Dear Tipsters,
a while ago Karl Wuensch mentioned that his students seemed to have problems telling which of two numbers was larger (Is p< .05, Mail: Sept 28th2012, http://www.mail-archive.com/[email protected]/msg08713.html) As probably most of you, I shook my head in disbelief, mentioned the story to my students in statistics 1 last semester during introductory sessions and got a few chuckles. But then I decided to collect some data and entered the following item in the final test: Warm-up 2: Sort the following letter-number-pairs in numerically increasing order (smallest number first, largest number last). Enter the sequence of letters that results as one word in the answer field. A: 1.2% B: 1/50 C:0.03 D: 0.05 E: 0.2% F: 1.2 The item was worth 1 of 64 points in a 2-hour final exam and intended as an icebreaker. 195 students participated in the exam, 172 of them used a laptop with spreadsheet software (Excel or Calc) for computations. As an exercise: How many students will answer this question correctly? Please take a note of your guess (to avoid hindsight bias). Well, 2 of the 195 participants failed on purpose (left all answers empty), 1 failed to follow instructions. Of the remaining 192 students, 142 sorted the numbers correctly (74%). 14 students (7.3%) answered FCBDEA (-1.2, -0.03, 1/50, 0.05, 0.2% , 1.2%), i.e.: ignored completely that the last 2 numbers were percentages. 6 students (3.1%) answered FCEBDA (-1.2, -0.03, 0.2%, 1/50, 0.05, 1.2%), i.e. showed confusion if percentages were used. 5 students (2.6%) answered CFEABD (-0.03, -1.2, 0.2%, 1.2%, 1/50, 0.05), i.e. could not order negative numbers. 3 students (1.6%) answered DBAECF wrong direction. The other students were distributed evenly on wrong answers, no pattern detectable. Item-total-correlation (part-whole corrected) was r=.34. I use percentages and decimal numbers interchangeably during lecture, and explain explicitly how compute the decimal number from a percentage (or fraction) and vice versa. Nevertheless I found several students who used percentages and decimal numbers wildly intermixed within the same formula (Bayes). So Karl: You are not alone. And tell your TAs to discuss percentages and small negative numbers too. Kind regards, Rainer Dr. Rainer Scheuchenpflug Lehrstuhl für Psychologie III Röntgenring 11 97070 Würzburg Tel: 0931-31-82185 Fax: 0931-31-82616 Mail: <mailto:[email protected]> [email protected] Web: <http://www.izvw.de> http://www.izvw.de --- You are currently subscribed to tips as: [email protected]. To unsubscribe click here: http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=25233 or send a blank email to leave-25233-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
