2. Hypothesis
Hypothesis:
• Hypothesisisconsideredasthe principal instrumentinresearch.
• Its mainfunctionisto suggestnew experimentsandobservations
Definition:
• A hypothesisisaconjectural statementof the relationbetweentwoormore variables.
(Kerlinger,1956)
• Hypothesisisaformal statementthatpresentsthe expectedrelationshipbetweenan
independentanddependentvariable.(Creswell,1994)
• Hypothesisneedstobe structuredbefore the data-gatheringandinterpretationphase of the
research
• The hypothesisgivesdirectiontothe collectionandinterpretationof data
Characteristics ofhypothesis
• Hypothesisshouldbe clearandprecise.
• Hypothesisshouldbe capable of being tested
• Hypothesisshouldbe limitedinscope andmustbe specific
• Hypothesisshouldstate relationshipbetweenvariables
• Hypothesisshouldbe consistentwithmostknownfacts,amenabletotestingwithina
reasonable time
Null and Alternative Hypothesis
• The Null hypothesis(H0) isa claimof “no Significance difference between2variables”
• The Alternative hypothesis(Ha) claims“H0 is false”.OpptoNull Hypothesis
Confidence interval
• It isexpressedasapercentage andrepresentshow oftenthe true percentageof the population
whowouldpickan answerlieswithinthe confidence interval.
• The 95% confidence levelmeansyoucanbe 95% certain;
• the 99% confidence levelmeansyoucanbe 99% certain.
3. • Most researchersuse the 95% confidence level.
Level of Confidence
• A confidence level referstothe percentage of all possiblesamplesthatcanbe expectedto
include the true populationparameter.Forexample,supposeall possible sampleswere selected
fromthe same population,andaconfidence interval were computedforeachsample.A 95%
confidence levelimpliesthat95% of the confidence intervalswouldinclude the true population
parameter.
Hypothesistestingof means
Mean of the populationcanbe testedpresuming
differentsituationssuchasthe populationmaybe
Normal or otherthan normal,itssample size maybe large or small,variance of the population
may be knownor unknownandthe alternative hypothesismaybe two-sidedorone sided.our
testingtechniquewilldifferindifferentsituations.
We may considersome of the importantSituations.
4. Hypothesistestingfor differencesbetweenmeans
The null hypothesisfortestingof differencebetweenmeansisgenerallystatedas H0 : m1 = m2 ,
• wherem1 is population mean of onepopulation andm2 ispopulationmeanof the second
population,assumingboththe populationstobe normal populations.
• Alternativehypothesismaybe of not equal toor lessthanor greaterthan type as statedearlier
and accordinglywe shall determinethe acceptance orrejectionregionsfortestingthe
hypotheses.
• There may be differentsituationswhenwe are examiningthe significance of difference between
twomeans,but the followingmaybe takenasthe usual situations.