Eval Sessions

From "A B C"
Revision as of 00:50, 3 February 2016 by Boris (talk | contribs) (Boris moved page BIO Quizzes to Eval Sessions without leaving a redirect)
Jump to navigation Jump to search

Quizzes

   

The in-class quizzes will test your completion of the assignments for last week. They will be 20 to 30 minutes long and given in class in the 5:00 to 6:00 slot on Tuesday. We will have eleven quizzes, one each week after the first.

Typical questions will ask about your experiences while going through the assignments, and your basic understanding of concepts - what you where doing and why you were doing it. I try to make these questions reasonably easy for anyone who has completed the assignments with an active mind. It should be possible to get a perfect score on all of them. However careful, structured thinking will be required.

The goal of the quiz is to test whether the readings and assignments have been completed with understanding and active mental participation.

There may also be questions on the concepts and methods, especially their computational aspects. You may be asked to give your opinion on possible improvements or generalizations and to think through some "problem solving" questions. It may be useful to discuss your understanding with your classmates in person or on the mailing list.

I really need you to think along with the course, and designing your own quiz questions is an excellent way to do this. On the Student Wiki you will find a page that collects quiz questions for every week, on which you can submit proposed questions/answers and critique question proposals. Four of your participation marks will depend on your contributions to these pages. On these pages I have provided a sample question for every week from previous quizzes so you can get an idea of what you may encounter.

We will pick up the quiz right after writing and mark it in class.


Marking

  • You will mark your own quiz.
  • You must bring a black or blue pen to write the quiz.
  • You must bring a red pen for marking.
  • You should write down and explain the right answer to yourself. Write legibly: if I like your explanation I might count the question as correct even if you made a mistake.
  • You must not make mistakes in your marking. If I find errors in my spot-checks I will mark the quiz as a zero.
  • Write down your quiz marks for immediate feedback on your course performance. I will email you in case marks change, so you can maintain a good idea of where you stand.


Missed quizzes

You may miss as many quizzes as you want, I will initially score these as zero. For your final grade, I will record the marks of the (lower) quartile of all your quizzes to replace the zeros. This value is easily calculated with the R summary() function. Since we are writing eleven quizzes, this means you will get partial marks for up to three missed quizzes. Conversely, if all your quizzes have perfect marks but you miss two, these two will be counted as perfect scores.


To illustrate:

# some vector of quiz results. Three quizzes have been missed.
x <- c(3.3, 0, 3.2, 3.6, 0, 2.8, 2.5, 4.0, 0, 2.9, 3.7)
 
summary(x)
#   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
#  0.000   1.250   2.900   2.364   3.450   4.000 
 
# Replace all zero values with 1st. quantile
x[x==0] <- quantile(x, probs=c(0.25))
 
x
 [1] 3.30 1.25 3.20 3.60 1.25 2.80 2.50 4.00 1.25 2.90 3.70

sum(x)
# [1] 29.75
 

# Almost the same vector. But four quizzes have been missed.
# I am removing the lowest mark: 2.5.
x <- c(3.3, 0, 3.2, 3.6, 0, 2.8, 0, 4.0, 0, 2.9, 3.7) 

summary(x)
#   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
#  0.000   0.000   2.900   2.136   3.450   4.000 
 
x[x==0] <- quantile(x, probs=c(0.25))
x
# [1] 3.3 0.0 3.2 3.6 0.0 2.8 0.0 4.0 0.0 2.9 3.7
 
# Note that in this case all zeros remain, since the lower quartile is zero.
# You will probably want to avoid this situation and thus miss as few 
# quizzes as possible to have a buffer for unexpected situations.

sum(x)
[1] 23.5  # Missing an additional quiz has dropped the total marks by 6.25 points.