Unfortunately, there were a couple of things I noticed right away. One is that the options to download data in rtf or html format are missing. You can run your code but I did not see any download option, unlike when I run the same code on a Mac or Windows computer.

What I did was take a picture of the output buy holding down the home and power button simultaneously (which takes a screen shot). I then cropped that screen shot in the regular iPad photo app. While that worked perfectly fine for my purposes, as you can see above, when I just wanted one small analysis, on a regular basis – for me – it would be a pain.

However, for something like doing a homework problem that was assigned, that would work fine. For example:

A researcher suspects irregularity in a dataset. She expects that in a survey with 51 items, there would be a higher proportion of missing data in the items that occur later in the survey. Generally, this is the case, but it does not appear to be so in the dataset. Use PROC SUMMARY to compute descriptive statistics for each item. Use PROC TRANSPOSE to create a dataset that has just the N for each item and correlate the sequence in the survey with the number of responses. Examine both the PROC SUMMARY results and the PROC CORR results. What do you think?

Well, here is the plus – since I had uploaded the data to the Projects directory before I left home (since SAS web editor is a SAS On-demand product only the professor can upload data), I was able to run the analysis above and more. My PROC SUMMARY showed about 3.5% missing data for every single item and as you can see above, just about 0 correlation between item location in the survey and percentage of missing data.

Any statisticians (or anyone else) want to weigh in on the probability of these data being legitimate?

The minus is that if I had *not* uploaded the data before I left home, there would have been no way of doing it.  It may be that SAS presumed (reasonably) that professors would have a computer and so they could use those to upload the data.

Certainly, I could select all of the log or code after I had run it and copy and paste it an email to turn in as an assignment.

Usually, I am interested in the students’ interpretation of their results, so they could certainly look at the results and write up their analysis. They could even, as I did, take screenshots, edit and paste into an email or other document.

My caveat for professors would be if you are using SAS web editor to teach, don’t assume you can do everything on your iPad and leave your laptop at home. For students, you should be fine. If your professor does insist on the output, if you have it working, all you need to do is find any computer anywhere, log on, run it and then you have your output within a minute, assuming you had your code running fine already.

I have run into a couple of really minor glitches. Sometimes when I am editing a program and deleting a line, it doesn’t really delete. Yes,  that is weird. I don’t know why that happens and I have had it happen both with and without a keyboard. In those cases, I copied the part I did want, pasted it into a new file and went on my merry way.

An even weirder thing that has happened once or twice when using the keyboard and ONLY when logging in, is that it will enter each keystroke twice. If I type Ann  it will have AAnnnn    and deleting doesn’t work because then it backspaces twice. Yes, it is weird. In that case, I’ve closed it, tried again later and it was fine.

In a nutshell, it runs pretty problem-free as long as you are aware of what you can and can’t do. No idea what the deal is with those occasionally random keyboard / typing problems but they happen so seldom they haven’t really annoyed me, and as a general rule, I’m pretty easily annoyed.

By the way, if anyone wants to weigh in on that missing data question, I’d be interested in your input.

Random bonus tip: they have free candy at the SAS publishing booth in the demo area. Warren says who needs candy and he doesn’t like chocolate. I suspect there is something seriously wrong with that man. Who doesn’t like chocolate?