerzy

Jerzy — Thu, 11 Apr 2013 00:14:00 +0000

Great idea for a stats course! I don’t have personal experience doing this with students, but high school math teacher Dan Meyer has some debrief notes after a similar project:
http://blog.mrmeyer.com/?p=556

And if your students are more comfortable with Twitter than flat data files, they could consider using “your.flowingdata” (by the FlowingData guy, Nathan Yau) to input and track their numbers, and download them as tab-delimited text files:
http://your.flowingdata.com/