{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

normal.testing

# normal.testing - A ONE SAMPLE TESTS OF THE MEAN OR OF A...

This preview shows pages 1–2. Sign up to view the full content.

ONE SAMPLE TESTS OF THE MEAN OR OF A MATCHED PAIRS MEAN DIFFERENCE Yes IS σ KNOWN No Yes N > 40 and dist’n unimodal Or N > 15 and dist’n approx. Normal Or Any N and dist’n Normal No Yes C.I. X +/- t* ( S / n ) N > 30 and dist’n unimodal Or N > 15 and dist’n approx. Normal Or Any N and dist’n Normal Yes No Transform Data Or Nonparametric A A.1. A.2. Transform Data Or Non- Parametric Test df df = n -1 HYP. T = x – μ 0 s / n df = n - 1 C.I. X +/- Z * σ / n HYP. Z = X μ 0 σ / n Test TEST STAT. TEST STAT . Note: Guidelines in A.1. come from MM&C Section 6.1 (p.366) and from class notes Part 7 (slide 8) and Part 9 (slide 20). Guidelines in A.2. come from MM&C section 7.1 (p.432-433) and from class notes Part 8 (slide 14) and Part 9 (slide 21).

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

View Full Document
TWO-SAMPLE TESTS OF THE DIFFERENCE OF MEANS (samples must be independent of each other) Yes Normality? Check box A.2. for the sum n 1 +n 2 . Also need n 1 > 5 and n 2 > 5. No Transform Data or Use a Non-Parametric Test Yes S LARGER 2 No UNPOOLED C.I. ( X 1 – X 2 ) +/- t*
This is the end of the preview. Sign up to access the rest of the document.
• Fall '10
• AndyMugglin
• Non-parametric statistics, Parametric statistics, Kolmogorov–Smirnov test, Anderson–Darling test, Nonparametric regression

{[ snackBarMessage ]}