Typos

[ghost]

Data Analysis Course, Module 2

PROGRESS: << (X) (X) (X) (X) ( ) ( ) ( ) ( ) ( ) ( ) >> (35% Complete)
Next, try calculating the variance for the data 'car$mpgCity'.

The name of the dataset is "cars" not "car".

[ghost]

Another issue in Data Analysis, Module 2:

PROGRESS: << (X) (X) (X) (X) (X) (X) (X) (X) (X) (X) >> (88% Complete)

Do large cars appear to have outliers?

1: yes
2: no
3: cannot be determined from this plot

Selection: 2

You almost had it, but not quite. Try again. Look to see if there are circles representing outliers above the 'large' box-and-whisker plot.

1: yes
2: no
3: cannot be determined from this plot

Selection: 3

Nice try, but that's not exactly what I was hoping for. Try again. Look to see if there are circles representing outliers above the 'large' box-and-whisker plot.

1: yes
2: no
3: cannot be determined from this plot

Selection: 1

You're the best!

There are clearly no circles/outliers above the large size car boxplot, so I don't know why it's insisting that there are.

[WilCrofter]

On Tue, 17 Dec 2013 08:31:11 -0800 hhambridge notifications@github.com wrote:

The name of the dataset is "cars" not "car".

Noted. Thank you!

[WilCrofter]

I certainly agree with you; large cars do NOT appear to have outliers. Thanks again. As is occasionally mentioned most of our recent effort (5 developers, 3 of whom are full time students) has gone into prototyping the next release, expected at the end of Jan. Effort on more and better course material is just getting off the ground.

[WilCrofter]

BTW, what would you like to see in terms of new course material? More R? More stat?

[ghost]

I am a biostatistician who uses R almost exclusively. I think it would be interesting to have lessons on learning R for programmers who are familiar with other languages. A lot of the people I work with use SAS almost exclusively and are traumatized when they have to use R for something. (I think this is typical of government and industry analysts/programmers/statisticians.) I think these sorts of people would benefit from courses that focus less on the statistics, which they already know, and help them learn the R syntax. These individuals are already competent programmers and just need to be walked through the R basics, but don't need nearly as much hand-holding as someone who is learning their first language.

[ncarchedi]

Thanks for the suggestions. A complete "Intro to R" sequence is in the works, which we hope will provide a friendlier introduction to the language. Once we've released the next version of the software in January, we'll have more time to focus on developing content. Please let us know if you have any other feedback.

[ghost]

It seems like you could easily adapt the Statistical Computing course at JHSPH for a comprehensive "Intro to R" course.

[WilCrofter]

On Tue, 17 Dec 2013 10:00:42 -0800 hhambridge notifications@github.com wrote:

courses that focus less on the statistics, which they already know, and help them learn the R syntax

Thanks once more. That is very gratifying to hear, since we had been forming a consensus precisely along those lines. Personally, I know of two labs, one at Johns Hopkins, one at NIH, who are moving to R because of cost, efficacy, freedom, etc.

FYI, I'm a nominally retired applied math guy with, maybe, 5 years R programming experience.

[ncarchedi]

@hhambridge That's the plan. See also: https://www.coursera.org/course/compdata