Wikipedia:WikiProject Medicine/Research/AnsweringQuestions
What effect does using different online sources have on a students score on standardized medical questions?
Comparison
[edit]Comparison will be a before and after format. Students will be given a number of questions without access to an online source. Than will be given a second set of questions with access to a specific online source (uptodate, emedicine, wikipedia). Will determine how much scores change with access to each source in question. Will want to have 1/2 the subjects do the test with the source first and half the subjects do the test without the source first. They will be randomly assigned to the group. Thus best to have even numbers of subjects.
Question sets
[edit]Pre medicine
[edit]A number of questions sets exist including Kaplans as found here on Amazon for MCAT [1] It is multiple choice which will make it easier to mark but is preclinical medicine. We have the "Secrets" book series which is clinical medicine but not multiple choice [2] While we have multiple choice ER questions here [3]
Medical school
[edit]If we are looking at Canadian medical students might be best to use test prep material for the MCCQE [4] with a book of questions on amazon here [5] Will need 200 questions in total. Another sample can be found here [6]
Residency
[edit]We could also look at family practice residents taking using test prep material for the CCFP exam. If Wikipedia makes a difference in test scores we could have a press release such as "Wikipedia Makes your Doctor Smarter"
Number of students
[edit]One thing we will need to determine is how many students do we need to generate statistically significant results. First we will need to determine what we consider a significant difference (5% maybe). If students have a 5% greater improvement with one source as compared to a second source we would deem that this is meaningful. May be 75-100 students? 100 questions on each test given 2 hours to complete. Thus 4 hours of student time with a break of an hour between tests.
Year of test subjects
[edit]There are a number of possiblities for year of test subjects. I propose that we use 3/4 year medical students as we would like to say that source A/B/C leads to X%improvement in scores and is thus appropriate / not appropriate for medical students.
Honorarium
[edit]As the students are going to be studying for the MCCQE anyway this should not be too much. We could look at cash or food? Should probably budget about $10-20 per study subject. Cost around $1000-2000.