Re: The reason of a notation (The unnecessary explanation).

From: "\"Luis A. Afonso\"" <licas_@xxxxxxxxxxx>
Date: Wed, 30 Aug 2006 18:56:49 EDT

We want to test
____________H0: pA <= pB
against_______H1: pA > pB
at 95% Confidence Level ( Z critical value = 1.645)

Case #1
Observed values
________pA = 30 / 50_____pB = 25 / 50
result
________pooled Z = 1.005_______not pool. Z = 1.010
Decision : cannot H0 be rejected

Case #2
Observed values
________pA = 120 / 200_______pB = 100 / 200
result
________pooled Z = 2.010_______not pool. Z = 2.020
Decision : we MUST reject H0.

This shows the so-called sample size effect: #2 have 4 times more data than #1. The observed proportions pA and pB are the same in the two cases.

___________licas (Luis A. Afonso)
.

References:
- The reason of a notation (The unnecessary explanation).
  - From: \"Luis A. Afonso\"

Prev by Date: Re: Luis A. Afonso's Impertinence and Disruptive Behavior should NOT be tolerated
Next by Date: Re: The reason of a notation (The unnecessary explanation).
Previous by thread: Re: Luis A. Afonso's Impertinence and Disruptive Behavior should NOT be tolerated
Next by thread: Re: The reason of a notation (The unnecessary explanation).
Index(es):
- Date
- Thread