Stats Final FAQ

As I get questions about the Stats Final, I will post them here, prepended by the date/time that I post them.  FAQs are in reverse chronological order (most recent first).

Monday, June 3:
  • 4:30pm  Regarding Question 5, Part 8 :
    Do you think these data satisfy the assumptions for regression? Say why or why not.Robyn brings to my attention that Parts 6, 7 are optional but that to accurately answer Part 8 you need to determine if the residuals are normal (hint hint). If you don’t do Parts 6,7, then you may choose to either skip Part 8 or speculate based on looking at the data distribtuion or your intuition about the data….

Sunday June 2:

  • 11pm:  I promised to give you a more precise way to interpret a chi-square test.  For the bird-foraging example from Week 7, we were able to reject our null hypothesis, and conclude:
    Birds forage preferentially.  In spring, they forage preferentially in trees, rather than in shrubs or on the ground, but in fall, they prefer ground for foraging.  Birds neither prefer nor avoid foraging in shrubs during either season (Chi-square, df=2 = 15.04, Chi2critical=5.991, p < 0.05).on the final, you need not give the Chi2critical
  • 11am: This question asked about Question 2, which is to “check the results of the t-test with resampling”.   Question:  There are multiple types of resampling, however, and I don’t know which one to use or how to use it in JMP. Can I have some clarification here?”
    – response: you will use resampling in Excel (we did NO resampling in JMP).  recall that a t-test essentially tests if two means are equal. the stats lab from week 3 ( look at Question 4) tells you how to do this with resampling in Excel.

Friday, 5/31:

  • Question 1 has an error on Part 2. It should read:
    2. State the null and alternative hypotheses for a statistical test of the claim that Malathion-treated plots have less beetle damage on oats.
  • On Question 6,
    Part 6 should read: Run the CHI-SQUARE (in JMP or Excel) and report your results.
    Part 7 should be deleted.

Wed.5/29, 10:30pm – Question: The first paragraph appears to be about cereal leaf beetles and the use of Malathion. question #2 is asking for the null and alt hypotheses on the weight gain of chicks eating high-lysine corn. question #4 is about confidence interval of beetle larvae. I don’t see a data file for the cereal leaf beetle/Malathion, The high-lysine file does exist.  Should JMP be run on this file?

Answer1:  Be aware that the spreadsheet has several worksheets.  I originally named the worksheets for the ordering of the questions in the first draft, but didn’t change the data when I uploaded the final draft of the exam, so the Qx on the worksheet does not agree with the Qx on the exam.   However, I believe all the datasets are there, in the spreadsheet.

On Thursday morning I will upload a new .xlsx file and have renamed the worksheets so they are correctly numbered AND explicitly say which problem corresponds to which data sheet.  Sorry.

Answer 2:  Whether you run JMP or Resampling in Excel depends on your preference and whether the data meet the requirements for a parametric ANOVA….  If you can’t figure out for sure whether the data meets the requirements, then feel free to run a JMP if you wish.  You’ll get credit for that part.  Note that a resampling (monte carlo) test (rather than parametric) is always OK even if the data set is parametric.  See why?