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: MGT 2250 —— Lesson 22
Introduction to Hypothesis Testing
July 21, 2011 ‘ Deﬁnition: Hypothesis: 1. A tentative explanation for an observation,
phenomenon, or scientiﬁc problem that can be tested by further
investigation. 2. Something taken to be true for the purpose of argument
or investigation; an assumption. 1.) Two competing hypotheses are proposed.
a.) The Null Hypothesis
The hypothesis to be tested. Represents the “status quo” —— no difference. Always includes the eguaiity.
This hypothesis is either rejected 6mm. %[ 6/17 6X” fl“; Pia/FEM
,o " ——
M u
s L A
lief/1 "‘ X SWM WII‘U‘
p .. * L. p
Ho ‘ /‘ 2: K
b.) The Alternative (Research) Hypothesis
Represents a “difference”. This hypothesis is accepted only if the Null hypothesis can be rejected.
Always is expressed as an inequality, never include a an e ualitv’. ., u
(jab: /‘ 7 A\<
Ha; /A 4" f,
Ha; # # k Since these are competing hypotheses, oniy one can he accepted. if we reject Ho we
must accept Ha. lfwe don’t reject Ho, we can’t accept Ha. W L22lntroHypothesisTesting  l — 2.) Examples of hypothesis tests a.) b.) Research: An automaker’s Model “A” averages 24 mpg. They
want to increase the mileage, so they install a new fuel
injection system on 100 of these cars. After a period of
having these cars driven, they measure the mpg on them
to see if there has been an increase. [ﬁll/Q: ﬂﬁg‘f Testing a claim: Orange juice cartons say they contain 32 ounces.
Consumer protection tests a sample of them to determine
if they actually contain less than 32 ounces. Flo‘ﬂ 331 Haj/4 432 Quality control:
Parts for a machine must be 2 inches long. If they are
longer or shorter, they are rejected. Mos/L 1Q
Harem L22 lntroH ypoth esisTestin g 3.) Types of Hypothesis Tests
21.) One—tail U per (lT—U) LZZIHH‘OHypotheSisTesting 4.) Potential errors
a.) Type I
We reject H0 when we shouldn’t. We “ﬁnd” a
difference that isn’t really there. Alpha (<7< ) = Level of Signiﬁcance = Probability of
making a Type I error. wf {34‘ (x, lob/5%: (0% 4 b.) Type II . ‘0 V
We don’t reject Ho when we should. we don’t “ﬁnd” a
difference that really is there. Beta (/5 ) = Probability of making a Type 11 error. U %2 {g Qéﬂﬁp/Dtéiwr Did 2X ( £< ’2
c.) Error Table (l {p OK [\
[+0 l H0 } ﬁ‘
raw F/Nﬁ ]
(me r~ (9.53:: ' H
QC J l \7‘ 66 [H J Fiﬂbm
jot/VT :rT’K’L‘pﬁ 1:1: l w
(+9 J Corleeef ﬁnite/L;
\_ ‘ 1 a
l . L22lntroliypothesiSTcsting ...
View
Full
Document
This note was uploaded on 02/05/2012 for the course MGT 2250 taught by Professor Milne during the Summer '08 term at Georgia Tech.
 Summer '08
 Milne

Click to edit the document details