H
Tb< p«T|i«:t bmUv1 tiofc Siffw? ArtMilfcr «l ^tlett*ar*«» <.»»• *» łt*au,vJ
tfe* mav be diffio.lt U*aiulYW (Creen cl *1 W) and « dlhltłftft ot the ongtmd CWKł* itui* to nrpeąicd wwwy wutufc iReimU- IWk hiMhcrcrmie, MdUminfl and foraUT 1 .!000) a. ftUfl łhm the «d«Ml pw»W* oi undcn*bn» ttie tVłph. *ur%<-y muy prrpi-mato of tbeca&cn*
untóuiwatMHi. at the-rescarchcrs try to lotce * consensus on the .respondent* In r«h ronini o! łhf M»tvny thi> smuuot. ibc «irvey feadus Hit lirtaJ tounds when Ojcre-Udlflfcuhym adnrroiK <W*«n^fftll»il the
rcscłrdwr*ar.- lorce.! 10 'wiUńta) dnom*1 (Greci. et Al IW p. 201). fbi* Icd Grccił« »1 to cooclmle itui, oiic V) (h*votumc ot data amilncd. th* rtwffrch le*m prodmęd ‘Us* dciajted. leUco*iiexi xpcxiłic lormulailons ihal couhi atitaci wnlcr a«rccrnan' (p 2*0)
Uurce is *Uó s pttJWcm malrndolofliód rigpOK**<tb lltlto consensuson wbethei thetechnuiiK: bąpuKww
łgrtUtóriYt in .i|>prruich nr drMftn (<ir«u er s! IW), lk*h ijuanuiatiye nul qtulUaO«- data analysik ICehaupMSf have becn spphcd to rbc-ootcumes of sonmyy {<«r«m<t <d 1999, Huyson et &1 200ft), IW btćk ot nutliudologtcal riaonr >> liiptvltj{hted by ft.c d*«m Owi giiklanceon o v$sXem#te approach u» and
utilUandn of Ddphi u bdOng In Ow liimiurr IHas&on eł al 20001. Iłowem; Iłte Ddphi sui vey may bc the only
Althoo«b «hc Delphi u.łimqiH' U .»vaiiabk> ‘n a r.mnher ol kitnu0.v. u loliowś j «iatulćdlted desigji fJhal allnws S.jr the .nlUcIlon <0 data r. laud to .rpinion» And emstimą Inmi» p«nrł t>( capiti* Ihe ><^e<m.hci ian form a deep undmrandios oj ihconctpt onder invc«tiR<u!pn b\ uamg Uciafton (repctrtkmi, .ummlJed ferdbacL anotiYinUY and «r»tl«<lcd a;ulvsii ot <xpcrl rwponses, Howticr. the procce ol ustng Ihe f>dphi tecluuque mav be mbjecuw and time uuksum.O}.. A topu of ddwle il wl^tljct all O! the >muo* itminj* in «lie tifvt and >.dtocque«: roumis ^an l»e aM«!preh«ut>dy .jprmcd ihron^li dcwript.se ^uiiytual atuiy^ta-.lhc qursłj«n »l*<. rrnraint as to thr •.'alidity ot the dattt cplU-Ufd ihrougb the Ddphi puKcv> llic Uata coliccilon process :»łpeiK«dlva iuhiectnt one. w Uhm ihe <nhens i8e obn-ćm c. qojni Uiabk. analyticai mefh.Kl fcf des£rłpbvc \ijiiiik' Ii. ałUltioo. Ihere are CpnyłUlaaBkwft io imdcrtaiatt}$aDelphiiUtWy wMh a mat-probahility wmplc. r. ii nuy not ho iqur»cntanvr otth» tarjęet ł«qm|un«m ofintereM.
Howowr. in muwnjjięswrck ihe rdauvdv frcqu«nt useo! the Ddphi Uxhnkpiei» supporlive ol ik etfccllwness m devtłopmę Mi understandlng ot a multliu.l.- ..I faue* that mfonn.nurting praetke thć Delphi IcohnUjuf l> eahuihle m tiul hunitttfcuity to jn tmdentandłn^ id £<*mpkx nuryin^.łmcrjuy m ahnatmnft whray ^^flAłńUtg«C4MuvittUi Mtw%<atęertt may n.K be featifcte oMiigolher merhods. swehm tndwdual and |.M-> gtrup Ua.-TV.CVY-, Ihe t>dphu»chniqiio ifiwnn n-orthwhile when ll is ut«l in jw<*ciaiinn w Uh other ij.tamUaUYo and ipuhiuiiw re®nch appowd.es. m ihi» tmy lead U. the devdn)nmtit of an eHec«ve and .omprrłw n*iv* nfuk- ^..uling-d nu. <iny. |'taUhx Omjdl. tlwr tHlplii technigue lllrm lor the <hsxk>pn«mi • •i u...! ..i eapetu rr^«K.»cx thv not [■ wsilłk |p tllujmnate hy c)tlu-r melhods