August 9, 2010 9:23 AM Subscribe

Interested in teaching yourself some statistics? Here is an excellent online and interactive statistics textbook developed at UC Berkeley, and also used at CUNY, UCSC, SJSU, and Bard. Here is the syllabus for the course at Berkeley. And here are some insightful reflections from the professor on developing Berkeley's first fully approved online course.

posted by AceRock (18 comments total) 172 users marked this as a favorite

posted by AceRock (18 comments total) 172 users marked this as a favorite

Thank you so much! I have been saying to myself I want a review of statistics because I think I've forgotten some of it. This will be helpful :)

posted by Librarygeek at 9:33 AM on August 9, 2010

posted by Librarygeek at 9:33 AM on August 9, 2010

There would be so much less dumbassery in the world if everyone had a decent grasp of statistics.

posted by Artw at 9:38 AM on August 9, 2010 [1 favorite]

posted by Artw at 9:38 AM on August 9, 2010 [1 favorite]

There's some really great stuff in that book on logical fallacies too, for anyone who's interested.

posted by l33tpolicywonk at 9:42 AM on August 9, 2010

posted by l33tpolicywonk at 9:42 AM on August 9, 2010

I was too daunted by the subject in college to take a course, and have regretted that ever since. Thanks much for the post, I needed it!

posted by bearwife at 9:42 AM on August 9, 2010

posted by bearwife at 9:42 AM on August 9, 2010

This is a really nice resource. Thanks for the post.

posted by wittgenstein at 9:51 AM on August 9, 2010

posted by wittgenstein at 9:51 AM on August 9, 2010

From Chapter 2:
*This is a hasty generalization.*

Um...interesting example.

posted by Philosopher Dirtbike at 10:02 AM on August 9, 2010 [1 favorite]

I want to estimate the fraction of free "adult" (commercial pornography) websites that are hosted in the United States. I run a search on a popular search engine for "free porn" and find a link to a list of 25,000 free porn websites, with user ratings. I look up whether each of those websites is hosted in the U.S. or elsewhere. Seventy percent are hosted in the U.S. Therefore, the majority of free porn websites are hosted in the U.S.

Um...interesting example.

posted by Philosopher Dirtbike at 10:02 AM on August 9, 2010 [1 favorite]

Oh, there's more!*This is still a hasty generalization from a sample of convenience.*

My Intro Stat students are always asking me for additional texts to help them study. This will be an interesting one to recommend...

posted by Philosopher Dirtbike at 10:23 AM on August 9, 2010

To try to get a better sample, I think of more search terms. I run searches on a popular search engine for "free porn," "free sex videos," "naked girls," "hot sex," "free intercourse videos," and 15 other terms. For each search, I record the first 100 links to free porn websites that the search engine returns, producing a list of 2,000 websites. I look up whether each of those websites is hosted in the U.S. Seventy percent are; therefore, the majority of free porn websites are hosted in the U.S.

My Intro Stat students are always asking me for additional texts to help them study. This will be an interesting one to recommend...

posted by Philosopher Dirtbike at 10:23 AM on August 9, 2010

Excellent, thanks for posting this AceRock!

posted by storybored at 10:41 AM on August 9, 2010

posted by storybored at 10:41 AM on August 9, 2010

The best math type professor I ever had used outrageous and witty examples - polymaths was the first time I ever really understood logic for that reason. He was brilliant, now if there's a complex argument I want to make it always involves porn, cats or sex with goats.

Sadly this doesn't work for explaining search engines (my*eyes*) but you can usually subsitute Lex Luthor.

posted by shinybaum at 10:50 AM on August 9, 2010

Sadly this doesn't work for explaining search engines (my

posted by shinybaum at 10:50 AM on August 9, 2010

Upper level statistics was the most challenging and rewarding math I took in high school and college. I think it's a shame more people never get to it; it feels to me much, much more profoundly useful in everyday life than any other math I took. Really.

So yay for this.

posted by ifjuly at 11:29 AM on August 9, 2010

So yay for this.

posted by ifjuly at 11:29 AM on August 9, 2010

Yes, yes, yes! Thank you!

posted by alms at 11:31 AM on August 9, 2010

For those who know a little and need a bit more help, including analysis examples for some stats packages (Stata, SAS, SPSS), here are a couple of very useful websites:

UCLA Statistical Computing

StatNotes (NC State)

posted by naturesgreatestmiracle at 12:58 PM on August 9, 2010

UCLA Statistical Computing

StatNotes (NC State)

posted by naturesgreatestmiracle at 12:58 PM on August 9, 2010

I generally stink at math but found statistics fairly easy all the way up to the graduate level.

It is almost all just understanding the concepts and logic and then using the stats program properly.

So if you are math phobe don't avoid this. Statistics requires far less math that you imagine.

posted by srboisvert at 3:12 PM on August 9, 2010

It is almost all just understanding the concepts and logic and then using the stats program properly.

So if you are math phobe don't avoid this. Statistics requires far less math that you imagine.

posted by srboisvert at 3:12 PM on August 9, 2010

This looks really good!

They did a really good job connecting the touching all of the important concepts.

One thing that I really don't like about every statistics course I've seen so far is the presentation of confidence intervals. Are confidence intervals important? Why not instead teach "posterior predictive distributions" instead. They're much simpler, more intuitive, and you can use them to find "confidence intervals" for a whole bunch of distributions.

posted by esprit de l'escalier at 4:50 PM on August 9, 2010

They did a really good job connecting the touching all of the important concepts.

One thing that I really don't like about every statistics course I've seen so far is the presentation of confidence intervals. Are confidence intervals important? Why not instead teach "posterior predictive distributions" instead. They're much simpler, more intuitive, and you can use them to find "confidence intervals" for a whole bunch of distributions.

posted by esprit de l'escalier at 4:50 PM on August 9, 2010

I must prefer these types of posts vs. Academic Earth and other video sites, another favorite textbook online of mine: Structure and Interpretation of Computer Programs @mit

Thanks for the quality post!

posted by godisdad at 9:03 PM on August 9, 2010

Thanks for the quality post!

posted by godisdad at 9:03 PM on August 9, 2010

This is terrible! What will the students do without a physical reference book for Z-tables?

posted by benzenedream at 6:58 PM on August 11, 2010

posted by benzenedream at 6:58 PM on August 11, 2010

esprit d'escalier: confidence intervals are just two numbers. posterior distributions are, well, distributions. They're more complicated, and therefore harder to teach. Not that we shouldn't teach them -- but they'd take longer and in an intro course you'd probably lose a lot of the students.

(And they were even more more complicated back before we had fancy computers.)

posted by madcaptenor at 9:43 AM on September 5, 2010

(And they were even more more complicated back before we had fancy computers.)

posted by madcaptenor at 9:43 AM on September 5, 2010

« Older "Fred is not a real kid, thank God: Fred Figglehor... | How Will You Measure Your Life... Newer »

This thread has been archived and is closed to new comments

posted by l33tpolicywonk at 9:26 AM on August 9, 2010 [2 favorites]