<p>Sure, maybe he does update his tests each term and maybe he doesn’t teach straight off the powerpoint screen provided by the publisher. </p>
<p>The good news is that that cheaters were turned in, or at least uncovered, by students SICK of cheaters. Good for them. </p>
<p>I had a very good friend in college who took the easy way through each of his classes. He majored in finance and had a series of jobs in financial services after college. I got a call one day wherein he needed help because he had to make a presentation to potential clients. So, even after college he was still skimming his way through. It was truly embarrassing. It was tempting to say gosh don’t you think you should have maybe studied a little bit you know back when we in school? Didn’t you realize that one day you would actually have to use this stuff? </p>
<p>Apparently not. </p>
<p>Last I heard he was unemployed and living off his wife who was employed. </p>
<p>Eventually, the ones who know what they are doing get rewarded one way or the other.</p>
<p>No excuse for stealing the test/answer key. I am not familiar with UCF, but an upper level course with 600 students for management with a multiple choice test sounds awful to me. If students behave like this it is no wonder so many people in business think it is ok to misrepresent the facts in financial statements, 10K reports, offering prospectuses and the like. There is a distinct lack of ethics. However it is worse than a lack of ethics, because some of the bad behavior is a violation of law. If a computer system was hacked, there were crimes committed. What will that person do next? Manipulate the poor financial data of the company s/he works for?</p>
<p>I would like to believe that it all evens out in the end Cecil, but I am not sure that I believe that.</p>
<p>Yes, Cardinal, I now know that the answer key appears to have come from a test bank. it wasn’t apparent that was the case until I read an article in the UCF campus paper. It doesn’t make what the students did right, but it does make it easier to understand how they got the answers. </p>
<p>I do wonder how the prof can know which kids cheated. He’s requiring everyone in the clas to retake the test. Those who admit cheating the first time will get their grade on the second test as their grade–seems lenient to me, so I don’t think they can complain. Those who do not admit to cheating and are not proven to have done so will get the higher of their two grades. (So, some of them could actually benefit from the retake.) Those who don’t admit to cheating but are later found to have done so will fail. </p>
<p>At the very least, this will convince some profs at UCF and perhaps elsewhere that giving tests provided by the publishers is not a good idea.</p>
<p>I go to UCF, but I’m not a Business major. One of the reason’s I decided not to pursue business through UCF is that the business college was too big. The most largest college, with the most popular major, at the 2nd largest university in the country. It doesn’t take a rocket scientist to figure out your business degree is probably watered down at UCF.</p>
<p>In order to separate myself for other students, I decided to major in Statistics. It’s a much smaller program at UCF. I can assure you the material isn’t watered down in the Statistics program.</p>
<p>Anyway, with my Stats background perhaps I can shed some light on the use of the Normal Distribution to determine if cheating went on. </p>
<p>The prof has given that class 4 or 5 other times, so he knows what the mean grade is historically. The larger the class size, the more accurate that mean grade will be. So when he saw “the average grade running 1.5 grade lines above typical”, he knew something was up.</p>
<p>You can take the mean grades from the cheating class and compare the mean grades to the non-cheating past classes. Using a z-table you can compute the probability of the mean grade being where it was. I’m not sure what the standard deviation of the exam was, but it’s fairly easy to compute with that info.</p>
<p>Take the mean of the suspect class, subtract historical mean, divide by standard deviation to get a “z-score”. Use z-score to determine probability of the event. If it’s above 3 standard deviations from the mean, the event is almost impossible without some kind of cheating.</p>
<p>Sure, but to what extent will the z-score predict that a given student has cheated? That’s the question being posed here, and it was discussed in great detail on this board a year or two ago when a similar incident was being discussed. Maybe someone else remembers that thread and can dig it up.</p>
<p>You can’t possibly use the scores alone to predict whether a given student cheated, in this particular case, because some honest students also can score well on the exam. So you can look at the data and say, Wow, too many 96%s, but you can’t tell of an individual whether he was one of the ones who actually scored 96% because he studied or he scored well by cheating.</p>
<p>The bimodal results show the cheating (of the group) vividly, though.</p>