PDB_Stat_100_Lecture_16_Printable

PDB_Stat_100_Lecture_16_Printable - STA 100 Lecture 16 Paul...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon

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

View Full DocumentRight Arrow Icon

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

View Full DocumentRight Arrow Icon

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

View Full DocumentRight Arrow Icon

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

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: STA 100 Lecture 16 Paul Baines Department of Statistics University of California, Davis February 9th, 2011 Admin for the Day I Homework 4 due today by 5pm! I !!! PLEASE TURN IN A PRINTED COPY OF YOUR R OUTPUT !!! I Hwk 4, Q1(f) – check the forum post ‘Question 1.f’! I Hwk 4, Q3(e/f) – what are the assumptions? independent? normal? same mean μ for each month? same SD σ for each month? I Hwk 4, Q3(e/f) – is it normal, are consecutive weeks independent? does it matter? Either answer is fine – just justify it! I Stop the prejudice – let means be free (and negative) , I Project details coming soon. . . References for Today: Rosner, Ch 6.5-6.8, Ch 6.5 (7th Ed.) References for Friday: Rosner, Ch 7.1-7.4 (7th Ed.) Computing Normal Probabilities Let X be normally distributed with mean μ and standard deviation σ , (i.e., X ∼ N ( μ,σ 2 )). We are often asked to compute things like P ( X > k ), or P ( X < k ), or P ( a < X < b ). We can do all of this with R , but what did people do before computers? They used normal probability tables. Did they have a separate table for every different mean μ and standard deviation σ ? No! That’s a lot of tables. . . , z-scores Instead, they used rules for linear transformations so they could just have one table. How? I Let X ∼ N (100 , 10 2 ) and let Z ∼ N (0 , 1 2 ). What is P ( X < 120)? = 0 . 977 What is P ( Z < 2)? = 0 . 977 I Let X ∼ N (20 , 5 2 ) and let Z ∼ N (0 , 1 2 ). What is P ( X < 12)? = 0 . 055 What is P ( Z <- 1 . 6)? = 0 . 055 If we can just convert everything to mean = 0, sd = 1, then we only need one table! z- scores The way we do this, is to form a z- score by subtracting the mean and dividing by the standard deviation. This z- score tells us ‘how many standard deviations away from the mean’ we are. If X ∼ N ( μ,σ 2 ), then: Z = X- μ σ , ⇒ Z ∼ N (0 , 1) You can check this yourself –find a and b and plug into the formulae for linear transformations! Using z- scores Example: Tax bills (or refunds, if negative) for UC Davis students are normally distributed with mean- $400 and standard deviation $250 i.e., X ∼ N (- 400 , 250 2 ). You only have access to a table of N (0 , 1) probabilities. Q: What proportion of UC Davis students will receive a refund? A: We want P ( X < 0). So. . . subtract the mean from both sides, and divide both sides by the standard deviation. . . P ( X < 0) = P X- (- 400) 250 <- (- 400) 250 P ( Z < 1 . 60) = 0 . 945 I Note that X- (- 400) 250 ∼ N (0 , 1)! I A normal distribution with mean = 0 and variance = 1 is known as a standard normal distribution I We use letter Z for standard normal distributions, i.e., Z ∼ N (0 , 1). About z- scores Much (but not all) of what we do with z-scores, and standardization of normal variables (i.e., transforming X from N ( μ,σ 2 ) to N (0 , 1)) is a legacy from pre-computer days, but. . ....
View Full Document

This note was uploaded on 03/09/2011 for the course STAT 100 taught by Professor drake during the Spring '10 term at UC Davis.

Page1 / 29

PDB_Stat_100_Lecture_16_Printable - STA 100 Lecture 16 Paul...

This preview shows document pages 1 - 8. Sign up to view the full document.

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