PDB_Stat_100_Lecture_20_Printable

PDB_Stat_100_Lecture_20_Printable - STA 100 Lecture 20 Paul...

Info iconThis preview shows pages 1–11. Sign up to view the full content.

View Full Document Right Arrow Icon
STA 100 Lecture 20 Paul Baines Department of Statistics University of California, Davis February 23rd, 2011
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Admin for the Day I Homework 5 due today at 5pm! I For Q2, please test at level α = 0 . 05 I Midterms are ready for pickup. . . I Midterm solutions are posted I Project groups assigned I Project Proposals due Friday! I Possible project datasets will be posted shortly References for Today: Rosner, Ch 7.1-7.7, 8 (7th Ed.) References for Friday: Rosner, Ch 8 (7th Ed.)
Background image of page 2
Midterm Feedback People generally did well on the midterm (better than midterm I), but there was fairly high variance. . . I Mean: 69 I Median: 72 I SD: 14.8 Histogram. . .
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
60 80 100 0 10 20 30 40 50 Histogram of Midterm II Scores Score Frequency
Background image of page 4
Grade Interpretation You can think of your midterm II scores as being curved, below is a general guideline for how well you did. . . General guidelines for interpreting your score: I 80-100: Excellent I 60-80: Good I 45-60: Fair I 30-45: Poor
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Exam Feedback I Make sure you know how to use a calculator! I Don’t write too much – many answers were unnecessarily long. . . I Please read the solutions – the final is cumulative!
Background image of page 6
Midterm Question 1 Sloths. . . I Estimates and CI – Excellent (except calculations) I Was it trustworthy? Check n ˆ p and n (1 - ˆ p ) conditions! I Hypothesis test – Mixed bag. . . 1. Hypotheses are about parameters not estimates! There are never any ‘hats’ in the hypotheses! 2. Don’t recompute your CI! The CI gives you plausible values – you just use those to check your hypothesis.
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Midterm Question 2 Malaria. . . I Estimates and CI – Excellent (except calculations) I For 2c – data is still Poisson, not normal! Use Poisson formulae. I For 2d – part (c) computed this for one week. You had 52 weeks of data, so remember to multiply by 52. I For 2e: 1. Assumptions are: (i) Independence across weeks, (ii) Each week has same mean, λ , and, (iii) Each week has a Poisson distribution. 2. Your estimate is not an assumption!!! 3. Your estimates would change, but this doesn’t directly address the question. . .
Background image of page 8
Midterm Question 3 Wine-making. . . I Excellent – parts (a), (b), (c) very similar to homework and practice midterm I Good – (d), (e), (f) recognize this is a linear combination I For part 3(f), this was a very tricky question. . . , 1. You might think that the probability of a blended wine being dull has to be between the two component wines – but. . . 2. The mean acidity level has to be inbetween the two means 3. The SD has to be inbetween the two SDs – but it’s nonlinear. . . 4. With proportion c of Chardonnay, the SD is p c 2 × 0 . 12 2 + (1 - c ) 2 × 0 . 04 2 . 5. If we choose c = 0 . 2 we can do better!
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Hypothesis Testing We have covered the basic methodology of hypothesis testing I Formulate a real-world hypothesis as a concrete statistical assumption (e.g., μ = μ 0 ) I Decide what alternative hypothesis we wish to consider (e.g., μ < μ 0 , μ > μ 0 or μ 6 = μ 0 ) I Go collect some data I Work out how likely it is that you would have got the answer you did, or even more convincing evidence against H 0 , if the null hypothesis were true (the p - value) I Compare that probability to how often you are willing to be wrong (is the p -
Background image of page 10
Image of page 11
This is the end of the preview. Sign up to access the rest of the document.

Page1 / 32

PDB_Stat_100_Lecture_20_Printable - STA 100 Lecture 20 Paul...

This preview shows document pages 1 - 11. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online