{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

10PowerFTest

# 10PowerFTest - POWER OF THE F-TEST Suppose C is a q p...

This preview shows pages 1–3. Sign up to view the full content.

POWER OF THE F-TEST A very common consulting question: “I’m planning a study to do .... ”. How many replicates (per treatment) should I use? I know 5 ways that can be used to determine an appropriate sample size As many as you can afford (time, money) n = 3 per treatment n so that s.e. (or s.e. of diff.) is ... n so that width of 1- α c.i. is ... n so that power of test (F or T) is ... for specified true difference Last is most common. e.g. NIH requires “proof that study is not underpowered” in grant applications. Often, need 80% power to detect an a-priori specified set of interesting differences We focus on the power approach. 1 α ci. width approach is the power approach with power = 1 α/ 2 for rank 1 tests. Copyright c 2011 Dept. of Statistics (Iowa State University) Statistics 511 1 / 21 Suppose C is a q × p matrix such that C β is testable. Earlier, we established that the quadratic form incorporating C ˆ β d has a non-central F distribution F = ( C ˆ β d ) [ C ( X X ) C ] 1 ˆ σ 2 ( C ˆ β d ) / q F ( δ 2 ) q , n k where δ 2 = ( C β d ) [ C ( X X ) C ] 1 σ 2 ( C β d ) Copyright c 2011 Dept. of Statistics (Iowa State University) Statistics 511 2 / 21 Non-centrality parameter of an F distribution The non-centrality parameter δ 2 = ( C β d ) [ C ( X X ) C ] 1 ˆ σ 2 ( C β d ) quantifies the discrepancy between C β and d with respect to Var ( C ˆ β ) = σ 2 C ( X X ) C . All else being equal, the noncentrality parameter increases as: the distance between C β and d increases, σ 2 gets smaller, the design, as defined by X , improves. (includes samplesize.) Copyright c 2011 Dept. of Statistics (Iowa State University) Statistics 511 3 / 21 1 2 3 4 5 7.5 10 10 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 + σ 2 = 1 Copyright c 2011 Dept. of Statistics (Iowa State University) Statistics 511 4 / 21

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

View Full Document
1 2 2 3 3 4 5 -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 + σ 2 = 2 Copyright c 2011 Dept. of Statistics (Iowa State University) Statistics 511 5 / 21 Let F α q , n k denote the 1 α quantile of the central F distribution with q and n k d.f. The power of the α level test of H 0 : C β = d is given by P ( F F α q , n k ) and is an increasing function of the noncentrality parameter δ 2 . Consider a study with 4 trts and 8 reps per trt. Design study around test of Ho: μ i = μ i . Consider 3 distributions: central F 3 , 28 (null), nc F 3 , 28 with δ 2 = 5 and nc F 3 , 28 with δ 2 = 10 Copyright c 2011 Dept. of Statistics (Iowa State University) Statistics 511 6 / 21
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

### What students are saying

• As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

Kiran Temple University Fox School of Business ‘17, Course Hero Intern

• I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

Dana University of Pennsylvania ‘17, Course Hero Intern

• The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

Jill Tulane University ‘16, Course Hero Intern