Академический Документы
Профессиональный Документы
Культура Документы
on the !asis that the higher the !est and "ice "ersa#
We used di&ension reduction to condense the $nternet %sage 'uestions into 3 factors as (e sa( fit# We (ere a!le to su!di"ide the 'uestions into 3 factors (hich (ere na&ed )ntertain&ent *users (ho use the internet &ainl+ for entertain&ent purposes such as do(nloading &o"ies, pictures etc,, $nfor&ation *users (ho use the internet to loo- at ./ progra& listings and go on chatroo&s,, and 0erious *users (ho use the internet &ainl+ for loo-ing for 1o!s and to !oo- trips,# .hese (ere the 3 factors (e (ere a!le to decipher fro& the rotated co&ponent &atri2# We also chec-ed their co&&unalities and sa( that the "aria!les3 data (as (ell represented in the e2traction# 0atisfaction 4 5o+alt+ When (e ran factor anal+sis on Q 64, 5, and 6 it did not gi"e us rotated co&ponent &atri2# We found that it is in fact one factor !ecause there is high correlation !et(een three "aria!les# And (e can sa+ that a person (ho is satisfied can also reco&&end and repeat purchase dell# Manufacturer $&portance7perfor&ance Ratings When (e run factor anal+sis on Q 6 8 (e found that there are t(o su! factors, one is con"enience and other is technical# 9on"enience are those (ho seefor ease of ordering, custo&i:ation, 'uic- deli"er+, co&petiti"e pricing, progra& run 'uic-l+, and readil+ upgrade# .echnical are those (ho see- for attracti"e design, no technical pro!le&s, appropriate soft(are, internet access, and eas+ asse&!l+#
;ricing
.he output suggests that 8 #<= of the "ariance is e2plained as sho(n !+ the co&&unalities data# .he co&ponent &atri2 sho(s a "er+ high correlation of 0# 4 !et(een the factor and the purchase li-elihoods in price increases !+ either 5 or <0=# )arl+ Adaptor Attri!utes Despite the fact that the 'uestion di"ides the responses into three &a1or factors, i#e# Mar-et Ma"en ite&s, inno"ati"eness and opinion 5eadership> !ased on the output it (as o!ser"ed that &ar-et &a"en ite&s and opinion leadership 'uestion responses (ere &ore or less the sa&e# ?actor anal+sis (as done a! initio using 3 factors (hich (ere later reduced to 2 (hen it (as felt that the latter ga"e a &uch &ore &eaningful result and confor&ing to e2pectations# *$n the for&er &ar-et &a"en ite&s and inno"ati"eness (ere e2hi!iting positi"e correlation (hich appears to !e fla(ed as a person (ho li-es to introduce ne( ite&s cannot possi!l+ thin- unpro"en ne( products are a (aste of ti&e, $t can !e o!ser"ed that the co&ponent < characteri:es respondents (ho are earl+ adopters (ho also share their opinions (ith others# @n the other hand, co&ponent 2 contains respondents (ho are inno"ators (ho li-e to tr+ different things# 0pecial attention should !e paid to Aunpro"en ne( product are a (aste of ti&e3# $t is negati"el+ correlated to other inno"ati"eness 'uestions (hich &a-es sense as a person (ho ans(ered agree completely to Ali-es to tr+ ne( things3 (ould &ost definitel+ &ar- disagree completely on Aunpro"en products !eing a (aste of ti&e3#
?actor of ?actor After factoring on all !loc-s (e run factor on factors (hich told us that (hat factors that are not rele"ant and (e should ignore it in &a-ing cluster anal+sis# We found that technical and e2pert opinion leader are gu+s (ho are al&ost sa&e for e"er+ co&ponents so (e ignored it for our cluster anal+sis Rotated Component Matrixa 9o&ponent < 2 3 4 )ntertain&en #028 D#0C< #886 D#033 t $nfor&ation D#024 #004 D#00< # 62 0erious D#050 #828 D#22 D#08 0atisfied #863 D#035 #0C8 D#0<2 9on"enience #C85 D#0<5 #06C D#244 ;rice #5C6 D#03 D#08C #<42 $ncrease $nno"ator D#038 #63< #453 #<62 )2traction MethodE ;rincipal 9o&ponent Anal+sis# Rotation MethodE /ari&a2 (ith Faiser Gor&ali:ation# a# Rotation con"erged in 5 iterations#
Rotated Component Matrixa 9o&ponent < 2 3 4 )ntertain&ent #0 0 #046 #84C #0C0 $nfor&ation D#<8 #C6 #065 D#08 0erious D#<02 #056 D#<28 #68< 0atisfied #866 #<28 #022 #003 9on"enience #80< D#<42 #05< D#008 .echnical #25C #530 D#2 2 #350 ;rice $ncrease #4 8 D#030 #026 D#<32 )2pert and @pinion #05C #63< #530 D#048 $nno"ator D#0C4 D#<03 #2C #C05 )2traction MethodE ;rincipal 9o&ponent Anal+sis# Rotation MethodE /ari&a2 (ith Faiser Gor&ali:ation# a# Rotation con"erged in 6 iterations#
After e2cluding technical, and e2pert opinion (e ca&e up (ith the follo(ing factors and (e can see that e"er+ factor is 'uite different for e"er+ co&ponent# 9luster Anal+sis Factor 1 Entertainment Satisfied Serious Information Convenience Price Innovation B B B 2 2 B
Cluster 3 4
We used the follo(ing color codes for each cluster# 9luster < is green# 9luster 2 is purple# 9luster 3 is !ro(n# 9luster 4 is orange color coded# 9luster < is !est in inno"ator and infor&ation (hereas cluster 3 is (orst in ter&s of inno"ation# 9luster < uses the internet for infor&ation (hereas cluster 2 does not use the internet for infor&ation see-ing purposes#
9luster 4 is &ost price sensiti"e !ut does not use the internet for serious purposes such as !oo-ing trips and 1o! searches# 9luster 2 is a "er+ serious user and is not "er+ price sensiti"e# 9luster 2 (ould !u+ e"en if
prices are raised# 9luster < (ants the &ost entertain&ent and "alues con"enience (hereas cluster 2 does not thinDell is con"enient# 9luster 4 is not a serious user nor an entertain&entDt+pe user, therefore cluster 4 uses the internet for infor&ation &ostl+#
9luster < is &ost satisfied and uses internet for entertain&ent (hereas cluster 2 is least satisfied#
9lusters < and 4 are price insensiti"e# 9luster 3 is &ost inno"ati"e !ut least price insensiti"e# .his sho(s that cluster 3 (ould !e price sensiti"e !ut also is the &ost inno"ati"e out of all the clusters# 9luster 2 is least inno"ati"e#
9luster 4 H 0atisfied
.his seg&ent contains consu&ers (ho surf the internet for entertain&ent purposes, re'uire a lot of con"enience and highl+ satisfied# .hese are sort of the !est custo&ers in the sense that the+ thin- Dell &eets their co&puting needs ade'uatel+# .he people in this seg&ent !elong to 40 D 44 age seg&ent, are college graduates and ha"e inco&e in the range I50,000 H IC4,