Hi all,
I got the February 2011 edition of the AMS Notices last week, and on the back of the second page is an advertisement for WebAssign. It reads "Find the derivative of y = 2 cos(3x - pi) with respect to x." It then gives the answer which is the typeset equivalent to the TeX code \frac{dy}{dx} = 6\cdot\sin 3x .
The copy in the advertisement is "...we decided to compare WebAssign with the other leading homework system for math. The results were surprising. The other system failed to recognize correct answers to free response questions time and time again."
So, what is "the other leading homework system"? And, is there an other leading homework system that would make the error that they cite in the advertisement? Obviously WeBWorK doesn't (though it does require that one parenthesize the argument of the sine function), though one could of course write a problem in WeBWorK that would fail to correctly evaluate this answer. Is there another system that they are comparing with? In the AMS survey the top three systems by far were MyMathLab, WebAssign and WeBWorK. The next most obviously targeting math were MapleTA and WileyPlus.
Can anyone speak to whether MyMathLab or WileyPlus would make this error? MapleTA wouldn't unless they've significantly decreased its effectiveness since I last used it.
I'm really curious whether this is correct advertising, deceptive advertising, or flat out false.
Thanks,
Gavin
I used WileyPLUS this Fall with Functions Modeling Change, 3rd ed., for my Precalculus classes. Not only did WileyPLUS cherry-pick the problems and turn many free-response problems into True/False or Multiple Choice, all of their problem templates seem to be hand-crafted (and WileyPLUS does not use a CAS). Their error rate was astounding --- e.g., for chapters 7--9 with 177 problems (out of 890 problems + 142 "Check Your Understanding" true/False items in the text), I reported 63 problems to Wiley as needing repair. (This was done 6--4 weeks before our classes reached those chapters --- Wiley did a rush job and fixed most in time for us to use.) On the other hand, I have extremely severe concerns about the technical expertise and editorial process used by Wiley and will need correspondingly severe penalty clauses before I would be willing to even consider use of their "new and improved" WileyPLUS for the 4th edition of Functions Modeling Change.
disclaimer: my checking of their chapter 7--9 problems was work-for-hire at the rate of $2.50 per examined problem (under $500 for about 150--200 hours). I agreed because we had a large group of students in three courses (College Algebra, College Trigonometry, Precalculus) being hostage to the faulty WileyPLUS product for FMC, 3rd ed.
disclaimer: my checking of their chapter 7--9 problems was work-for-hire at the rate of $2.50 per examined problem (under $500 for about 150--200 hours). I agreed because we had a large group of students in three courses (College Algebra, College Trigonometry, Precalculus) being hostage to the faulty WileyPLUS product for FMC, 3rd ed.
I need to correct my careless statement "WileyPLUS does not use a CAS".
Some problems in WileyPLUS mention they are graded by Maple (probably a subset of MapleTA). My testing showed that most of those problems would handle simple alternative forms, e.g., shift a sine to get a cosine, replace an arccos computation with an equivalent arctan expression.
On the other hand, intelligent use is a bit spotty. E.g., a problem involving exponential growth with correct answer of the form 3720 * (0.7^x) or 3720 * (0.7)^x systematically rejected 3720 * (.7^x) or 3720 * (.7)^x [i.e, leading zero was required for this problem, 0.7 instead of .7, but no message to students identified the form which the system would require for an answer to be scored as correct].
Some problems in WileyPLUS mention they are graded by Maple (probably a subset of MapleTA). My testing showed that most of those problems would handle simple alternative forms, e.g., shift a sine to get a cosine, replace an arccos computation with an equivalent arctan expression.
On the other hand, intelligent use is a bit spotty. E.g., a problem involving exponential growth with correct answer of the form 3720 * (0.7^x) or 3720 * (0.7)^x systematically rejected 3720 * (.7^x) or 3720 * (.7)^x [i.e, leading zero was required for this problem, 0.7 instead of .7, but no message to students identified the form which the system would require for an answer to be scored as correct].