hypothesis testing chandru
TRANSCRIPT
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 1/17
HYPOTHESIS TESTINGHYPOTHESIS TESTINGHYPOTHESIS TESTINGHYPOTHESIS TESTING
Presented by:
CHANDRA SHEKAR B.M.
Research Scholar
Dept. of commerce,
Bangalore University.
Under the Guidance of
DR. K. NIRMALA
Faculty,
Dept. of commerce,
Bangalore University.
1
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 2/17
• Meaning of Hypothesis
• Hypothesis Testing
• Related Terminologies
• Type I vs. Type II Error
CONTENTS
• αααα Level of Significance• Acceptance Region and Rejection Region
• Critical Value
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 3/17
WHAT’S AN HYPOTHESIS
• HYPOTHESIS
• An educated guess or the best answer to a
question, based on the most reliable factsavailable; a guess that will be TESTED
• In statistics, it is a claim or statementabout a property of a population
• It is a proposition the validity of which needs tobe tested
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 4/17
WHAT’S THIS ALL ABOUT?
• HYPOTHESIS TESTING
•• It is a standard procedure for testing a claim aboutIt is a standard procedure for testing a claim abouta property of a populationa property of a population
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 5/17
• STEPS IN HYPOTHESIS TESTING
1.1. GivenGiven aa Proposition,Proposition, statestate thethe NullNull HypothesisHypothesis andand
Alternate Alternate (Research)(Research) hypothesishypothesis inin operationaloperational termsterms
2.2. SelectSelect thethe Appropriate Appropriate statisticalstatistical testtest
3.3. SelectSelect thethe desireddesired levellevel of of significancesignificance
4.4. ComputeCompute thethe appropriateappropriate statisticstatistic fromfrom thethe samplesample datadata
5.5. StateState thethe conclusionconclusion of of aa hypothesishypothesis testtest inin simple,simple, nonnon--technicaltechnical termsterms..
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 6/17
NULL HYPOTHESIS VS. ALTERNATIVE HYPOTHESIS
NULL HYPOTHESIS
• Statement about the valueof a population parameter
ALTERNATIVE HYPOTHESIS
• Statement about the value of apopulation parameter that must betrue if the null hypothesis is false
6
• Represented by H0
• Always stated as an
Equality
• Represented by H1
• Stated in on of three forms
• >
• <
• ≠≠≠≠
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 7/17
STEPSTEP 11:: GivenGiven aa Proposition,Proposition, statestate thethe NullNull HypothesisHypothesis andand Alternate Alternate(Research)(Research) hypothesishypothesis inin operationaloperational termsterms
PROPOSITIONPROPOSITION :: ChangesChanges inin production production process process will will resultresult inin betterbetter
product product (( IncreaseIncrease lifelife of of TyresTyres))GivenGiven:: MeanMean lifelife of of oldold TyresTyres == 4040,,000000kmkm
MeanMean lifelife of of New New TyresTyres == 4141,,200200kmkm
NULLNULL HYPOTHESISHYPOTHESISHH00 == ChangeChange inin production production process process will will notnot changechange lifelife of of TYRESTYRES
ALTERNATE ALTERNATE HYPOTHESISHYPOTHESIS
HH11 == ChangeChange inin production production process process will will changechange (improve)(improve) lifelife of of TYRESTYRES
HH00 : µ = 40,000 km: µ = 40,000 km
HH11 : µ > 40,000 km: µ > 40,000 km
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 8/17
STEPSTEP 22 :: SELECTSELECT THETHE APPROPRIATE APPROPRIATE STATISTICALSTATISTICAL TESTTEST
PARAMETRIC TESTS & NON PARAMETRIC TEST
(Assumes Parameters such as
mean, SD etc)
Z-TEST
(Do not assumes Parameters )
CHI-SQUARE TEST
T – TEST
F-TEST
ANOVA etc
MC NEMAR TEST
RANK SUM TEST etc
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 9/17
STEPSTEP 33 :: SELECTSELECT THETHE DESIREDDESIRED LEVELLEVEL OFOF SIGNIFICANCESIGNIFICANCE
Denoted byDenoted by αα
TheThe probability probability thatthat thethe testtest statisticstatistic will will fallfall inin thethe
criticalcritical regionregion when when thethe nullnull hypothesishypothesis isis actuallyactually
truetrue..
CCommon choices are 0.05, 0.01 and 0.10ommon choices are 0.05, 0.01 and 0.10
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 10/17
The The testtest statisticstatistic isis aa value value computedcomputed fromfrom thethe samplesampledata,data, andand itit isis usedused inin making making thethe decisiondecision aboutabout thethe
rejectionrejection of of thethe nullnull hypothesishypothesis..
STEP 4STEP 4 :: COMPUTE THE APPROPRIATE STATISTICCOMPUTE THE APPROPRIATE STATISTICFROM THE SAMPLE DATA FROM THE SAMPLE DATA
z =
x - µ x
σ
n
Test statisticTest statistic
for meanfor mean zcal
= 3.2
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 11/17
STEP 5STEP 5 :: DECISION &DECISION & STATE THE CONCLUSION OFSTATE THE CONCLUSION OFHYPOTHESIS TEST IN SIMPLE, NONHYPOTHESIS TEST IN SIMPLE, NON--TECHNICAL TERMS.TECHNICAL TERMS.
αα = 0.05= 0.05
Acceptance Region Reject H0
zCal = 3.2
µ
z (Table value) = ± 1.64
0 1.64 3.2
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 12/17
As seen above the computed value of z=3.2 falls in the
rejection region.
Thus we reject the null hypothesis that u=40,0000 kmsand alternate hypothesis is accepted
DECISIONDECISION
ALTERNATE ALTERNATE HYPOTHESISHYPOTHESIS
HH11 == CHANGECHANGE ININ PRODUCTIONPRODUCTION PROCESSPROCESS WILLWILLCHANGECHANGE (IMPROVE)(IMPROVE) LIFELIFE OFOF TYRESTYRES
HH11 : µ > 40,000 km: µ > 40,000 km
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 13/17
REJECTION REGION / CRITICALREJECTION REGION / CRITICALREGIONREGION
Set of all values of the test statistic that would cause aSet of all values of the test statistic that would cause arejection of the null hypothesisrejection of the null hypothesis
CriticalRegions
5%5%
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 14/17
CRITICAL VALUECRITICAL VALUE
Fail to re ect H Re ect H
Any value that separates the critical region (where we reject Any value that separates the critical region (where we rejectthe null hypothesis) from the values of the test statisticthe null hypothesis) from the values of the test statistic
that do not lead to a rejection of the null hypothesisthat do not lead to a rejection of the null hypothesis
Critical Value( z score )
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 15/17
PP--VALUE VALUE
TheThe P P --value value (or(or p p --value or probability value) is the probability value or probability value) is the probability
of getting a value of the test statistic that isof getting a value of the test statistic that is at least at least as extreme as extreme
null hypothesis is true. The null hypothesis is rejected if thenull hypothesis is true. The null hypothesis is rejected if the
P P --value is very small, such as 0.05 or less. value is very small, such as 0.05 or less.
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 16/17
TwoTwo--tailed Testtailed Test
H H 00:: ==
H H 11:: ≠≠≠≠≠≠≠≠
αα is divided equallyis divided equally between thebetween thetwo tails of the criticaltwo tails of the critical regionregion
Means less than or greater thanMeans less than or greater than
Values that differ significantly from HValues that differ significantly from H00
7/29/2019 Hypothesis Testing Chandru
http://slidepdf.com/reader/full/hypothesis-testing-chandru 17/17
1