4-ARM-1 h

7/26/2019 4-ARM-1 h

http://slidepdf.com/reader/full/4-arm-1-h 1/22

Chapter Four

The research processTheoretical framework andhypothesis development

BY RAJA. MA!AR !AM""#

1

7/26/2019 4-ARM-1 h


The need for a theoreticalframework

A theoretical framework is thefoundation of hypothetico$deductive research.

Represents your %elieves on howcertain phenomena are related toeach other.

2

7/26/2019 4-ARM-1 h


3

7/26/2019 4-ARM-1 h


The process of %uildin& a theoretical frameworkincludes'

(. )ntroducin& de*nitions of theconcept or varia%le in yourmodel.

+. #evelopin& a conceptual modelthat provides a descriptiverepresentation of your theory

,. Comin& up with a theory thatprovides an e-planation forrelationships %etween the

varia%le in your model. 4

7/26/2019 4-ARM-1 h


aria%les'

• A varia%le is anythin& that cantake on di/erin& or varyin&values.

• "-amples of varia%les are'production units0 a%senteeismand motivation.

5

7/26/2019 4-ARM-1 h


Types of varia%les '

($ #ependent varia%le

The dependent varia%le is thevaria%le of primary interest to theresearcher. Throu&h the analysisof the dependent varia%le ispossi%le to *nd answers or

solutions to the pro%lem

6

7/26/2019 4-ARM-1 h



+$ )ndependent varia%le

The independent varia%le is &enerallycon1ectured 2an opinion or idea formedwithout proof or su3cient evidence4 that anindependent varia%le is one that in5uencesthe dependent varia%le in either a positive orne&ative way. That is0 when the independentvaria%le is present0 the dependent varia%le is

also present0 and with each unit of increase inthe independent varia%le0 there is an increaseor decrease in the dependent varia%le.

7

7/26/2019 4-ARM-1 h



,$ Moderatin& varia%le

The moderatin& varia%le is thepresence of a third varia%le thatmodi*es the relationship %etweenthe independent and thedependent varia%les.

8

7/26/2019 4-ARM-1 h



6$ Mediatin& varia%le

The mediatin& varia%le is one that surfaces%etween the time the independent varia%le

start operatin& to in5uence the dependentvaria%le and the time their impact is felton it. The Mediatin& varia%le surfaces as afunction of the independent varia%le

operatin& in any situation0 and helps toconceptuali7e and e-plain the in5uence ofthe independent varia%le on the dependentvaria%le.

9

7/26/2019 4-ARM-1 h


Theoretical framework

• The theoretical framework is the foundation on whichthe entire research pro1ect is %ased.

• aria%les are relevant to the pro%lem situation andidenti*ed throu&h8 interviews0 o%servations andliterature review.

• "-perience and intuition also &uide the developmentof theoretical framework after identifyin& theappropriate varia%le0 the ne-t step is to ela%orate thenetwork of associations amon& the varia%les0 so thatrelevant hypotheses can %e developed andsu%se9uently tested.

• Based on the results of hypothesis testin& the e-tendto which the pro%lem can %e solved %ecomes evident.

10

7/26/2019 4-ARM-1 h


Theoretical framework

• The theoretical framework'oRepresents and ela%orates the relationship

amon& the varia%les.

o"-plains the theory underlyin& these relations.

o#escri%es the nature and direction of therelationships.

Just as the literature review sets the sta&e

for a &ood theoretical framework0 this inturn provides the lo&ical %ase for developin&testa%le hypothesis.

11

7/26/2019 4-ARM-1 h


Basic features in any theoreticalframework'

The varia%les considered relevant to the studyshould %e clearly de*ned.

A conceptual model that descri%es the relationships%etween the varia%les in the model should %e&iven.

There should %e a clear e-planation of why wee-pect these relationships to e-ist.

• Both a schematic dia&ram of the conceptual

model and a description of the relationships%etween the varia%les in words should %e&iven0 so that the reader can see and easilycomprehend the theori7ed relationships.

12

7/26/2019 4-ARM-1 h


!ypothesis development

• !ypothesis can %e de*ned as lo&icallycon1ectured 2an opinion or idea formedwithout proof or su3cient evidence4relationships %etween two or more

varia%les e-pressed in the form of testa%le

• :tatement' By testin& the hypothesis andcon*rmin& the con1ectured 2an opinion oridea formed without proof or su3cient

evidence4 relationships0 it is e-pected thatsolutions can %e found to correct thepro%lem in countered.

13

7/26/2019 4-ARM-1 h


:tatement of hypothesis 'Formats

($ )f$Then statement

To e-amine whether or not thecon1ectured 2an opinion or idea formed

without proof or su3cient evidence4relationship or di/erences e-ist0 thishypothesis can %e set either propositionsor in the form of )f$Then statement.

"-ample')F the employees are more healthy0 T!";they will take sick leave less fre9uently.

14

7/26/2019 4-ARM-1 h


:tatement ofhypothesis ' Formats

+$ #irectional and non directional hypothesis

)f0 instatin& the relationship %etween two varia%les orcomparin& two &roups0 terms such as positive0 ne&ative0more then0 less then0 and the like are used0 then these are'

Directional hypothesis: %ecause the direction of the

relationship %etween the varia%les 2 positive < ne&ative4 isindicated.

"-ample'

The &reater the stress e-perienced in the 1o%0 the lower the 1o% satisfaction of employees.

Non directional hypothesis: there is a si&ni*cant relationship%etween two varia%les0 we may not %e a%le to say whetherthe relationship is positive or ne&ative.

"-ample'

There is a relationship %etween a&e and 1o% satisfaction.

15

7/26/2019 4-ARM-1 h



,$ ;ull and alternate hypothesis

;ull hypothesis may state that thecorrelation %etween two varia%les is e9ual

to 7ero.The null statement is e-pressed in terms ofthere %ein& no relationship %etween twovaria%les.

The alternate hypothesis0 which is theopposite of the null0 is a statemente-pressin& a relationship %etween twovaria%les.

16

7/26/2019 4-ARM-1 h



The steps to %e followed inhypothesis testin& are'

(. :tate the null2if re9uired0 usually

not necessary i.e.0 hypothesis4and the alternate hypothesis.

+. Choose the appropriate statistical

test 29ualitative or 9uantitativeresearch methodolo&ies4.

17

7/26/2019 4-ARM-1 h


Cross-Sectional Research

• Cross-sectional research is used to examine one variable in diferent groupsthat are similar in all other characteristics

• #e*nition of Cross$:ectional Research

• If you wante to !now if the "ercenta#e of wo$en ia#nose with %reastcancer increases with a#e& how woul you #o a%out answerin# this 'uestion()ne way you coul *n the answer is to loo! at three #rou"s of wo$en who

are si$ilar %ut of i+erent a#es, et.s say your three a#e #rou"s are 20-35years& 36-50 years& an 51-65 years, /ou can then calculate the "ercenta#e ofwo$en in each #rou" that hae %een ia#nose with %reast cancer, hisinfor$ation can then %e use to answer your 'uestion,

• his is an ea$"le of cross$sectional research, Cross-sectional researchinvolves using diferent groups o people who difer in the variable o interestbut share other characteristics, such as socioeconomic status, educationalbackground and ethnicity ,

• In the ea$"le a%oe& the aria%le of interest was a#e %ecause you wante tosee if any chan#es were notice in #rou"s of i+erent a#es, y loo!in# atsi$ilar wo$en in i+erent a#e #rou"s you can assu$e that any i+erences%etween #rou"s can %e attri%ute to a#e i+erence rather than another factor,

7/26/2019 4-ARM-1 h



• Cross-sectional research stuies are %ase on o%serations that ta!e "lace ini+erent #rou"s at one ti$e, his $eans that there is no e"eri$ental"roceure& so no aria%les are $ani"ulate %y the researcher, Instea of"erfor$in# an e"eri$ent& you woul si$"ly recor the infor$ation that youo%sere in the #rou"s you are ea$inin#, ecause of this& a cross-sectionalresearch stuy can %e use to escri%e the characteristics that eist in a

#rou"& %ut it cannot %e use to eter$ine any relationshi" that $ay eist, his $etho is use to #ather infor$ation only, he infor$ation $ay then %euse to eelo" other $ethos to inesti#ate the relationshi" that iso%sere,

• et.s use the "reious ea$"le to unerstan how this wor!s, In the cross-sectional stuy to ea$ine if there are i+erent "ercenta#es of wo$enia#nose with %reast cancer at i+erent a#es& you *n out that the"ercenta#es are hi#her as the a#e #rou" increases, his infor$ation oes nottell you why %reast cancer ia#nosis increases with a#e only that it oes, Ifyou co$%ine this infor$ation with other research& you coul use it to eelo"a hy"othesis a%out why %reast cancer ia#nosis increases with a#e, /ou woulthen nee to use other research $ethos to test your iea,

7/26/2019 4-ARM-1 h



• Cross-sectional research stuies all hae the followin#characteristics

• ta!es "lace at a sin#le "oint in ti$e

• aria%les are not $ani"ulate %y researchers

•

"roie infor$ation only o not answer why• In "sycholo#y& cross-sectional research stuies are often use

%y researchers stuyin# eelo"$ental "sycholo#y, heyoften o this to ea$ine chan#es in si$ilar #rou"s at i+erenta#es, oweer& a cross-sectional stuy $ay also co$"are#rou"s fro$ i+erent %ac!#rouns, It woul %e a cross-sectional stuy if you $easure the reain# a%ilities of si-year-ols in low& $ile& an hi#h-inco$e fa$ilies, /ou coulo%sere any i+erences in reain# a%ility at that a#e for thearious inco$e #rou"s in a co$$unity,

7/26/2019 4-ARM-1 h


on#ituinal Research

• Longitudinal research is used to study individuals at diferent stages in their lives.One group is studied over many years

• #e*nition

• ae eer wonere how eelo"$ental "sycholo#ist stuy chan#es oer thelifes"an( ow o chilren.s career as"irations eelo" oer ti$e( ow oesiorce a+ect chilren( ow sta%le is "ersonality( hen o eelo"$entally

elaye chilren receiin# s"ecial serices catch u" with their "eers& if they o(hat are the "hysical an $ental health conse'uences of %ein# in a war one(on#ituinal research can answer those 'uestions an $ore,

• lon#ituinal stuy is correlational research which follows one #rou" ofiniiuals oer a lon# "erio of ti$e& "erha"s ecaes, Correlational researchoes not esta%lish cause an e+ect& %ut only a relationshi" %etween two factors orcharacteristics, o %e e*ne as lon&itudinal research& researchers $ust

ealuate the su%:ects at a $ini$u$ of two i+erent ti$e "erios so they can %eco$"are, ;re'uently& researchers $eet with the su%:ects $any ti$es on a re#ular%asis& for ea$"le eery two $onths or eery *e years, he len#th of ti$e ise"enent on the to"ic of the research& the len#th of the stuy& an the a#e of thesu%:ects, he ientifyin# feature of lon#ituinal research is stuyin# one #rou" ofiniiuals oer ti$e,

7/26/2019 4-ARM-1 h


4-ARM-1 h

Documents