Archived-Announcements-2013

From Statistics for Engineering
Revision as of 13:24, 6 May 2013 by Kevin Dunn (talk | contribs)
Jump to navigation Jump to search

From most recent to earliest


  • As you know, I value your input on all aspects of the work I do here at McMaster. I'd like some feedback from you by filling out this survey.
  • 27 Apr: this is an announcement more for the people in my 4N4 class. It's with great sadness that I wish to let you know that Dr. Don Woods passed away on Friday evening. He was a phenomenal instructor in our department, and we all still benefit from the work he did.
He developed innovative teaching methods, particularly focused on problem-based learning; he was known and recognized nationally and internationally. Details about his life and funeral.
  • 25 Apr: some personal things have come up that prevent me from being at the exam this evening. As with all my exams, if anything seems unclear, simply state your assumption and work with it. As long as it is reasonable, I will consider your interpretation.
  • 24 Apr: All grading issues with Quest (especially Quest 3) should be fixed. All 6000 questions have now been auto-graded. Please sign in again to double check.
The overall Quest grade has also been transferred over to Avenue. It has a weight of 22% in the course grade.
Thanks for your patience with the Quest system, and your help with finding all the bugs. Classes in future years will benefit from it.
  • 24 Apr: The grades for assignment 7 are loaded into Avenue. Please note, Avenue does not correctly calculate the Assignment weight for the course: I will only use the best 6 assignments. Please calculate your own grade; or wait for me to do it after the exams, when I upload the final grades.
  • 23 Apr: Finally! The grades for all Quests are available. Who knew that writing code to automatically grade would be so intricate? Please check your grades, and see if there are any errors. The tolerances on the numeric values are very generous; no grades were deducted for too many significant figures.
  • 23 Apr: as promised, the true optimum from the optional response surface mini-project is now available. Sign in to see how close you came to solving it. If you get an error message, just reload the page.
  • 21 Apr: the RSM project now shows a leader board, to see how you rank relative to the other teams. Only 2.5 days left to complete it! Remember, you can only run 8 experiments in 24 hours, so plan ahead.
  • 19 Apr: grades for the Quests are available for multiple choice questions. I'm working on the grading code right now for other question types (i.e. short answer and numerical answers).
  • 18 Apr: an update on the final exam details, and that assignment 7 solutions are posted (which will help with the RSM bonus project).
  • 15 Apr: a current job opening in London, ON, requiring the use of DOEs and response surface methods.
  • 09 Apr: an article came out this month on DOEs as used in companies. For example, it describes why BASF used DOEs to create an effective dishwashing soap.
  • 09 Apr: the notes on latent variable methods are posted. There is no need to print them for class; however you might prefer to do so, if you have time.
  • 08 Apr: the final assignment is now due on 10 April at class if by paper; or by noon on 11 April if electronic submission. Solutions will be posted after that, and no further late submissions will be accepted.
  • 08 Apr: good to know: the optional response surface mini-project is now available. It can be used to boost your grade; it should take little time to complete in between exams, and will help you study for the final exam.
  • 05 Apr: additional R source code for analyzing response surface models has been posted on the DOE page.
  • 05 Apr: an interesting tribute to George Box: to understand his contributions to statistics.
  • 05 Apr: the final response surface competition will be optional. You can earn an additional 5% of the course grade (i.e. bonus marks) this way, to boost your grade. The competition actually doesn't require much time, and it is a great way to study for the DOE section of the final exam.
  • 02 Apr: R source code for fractional factorials is posted on the DOE page.
  • 02 Apr: Solutions to assignment 6 were reposted; you don't need a username and password now.
  • 01 Apr: course evaluations are running. The comments and feedback you provide are extremely helpful to improve this course (even after teaching it 4 times, there are still improvements to be made).
  • 01 Apr: by request, the project due date in now moved to Friday, 05 April. Due at noon, electronically. Thanks.
  • 26 Mar: About Quest grading. You can sign into Quest any time after a Quest and it will show you a full solution to your questions. This feature has been there since the beginning, but it appears many don't know about it. You can estimate your grade to a high accuracy as follows:
    • You will get full grade if you answer the question correctly.
    • For multiple choice questions, subtract off 0.5 marks for every checkbox that should not be checked.
    • For numeric answers, if your answer agrees with about 1% tolerance, you will get full grade.
    • If you have too many significant figures, you will get deductions of 0.25 marks per box.
    • If there are 4 grades for a question, and there are two numeric values to fill in, then each box has equal weighting, i.e. 2 + 2.
    • Hope that helps clarify it.
  • 25 Mar: the final assignment 7 is posted. Due on 09 April (no late submissions). Also, assignment 6 solutions are available. Grading breakdown for assignment 5:
    • Q1: 12 = 1 + 5 + 2 + 4
    • Q2: 10 = 2 + 2 + 2 + 2 + 2
    • Q3: 15 = 4 + 2 + 2 + 6 + 1
    • Q4: 8 = 2 + 6
  • 20 Mar: grades for assignment 4 have been added to Avenue. Please confirm your grades are correctly recorded.
  • 19 Mar: please read slides 11, 12 and 13 before class tomorrow. We will review them, then continue on with the rest of the class. Also, there are some minor updates to the newer slides; slide 65 is new however.
  • 18 Mar: the class will slightly early, at 1pm on Tuesday, as I have to attend a university meeting.
  • 17 Mar: please note that the DoE project due date has shifted to 04 April to accommodate the 3 missed classes (snow day, Kipling, and Easter Friday).
  • 15 Mar: job opening with Kevin over the summer. I'm looking for a student to work with Dan Wright and myself for 2 months (full time), or 3.5 months (part time) to help redesign and implement the ChemEng website in Drupal. Experience with webservers, and Linux is a plus; knowledge of Drupal is mandatory. Work will be to evaluate other websites, take all user requirements into account, and convert existing pages over to the Drupal CMS. Email me by the end of March.
  • 15 Mar: please go through the three factor example and ensure you are able to derive the main effect and 2 factor interactions (CT, CS, TS) by hand.
  • 14 Mar: the grading rubric (schedule) for the course project.
  • 14 Mar: Solutions are posted for assignment 5.
  • 11 Mar: important update: please re-download assignment 6, there was a small error in part 2 of question 3 that's been fixed now.
  • 10 Mar: a short assignment 6 is posted: three quick questions. Due by 15 March.
  • 07 Mar: a summary of the course evaluations. Thank you for your valuable feedback.
  • 06 Mar: please take a look at the software tutorial to help you use integer variables in your linear model.
  • 05 Mar: Project proposal emails to me must include:
    • What is your objective: describe it clearly; describe how you are going to measure it; why are you interested in this objective?
    • What are all possible variables that can influence your objective?
    • Which variables are you going to keep fixed in your experiments?
    • Which variables are you going to vary? Tell me the low and high levels you are going to use for each variable.
    • Please just send it to me in an email; no Word document attachments. Thanks.
  • 04 Mar: Dr. Thompson is looking to fill a summer position in acoustic emissions testing. He would like to get a person with MATLAB skills and a more process control or stats focus. The project is about detecting a specific failure mode in plastics with acoustic monitoring so that we can catch the problem faster. But the focus will be on handling data. At 5 MHz, we collect A LOT of data very fast and we need to related a change in the acoustic spectrum to the formation of a microcrack. A process control oriented person or person comfortable with PCA would be most desirable.
    • I will be helping Dr. Thompson with the supervision and the data analysis part.
    • Please reply to Dr. Thompson (maybe CC me) to express your interest in this project.
  • 04 Mar: solutions to assignment 4 are posted.
  • 03 Mar: there is a description of the course project posted. Please read through it, and start thinking about some ideas. We will also talk about it in class later this week.
  • 03 Mar: The next weekly test is posted, due by Tuesday, 5 March, around noon time. Very short, only on least squares.
  • 28 Feb: assignment 5 is posted. It is due on 08 March 2013.
  • 01 Mar: a guest lecture by Emily Nichols, on designed experiments is coming up on 08 March. Please aim to be at class for this interesting talk and discussion afterwards.
  • 01 Mar: assignment 5 has an interesting extra question 4 added to it now.
  • 28 Feb: grades for assignments 1, 2 and 3 are posted in Avenue for 400-level students. Please bring any discrepancy to my attention right away.
  • 26 Feb: the slides for the Designed Experiments part of the course are posted. Don't print them yet. They are more for you to get an idea of the coming section, so you can see some context regarding the course project.
  • 24 Feb: there is a new Quest test available. You should have received an automated email about it earlier today.
  • 17 Feb: assignment 4 is posted; due on 27 February.
    • Please note that Q4 was updated, there are now 5 parts to the question.
    • Also take a look at the software tutorial for sample R code that will help you with this assignment; particularly sections 15 and 16.
    • Do not use R libraries, like qcc to do the work for you on building monitoring charts. The purpose of the assignment is to show that you can do the necessary calculations (you'll need to show that anyway in the final exam).
  • 15 Feb: an important class on least squares was held today. Please review the video if you missed it. We will resume from slide 44 onwards, after the midterm break. Enjoy the time off.
  • 14 Feb: the software tutorial for R has been updated with new sections 16 to 22, on building and using least squares models in R.
  • 12 Feb: the notes for the least-squares section are posted.
  • 12 Feb: weekly test 5 deadline extended to 12:15 today. A few people haven't got around to completing it.
  • 10 Feb: assignment 4 will be posted this week, but the due date changed to after reading week. This is as a result of the bump in dates from the snow day.
  • 08 Feb: Snow day: class for Friday is cancelled, and assignment 3 is now due on Tuesday, 12 Feb, paper or electronic, at class. No late hand-ins, we will post solutions on 12 February.
  • 07 Feb: Important note: A large number of students are submitting assignments electronically, which we encourage and find easier to grade. However, you must follow the electronic submission instructions in order for us to grade your work. If you prefer not to follow these instructions, please submit in paper form instead.
  • 06 Feb: quite a few emails asking if enough information is provided in assignment 3, etc. General rule: if information isn't provided, and you think it should have been, use an assumed value, but clearly justify why you've assumed the value, and why you picked the value you did.
  • 06 Feb: regarding question 2 and 3 in assignment 3: the problem is sufficiently specified, it is just not obvious what you need to do. That's typical ... an industrial statistics problem doesn't present itself to you in the form: "Calculate a confidence interval ...."; you have figure out what you want to determine (the aim), and then what you need to calculate. It is an important skill to learn how to translate an industrial requirement into a statistical calculation.
  • 06 Feb: the weekly tests are going well so far, so the midterm for 4C3 is cancelled.
    • Expect the grading part of the website to be ready during reading week.
    • Expect a slightly longer than normal weekly test at the end of reading week, going into the next week. That test will cover all the course material.
  • 05 Feb: solutions to assignment 2 are posted (PDF).
  • 03 Feb: the next weekly test is posted. You must complete it by 11:30AM on Tuesday, 5 Feb.
  • 01 Feb: assignment 3 is posted; due on 08 February.
  • 30 Jan: another visually interesting analysis of data by Hans Rosling.
  • 30 Jan: Please print the revised slide 20 from the process monitoring section. Also try the example on that slide before the next class.
  • 28 Jan: the weekly test has been emailed to you; please complete it before 16:30 on Tuesday, 29 Jan.
    • The problem with questions randomly being reordered has been fixed. Thanks for reporting that.
    • Background/interest: an interesting publication that demonstrates the effectiveness of spaced testing.
  • 27 Jan: some delays with the weekly test and the server ... expect it on Monday (28 Jan).
  • 25 Jan: I forgot to add in class today that if you'd like the R code for any drawing in the course textbook or slides, just send me an email with the page or slide number.
  • 25 Jan: please note that class on Friday, 01 February, will be cancelled to accommodate the 4W4 students going on the XRCC plant tour (and I will be joining you).
  • 25 Jan: the R code demonstrated in today's class is available
  • 24 Jan: the assignment 1 solutions are posted.
  • 21 Jan: the second assignment is posted; due on 30 January.
  • 21 Jan: Another great R tutorial video, this one is on plotting data in R. He doesn't use RStudio, but the plotting commands are the same.
  • 20 Jan, 21:30: the weekly test is posted. You should receive an automatic email. If not, please request a link at http://quest.mcmaster.ca
    • The automatic grading part of the software isn't written yet. But the website does display full solutions after the testing period is over. You can use this to estimate your grade for now.
  • A good overview of understanding plotting in R is given in this YouTube video.
  • 17 Jan: 6C3 students: please email me with your full name and student number. I do not get your registration information from the university. It is your responsibility to get this information to me, so that I can set you up for the weekly tests.
  • 17 Jan: Please follow these instructions to submit an assignment electronically. Whether you choose paper (in class) or electronic, assignments are due at 12:30 pm on Friday.
  • A tutorial for the course software is now available. It shows how to download and install the software and steps to get started. You'll need this for assignment 1.
  • A copy of the tables for the normal and \(t\)-distribution can be downloaded and printed. Bring to the next 2 or 3 classes.
  • If you want to receive these announcements on your cell, but don't have a Twitter account, text: follow stats4eng to this number 21212 and you'll get a text message instead of a tweet.
  • Solutions for the first test are now available: sign in http://quest.mcmaster.ca/ to compare your answers. In a week or two the system will also automatically grade you (where possible); that feature isn't developed yet.
  • Another interesting seminar on campus: John Bianchini, CEO of Hatch, will be talking at the free Chemical Engineering seminar on Thursday, 17 January, 10:30 to 11:20, JHE326H: moved to 28 March.
  • Assignment 1 is posted; due 18 January, in class on paper, or electronically. Please do not use the Chem Eng dropbox for assignment submissions.
  • The weekly test closes at 6AM on Tuesday morning, 15 January; a few students were under the mistaken impression that they can do the test during the day on Tuesday.
  • The weekly test has been emailed to your McMaster email address. Please follow the link and instruction to complete the test. Let me know if you experience any technical difficulties.
    • If you didn't receive the automated email, you can re-request it from http://quest.mcmaster.ca/
    • There was a technical issue that sent a second sign-in email around 11:07 on Monday morning; please ignore that email.
    • If you get the message "Sorry - you are not registered to use the Quest website", then you are not on my official class list. Please email me your full name, student number and McMaster email address.
  • Notes for the next section, univariate data analysis, are available.
  • Notes for the first section of the course are available: Visualizing process data
  • You should probably wait until after the first class before printing the course textbook. I have some advice on what to print.
  • The first class is on Tuesday, 08 January, in T13 room 127 at 12:30.
  • The course outline is available.