Online Homework, In Class Quizzes, and Google Forms

There are some Web tools I just like.  I'm not sure why.  The Google Form tool attached to a Google Spreadsheet is one of those. It is drop dead easy to make a form.  The data come back that way I want.  (For example, I just tested that if you ask people to put in a url in a text box that when you get the data in the spreadsheet it is linked to that address.)   And it is fairly simple to embed the form in another Web page.  (I've found that first, the place where you want to embed the form has to be wide enough and second, that you have to muck around with the height parameter (make it bigger) to get rid of the scroll bar.)  This makes it quite usable in a variety of contexts, some of which I'll talk about below. 

Before I do, here's a slight diversion that got me thinking this way.  I suppose it is the contrarian in me, but my main reaction to all the flurry there's been of late about Udacity and Coursera (on several lists I participate in this David Brooks column got a lot of attention) is that this is not the new market to focus on.  This is really process innovation (doing what we've always done but doing it cheaper).  We need product innovation (doing something new).  What I have in mind isn't really all that new.  We did it in the 1990s - online office hours staffed in the evening.  The really new part is to make this a cross campus offering - available to anyone on the globe - and hence not tied to a particular course. Then have it focus on a disciplinary area, but otherwise not restrict what the questions are about. 

Why do this?  First is to show there is substantial demand for it and hence demonstrate the student needs which should drive further innovation.  Most students don't need additional courses from which to choose, at least not during the regular school year.  Many of them, however, need help with the courses they are taking.  They get stuck.  Then they fret about it.  But they don't now how to get help.  Some look online for help, which I know because I've produced videos for microeconomics that are in YouTube and the hits those videos get are driven by students trying to fill this need.  This is fine, but it is not sufficient.  With pre-produced online content students may still not see the light.  They need to ask their own questions and get somebody in the know to respond to them in a serious way.  Second is to make a scale economy argument that I attribute to Burks Oakley, a point he made way back when.  Lurkers can benefit from a posed question that has generated a serious response.  Thus the archive of such questions and responses can have substantial value to the students.  From the lurkers point of view the archive looks a lot like pre-produced content.  But the difference is that it is framed from the student's perspective. This is the need that is not being satisfied at present, in my view. 

Third, there is an issue of how to generate credible response.  In the late 1990s, I used undergraduate TAs (they had taken my class previously) to provide the response.  But the set of question where about homework problems from the textbook and I had prepared solutions for those problems that the TAs had.  I also coached them in the type of response I was hoping for.  In that limited domain my TAs could function quite well, for the most part. Now, however, the domain of possible questions will be much larger, so whoever has the responsibility to provide responses must have much more expertise - an ability to work through the question on the fly or with a minimal amount of down time to research the solution.  So the question of who will do this remains open. 

Not having a general answer to that question, I want to try it myself to see how it goes and learn from the experience.  Yesterday, I set up a site for undergraduate Microeconomics called Ask The Prof.  At your convenience take a look.  Then if it interests you, submit a question to help get the site started.  The embedded Google form for that purpose is on the third Tab.  I've promoted the site on my YouTube channel and in specific videos.  But students might be reluctant to post their own questions if there isn't already some history of prior questions and responses on the site that they can view.  So I'd like to provide that history and do so without making canned questions I generate myself.

Now let me turn to using Google Forms as a homework submission tool.  I've not yet done this but this is what I'm thinking.  I make assignments (in Excel of course, that's the tool I know) that illustrate the math models for my class.  While the underlying model is the same for each student, parameter values will be student specific so each student will have a unique homework set.  Each student will be assigned an alias (the name of a famous economist concatenated with the course name).  The will copy the solutions they get from Excel and paste those into the Google Form.  They will also select their alias from the list and that is what will be used as identifier of the student.  I believe that should handle all the FERPA issues.  I can then auto grade these submissions using Excel via a straightforward comparison of their submission to the correct answer.  I can also include a paragraph question or two to have them explain their thinking in doing the math.   Obviously that can't be auto graded, but it means the amount of discretionary grading can be kept down.

How does this compare to using the LMS for the same purpose?   For the LMS that I know about, using the LMS quiz engine one can't make a rather extensive problem which has many subsidiary questions, with the parameter values unique to the particular student, but invariant from one subsidiary question to the next.  So the LMS quiz function simply isn't good enough for what I'm after.  Last year, I tried something similar with the LMS drop box.  I had an "answer sheet" in Excel and indeed Excel had graded the homework before the students submitted it.  The drop box was used simply to verify that they had done the homework correctly.  The students were instructed to save the answer sheet as a csv file  (which saves only the answer sheet and not the rest of the workbook), naming the answer sheet in a certain way that would include their identity in the name, and then submit that.  This mechanism was far from foolproof.  Some students didn't see csv among the formats available when saving the file.  Others screwed up the file name.  Further, I'd have to manually given them credit for the submission.  The approach with Google Forms would substantially reduced the workload for me. 

The downside I see as possible with this approach is that, since their is no authentication by the students required in the process, one student could submit as another to sandbag that person.  In a small class I don't see that as an issue.  In a larger class, perhaps there needs to be an additional step of assigning a unique identifier for the particular homework and the particular alias and sending that to the students in advance.  Managing that would be an additional headache but maybe it wouldn't be too hard if done in conjunction with mail merge in Word.

If this can work for homework, then I'd want to use it in class as well.  It could be done as a more sophisticated version of how instructors now use clickers.  And perhaps you could actually administer quizzes this way.  The issues with this are twofold as I see them.  Each student would need a device of some sorts.  I did verify that the form works on my iPad and iPhone, but on the phone version especially, that can be clunky (at least for me) to enter responses.   Second, to the extent that the devices uses the campus wireless network, there could be slowdown/congestion effects that you don't want to see, especially in a high stakes testing setting.  So far, I don't have that many students registered for my class in the fall.  If the numbers remain modest, I may experiment with this a bit to see how it works in that setting.