This preview shows pages 1–5. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Todays Outline Administrative stuff Students t ConFdence intervals The pooled ttest Error ranges Inquiry our Administrative Stuff FRI Stream allocation notication yesterday Allocation was 69%, 17%, 9%, 4%, 0% for 1st, 2nd, 3rd, 4th, and 5th preferences respectively. Some students from other sections were allocated 5th preferences, or a stream they hadnt listed. There will be a stream change request form announced soon. Wait until you see what else you are able to register for before requesting a stream change. If you do submit a stream change form, be as specic as you can about what credit you would accept and what streams you would consider. 1 2 3 Inquiry Four Last week I moved the assignment of Inquiry Four to today However, my highest priority today is getting through the statistics needed for inquiry three. Depending on how long this takes I may or may not get to inquiry four today. I am moving the rst weeks requirement for inquiry four to THURSDAYS LECTURE to remove the time crunch and make sure the inquiry is doable in the time you have left. For this reason, it is IMPORTANT to attend the next lecture. Making things hard to read can boost learning http://www.bbc.co.uk/news/world11573666 Students t The Ztest was for a situation where the population mean and standard deviation are already known. This is rare. More usually, we do not know and . We deFne a new variable Students t, similar to z, but using standard error instead of population standard deviation: t = x s m = x s x n x z = ) x s x n ( = 4 5 6 The tdistribution is more spread out, fatter than a normal distribution. This is because it is more uncertain. z distribution larger sample t distribution t distribution The spread oF the tdistribution is dependent on the sample size. The more samples, the more condent we are that s m is close to and the closer t is to z. smaller sample Students t is tabulated For diFFerent condence levels, and diFFerent degrees oF Freedom (= n 1) William Sealey Gosset, a.k.a Student William Sealey Gosset, a.k.a Student Chemist & Statistician at: 7 8 9 We can use the value of students t to express a confdence interval We can use students t to predict the range in which is expected to lie with a given level of conFdence. As n increases, the value of t, and hence the confidence interval, decreases, meaning will lie within a smaller range. t = x s x n Rearranging: = ts x n x = ts m x Confidence Inter val What range contains the true value with a certain confidence level? ttest example one.....
View
Full
Document
This note was uploaded on 01/06/2012 for the course UGS 303 taught by Professor Foster during the Fall '08 term at University of Texas at Austin.
 Fall '08
 FOSTER

Click to edit the document details