V . V; .. ... . .. ...A~V...VV..VVV:V.. . . a... $5..» k... .Vsfimmnfi. 31.3.33...» . .E‘x‘d-ifl“; \o‘ If A .3; V . s in am. 7% .finmwalu mug}? :V um. 13...); 3.. .3... 25. .. u n V . . . VV .. n. v” .VVVV.V.V... Er... . VV. 5. . V V 35.11... {.1595 fish»... . -3fi3 bummmbvmwnmwe 25.2 .2 tV .. BK... I . I guns; «I\ .3. . V : \Ixxvflwfluflrfi )EJN‘fl-Ifly hzfifii) . 3. . fi..a....n..hVV..VVV4.V.. :2. %fl .1... V. . I 1 1:! .Q’ ‘. .1 ‘4 3 )1, . .VVVV V .qu m? VHmanwVVVwmmmu§mr .3. $5.63. .12... A V V V . finwwwwum 3. . V \ I .I\.. )isx .33.. :\VV...V V A #33:.va V . .5. \IIIYT V n. a. VmexahbfiAfia u.) . L...1....5. V . V... 3.39% r Mega. . . . . s... .u n.1g . 0' . "‘ .2! .'.‘ 9...}, 1...... 1. V1 73's.... V i0.. 1.361.. .3 V {In}: ibhiflk A 2.533.... fifiwnfi. V s ...V.\.Zx.5?.d.. .V 1. 1.3V . .VV QWVVWNTJ; :03 I: I1 N. .n . V rant). . l .1) .49.: A 3%.. Nan. I I... a x N V . V1... hum! V fibrnw V4 Wag... “an. L3 V 2. V . . VV . . V A . I3. . 2 hangahwam Ignaa 52.... V. .2}... .2 A .“fla‘a‘xnfi hmm.‘ .1stqu1 waafldmmfiafi 1V3: ILA. 9“! 1%. iii-$1115 I. z??? 17. V.I(.1V £1) 31.: A. 13211. I! «I . 33......an .VVMmV a78wwunw¥a3hfimw ting .V V .. V135». I x. . fiqu§2 $9; 3.. V is... \IIIII . MV .. vi. 5.9W1Iwrl..onffi 71 11.. V. : . A1 :\ XVIVHVNJ V.V rfixxxIVIHazfifiIiI . x )«.K.VV.4~.. . . I \VtVVvV‘...V..sVIx.t. I 33 2:. V}i1\.\»§.43131’\. .1313.le 1 91).... [1 fig: 2.. . ”Tux A1... 11.0 I: I. .w. 1:.) .33...) V25 a V. . V. 3. VV V... x...» g.&F?.1.A§wfi 3am? gsiufié. {Wuhanbd ‘0» . I W. .. .V. VV V 3 V .VV. .3. Venn V . a... .Vpfififiz? V *§' ‘15: . a V 1.... 13:...3. jg 35.: V )V ., .V 1. #0.“ T31.) It. .52}... .351: 3% . -VV (.1 Q2211I31iuic. [1'3 .1} V131Iv...~7 .Iiliithfi 11" 3.5 1 I} V. .K 0“ fi11..in.1..VlV-y3.£¢.13\.) .. Nu- ... 3.3:}... ixuh $3.35} in.“ I l. w 75' 4.! 11th.. JIL’VXI lxvzauk. :03.) .SkuV..\5>yb. 15.5.. .7713}..th1 altiflumxglnfiigxa )3 V \ V V {inakilttfi a . .. m 2 V . V x .3 13.51.. v.5 15m: .. £52.53fluh .33.. V s .2333} .1, us . x I 31313.3"... «mfiah . atixfiil)il\1. I...“ . Jul: 2 L 4 .35 V V. 2» 35.. V.. V. .b. :t 51.2.. is 3. I1 skin‘s {A V I) . 7 X ~«l. 9.. (I . , . '1’ ;"‘1; AI: . .93.}... . hi d. Mmfizkng .V. . L 5.. 1V»... 1.5.1.9. Vfiafinw..u1 In: Ahflhufi IihhafixmkI‘I S 33.x...) 533111.39. . IMLAWPI. II... 31%.! th. ”A. u... 10“....le3 C Kw.” be”. .V . . . jinn. 11;...» W>Ixflanaa3zx10t I “\II.MI1I..1§J§ ‘ 1.4. 7353.. . . V H... A. .. i .. TR: .2 a ..I Z... {Silks V. V ALLY}. . VV .V V r$1.wau_hu.lxv1xh1nxnwfifivvli2uni. N13141:? . 1:... .11....1 t§$§ V .1 ... a .11.. 1 V VV L... . c I In??? V V . V. . . 1.... .1)... .. 3.1V 33.. inf“. Sexlrzg $4,311 . Y’Itv a. it!!! £53.99? V“: we. 3 x 1 I3..." s. . V H.131.‘.3\1..\>313I Vuimxtfl. 1L... 1 fr? 53131.3“..31lt4 V V 35?!!! l: . a . LII. V r9047 3...... . .V A :3 V I! . 1......12%\Vul\IIIIL\))IX1{I\ 11:1... abikflfli :IX‘JI 1. V1.4 \: 1733‘7; élsuu‘ at .- ...~V V 1.. Idea“. .. . :7 .84.... V . I 551113....) Ihfiv 3idUflN§fl343Inuufithmflfitux\\1:1 $31... \7 ngzIHLISIHV V dhkuflwui: \ I Z.- igtz‘iilk, A . 1:3... .. . . 5.3V x 3.. 5. .r V. .V 15 II. 11V 3 In“... 2.5.! ._ I1 .5. 5.. If it n.3,: .. rxxufln 5.1.. .1347: V..I(c.x.:I..V\33 Gr}... ...V r..xatxuxmas.... \vtxjysis.sxxss\3)s..x 19%... 1904 .V 1 . I V5.3...I1yluifi 333.“.va 3. 2 1.5.3.“. I . i}. , .3! VIIQSYOJ... 1 \.u>\31)\fli III);I..1\.IX..I.$N.HI11.VHII I fit. ...~Lv7nfi.v..vufl“xi:fitofi§11 :4.qu .. I! .. x15...) 7.. 17 II 1 V1)»; :51 )kfv.‘ I 3 III. “ :3 it Va: 7% I\w )2 1.. V V I ..VV . V: V. . V. x % 1.3V . V V .r 3§¥ a? V...Afi1.VV...V.V:VVVVS...V.VV......2.V%:.2VV.V..LV;. 2. V.....V2V-A.AV.._.V.?.V...V.. a... 3.... .flnfifwkvnpania 13.4 V. .. a. V a... it: ~ x11171xsp§1xi1fi§a§t1i3f€fl .. _ V. I)II.I..1)I1_...>: 7\.(....:V:.§)IV1..1..I...VFL.VS 1511;31:333‘; 3.31:}. 3&5qu Q. l. 3113. \ .30 V u (Vt-t; .4 )1} 3 :3... .1. IzaIAJVaIxIa _ «. V..VI.:..:V3..I1H .. 1...»)..VIIV51 13.11:» x. 3. 31:111.“? \N‘gx xxx: . 21143.5... 3)? Ito‘vfjft; n. I??? E V .331 .V..1.5:“H\3!«. 17.} .00.:111571II15 VV. \H?:51711\)x.: 1...]: ((1)3)>17)V.«u...1:;123 .\ .«VIHI‘SWXI. 7‘. J\ )uv I; .IV... 5 .la \fhh’ .a \l V»: Inaug- x. .41. 31.1.3.1? SI}.11 5......3...‘ 51.15 ..5\2133333$3.V...f?{\. )Is ‘5‘}...(i1Itxsxnt. . .33.}. r! 31.191313: .2.) I: 3! VII...) 19‘ . Ivan“ V w. I a»... {vb 13393.I>s$11.)1\\5.u1}1~u .VV x175..“?1.€ 1\.3\n2k£§131111)¥t\3€.51 $51!: ‘1 IXVVIfgsw>z [\iVI‘i 1‘13: n V I «3.91.971 .4 I! 1 .. «VS. 7.}. £535 1.)...IV\\...~£€1.:VVXI I H.1..)..A....r.fix..1.75\ )11I‘7IdfiXsaflfiz; 3.3.331 S¥1§Wi§$lva 31.31;}; r1312 V V J‘s. : 1 rvi. .. .V 3.133.... 2. .. .....\11....K I. .31....{1 )2. 323V :. ;.....!{I.I]..)..V.V 1A).. .1:31..I.V1......V1..7:s..{£113.}; IahIaixujthIIétw. It! s ‘11 f....f.l'..f€ 1...». .....v.:2X&.k-.1. 3.21 In”; \ )Abu) x121 \11.3In..u.aUD§\.:(:l§ I. H)!....XIIXT5)III>)IIL\$)II11....111I 5:11.1312...9\I1131(\1.3:.3193344255355.513 55er :35 If: . i}: buy...- l‘! 1”“. IV in.) 1V3: 331353333. .3913 I.) .1)! _.\kxfix.}t.IOifi}4TimsV:I1 .3. VI fi.x.3<§1I17.5.{I.V)t\~fi!aaa 2773.5...7 .1 §I52tx§dzla3vl3¥ . .rV PP! . 054.13... . . 3531...“. v 1! so. .3 .. .fl- . :W I.\)I 3‘ h... IKIIBIXII..V..‘1.. fikhbuvwaffliaxxfix ..VV:V1)..JUI.V.VI 31..va I332, In VIVII‘QQWWJ. Y 1:.) :11 .1 2 I » .L L“ .9: .II ..V r‘ashnfl . . I. .f . .ofi auto“ V .0 3 .fl V I. .V . VI» ... .. .rVV. :51 .I “fin. . M??? I .l. T .l: . \(Iv . . :rIV. . 7112:.I.xv§¥..bus I. r .. I 653 \ IVQIIHQVJI115 I‘mxhfi if...) \a\,I.ox\$.u«.3I1.h..QVH1\V.-¢! 2,2113...)xxstISflRIUU.§HHIV.Is181....Aanwxiliuia...Luvfidnuiaxdwzixxd. I191}; 30.12 21: . V I}? EMILIWXLL .v 1’45: fa. Jr x In». 1151.}; V 131.11.01.1(7 31!}? 15\1\laxi.\5\m t..)..-€€.‘~)lx $1th)..131IIIKPY..J\§V.!11)A..\...fi.)3:)lx 1;“..157 if V 1‘? V 0.7M. 1J1??? V. 30$. ‘1. t t .13.“; t: 311‘: 3.3. 133:“)...{31} .II. \V.S..I))I..\.k.1\ 112.1191}...IlelvxmvumaxIIE \Iitaixsilest‘xflvfi. I.) . kg? fit I: .51! . .I III... «Vii 7 3.3.13.1! 3...)? 2.3)..) If 25"! {ISImsatluxax 1351...: I)IIX3VV$.\1II$37¥¢1 \Ts.ix:)..x1 . \ Tn??? . . .191 V I». 1.31.... #631 .p $1.2 1:93:31)! I}. Ilia-1.3.31.1! .YVIIIIAISI. .«1I...1..I.v..4?.)..!1\ 1:11;. o ..r. I 1. WIN“. , $3.. I V H 5.2:). 2.21.93. xgxlsnsia? :1... 531157;... 352.311... 3.3,... . Int If: . £11.11... .. )VJLN. 'lhnn‘l. LI33171IX1I€¢1VI 1.... .11..«.).hp.s\.\.{ns\ .34 . .l. .ftuvlvz.f: a”. .4 V 3... S I 343‘: 3‘\r.(5p $77.13“? Juvinali$31§dfiizli an 0 .V. V. V .guwja 3' 2. .Vi 30.. 111.. , xtflnufistIthIflrxS‘axaxh:Ir}.315x13: V r 2.511: «n: z 133 . I}... V. V .13» . .. r.é.0.u.:{t.iufi.....¥ Id... fix“! 1.: 1!). .. .thIVVEV.bVVI..V:V.V:V 2. tffifluhlv‘fn It: .34 . . . .. . . g 511).. .. 3 ) 1.4111,...31 1 lilo}! : I €1.13 .5 v £13!) . 3., . a)... x V _. .2 AV. ... 3.2....) .2. V2,... fifluhflfi; .54.... unfit: Lu. “5,2..ch 2 u. I 73:". 5.337... A? E3133»... . “5.31..” )3! VI .. V V.v .. . 5 1 I .t .u '16:}.- 1393 . a..(vs.a.uusxéhsavxzfler.. 51V; «95...». 9.5. I... 1;)? IIV;xiii!)211...!\I)..!?.13..IR...V.VIIfl. _ )IV..\ I 7.3.1.. 3.2.5.1.: .. A 33.5.3.2... 3.9 .V 3 ERIE... 1.131... . .5. V. géil‘l‘ig 2);.(‘52 .5. V2. 30% .TiauALLzhufit. ..: . i. 1.71.5.5: la V 1“]t Qt 51.7.89 1.13... . VI. 2 I ‘Ifi’tl‘ lllll . llgafaVufihahfwifiimuvzzhmmws . {Inn 0.1 Ef‘...( A \\;‘I.\’.IUI§JI.OD” >,-Iflu}fi§ gut-“.91.. . VV ‘4; ‘ _‘ 3.51'v‘z’l; . IXKJI)?‘ 1...)?“ 1..\.:.3>1..vII.)3I111v1\ 3.. V! 1:31? xiég )3 V .f5f‘l‘iitll‘ i P):’).MH$.£§I. .flfi?‘xfi)i)1} V15!) 5K.<731J)1)‘I~1? 1.)... t . . I 1 uni}? IHLNV:I..1rIItinIIuFIH I .III.1..I{)II)...V dawn.” 1......“ I “11.4%! )N. 35.1.1.3.“ Ir h.:.flflu..01.h...;.>.3. Ink? ahv.V...x_71..I,VI..I.\.&P’. r... 3 I: . Juana... a k 11.21»... . . ‘3) 7’ IXLrVHwUMHSAtV (.tu Hi... I. 1 .1 llisi;\3$1hxhxx.l.ttxrIIISISt . r1...IV«I:V...\.V\l I131. 1...I3:I..II\.JI1IV)51V|\1 n3 I. .7)... 1V): 2311?"? g. 5.. 1 hr iv 1! .ign‘é. 3.1.55.3. ~11 32.15“)“ t 1.1.23... ‘3 ) ).L}213x5.1.111:5fiL§111>V 7-x.SV1wax? fiqwlugaiugja. 3.11.17 «sifripgk V 2:. L. I? I V a .1316. . )1. fun. LIP §.§§1P th$§I31.1732811%\ \uVIUIdH. 3 I 3.: \JIVV.5V.\33.))~....V url‘lntdua J . 3n...L.N.I..\Y£.V.¢1 x7 13.. £5 13.. 1 1| iiiLttirilfL b I 1.. 2... .3113 I.) V.. 1.1.. I . V ,VV .1: V. A. 1.. . . V . . . I... 1553.3. 3.1.1.1.. 13. 1131.104113.’ V .V V... \flqiaraLNIIISII 081...")... . 112.41.... VV I. .9 \‘ifilaz!’ ...V tr. ix... fill-.19 VILfLrIIIAtxfi V \ II. \\§12kv§\a)fi.!u.fil.xa .7 . a . 3.1.1.... IZVWXVIL. 3131‘ v5}..\\ I}! ll 51‘ . ilt’blfl:\tlhq 3311:133- , 12.33.}.1-{5x 11g??? 3"“.3111 .I Y‘. ‘I l: . («III-ll‘tsvkr A V.. . V V..V...§1¢(1.I..\V§v.. $3.1 At... t . . t 31).. . ..MHM..K3i3g§3i51 77.“. V . P1?!1.r..(rv 1.. V 317‘ IIk.11..I!?17 brixz.‘ 4 1&5) Al 1.. I ll: ‘- trI i 534).- 1 V: I}... ;w\1fIIIsV’171xi\31;x ViixfixnflIJI<4s 12...), I . (fail); $\513§1.II¢§$I$I)IFII . ...11..a5)x:1vvi)21l11!}tt. V.1z§l§..asi V113!” Ii?) .1 )IV.IVcll.3111.\.I3..1I(.fl.I.K . . I .VISIIIII>II.:. . . .. II .a 1.7! 13%;; x V V . V V . . t 13.. 115 II .. ,3; aggix}. 3.91.. .023. . 2:12.119?) .1 LA‘E.‘ I 7!}. .rvfl x: .flcViIxzai I..- :3 i V. V V V’. .run!1.l.6.....l.tl.1 I... .itc‘ . Ivgwf V 5...}: . . 4.1..) 31;; V I! . CiIIUrV .Vflku... 1:... A ginger a... .n V t. 3%....th ILL: V ”$3 V...xV&§§L.VuVI}.VmWflnia—Nr¥£t. . sixzxazflsv tram”). « fit 5.5.9:: 5 .. 7!! .. 4"}... I rlItrr.bk.Afk . Iiihngi iv .3. T V 9.55:5). NHL-Yttfl‘lfuvn (tut. :flli’flafl. .1! $4vlialLAtlztrxfr... .5...}A~ d. .1} VI 1- 6.3. II )Jéxtafavfi it»? c giviri. rV.1}~9.V 31...?! izibflb finflnufi‘\ 021......” V?xaa..zms1m$x12¢bfiafltpxglh.1 33%;. Virgil w it”... if. I in... L . 1 . a. $.Nrrriifitl.i.mamo.£t {nu 77....an..gVV. 1.5....» V. .V 17...) I . . 91 r: 1% .II...V.. V V . I. .1». I...Ii.uu....IlaJ:dfl«.._.¢ )ifllfiztiaxi gun”- I. x I I L. VJ) féjokklhnlrQVIIiowI‘nflIltk II...YV.)!J«.JDIN.14.¢.Ifilfliw)~§§2} : 1 5. V I V I. 7119....5023ul62-{fil9 .Enlinalskxj V V 114.5123 IHPPQLIJIifiv‘zV $14.. lit. I. .VV.VA..A.I.I.I!LIILI&V_II.V V. .leiiIaIIi I.“ ‘(l\xf1}11§s!§ .- A f... \IrrIIvII .. . ..\91\\§J~w.fi\u¢.‘ V3 V3. 3.. \LJVIIAI ...I.\Ht . .a.wini}!€)isla3’viifi)1}7§ . >1. u V a! .y V . .. hudrrhhivfixixxiii 11:”? r... r! i\ .5 V. V . A3 .sévaiaur a. . Luur z . . {Liftltkrit . V...r..Ix.t.V.V...rV..i-V. i VV V1.2...I A V . LVVIIIVS V w. 7h (.3... 1n. V V :53...» . A V was}. . . it: )1..r.¢t.1§1 $70; $9. . is Sfl‘sz...» .. XgflVflil 1),... 2.2:. V .V .1 A fl». . . ha... Vs... V . V I. . V V): >VVV {.31 iv :1: k: Ishééunb a... . 93.3.31... .8... 7).}?‘Kflhu V .Jvizftvuxz. 1.!1TX!I\HI1V x .I .. 1 (4‘; g! .54 V 11.. i? V V I V 1mg“.1’lt’ . 1 . VL.1V.I$.N:LI a 1,135.3. If iii}?! 71 2 7 5V II V V. I .51... «a. .. It”?! ...Vf.I.V. I V‘cxlr‘ it LV V .V Luigi. KE1I. . .1... :s . .V. 3.. .gxafinflrzrfwfl tag 1...: AI11¢IAR |.' t .11 IA ‘05:: AM}, . I 7%., \ nigg“;htf3 V lash .. _ L7 ‘3“ 4 .I’ gall . r {1% Ilioulma .\ 7w {haAfzi . :VL“.H..VLD..VH.xfl.PfilnnnNMerA|bIV;%2 i¢5§kfiia V .I‘\.1Pq.1-f .0. . . . 1 V. '.iaruVV.IL.-Ivll . vl‘il‘VV f «.1! .7! V V V. V . V VA... .WUFLAVZVIINQWréVhUg} . .V V . V V Vi...) ()2...le . AEILEI 31;...“1ffifflg. 1.4.3 . I . V 3 £151.. fill .. V 'hgfigl‘téu- :- V It. 1' . §.I..M‘p lhlwpmmflmwwvrm (AI: I‘v’l ‘X‘ . If? .\ A ‘3-1 . ”Purl... figgafliurqi}! .fldiiki at}!!! gigs; hobrfgmrl. fifimfi. I)?“ fig .. anfitLliwfl.‘ I. In! . . 525.13 .Ii{n?s v L.,.UZI\..AAL£\A7~L1VVA1 It: I all. 2! .~ ’19:}...2 V . 52.1.2 3:? .{t . VV .5: I uni..$.H..file: :Rnhménwrfi View? V . gillixvv étitng: .ZFIL-trlf 18.15:..- . $2.315; .1..."qu t... . V. V WWII... “manuflws.kht.fii.f.hkirfiamm£11¢m.i&uluuvflflvnn3 “unvffiaflntt.!2:flti )ILI. .. 21:15.. .J‘u Hm“; I...“ «I’M... drag: V 19?? :1 .. .‘lfiét. 02.81111: t»§...17‘i’1 A . . ertiti.?15$¥.h~k.t‘ lI . ~ I .1141}?! 52!:1; 7 gr 5151‘ . stov§~ix..Vd V 5:13.... 1. I... hung; .1?st .r . V V V. V...» . Vii. . V I .. .V {1.}.Vte? 152..-..- 1.1.5.1... Irkziiffifitu ti... 2 III....?..I>I£I.{;IKI:\LI fflnk’tlrib xiii I§~.HNF§S§LRX)HII§§;P.35)..7‘73 3 . V. . 1 £3.22}. . 52 t. 1-1.21... .2. . V2 ..h.1..tu..f¢ if! .2225}- .V...V.....V.z.i..i 2...“. .111 Vée‘..){.nLfisz?Lflu (2.3.1.22... .VII.’qx.n.Ll.7v?IquI-. viiiIIXilzl '1 LP), 12).: 57.5.53?! 11:31.2... t. ushefiuzifnzfii h.fin¥r£.:§2h$.§8§?3fiflftxr .5 Est.- I.V .{l 345‘ V {titrli V ‘ir ...... I}-.t..V!91{§V:if:¢dnm2iin§ 2:52:32... «4.3.1... s S’FKV u}. L- zit-«oh. .1: a. I v 7. . VVIIxVI.Viat‘St+a5u..f$3r..§V§ Ilit§zfluufifit sefirknwfi 8F .:1§:52V.u....2-3£ VisVIIsziiixs: zéiiiisi : §i¥‘zt1{illvl.i§1lillzflkl V i 3...!» 1.32.! fiHxVVntxh... .. 2 . . .. . LIBRAR Y Michigan State Univusity Ill!lllllljflllgllllulllllllzllllllllllllijlzlll THemS This is to certify that the thesis entitled AN EMPIRICAL INVESTIGATION OF THE LATENT AND AFFECTIVE STAGES OF THE CONFLICT PROCESS IN CHANNELS OF DISTRIBUTION presented by Daniel C. Bello has been accepted towards fulfillment of the requirements for Ph.D. Marketing degree in Major profes7/\ \\_-/ Date 0-7 639 OVERDUE FINES ARE 25C DER DAY PER ITEM Return to book drop to remove this checkout from your record. © 1979 DANIEL CUSTER BELLO ALL RIGHTS RESERVED AN EMPIRICAL INVESTIGATION OF THE LATENT AND AFFECTIVE STAGES OF THE CONFLICT PROCESS IN CHANNELS OF DISTRIBUTION By Daniel C. Bello A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation Administration 1979 AN EMPIRICAL INVESTIGATION OF THE LATENT AND AFFECTIVE STAGES OF THE CONFLICT PROCESS IN CHANNELS OF DISTRIBUTION By Daniel C. Bello A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation Administration 1979 ABSTRACT AN EMPIRICAL INVESTIGATION OF THE LATENT AND AFFECTIVE STAGES OF THE CONFLICT PROCESS IN CHANNELS OF DISTRIBUTION By Daniel C. Bello The dissertation investigated the nature of the conflict process occurring among members of a distribution channel. A model of the conflict process was developed by integrating prior theoretical and empirical work in the conflict and attitude areas. The model's scope included the latent and affective stages of conflict. The latent stage was represented by variables thought to cause conflict among social entities. Incompatible goals, domains, and perceptions of the parties comprised the latent preconditions to a conflict relationship. Representing the affective stage was Fishbein's two component compensatory multiattribute attitude model. In his model, overall affect toward an object is a function of a respondent's beliefs about the object and his evaluative responses associated with those beliefs. By integrating the variables comprising the latent stage with the attitude model representing the affective stage, an integrative model serving as the basis for the dissertation research was developed. The central issue investigated was the nature of the mechanism whereby the latent variables lead Daniel C. Bello to the affective stage of conflict. The question arises as to whether a channel member's goals, domains, and perceptions lead to affective conflict toward another member by influenc- ing the belief component or the evaluation component of attitude. In order to determine which attitude component is a mediator between the conflict stages an empirical test was performedijlthe distribution channel for new automobiles. A measure of the intensity of affective conflict held by Ford dealers in the state of Michigan toward Ford Motor Company was obtained. Additional measures were obtained from the dealers for the other variables in the model including goals, domains, perceptions, beliefs, and evaluations. The data collection procedures included telephone pre- notification and a three page mail questionnaire. A cover letter further explaining the academic nature of the project and containing the endorsement of the Michigan Automobile Dealers Association was sent to the dealers along with the questionnaire. These procedures resulted in usable responses from sixty—seven percent of the Ford dealers in Michigan. The principle findings were, 1.) the latent variables were associated with the intensity of affective conflict, 2.) evidence was found for the mediating role beliefs play in linking the latent and affective stages, 3.) the evalua— tion component did not serve as a mediator in the conflict process. Daniel C. Bello The limitations of the research were, 1) the results may not be generalizable beyond the particular channel system examined, 2.) the field research approach did not allow for the control of all variables influencing the conflict process, 3.) the statistical analysis utilized measured the degree of association among variables but did not imply causation. The final section of the dissertation examined the pos- sible managerial implications of the integrative model. Specifically, the section discussed the potential uses of the model for establishing criteria for the selection of conflict reduction strategies. To the memory of my father, Samuel, for all his help in the past. To my mother, Catherine, for her continuing support. iii ACKNOWLEDGMENTS The design and execution of this research required the cooperation of many individuals. Dr. Donald A. Taylor was chairperson of the dissertation committee at Michigan State University. His guidance and encouragement were important factors in the formulation and completion of the project. His role as advisor is gratefully acknowledged. Drs. M. Bixby Cooper and Gilbert D. Harrell also served on the dissertation committee. Dr. Cooper provided guidance in the areas of conflict and channels of distribution. Dr. Harrell provided counsel in the area of attitude theory and measurement. Their advice was always constructive and is greatly appreciated. In terms of the data collection process, the coopera— tion of the officers and members of the Michigan Automobile Dealers Association was crucial. The success of the field research depended on their willingness to complete an anony— mous lengthy questionnaire. In addition, the advice of the association's Executive Vice President, Gilbert L. Haley, was important in completing this phase of the project. The association's support of this academic research is gratefully acknowledged. iv The final acknowledgment is to Ethel Kendall who typed, in an efficient and competent manner, the entire manuscript. Her typing skills and understanding greatly aided the pro— ject's final step. TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . ix LIST OF FIGURES . . . . . . . . . . . . . . . . . . . xi Chapter I. INTRODUCTION . . . . . . . . . . . . . . . . . . 1 Research Objectives JuStification for the Study Scope of the Study Overview II. LITERATURE REVIEW. . . . . . . . . . . . . . . . 12 Channels of Distribution Social Systems Cooperation and Conflict. Channel Conflict Conflict Conditions Attitude and Behavior Models of Conflict Stages of Conflict Latent Conflict Perceived Conflict Felt Conflict Manifest Conflict Effects of Conflict on the Channel Dysfunctional Effects Functional Effects Threshold Concepts Relationship of Conflict to Attitude Attitude Models Components of Fishbein's Model Recapitulation III. CONCEPTUAL MODEL AND HYPOTHESES . . . . . . . . 59 Model Development Conflict Causes Integrative Model Hypotheses Hypothesis Set I Hypothesis Set II Hypothesis Set III Hypothesis Set IV Hypothesis Set V Recapitulation vi IV. METHODOLOGY . . . . . Selection of the Channel Candidate Channel Systems Michigan Ford Dealers Development of Measures Goals Goal Statements Domains Domain Statements Perceptions Perception Statements Beliefs Belief Statements Evaluations Evaluation Statements Data Collection Procedures Initial Research Activities Questionnaire Development Statistical Analysis of Data Limitations of the Study V. ANALYSIS OF RESULTS Analysis of Sample Non-respondents Profile of Respondents Testing of Hypotheses Hypothesis I Hypothesis II Hypothesis III Hypothesis IV Hypothesis V VI. SUMMARY, CONCLUSIONS, AND IMPLICATIONS. Summary of Results Hypothesis I Hypothesis II Hypothesis III Hypothesis IV Hypothesis V Study Conclusions Findings Discussion Implications for Channel Management Conflict Resolution Strategies Criteria for Strategy Selection Implications for Future Research APPENDICES. . . . . . . . . . . A. Initial Cover Letter B. Initial Questionnaire vii Page 70 101 132 152 C. Follow-up Cover Letter D. Short Form Questionnaire BIBLIOGRAPHY viii Page res. ‘. . LIST OF TABLES Table 1. 10. 11. 12A. 12B. Analysis of Respondents and a Sample of Non—responding Dealers Completing an Abbreviated Questionnaire Geographic Location of Dealerships and Responding Dealers in State of Michigan Number of Years Responding Dealers have been Ford Dealers Annual New Car and Truck Unit Sales Volume of Responding Dealers Responding Dealers' Perception of Their Profitability Compared to All Michigan Ford Dealers . . . . . . . . . . Responding Dealers' Perception of Their Relationship with Ford. Matrix of Correlation Coefficients for 15 Latent Variables Factor Analysis of Latent Variables Factor Names and Variables Loading on Factors Regression Equation: Attitude (Criterion) and Latent Factors (Predictors) Summary of Multiple Correlation Coefficients: Belief (Criterion) and Latent Variables (l5 Predictors) Regression Equations: Noncoercive Beliefs (Criterion) and Latent Variables (l5 Predictors) Regression Equations: Coercive Beliefs (Criterion) and Latent Variables (15 Predictors) ix Page .103 .104 .105 .106 .107 .108 .111 .114 .115 .118 .120 .121 .122 13A. 13B. 14. 15. 16. Regression Equations: Noncoercive Beliefs (Criterion) and Latent Factors (4 Predictors) Regression Equations: Coercive Beliefs (Criterion) and Latent Factors (4 Predictors) Summary of Multiple Correlation Coefficients: Evaluation (Criterion) and Latent Variables (l5 Predictors), Latent Factors (4 Predictors) Summary of Multiple Correlation Coefficients: Attitude (Criterion) with Beliefs, Evaluations as Predictors Three Regression Equations: Attitude (Criterion) and Beliefs (10 Predictors) Page .124 .125 .126 .129 .130 LIST OF FIGURES Figure 1. Conflict Portion of Cadotte and Stern Model 2. Integrative Model of Latent and Affective Conflict Stages. . . . . . . . . . . 3. Five Goal Topics and Related Questionnaire Statements 4. Five Domain Topics and Related Questionnaire Statements 5. Five Perception Topics and Related Questionnaire Statements 6. Standard Presentation Used in Telephone Pre— notification xi, Page 61 63 77 80 83 90 CHAPTER I INTRODUCTION Historically, channels of distribution have been primar— 1 However, channel systems ily analyzed as economic entities. are more than a series of markets and a full understanding cannot be obtained by focusing solely on economic exchanges between channel members. Increasingly, channel researchers are viewing channels from an overall social system perspective, treating economic relationships as one important subsystem.2 Such a perspective focuses analysis on additional social and behavioral variables critical to the functioning of distribu— tion channels. Of particular interest is the role conflict plays in distribution systems. Although it would appear that channel members, sharing a common fate in system survival and growth, would benefit through cooperation, conflict often does occur. In fact, Rosenberg and Stern note, ”The process of conflict, however, is pervasive and inherent in channel systems characterized by behavioral interdependence for mutual ends.”3 The variables creating the latent or underlying condi— tions for conflict and the manner in which the later affective and behavioral stages of conflict find expression have been treated by previous researchers and will be reviewed in the l 2 next chapter. However, questions remain as to the mechanisms that contribute to conflict's pervasive and inherent nature. In the following discussion of the dissertation's objectives, the purpose of this research in terms of these unaddressed issues will be specified. Research Objectives The dissertation will attempt to accomplish the following three interrelated objectives: 1. Analyze evidence for a theoretical link between hypothesized latent conflict variables and a later stage of conflict. 2. Investigate the mediating mechanism between the latent and affective stages of channel conflict. 3. Illustrate the managerial implications of estab- lishing criteria based on the mediating mechanism for evaluating conflict reduction strategies. The first goal represents a contribution to distribution channel research by empirically demonstrating the relationship between variables posited to cause conflict and a measure of affect toward other channel participants. As discussed in the Latent Conflict section of Chapter II, theorists have hypothesized a variety of antecedent conditions underyling the channel conflict process. Of particular interest is the typology of conflict causes developed by Stern and Heskett.4 The authors conceptualize incompatible goals, domain defini— tions, and perceptions as the fundamental preconditions for the conflict state between members of distribution channels. Empirical evidence provided by Rosenberg demonstrated these 3 three variables tend to vary for different structural levels of the channel and are somewhat related to satisfaction with reciprocal channel member performance.5 However, the linkage between the hypothesized conflict causes and the degree of affective conflict held toward another channel participant remains largely untested. In addressing the first objective the dissertation research will explicitly test the degree of association between the latent variables and an attitudinal measure reflecting the felt stage of conflict. The second dissertation goal is an attempt to move chan- nel research beyond descriptions of relationships between con— flict stvges and focus on the mechanisms linking the stages. By investigating the nature of the impact latent variables have on a later stage of conflict, the research will address the manner in which the conflict process develops from the underlying covert phase to expressed overt conflict. Speci— fically, the analysis will attempt to determine which component of a social psychological model of the attitude construct is systemically related to Stern and Heskett's typology of con— flict causes. By identifying which attitude component is in— fluenced by the latent variables, an improved understanding of the mediating mechanism between the latent and affective stage of conflict can be gained. Because the model has not previously been applied in a channel conflict context, its advantages and theoretical justification are discussed in the Attitude Models section of Chapter II. The third goal of the research is to illustrate the 4 managerial relevance of identifying which attitudinal compo- nent is influenced by the latent variables. Although manage— ment has a variety of conflict reduction strategies available, theoretically based criteria for strategy selection is notably absent. The accomplishment of the dissertation's first and second objectives provide a basis for specifying criteria use— ful to management in evaluating various techniques for reducing the level of conflict in channel relations. As investigated under the project's first objective, the overt conflict manage— ment observes in other channel participants is hypothesized to be associated with the latent conflict variables. Therefore, any conflict reduction strategy must address the latent condi— tions that underly the conflict. Failure to consider the causal variables may lead to a state of chronic conflict be- tween the parties since managerial maneuverings are likely to focus merely on the symptoms and not the cause of the channel antagonisms. Unfortunately, even a management cognizant of the under- lying causal conditions may lack sufficient information to effectively evaluate the appropriateness of conflict reduction strategies. For as analyzed under the dissertation's second objective, the latent variables may impact felt conflict through either of the components comprising the attitude construct. Since the particular attitude component influenced may vary for different situations, management's criteria for evaluating conflict reduction techniques should explicitly include this dimension. Ignorance of the mechanism linking 5 the latent to later conflict stages could result in insuffi— cient criteria for strategy evaluation. Justification for the Study The term conflict is derived from the Latin confligere meaning to strike together or to clash. However, as Alderson, Mallen, and other marketing writers stress, channel systems function best under conditions of sustained cooperation rather than when the objectives and activities of participants clash and members fail to work together as an integrated system.6 Yet as Rosenberg observes, conflict is a fundamental process of channels and should be studied carefully:7 Because conflict is an inherent condition of channel systems and can disrupt performance organizations in a given channel need to under— stand what conflict is and know how to deal effectively with it. The justification for the current research rests on the need to better understand the functioning of the conflict pro- cess in distribution channels so that management can effec- tively develop strategies to manage it. This need guided not only the specification of the research objectives but also the development of the research strategy required for their accomplishment. As discussed hithe literature review contained in the following chapter, the three specific research objectives are based on viewing a channel as a social system. The social system's perspective recognizes the inherent interdependency 6 of channel participants leads to a variety of social and be— havioral interactions. Although such relationships were largely ignored by the earlier economic studies of distribu— tion channels, these social system variables have the poten— tial for dysfunctional effects on a channel's performance. However, the effect these variables have on performance and their cumulative impact on each other is complex and not well understood. In View of these considerations, the current research program was designed and executed to contribute to the know— ledge of the channel conflict process available to market— ing scholars and practitioners. Scope of the Study As indicated in the discussion of the dissertation's objectives, the thrust of the research is theoretical. The rationale for the project is to develop and empirically test fundamental channel relations having significance for the management of distribution channels. In order to accomplish this the nature of the channel and conflict constructs under investigation must be determined. As noted by Bucklin, consumers are excluded from the do- main of the commercial channel; consequently, the study will focus on the relationships between business organizations in the channel setting.8 Since these organizations comprising the channel specialize in performing functions, there exists 7 a degree of operational interdependence among the channel firms. As discussed in the Social Systems section of Chapter II, the higher the degree of interdependency the greater the tendency for conflictual relations among firms. Therefore, the channel selected for analysis should reflect a high degree of functional dependency among organizational participants. In addition, member firms should have a high stake in the particular channel system and to its product—market commitment. Without a sufficiently high stake in the channel's operations a firm might leave the system when a conflict situation deve- lops, thus eliminating the opportunity to study the conflict phenomena. Since the degree to which relations among firms are struc— tured can range from the formal patterns existing in corporate channels to the loose, ephemeral patterns existing in some traditional systems, the channel selected should contain a balance between interdependency and stake in the system with each firm's ability to pursue its own objectives. While mana- gers in a corporate channel are employees of the same organi— zation and those in an unstable, fragmented channel may lack stake and interdependence, administered and franchise systems on the other hand, may provide an adequate balance between autonomy and commitment necessary for conflict to exist. Consequently, the dissertation will examine the conflict phenomena for a franchise distribution system. Research on the relationship of channel type to conflict intensity indi— cates a franchise system is a representative vehicle for the 8 study of the conflict construct. Conclusions from Kelley and Peters empirical study of various types of channel systems indicate channels representing contractual and independent system categories had about the same level of conflict inten— 9 In addition, Assael and Lusch have used franchise sity. systems to investigate conflict and other behavioral dimen— sions of distribution channels.10 Besides delineating the scope of the channel investigated, the nature of the conflict phenomena under consideration must be specified. The range of topics discussed in connection with the conflict construct is wide and often includes ante- cedent conditions, phases, economic and behavioral outcomes, and resolution strategies. Given the dissertation's objec— tives, the study will focus primarily on the relationship of the antecedent conditions to selected conflict phases. In particular, cognitive and affective phases of conflict will be stressed. Considering the complexities involved in these variables and the resource limitations of the project, neither the behavioral dimensions nor modes of conflict resolution will be explicitly treated. In this connection, it should be stressed the third objective of the project involves illustrat— ing the importance of establishing criteria for evaluating conflict reduction strategies. The objective involves the specification of a hitherto unaddressed set of theoretically based criteria for the evaluation of techniques used by manage— ment to reduce channel antagonisms. Application and empirical verification of such criteria to the myriad of conflict 9 reduction methodologies available is beyond the scope of the present study. The tOpic is taken up again in the final chapter's section on Implications for Future Research. Beyond the dictates of proper scientific process requiring that empirical development and verification occur on separate sam— pling units, the accomplishment of the dissertation's primary objectives require the scope of the conflict process consid- ered be delimited. As a final comment on the scope of the study, several limitations inherent in the project should be stated. Since empirical verification of the research hypotheses will be restricted to a single channel system, caution must be exer— cised in generalizing conclusions to channel systems differ- ing in structure and mission. In addition, several methodo— logical restrictions further limit the potential contributions Of the study. As explained in Chapter V, the methodology employed is survey type field research. Consequently, the rigorous con— trol over variables available in experimental laboratory re- search is not present in the study. However, field research does offer the advantage of studying the conflict process in its natural channel setting, thus avoiding the contrived and artificial environment often associated with the experimental approach. If carefully designed and executed, the field re- search should yield results that are both interpretable and realistic. Additional methodological limitations associated with the specific statistical tests utilized in the data 10 analysis phase of the research will be discussed in Chapter V. Overview While this chapter briefly stated the dissertation's objectives, justification, and scope, the following chapter will review the literature relevant to the study. Prior con— ceptual and empirical work in the channel, conflict, and at— titude areas will be reviewed for later development in a model of the conflict process. Chapter III will introduce the conceptual conflict model that will underly the disser— tation's empirical research. In addition, the chapter will specify the hypotheses that will be tested. The following chapter will describe the methodology used to verify the hypo— theses. .Topics discussed in this chapter include selection of the distribution channel, development of the research in- strument, and data collection procedures. Chapter V analyzes the research results. The data will be applied to the hypotheses in order to test the posited relations. Based upon the outcome of the empirical analysis, the final chapter will specify the conclusions that are warranted by the dissertation. In addi- tion, the implications for channel management and for future research generated by the study will be discussed. 11 Chapter I Footnotes lBert C. McCammon, Jr. and Robert W. Little, ”Marketing Channels: Analytical Systems and Approaches,” George Schwartz (ed.), Science in Marketing (New York, John Wiley and Sons, Inc., 1965), p. 322. 2Louis W. Stern, Distribution Channels: Behavioral Dimen— sions, L. W. Stern (ed.), (Boston, Houghton Mifflin Co., 1969), p. 5. 3Larry J. Rosenberg and Louis W. Stern, ”Toward the Analysis of Conflict in Distribution Channels: A Descriptive Model,” Journal of Marketing, 34:40 (October, 1970). 4Louis W. Stern and James L. Heskett, ”Conflict Manage— ment in Interorganization Relations: A Conceptual Framework,” in Stern (ed.), op.cit., p. 293. 0Larry J. Rosenberg, ”An Empirical Examination of the Causes, Level and Consequences of Conflict in a High Stake Distribution Channel,” Dissertation, Ohio State University, 1969. 6Wroe Alderson, Dynamic Marketing Behavior, (R. D. Irwin, Inc., Homewood, Illinois, 1965), and Bruce Mallen, ”A Theory of Retailer-Supplier Conflict, Control, and Cooperation," Journal of Retailing, 36:24—32 (Summer, 1963). - 7Larry J. Rosenberg, "A New Approach to Distribution Conflict Management,” Business Horizons, 19:67 (00Kter,1974L 8Louis p. Bucklin, A Theory of Distribution Channel Structure, Institute of Business and Economic Research, Uni— versity of California, 1966, p. 5. 9J. Steven Kelly and J. Irwin Peters, "Vertical Conflict: A Comparative Analysis of Franchise and Distributors,” in Contemporary Marketing Thought, B.A. Greenberg and D.N. Bel— lenger (eds.), (American Marketing Association, 1977). 10Henry Assael, Dealer Reactions to Factory Policies and Actions: A Case Study of The Automobile Dealer, Dissertation, Columbia University, 1965 and Robert F. Lusch, An Empirical Investigation of Power, Conflict and Satisfaction in a Fran- chisor—Franchisee Channel of Distribution, Dissertation, University of Wisconsin, 1975. CHAPTER II LITERATURE REVIEW In order to address the dissertation's research issues, the sources of relevant concepts for the study of channel conflict must be identified. In this chapter the literature related to distribution channels, conflict, and attitudes will be reviewed and analyzed for the subsequent development of a model of the channel conflict process. Channels of Distribution Alderson has shown that the essence of the marketing process is sorting meaningless heterogeneity found in nature into meaningful heterogeneous assortments.1 Distribution channels are central to the sorting process since the commer— cial channel consists, according to Bucklin, of the set of institutions that perform all the activities required to move a product and its title from the point of production to con— sumption.2 The important role of channels in the sorting process has been demonstrated through the commodity approach for specific products, the institutional approach for speci- fic marketing institutions and agencies, and the functional approach for specific marketing activities and flows.3 Fur— thermore, the complexity of the marketing process has resulted 12 13 in many channels evolving beyond a loose coalition of insti- tutions to a type of system organized along corporate, administered, or contractual dimensions.4 Traditionally, the various types of distribution channels have been studied as economic systems since the, ”...laws of economics have been thought to coordinate the behavior of the firms comprising the channel."5 The channel according to this view, is considered solely an economic entity consisting of a series of levels or markets. Firms only engage in transac— tions with firms at adjacent levels and are unconcerned with the institutions and functions occurring at more distant levels. Not only are cost and revenue considerations thought to be the motivating factors explaining the structure of in— terfirm alignments, but as McCammon and Little point out,6 Stigler, Coase, Bucklin, and others who work within the context of formal economic theory ’ postulate that channel linkages are forged by rational profit—maximizing entrepreneurs who are uninfluenced by social, political, and psychological considerations. However, such a myopic perspective can lead to an over- simplification of channel relations and a misunderstanding of the forces at work in channel systems.7 Social Systems Although extremely important, economic relationships represent only a single type of social interaction present in distribution channels. Stern and Brown explain, ”Because 14 social relationships underlie economic connections, the anal- ysis of distribution channels as social systems should pro- vide additional insights into the wide variety of interac— tions within such channels."8 Viewed as a social system, aspects of channels such as power, conflict, and communica- tion become as useful as economic variables in explaining channel phenomena. Parsons and Smelser define a social system as, ”...the system generated by any process of interaction, on the socio- cultural level,*between two or more actors.”9 For distribu— tion channels, Stern and Brown emphasize the business unit is the actor of concern; ..."when the commercial channel is the relevant system, the individual firm is the lowest level of the system - a component."10 Thus, viewing channels as social systems involves studying the process of interaction between individual firms. In a broad sense, the firm and its relationship to the environment become the focus of the anal- ysis since as Hall and Fagen observe any system consists of a set of objects with relationships between objects and their 11 This perspective is important in understanding attributes. distribution channels since the behavior of a member in a social system influences not only the condition of adjacent components but the state of the system itself. From a marketing point of view, channels can be consid— ered social systems because of the interdependency among mem- bers relative to the performance of distribution activities. In fact, Stern and El—Ansary have characterized a channel as, 15 "...a set of interrelated and interdependent components en— gaged in producing an output."12 There exists much opera- tional interdependence among firms in a channel because members specialize in performing functions. The functional specialization causes firms to depend on each other to ac- complish tasks. For the network of firms comprising the channel, Rosenberg and Stern conclude, "...two or more units must take each other into account if they are to accomplish their respective goals. This line of thinking leads to the acceptance of a distribution channel as an operating beha— vioral system.H13 Besides noting member firms are interdependent, McCammon and Little identify five additional characteristics which indicate a channel is a social system.14 They recognize a single firm often coordinates most of the functions, functions are performed sequentially, firms recognize the need for coop- eration, sanctions are used to enforce behavior norms, and participation in channel is voluntary. Similarly, Reve notes that channels meet important criteria of social systems:15 1. Activities among members are aimed at attaining both collective and self-interest goals. 2. There is a division of functions and tasks resulting in interdependent processes. 3. Integrated action takes place which results in a unique identity separate from its members. Based on his list of characteristics, Reve concludes most channels are “organized forms of collective behavior,” 16 while McCammon and Little conclude a channel is an ”organized behavior system.”l6 Likewise, Stern and El—Ansary state distribution chan— nels are what Alderson considered organized behavior systems because of the explicit rules and policies existing in chan— nels that result in routinization among firms.17 They explain that through various forms of transaction routines involving delivery, billing, order size, and other factors, rules evolve which result in channels approaching the formalization existing in individual organizations. In fact, these authors feel formalization occurs to such a degree that channels are really superorganizations since, ”...'superorganization' im— plies that channels have characteristics of all complex organ— izations, even though channels are comprised of collectivities rather than individuals."18 Viewing a network of firms as a social system or super— organization is also supported by Carlson and Kusoffsky who consider an established relationship between a producer and distributor as forming an organizational structure with the firms being the organizational units.19 In a related manner, Mallen developed an organizational extension concept where a channel is considered to be an extended version of each mem— ber firm's organization.2O Describing a channel in organizational terms is further justified by March and Simon who state,21 The distinction between units in a production— distribution process that are 'in' the organi— zation and those that are 'out' of the 17 organization typically follows the legal definition of the boundaries of a particu— lar firm. We find it fruitful to use a more functional criterion that includes both the suppliers and distributors of the manufacturing core of the organization. Using Weich's summary of organization attributes, Stern and El-Ansary continue this line of reasoning by identi- fying several aspects of organizations common to distribution channels.22 They note channels have structural complexity, a division of labor among institutions, rules that routinize transactions, collective goals, and a chain of command. Given these factors, Drucker's argument against Viewing a channel as merely a series of legal entities is reinforced.23 Furthermore, emphasizing a channel as a marketing.organization even though it often is not a legal entity is critical since, ”...the relevant unit of competition is an entire distribution system comprised of the entire network of int-rrelated insti— tutions and agencies.”24 Besides recognizing the channel as the important unit of competition in the economy, McCammon and Little identify several other advantages of viewing channels as social systems.25 They note a systems approach recognizes a channel is not a random collection of firms but is purposeful and ra- tional, that both cooperative and antagonistic behavior is present, the social interactions reflect the aspirations of participants, and that a systems view allows identifying the dysfunctions that are system generated. These benefits and their implication form the foundation for a more complete 18 understanding of the processes inherent in distribution channels. Cooperation and Conflict Mallen while finding both cooperative and antagonistic interactions fundamental behavioral processes in channels, concludes channels are, "...best off under conditions of op— timum cooperation leading to a consumer and profit satisfac- tion.”26 Similarly Alderson states,27 Marketing channels cannot function without sustained cooperation in which each party knows what to expect from his opposite member. Economic processes involving marketing re- quire cooperation and close coordination... Stern and El—Ansary attribute the importance of coopera— tion to two specific effects it has on distribution channels.28 First, it allows members to coordinate their planning, infor- mation requirements, and decision-making. Secondly, coopera- tion justifies joint goals since it has the effect of arrang— ing payoffs so members can pursue common goals based on indi— vidual criteria. The need for coordination and the pursuit of common goals are critical for overall channel effectiveness because as Stern points out several forces operate in channel systems:29 1. Each member is dependent upon the behavior of other channel members, 2. A behavior change at any point causes change throughout the channel, 3. The whole channel must operate effectively 19 if the desires of any one member are to be realized. However, McVey stresses that such channel—wide coopera— tion is rare among channel members.30 Rather than concern themselves with entire channel systems, members are only interested in their immediate suppliers and customers. Con- sidering himself as an independent market, a member performs only, ”...those functions he deems inescapable in the light of his own objectives, forming policies for himself wherever he has freedom to do so...many of his choices are indepen- dent."31 Labeling a channel as a ”derivative operating system," McCammon and Little recognize members are not primarily con— cerned with channel growth or survival, but only with their internal operations and rate of growth.32 Significantly, they conclude the systems concept must be modified when applied to distribution channels since coop— eration is not always present.33 The contention that marketing channels are organized behavior systems should not ob- scure the fact that they often consist of independent and idiosyncratic intermediaries. Alderson and McVey both emphasize this po- tential source of channel disharmony in their analysis. Alderson, for example, argues that a channel is a group of firms that con— stitute a loose coalition engaged in exploit— ing joint opportunity in the market. The behavior of firms in this loose coalition is ex- plained in part by Stern and El—Ansary's observation that channel members have mixed motives.34 On the one hand, 20 members desire to cooperate since they recognize their inter- dependencies, while on the other they want to maximize their autonomy. Autonomy being used to refer to the tendency of firms to operate in their environment in an independent and dynamic manner. Therefore, it is not surprising that while Rosenberg and Stern feel existing channels reflect mostly cooperative rela— tions they note, ”The process of conflict, however, is perva— sive and inherent in channel systems characterized by behav- ioral interdependence for mutual ends.”35 Similarly, Stern and El—Ansary state, ”It is, therefore, possible to predict with certainty that distribution channels will exhibit evidence of conflict."36 Channel Conflict Using a social systems perspective, Stern and Gorman ob— serve a distribution channel initially in stable equilibrium can experience a change in either its environment or within a member's organization such that members perceive such change 37 In their attempt to resist or reverse the as frustrating. change in channel conditions conflict can result. However, the emergence of a conflict state is not an automatic result of system disequilibrium. Conflict Conditions There are several conditions necessary in the relations among channel members before a state of conflict exists. Mack 21 and Snyder have identified five empirical preconditions to social conflict.38 First, at least two analytically distinct units must be involved. For marketing channels, PalamountxUIhas pointed out the nature of the opposing unit is important in differ— entiating different types of conflict.39 Intertype conflict occurs when the opposing unit is another type of channel system in the environment. If the unit is a competitive firm of the same type within the channel, a horizontal conflict exists. When the unit is in the same channel but at a dif- ferent structural level, Palamountain refers to this as a vertical conflict. The second condition Mack and Snyder specify is that the conflict arises from opposing values between the units of the conflict. A major reason competing values can lead to con— flict is the existence of resource scarcity. In fact, Bould— ing adOpts the View that all conflicts are resource conflicts.4O ...only when there is scarcity in the economists sense can there be issue conflict, for the only thing that conflicts can rationally be about is the distribution between two or more parties of some good which is both scarce and valued. The third condition is that the units can gain only at each others expense. This condition is central to Goldman's conceptualization of the conflict construct. He defines con— flict as a state where there is:41 ...a social relationship between two or more parties in which at least one of the parties 22 perceives the other as an adversary engaging in behaviors designed to destroy, injure, thwart, or gain scarce resources at the expense of the perceiver. The fourth condition requires that conflict involves an interaction where each opposes the actions of the other. Again, Goldman's emphasis on viewing the other unit as an adversary illustrates the importance of this condition. Mack and Snyder's final stipulation is that conflict involves attempts to gain control of scarce resources or po- sitions or to influence the other units behavior. Beier and Stern have demonstrated that the attempt by one channel member to gain control or influence the behavior of another is common to distribution channels.42 An understanding of why such behavior is typical is illustrated by Berg's analy— sis of the strategic options open to channel members. Berg states that members should try to,”...accumulate market power which will then be exercised unilaterally to extract the com- pliance of middleman.H43 Not surprisingly, such manipulative attempts to control other members can result in frustration. Thus, if a disturbance upsets the equilibrium of a chan— nel system such that these five conditions are present a state of conflict exists. When this state occurs, Fink notes mem— bers would be, ”...linked by at least one form of antagonistic psychological relation or at least one form of antagonistic interaction.”44 Such interactions occur because as El-Ansary and Robicheaux observe, conflict, ”...is, in essence, a state of frustration of a channel member who perceives another 23 member to be restricting his performance and his ability to achieve his goals."45 Stern and Gorman further explain such restrictions in- volve the blocking of a member's instrumental behavior pat- terns.46 These patterns refer to the specific activities a member performs in order to accomplish his goals. The conflict state involves restricting these instru- mental behaviors because during the disturbed equilibrium the desires of specific components within a given channel system are not sufficiently integrated. As March and Simon note, such a state represents a ”malfunction of the system” since there is a breakdown in the standard mechanisms of de— cision making. 47 This view is shared by Dahrendorf who concludes that in terms of motivations, all the relations between conflicting parties, ”...involve an incompatible difference of objective.”48 Walker shifts the emphasis from motivation to behavior in explaining the breakdown. He states it is the, ”...actions by one member which are incon— sistent with the goals or objectives of some other member."49 Attitude and Behavior Dahrendorf‘s focus on the motivations of members and Walker's emphasis on their behaviors as an explanation for the breakdown in a system's standard mechanisms represents a general difference in focus existing in the conflict liter- ature. Fink refers to those authors defining conflict as essentially attitudinal in nature as the motive—centered school, while those viewing overt activity a precondition as 24 the behavior—centered school.50 Mack and Snyder represent the behavior—centered school by defining conflict narrowly as active opposing behavior.51 In a related manner, Schmidt and Kochan consider conflict to be explicit interference arising out of, "...a process in which one unit seeks the advancement of its own interests in its relationship with others.”52 Fink identifies Wright, Kerr, and Coser as supporting the behavioral orientation. However, as Stern and Gorman emphasize, for a member to be frustrated it is only necessary that he perceive the be- havior of another as restricting his instrumental behavior patterns. This View is supported by Goldman who states con— flict is a condition, ”...in which at least one of the parties perceives the other as an adversary engaging in behaviors designed to destroy..."53 In considering their concept of conflict, Stern and Gorman note it, ”...does not attempt to tie overt responses to frustration, since it does not associate, on a one-to—one basis, conflict with the display of threatening behavior on the part of the frustrated member.”54 Fink notes many au- thors (i.e., Aubert, Galturg, and Stagner among others) do not associate conflict, which he defines as psychological antagonisms, directly with behavior.55 Rather, these motive- centered authors consider conflict attitudinal and any beha- vior a result or an after effect of the conflict. He states Shepard and others, ”...restrict the terms 'conflict' to the psychological patterns, referring to the consequent behaviors as 'struggle', 'active opposition'.56 Singer, for example, 25 cites Simmel as considering an actual behavioral struggle as a form of solving the conflict rather than as an instant of it.57 However, Fink concludes that the differences between the motive and behavior-centered authors fails to have a sub- stantive impact on theory since their opposing views repre— sent, ”...merely a shift in the system for labeling different stages of the same process..."58 Models of Conflict The debate between the motive and behavior—centered con— flict theorists suggests the conflict phenomena is too complex for single dimension definitions. It would appear a valid conceptualization of the conflict construct would have to ac- count for both its attitudinal and behavioral components. Wright develops such a framework by identifying several sequential states in the conflict process.59 He breaks the total conflict process into four distinct phases: 1. Inconsistencies in purpose or orientations between the parties. 2. DevelOpment of social tensions. 3. Process of resolving the inconsistencies. 4. Open Violence. Wright notes open violence, the behavioral stage, would only occur if there was a failure to resolve the inconsisten— cies developing from the social tensions, the affeCtive stage. By making explicit the relationship between attitude and 26 behavior, his framework helps to structure the conflict phe— nomena. Rosenberg and Stern offer a model of the channel con— flict construct which also adds considerable insight into the complex conflict process.6O Basically, the model states the structural alignment of firms is associated with certain causes which generate a level of conflict. Causes Level Outcomes Channel Behavioral Structure Reactions The level of conflict, in turn, results in behavioral re- actions which lead to certain performance outcomes. The model is a contribution because it highlights the relationship of channel structure (manufacture, wholesale, retail levels) to attitude and behavior. Rosenberg, in a later article, expanded upon this model by making explicit the causal chain operating within the Cause- Level-Outcome paradigm.61 He notes marketing goals guide the implementation of planned actions and unless there is agree- ment and cooperation among members, a level of tension is created. With members holding antagonistic attitudes an in- cident may occur which could trigger a dramatic flare—up. Such potentially damaging behavior between the parties leads to conflict resolution attempts. Certain financial or behavioral outcomes may, in turn, shape future goals and tensions between the channel members. 27 Other authors treat conflict along with other elements of channel sentiments such as satisfaction, power, and leader— ship in attempts to explain, in a wholeistic manner, the re— lationships between channel structure and effectiveness. In developing their models these authors not only focus on the various stages of the conflict process but note the relation- ship of these stages to other aspects of channel relations. Reve, in his model of distribution channel functioning, emphasized the effect the political substructure can have on channel effectiveness.62 Various inter—organizational rela- tionships, including conflict, are shown to affect the economic performance of channel members as well as the competitive effectiveness of the channel system. Reve, like Rosenberg, emphasizes the relationship of structural dimensions to member sentiments and behavior. El-Ansary and Robicheaux develop a broad general model for understanding channel member behavior.63 Although the authors do not make explicit the functional form of relation— ships they do organize a large number of behavioral constructs and indicate possible linkages among them. Significantly, the authors note the important role conflict plays in interorgan- ization relationships and the potential impact conflict can have on the viability of the channel. In their discussion of the conflict construct they state: Conflict between organizations is an inevitable outgrowth of functional interdependence and the scarcity of resources...When a channel member perceives the behavior of another member as ob- structing the progress of his goal attainment or 28 behavioral patterns, an atmosphere of frustration prevails. A state of conflict may, therefore, exist when two or more channel members become objects of each other's frustration. In a similar vein, Cadotte and Stern highlight the im— portance of conflict in their process model of dyadic inter- organizational relations.65 They incorporate a causal link between variables which are posited to create a potential for conflict. Once members perceive they are in a conflict state, the conditions exist for various behavioral responses. However, unlike Cadotte and Stern and other authors who develop general schemes for explaining channel member beha— vior, Pondy develops a detailed typology of the conflict pro- cess. His analysis is based on the recognition that each conflict episode, ”...exhibits a sequence or pattern of de— velopment, and the conflict relationship can be characterized by stable patterns that appear across the sequence of epi— sodes."66 Specifically, Pondy identifies five stages constituting a stable pattern appearing in conflict episodes.67 1. Latent conflict (conditions) 2. Perceived conflict (cognition) 3. Felt conflict (affect) 4. Manifest conflict (behavior) 5. Conflict aftermath (conditions) Although he stresses the process may not always proceed in a sequential manner and the entire process may stop at any 29 stage, his analysis is extremely useful in structuring the conflict construct and integrating the views of the motive and behavior-centered conflict theorists. By explicitly noting antecedent conditions underly both the perception and feeling of tension, Pondy's framework provides a basis for an improved understanding of the full range of behavioral responses to conflict. For whether a specific conflict re— sults in passive behavior or overt aggression, its origin lies in much earlier interactions between the parties. Stages of Conflict Latent Conflict A search of the literature reveals a plethora of antece— dent conditions which theorists posit as underlying causes of conflict. However, Assael in his nine industry study of con- flict expresses a fundamental condition common to the views of many writers:68 The underlying problem in these distribution conflicts is one inherent to channel manage— ment: the basic interdependence between organizations that are nominally independent. This interdependence must result in both co— operation and conflict; cooperation for survival, conflict because of different economic goals and ideological motives. While recognizing the need for interdependency, March and Simon, in a manner similar to Assael, recognize that a difference in goals is often a major cause of conflict be— ( tween groups.63 In addition, the authors consider a 30 difference in perceptions of reality and the need for joint decision-making as critical conditions. Little also empha- sizes the failings in the joint decision—making process, but adds the problems created by misunderstood communications and divergent functional specialization in the channel struc— ture.70 After a review of the conflict literature, Pondy con- cludes there are three basic types of latent conflict. He identifies competition for scarce resources, drives for au— tonomy, and divergence of subunit goals as the primary causes of conflict.71 Schmidt and Kochan similarly reduce a long list of possible causes to a few primary conditions.72 They consider the perception of goal incompatibility and the oppor— tunity for interference as the most basic preconditions. For distribution channels, Mallen has perhmxsgone the farthest in isolating the underlying causes of conflict by stating the exchange act itself is the primary source.73 Although it is useful to reduce long lists of causal variables to fundamental conditions, it is also important for diagnostic reasons to be able to identify specific con— flict causes. Along this line, Stern and Gorman specify seven conditions which could cause conflict between channel members:74 1. Role deviance from the expectations of others. 2. Disagreement over allocations related to scarcity issues. 3. Difference in perceptions leading to di— vergent behaviors. 31 4. Expectations related to contingencies which fail to develop. 5. Decision making attempts in another member's domain. 6. Interdependency of member's goals. 7. Inadequate communication leading to uncoor— dinated behaviors. The authors developed their list from theoretical work in social psychology, sociology, economics, and other disci- plines. Assael, on the other hand, developed a list of cau- sal factors from an empirical study involving nine major industries.75 Based upon his data, Assael concludes conflict is caused by (1) ideological differences between firms involved in as— pects of production and those involved in distribution, (2) functional absorption by one party of functions held by another, and (3) differing economic objectives of parties leading to different interpretation of roles. While conflict causes developed empirically and based on theoretical writings in other disciplines represent impor- tant contributions to understanding the latent stage, a typo- logy of causes which could serve as a guide to research would be especially important. As indicated by Rosenberg, Stern and Heskett's treatment of the antecedent condition is particularly well suited for empirical research.76 Stern and Heskett propose the dependency existing in channel systems is expressed as conflict when there are:77 1. Incompatibility of goals among system members. 32 2. Role incongruency within the system. 3. Differences in perceptions of reality. As will be shown later, these three conditions can be used to empirically investigate the latent stage of the con— flict process. Perceived Conflict Perceived conflict is defined as the stage where the parties become aware of the conflict situation.78 It is a cognitive state. As indicated by Pondy, some latent conflict situations may fail to reach the perceived stage because or- ganizations are faced with an overwhelming number of con— flicts.79 As a result, less critical conflicts do not gain attention since organizations focus only on important ones. As Boulding notes, situations important to an organiza- tion are not necessarily determinable based on objective cri- teria. Rather, he states, ”...it is not the 'objective' sit— uation which matters, but the subjective images of partici— pants."80 Along this line, Pondy discusses the effect a sup— pression mechanism may play in influencing the subjective images.81 He notes individuals tend to block out conflicts that are only mildly threatening, but when they relate to values central to an individual, such conflicts must be acknow- ledged. Felt Conflict This stage refers to the affective dimensions of conflict. 33 Authors such as Raven and Kruglanski View affect as central to the conflict construct. They state, ”Social conflict is defined as tensions between two or more social entities."82 To LeVine, social conflict is "...regularly accompanied by certain feelings and beliefs of individuals who are partici- 83 pating in the conflict." Similarly, Rosenberg notes, ”... conflict is felt when it makes the participants tense or anxious and affects his attitudes toward the other party.”84 LeVine identifies two attitudinal concomitants of con— flict.85 First, conflict has a significant hostility element since the party engages in symbolic or fantasy aggression toward the conflict object. Second, the party holds negative images and unfavorable evaluations of the conflict object. These lead to further distorted views of the other party and focusing on negative traits. However, conflict may be perceived but not result in the tension, stress, hostility, and anxiety associated with the felt stage. As stated by Lusch, ”An organization may perceive that it is in disagreement with another organization over some policy but this may not create an affective state of conflict within the organization.” Pondy notes that in order for a perceived conflict to be ”felt” it must be personalized, that is, the individual must feel personally involved with the identified conflict situa- tion.87 34 Manifest Conflict Brown considers the manifest or behavioral state to arise when conditions in or outside the channel system exa- cerbate affective tension to create apathetic, non—coopera— tive, or aggressive behavior.88 Rosenberg in an empirical study of executives from manufacturing firms, concludes man— agement often equates the entire conflict process with the, “...flare—up phase when conflict is disruptive and demands attention.”89 His results indicate management tends to ignore conflict before and after it erupts. However, the tendency to disregard conflict during its ”invisible period” unduly hampers effective conflict management since most of the con- flict process is covert. As Coser notes, hostile attitudes are predispositions to engage in conflict behavior.90 However, social norms limit their expression as open aggression such as physical and verbal violence. Rather, behavior is likely to take the form of try- ing to block opponent's plans, apathy, and rigid adherence to rules. Schmidt and Kochan identify two points at which blocking 91 Interference at the point of resource behavior could occur. attainment occurs if one party blocks the other from gaining resources. In addition, interference could occur where ac— tivities of the conflicting parties are interdependent. Pondy's conception of overt conflict supports these views, as he defines it as behavior that intentionally frustrates the goals of the other party.92 35 Effects of Conflict on the Channel The residual impact of manifest conflict on the channel system constitutes the conflict aftermath stage. Pondy iden— tifies three dimensions of organizational systems that are 93 affected by conflict. Conflict can have an impact on: 1. Productivity or efficiency in the system's economic activities. 2. Stability in terms of organizational cohesiveness. 3. Adaptability or ability to adjust to changing internal and environmental pressures. However, these three dimensions are themselves interde— pendent since any conflict situation yielding a reduction in system stability would concomitantly increase system adapta- bility. Pondy observes such dependency makes it difficult to determine whether the overealleffects of conflict are func— tional or dysfunctional:94 It is argued here that a given conflict episode or relationship may have beneficial or delete— rious effects on productivity, stability, and adaptability. Since these values are incom— patible, conflict may be simultaneously func- tional and dysfunctional for the organization. Dysfunctional Effects Stern and El—Ansary identify conflict as destructive or dysfunctional when, ”...pathological moves are made that ulti— mately impede the performance of the conflicting parties and n95 the system itself. Stern and Heskett describe the sequence of pathological moves resulting from high levels of conflict.96 36 By reducing cooperation between the parties conflict leads to a duplication of efforts which ultimately reduces per- formance for the channel system. Similarly, Mallen finds that if conflict outweighs cooperation the effectiveness of the channel is reduced and distribution becomes less effi- cient.97 In addition, Stern and El—Ansary comment that a loss of cooperation and coordination can result in reduced channel effectiveness, efficiency, growth, and long—run survival. They conclude increasing degrees of independent behavior by individual firms, ”...serves as a detriment to the viability of the total interorganization network.”98 In an empirical study of the channel network for several consumer products, Rosenberg identified three dysfunctional aspects of conflict.99 1. Reduced efficiency in distribution and increased costs. 2. Increased resistance to resolving future conflict issues. 3. Emotional disruption and damage to business relationships created through distorted judg— ments. A possible explanation for these empirical results can be found in Stern and Gorman's observation that organizations faced with a frustrating situation are likely to respond in two general ways.100 A likely reaction to conflict is the exercise of power by attempts to influence the behavior of the other party. A second response would be some form of intraorganizational change on the part of the frustrated 37 member. However, as Stern and Heskett point out,101 In important conflict situations which have become overt, frustrated parties are more likely to exercise influence or power than to change their own goals or operations. The tendency of firms to re—establish channel equili— brium by attempts to control other members behavior rather than adjusting their own activities, partially accounts for the pathological reactions to conflict.102 However, empirical tests of the effect conflict has on various measures of channel performance produces equivocal results. Pearson, in a study of fifty-eight channel seg— ments in the grocery channel, found no statistical difference in the operational performance of segments characterized by conflict or cooperation.103 He reports the direction of his data, although not significant, showed conflicting segments as having poorer performance, using a product turnover mea— sure. Similarly, Kelley and Peters, in a study involving five different types of channel systems, report only weak evidence that channel performance is negatively correlated with conflict.104 In spite of the lack of clear cut empirical conclusions, it would seem that conflict has the potential for a destruc— tive impact on channel systems. For as Mack and Snyder warn, ”...conflict is, on balanace, dysfunctional to the extent its positive functions are impaired or neutralized under cer— tain conditions.”105 38 Functional Effects As Pearson observes, many authors including Palamountain, Stern, and Galbraith, hold that some conflict is not neces— sarily dysfunctional and can actually have positive system effects.106 Supporting this View, Mack and Snyder identify several aspects of social systems which serve as inherent limitations to the destructive aspects of conflict.107 Con— flict's dysfunctional effects are limited because of the, 1. Regulating effect of institutionalized norms and procedures. 2. Functional interdependence and the continuing need for communication. 3. Inertia and organizational inefficiency of the parties. 4. Cost of conflict and the feasibility of cer— tain modes of resolution. In conjunction with these factors, Stern and Gorman's observation that intraorganizational change is a possible reaction to conflict partially accounts for conflict's posi- tive effects.108 Faced with conflict, a channel member could change his behavior instrumental to attaining his goals in order to alleviate the frustration. Since the new behavioral patterns could be more effec— tive and efficient, the conflict could yield highly functional system-wide effects. Functional results similar to shifts in instrumental behavior patterns could occur even if channel members made other internal changes in attempts to resolve frustrating conflict. Although infrequent, channel members 39 could reappraise and change their own goals which contribute totflmiconflict situation. In addition, spontaneous changes could occur within the firm such as selective recall which would tend to alleviate the most frustrating elements of a conflict situation. Palamountain has pointed out vertical conflict can lead to innovations that lower economic and social costs.109 In an empirical study, Rosenberg identified several forces lead— ing to functional results for channel conflict. He found con— flict to be,110 ...a strong motivating force for management to remain active in reviewing and upgrading its activities...a means of indicating the need for change and evaluating management performance. Similarly, Mack and Snyder stress several benefits of conflict.111 They note conflict strengthens group cohesive— ness, clarifies objectives, and establishes group norms. Threshold Concept From the theoretical discussions and empirical studies reviewed, conflict would appear to have both functional and dysfunction effects. Stern and Gorman indicate this may be the case since they note there are two fundamental forces re- storing equilibrium once conflict has destabilized the sys- tem.112 The tendency of parties to engage in behavior that breaks their dependency relationship represents a force to— ward system disintegration. On the one hand, there is force 40 toward system disintegration. On the other hand, there is force toward system unification which can result in the new equilibrium falling anywhere between the extremes of inde- finitely maintained frustration to complete elimination of frustration. The possibility of a threshold level of conflict above which disintegration forces overcome the pressures toward unification has been suggested in the literature.113 In ex— ploring the possible application of the threshold concept to channel conflict, Stern and Heskett refer to Boulding's em- phasis on the need to,114 ...identify the level of hostility above which conflict processes will be malign (all system members are worse off) and below which they will be benign (all system mem- bers increase their welfare). Along this line, Rosenbloom has developed four graphical models showing the possible relationship between the level of conflict and channel efficiency.115 Three of the models show efficiency as either decreasing, increasing, or remaining constant as conflict increases. The fourth model represents the general condition where efficiency remains unchanged at low levels of conflict, increases as conflict becomes more intense, finally a point is reached where greater levels of conflict reduce channel efficiency. Lusch examined these relationships empirically but found no support for the claim conflict increased operating perfor- mance nor did he find any evidence indicating the existence 41 of a threshold effect.116 However, the hypothesis that greater levels of conflict were associated with reduced operating performance received partial support. Pearson, in a related study, was unable to find any empirical support for a relationship between conflict and channel performance.117 However, the inability to empirically demonstrate a con- flict—performance relationship may reflect more the methodo— logical deficiencies in performance measures used than the absence of an underlying relationship. Although Reve notes such indicators of a channel's economic performance as produc— tivity ratios, return on asset measures, and perceived economic effectiveness have been used, no generally agreed upon performance measures exist and measurement problems re— main largely unsolved.118 Besides, as El—Ansary and Robicheaux point out, channel performance is a function of a great many other factors in addition to the level of conflict present.119 Although they accept performance as a function of conflict intensity, they add that the degree of satisfaction and the outcome of chan- nel control also determine the level of performance actually achieved. Regardless of the degree of influence conflict has on operating performance, the conflict is likely to have a con- tinuing impact on channel relationships. For as Coser ob— serves, social conflict has no precise termination point since the conflict process will continue to operate if no explicit provision for stopping it is made.120 Similarly, Mack and 42 Snyder note there may be a ”chronic recurrence of unsettled issues” in the absence of agreed procedures for reviewing relations.121 This may often be the case, since as Pondy states, "...if inducements for participation are sufficiently high, there is the possibility of chronic conflict in the context of a stable relationship."122 El-Ansary and Robicheaux indicate this may be especially true for conflict occurring between members of a distribution channel.123 Stern and Gorman show the maintenance of con- flictual stress may be a continuing feature of channel sys- tems since a possible conflict outcome is the,124 ...elimination of conflict behavior without elimination of frustration. The parties to the conflict decide that their other alterna— tives cannot provide as much goal satisfaction as that experienced in their present alignment, and the relationship is sustained even though the conflict has not really been fully resolved ...the maintenance of frustration indicates potential for further conflict behavior. Relationship of Conflict to Attitude As previously indicated, Pondy's treatment of conflict places affect as central to the conflict construct. The stages of his model are conceived such that conditions and cognitions lead to "felt conflict" and behavior and conditions result from it, indicating conflict is essentially affective with certain conditions preceding and following. This View is reflected in the marketing literature, since many defini- tions of channel conflict have a substantial affective 43 emphasis. Studies defining conflict as a feeling of stress, tension, or hostility toward another channel member, perceiv- ing another as an adversary, or when a member is in a state of frustration illustrate the importance of affect.125 These definitions employed by channel researchers are similar to the social psychologist's view of attitude. To Fishbein attitudes are, "...conceptualized as learned predis- positions to respond to an object or class of objects in a consistently favorable or unfavorable way.”126 Even Allport's classic definition of attitude reflects the View toward af- fective conflict adopted by some channel researcherslz7 An attitude is a mental and neural state of readiness, organized through experience, exerting a directive or dynamic influence upon the individual's response to all ob- jects and situations with which each is related. Therefore, when Pearson had retail grocery buyers eval— uate their suppliers by rating them on a Staple scale along bickering, bitterness, resentment, and similar dimensions, he not only obtained his intended measure of channel conflict but also of attitude.128 This occurred because an attitude represents a respondent's general feeling of favorableness or unfavorableness toward some stimulus object. Such a View is consistent with Fishbein and Ajzen who intend, "...attitude to refer solely to a person's location on a bipolar evalua— tive or affective dimension with respect to some object, ac— . 129 tion, or event.” 44 Attitude Models Although there are over a dozen important attitude theories, two theoretical bases are especially useful in understanding the construct.130 Learning theories of atti— tude are based on the stimulus response approach of behavior theory, while the second major school of thought, cognitive consistency theories, are based on the cognitive approach of field theory.131 Fishbein's attitude model represents the first approach since it was developed within the framework of behavior theory. The model accounts for the relation between beliefs and attitudes in terms of the conditioning process. Fish- bein's model is based on the proposition that a person's at- titude is a function of an expectancy or belief component and a value or evaluation component.132 According to the theory, a person's attitude toward any object is a function of his be- liefs about the object and the implicit evaluative responses associated with those beliefs. The central equation of the theory can be expressed as follows: where, A0 is the attitude toward some object, O; bi is the belief 1 about 0, i.e. the subjective probability that: 0 is related to attribute i, ei is the evaluation of attribute i and n is the number of beliefs. Rosenberg's attitude model, on the other hand, is based 45 on cognitive theory. His model is based on the need for cognitive—affective consistency. For,as Rosenberg notes, an attitude is, ”...accompanied by a cognitive structure made up of beliefs about the potentialities of that object for attaining or blocking the realization of valued states.”133 A0 = PI-VI 1 ”MD i where, A0 is attitude toward an object; a stable affective response; PI is the perceived instrumentality of the object in attaining or blocking the in- dividual's goals; VI is the value importance of the goal; the amount of affect attached to the goal; and n is the number of beliefs. In spite of different theoretical bases, both Fishbein and Rosenberg maintain, "...that attitudes are affective and correlate with 'evaluative beliefs' as additive compo— nents.”134 Furthermore, Harrell notes,135 In essence, Rosenberg's formulation of atti— tudes toward objects could be considered almost identical conceptually to Fishbein's AO formula with the exception of the specificity of the belief elements that are measured. The exception refers to Rosenberg's use of thirty—five valued states as belief elements for all attitude objects, while Fishbein varies the specific belief elements depending on the attitude object. However, even this difference may 46 not apply when Fishbein belief elements have high evaluation since they would be quite similar to Rosenberg's valued states.136 In view of the similarity of the models and the addi- tional flexibility inherent in Fishbein's treatment of belief elements, the dissertation will utilize Fishbein's model of attitude toward an object. Furthermore, since attitude and affective conflict were shown to be similar constructs, the AO element of Fishbein's model will be used to represent a channel member's affective conflict toward another channel participant. In this regard, the two components of the model become central to the dissertation's research issue. Components of Fishbein's Model and e- are structured The model's two components, b 1, 1 according to a compensatory, miltiattribute approach. Their structure is compensatory because of the manner in which the attributes are combined and multiattribute since more than one attribute of the object is included. Wilkie and Pesse- mier explain the logic underlying the structure of the general class 0f compensatory attitude models:137 Overall affect is posited to reflect the net resolution of an individual's cognitions (beliefs) as to the degree to which given objects possess certain attributes weighted by the salience of each attribute to the in- dividual. In terms of Fishbein's model, the bi component refers to the cognitive belief that the attitude object possesses 47 certain attributes i, while the ei component refers to the evaluative aspect of each attribute. Referring to the bi component, Fishbein has stated:138 A belief about an object may be defined as the probability or improbability that a particular relationship exists between the object of belief and some other ob- ject, concept, value, or goal. In other words, a belief is a relational statement, link— ing an object to some attribute. The component is cognitive since it refers to the information a person has about the attitude object. Similarly, other writers have stressed the role of cog— nitive elements in the attitude structure. Day has stated, "...most theorists regard affect as the core of the attitude concept and derived from the more specific cognitive compo- nents.”139 However, Fishbein makes a distinction between belief content, the identifying cognition that an attribute is as— sociated with an object, and belief strength. Belief strength refers to the subjective probability the object is associated with an attribute. Fishbein explains,140 ...the belief per se is the position that an individual ascribes to the statement on the probability dimension i.e., the pro— bability or improbability that the particu— lar relationship expressed in the statement does exist. In conjunction with Ajzen, he explains the implications of this View of belief strength:141 48 With respect to any object—attribute associa— tion, people may differ in their belief strength. In other words, they may differ in terms of the perceived likelihood that the object has (or is associated with) the attri- bute in question. Thus we recommend that ”belief”, be measured by a procedure which places the subject along a dimension of sub— jective probability involving an object and some related attribute. The model's second component, ei, refers to the evalua— tion of the attributes in a good—bad sense. The component is intended as a measure of the desirability each attribute holds for the subject. Although each attribute will have an evalua— tive response, its contribution toward overall affect is de— pendent on the strength of the belief that the object is asso— ciated with the attribute.142 However, the results of several empirical studies have raised doubts as to whether the evaluation component should be included in the model.143 Certain studies have found that inclusion of the component appear to have either a suppression effect, where the explanatory power of the model is reduced, or fails to raise the model's explanatory ability. After re- viewing several of these studies, Wilkie and Pessemier con- clude, ”...it appears that the importance or value component ”144 should continue to be included in the model. In addition, Scott and Bennett conclude that the evaluation component is important for cognitive models of attitude structure.145 In commenting on several of the empirical studies ques- tioning the inclusion of ei component, Fishbein notes a cri— tical methodological error which seriously confounds the 49 actual role of the evaluation component. Fishbein and Ajzen draw a distinction between what the ei component theoretically represents, attribute desirability, and what some researchers have misinterpreted it to mean, attribute importance:146 ...measures of perceived importance of an attribute have sometimes been substituted for evaluation of the attributes (e.g., Bass and Talarzyk 1972; Sheth and Talarzyk, 1972; Hansen, 1969)...this measure of im- portance is not equivalent to evaluation; rather it will tend to be related to polarity of evaluation. They conclude that except for unusual situations where evaluations are all positive or all negative, better estimates of attitude are obtained by including the evaluative compo- nent. Recapitulation This chapter has reviewed the literature related to dis- tribution channels, conflict, and attitude. Due to the inter— dependency among members, a distribution channel was recognized as a social system with behavioral as well as economic varia- bles playing key roles in the interactions between channel participants. Conflict was found to be an inherent aspect of intra-channel relations and as having the potential for des— tructive as well as functional effects on channel performance. A similarity was observed between the felt stage of conflict and the social psychologist's attitude construct. Specifically, it was found that Fishbein's model of attitude toward an ob- ject could represent what channel researchers consider to be 50 felt conflict toward another channel member. The following chapter will integrate these observations from the literature in developing a conceptual model_of the latent and affective stages of channel conflict. 51 Chapter II Footnotes 1K. Tim Hostiuck and David L. Kurtz, ”Alderson's Func- tionalism and the Development of Marketing Theory,” Journal of Business Research, 1:148 (Fall, 1973). 2Louis P. Bucklin, A Theory of Distribution Channel Structure, Institute of Business and Economic Research, University of California, 1966, p. 5. 3Robert Bartles, History of Marketing Thought, 2nd ed. (Columbus, Ohio, Grid, Inc., 1976), pp. 145—148. 4Bert C. McCammon, Jr., ”The Emergence and Growth of Contractually Integrated Channels in the American Economyj'in Peter D. Bennett (ed.), Marketing and Economic Development, (American Marketing Association, 1966), p. 452. 5Louis W. Stern and Jay W. Brown, ”Distribution Chan— nels: A Social Systems Approach,” in L. W. Stern (ed.), (Boston, Houghton Mifflin Co., 1969), Distribution Channels: Behavioral Dimensions, p. 12. 6Bert C. McCammon, Jr. and Robert W. Little, ”Marketing Channels: Analytical Systems and Approaches,” George Schwartz (ed.) Science in Marketing (New York, John Wiley and Sons, Inc., 1965), p. 342. 7Louis W. Stern, Distribution Channels: Behavioral Di- mensions, p. 5. 8Stern and Brown, p. 17. 9Talcott Parsons and Neil J. Smelser, Economy and Society: A Study in the Integration of Economic and Social Theory, (New York, The Free Press, 1956), p. 8. 10Stern and Brown, p. 18. 11A. D. Hall and R. E. Fagen, "Definitions of a System,” in L. Bertalanffy and A. Rapoport (eds.), General Systems (Ann Arbor, University of Michigan Press, 1956), p. 18. 12Louis W. Stern and Adel I. El-Ansary, Marketing Channels, p. 11. 13Larry J. Rosenberg and Louis W. Stern, ”Toward the Anal- ysis of Conflict in Distribution Channels: A Descriptive Model,” Journal of Marketing, 34:41 (October, 1970). 14McCammon and Little, p. 329. 52 15Torger Reve, ”Interorganizational Relations and Dis- tribution Channels," A paper presented at Eight Annual Albert Haring Symposium, Indiana University, March 1978, p. 3. 16Reve, p. 3, and McCammon and Little, p. 330. 17Stern and El-Ansary, p. 10. 18Ibid, p. 8 19Bjorn Carlson and Bertil Kusoffsky, ”Distributor Brands and Conflicts In Distributive Organizations, in Stern (ed.), op.cit., p. 188. 20Bruce Mallen, "A Theory of Retailer—Supplier Conflict, Control, and COOperation,” Journal of Retailing, 36:27 (Summer 1963). 21James G. March and Herbert A. Simon, Organizations (New York, John Wiley and Sons, Inc., 1958), pp. 89—90. ZZStern and El-Ansary, p. 9 23Mallen, p. 27. 24Stern and El—Ansary, p. 6. 25McCammon and Little, p. 330. 26Mallen, p. 51. 27Wroe Alderson, Dynamic Marketing Behavior (Homewood, R. D. Irwin, Inc., 1965), p. 239. 28Stern and El—Ansary, p. 282. 29Stern, p. 2. 30Philip McVey, “Are Channels of Distribution What the Textbooks Say?” Journal of Marketing, 24:65 (January, 1960). BlIbid. 32McCammon and Little, p. 335. 331bid, p. 334. 34Stern and El—Ansary, p. 282. 35Rosenberg and Stern, p. 40. 36Stern and El—Ansary, p. 282. 53 37Louis W. Stern and Ronald H. Gorman, ”Conflict in Distribution Channels: An Exploration,” in Stern (ed.), op. cit., p. 172. 38Raymond W. Mack and Richard C. Snyder, ”The Analysis of Social Conflict—Toward an Overview and Synthesis," Journal of Conflict Resolution, 1:218 (June, 1957). 39Joseph C. Palamountain, Jr., The Politics of Distribu— tion (Cambridge, Harvard University Press, 1955), pp. 1—57. 4OKennett, E. Boulding, "Economics of Human Conflict,” in Elton B. McNeil (ed.), The Nature of Human Conflict (Engle- wood Cliffs, Prentice—Hall, Inc., 1965), pp. 172—173. 41Ralph M. Goldman, "A Theory of Conflict Processes and Organizational Offices,“ Journal of Conflict Resolution, 10:335 (December, 1968). 42Frederick J. Beier and Louis W. Stern, ”Power in the Channel of Distribution,” in Stern (ed.), op.cit., pp. 92-113. 43Thomas L. Berg, ”Designing The Distribution System,” in W. D. Stevens (ed.), The Social Responsibilities of Market- ing (American Marketing Association, 1961), p. 490. 44C. F. Fink, "Some Conceptual Difficulties in the Theory of Social Control,” Journal of Conflict Resolution, 12:456 (December, 1968). 45Adel I. El-Ansary and R. A. Robicheaux, ”General Model for Understanding Channel Member Behavior,” manuscript, Loui— siana State University, December, 1974, p. 9. 46Stern and Gorman, p. 156. 47March and Simon, p. 112. 48Fink, p. 432. 49Orville C. Walker, An Experimental Investigation of Conflict and Power in Marketing Channels, Dissertation, Uni- versity of Wisconsin, 1970, p. 18. 50Fink, pp. 412-460. 511hid, p. 432. 52Stuart M. Schmidt and Thomas A. Kochan, "Conflict: To- ward Conceptual Clarity,” Administrative Science Quarterly, 17: 365 (1972). 53Goldman, p. 335. 54 54Stern and Gorman, p. 155. 55Fink, 434. 434. p 56Ibid, p. 439. 57Ibid, p p 581bid, 436. 59Ibid, p. 435. 60Rosenberg and Stern, p. 42. 61Larry J. Rosenberg, "A New Approach to Distribution Conflict Management,” Business Horizons, 19:69 (October, 1974). 62Reve, p. 4. 63El-Ansary and Robicheaux, p. 3. 64Ibid, p. 10. 65Stern and El-Ansary, p. 292. 66Louis R. Pondy, ”Organizational Conflict: Concepts and Models,” Administrative Science Quarterly, 12:298 (Septem— ber, 1967). 67Ibid, p. 300. 68Henry Assael, ”The Political Role of Trade Associations -in Distributive Conflict Resolutions,” Journal of Marketing, 32:23 (April, 1968). 69March and Simon, p. 121. 70Robert w. Little, "The Marketing Channel: Who Should Lead this Extracorporate Organization,” Journal of Marketing, 34:32 (January, 1970). 71pondy, p. 300. 7ZSchmidt and Kochan, p. 363. 73Mallen, p. 25. 74Stern and Gorman, pp. 159—161. 75Assael, p. 22. 76Larry J. Rosenberg, "An Empirical Examination of the Causes, Level and Consequences of Conflict in a High Stake Distribution Channel,” Dissertation, Ohio State University, 1969, p. 37. 55 7 Louis W. Stern and James L. Heskett, "Conflict Manage— ment in Interorganization Relations: A Conceptual Framework," in Stern (ed.) op.cit., p. 293. 78Robert F. Lusch, ”An Empirical Investigation of Power, Conflict and Satisfaction in a Franchise-Franchises Channel of Distribution," Dissertation, University of Wisconsin, 1970, p. 34. 79Pondy, p. 301. 80Kenneth E. Boulding, ”Organizations and Conflict,” Journal of Conflict Resolutions, 1:131 (June,.1957). 81Pondy, p. 301. 823. Raven and A. Kruglanski, ”Conflict and Power,” in Paul Severgle (ed.), The Structure of Conflict (New York, Academic Press, 1970), p. 70. 83R. A. LeVine, "Anthropology and the Study of Conflict: An Introduction,” Journal of Conflict Resolution, 5f7(dune,1962y 84Rosenberg, ”An Empirical Examination...”, p. 42. 85LeVine, p. 7. 86Lusch, p. 34. 87Pondy, p. 302. 88James R. Brown, ”Toward Improved Measures of Distribu- tion Channel Conflict, in B. Greenberg and D. Bellenger (eds.), Contemporary Marketing Thought (American Marketing Association, 1977), p. 385. 89Rosenberg, ”A New Approach...”, p. 70. 90Fink, p. 433. 91Schmidt and Kochan, p. 365. 92Pondy, p. 304. 93Ibid, p. 308. 94Ibid. 95Stern and El—Ansary, p. 285. 96Stern and Heskett, p. 293. 97Mallen, p. 24. . 56 98Stern and El—Ansary, p. 8. 99Rosenberg, ”A New Approach...,” p. 73. 100Stern and Gorman, pp. 161—162. 101Stern and Heskett, p. 292. 102Stern and Gorman, p. 161- 103Michael M. Pearson, "The Conflict — Performance Assump— tion, ” Journal of Purchasing, 9:67 (February, 1973). 104J. Steven Kelley and J. Irwin Peters, ”Vertical Con- flict," A Comparative Analysis of Franchise and Distributors,” in Greenberg and Bellenger (eds.), op. cit., p. 384. 105Mack and Snyder, p. 228. 106Pearson, p. 60. 107Maok and Snyder, p. 245. 108Stern and Gorman, p. 163. 109Stern and Heskett, p. 292. 110Rosenberg, ”A New Approach...," p. 73. 111Mack and Snyder, p. 228. 11ZStern and Gorman, pp. 170—171. 113Many authors discuss this point; see Boulding (1965) Lusch (1976), Pearson (1973), Rosenbloom (1973), Rosenberg (1969), Stern and Heskett (1969). 114Stern and Heskett, p. 292. 115Bert Rosenbloom, ”Conflict and Channel Efficiency: Some Conceptual Models for the Decision Maker, Journal of Marketing, 33:27-29 (July, 1973). 116Robert F. Lusch, ”Channel Conflict: Its Impact on Re— tailer Operating Performance,” Journal of Retailing, 52:11 (Summer, 1976). 117Pearson, p. 66. 118Reve, p. 12. 119El—Ansary and Robicheaux, p. 4. 120L. A. Coser, ”The Termination of Conflict,” Journal of Conflict Resolution, 5:357 (June, 1962). 57 121Mack and Snyder, p. 243. l22Pondy, p. 296. l23El-Ansary and Robicheaux, p. 13. 124Stern and Gorman, p. 171- 125Robert F. Lusch, "Sources of Power: Their Impact on Intrachannel Conflict,” Journal of Marketing Research, 13:383 (November, 1976), and El-Ansary and Robicheaux, p. 97. 126Martin Fishbein, ”A Behavioral Theory Approach to the Relations between Beliefs about an Object and the Attitude Toward the Object,” in M. Fishbein (ed.), Readings in Attitude Theory and Measurement (New York, John Wiley and Sons, Inc., 1967), p. 389. 127Gordon W. Allport, “Attitudes, in C. Murchison (ed.), Handbook of Social Psychology (Worcester, Mass., Clark Univer— sity Press, 1935), p. 810. l28Pearson, p. 63. 129Martin Fishbein and I. Ajzen, Belief, Attitude, Inten— tion and Behavior: An Introduction to Theory and Research (Reading, Mass., Addison-Wesley Co., 1975), p. 316. 130D. Gilbert Harrell, Modeling Physician Prescribing Behavior, Dissertation, Pennsylvania State University, 1972, p. 13. 131Fishhein and Ajzen, p. 21. 1321bid, p. 29. 133Ibid, p. 31. 134Harrell, p. 13. 135Ihid., p. 17. 1361hid, p. 19. 137William L. Wilkie and Edgar A. Pessemier, ”Issues in Marketing's Use of Multi-Attribute Attitude Models,” Journal of Marketing Research, 10:428 (November, 1973). 138Martin Fishbein, "A Consideration of Beliefs, and Their Role in Attitude Measurement,” in Fishbein (ed.), op.cit., p. 259. 139G. 8. Day, ”Theories of Attitude Structure and Change," in S. Ward and T. Robertson (eds.), Consumer Behavior: Theore— tical Sources (Englewood Cliffs, Prentice-Hall, Inc., 1973), p. 309. 58 140Fishbein, "A Consideration of...”, p. 259. 141Fishbein and Ajzen, p. 12. 142Fishbein, ”A Behavioral Theory..., p. 390. 143Wilkie and Pessemier, p. 436. 144Ibid. I 145Jerome E. Scott and Peter D. Bennett, "Cognitive Models of Attitude Structure: 'Value Importance' is Important,” in Fred C. Allvine (ed.), Combined Proceedings: 1971 Spring and Fall Conferences American Marketing Association, 1971, p. 346. 146Fishbein and Ajzen, p. 228. CHAPTER III CONCEPTUAL MODEL AND HYPOTHESES Using concepts identified in the literature, this chapter will develop a conceptual model of the channel conflict pro— cess. By integrating conflict and attitude constructs, a model having implications for the management of the conflict process in distribution channels will be specified. In addi— tion to model conceptualization, the chapter will delineate specific research hypotheses which will be subject to empirical verification in subsequent chapters. Model Development Conflict Causes As previously noted, Stern and Heskett have identified three basic causes of conflict between channel members. They hypothesize the potential for conflict exists whenever members hold incompatible goals, domain definitions, and perceptions. Rosenberg empirically tested these latent stage variables in a channel system for an expensive household dur— able good.l Importantly, the study demonstrated different levels of the channel (i.e., manufacturer, distributor, and dealer) hold, in general, different goals, domains, and perceptions. Unfortunately, the study did not explicitly attempt to link latent variables to the later perceived, felt, or manifest stages of conflict. In terms of conflict aftermath, the study 59 60 attempted to identify a relationship between the latent variables and measures of financial performance. Rosenberg concludes his data failed to indicate either a threshold effect or any relationship to financial performance.2 However, he did find a limited relationship between the latent variables and satisfaction with the reciprical member's per- formance. Although significant for only two of the six dyadic relations (i.e. 3 channel levels, 2 relationships per level), it indicates the three latent variables tend to be related to a subjective measure of channel performance.3 In spite of the limited evidence on the linkage between the latent variables and later stages of conflict, several authors have conceptualized a direct relationship. In Stern and El—Ansary's recent text, Marketing Channels, the authors state conflict is a function of the three causal variables.4 The degree to which the behavior of one channel member could potentially destroy, thwart, or hinder the goal attainment of another is a function of goal incompatibility, domain dissensus, and differences in percep- tions of reality between them as well as the extent of their interdependence. In addition, the authors reproduce Cadotte and Stern's, ”Process Model of Dyadic Interorganizational Realtions,” which specifically incorporates a direct relationship between the three causal variables and conflict.5 61 Goal Domain Perceptual Incompatibility Dissensus Incongruency l J + Conflict Potential Between A & B + Perceived Conflict by A + Resultant Force on B's Goals, Domains, Perceptions i Conflict Aftermath Figure 1 Conflict Portion of Cadotte and Stern Model. Figure 1 illustrates the conflict portion of the model in which the three latent variables are shown to create a potential for conflict between channel members A and B. Member A will recognize the conflict state when the potential for conflict is sufficiently intense to the extent his induce- ment—contribution balance makes the channel relationship important. Depending on the power balance between the two members, member A's perception of conflict with B will lead to pressures on B's goals, domains, and perceptions to make them congruent with A's. The actual aftermath of the conflict depends on the resolution of the pressures on B's goals, 62 domains, and perceptions precipitated by the conflict with A along with other internal and external forces which affect B. The model posits a series of causal linkages between stages of the conflict process. It is structured largely with— in Pondy's five stage typology since the three causal variables form the latent condition which in turn lead to perceived conflict and the conflict aftermath condition. However, the model does not explicitly treat the affective or manifest stages, focusing instead on the forces that feedback to alter the latent variables. Since the review of the literature revealed the conflict construct to have important attitudinal dimensions, a model explicitly incorporating the felt conflict stage would be use- ful for understanding and analyzing channel conflict. Similar to the Cadotte and Stern model, a new model should show the vcausal connections betweentflmaconflict stages. Of particular interest would be the mechanisms whereby the latent variables lead to affective conflict. Integrative Model As previously discussed, the channel researcher's concept of felt or affective conflict is similar to the social psycho— logist's attitude construct. Specifically, it was noted Fishbein's model of attitude toward an object could be applied in a channel context to model the degree of affective conflict one member holds toward another. Such an application, in conjunction with Stern and Heskett's typology of conflict causes, 63 yields an integrative model of the conflict process. The model represents the latent, perceived, and felt stages in Pondy's description of the conflict process. Represented in Figure 2, the model integrates conflict and attitude constructs in the following manner: Pondy's Conflict Stages Components to be Integrated Latent Conflict (conditions) Goals, Domains, Perceptions Perceived Conflict (cognition) Expectancy or belief component, bi Felt Conflict (affect) Attitude Toward Object, AO In the model, Stern and Heskett's typology is incorporated with Fishbein's two component, multiattribute model of the atti— tude construct. The resulting model will be used to address the dissertation's three research objectives discussed in Chapter I by investigating the question: "Which component of affective conflict, belief or evaluation, is affected by the latent conflict variables?” Goals, Domains, Perceptions Beliefs Evaluations about other Channel of Member's Characteristics Characteristics Conflict Attitude toward other channel member Figure 2 Integrative Model of Latent and Affective Conflict Stages 64 The model contributes to the current level of knowledge regarding the conflict process by examining the mechanism in which one stage has an impact on another. This approach advances the prior empirical work of Rosenberg and conceptual development of Cadotte and Stern by focusing explicitly on whether the three causal variables influence attitudes toward other channel members by affecting beliefs about the channel member or evaluations of the member's Characteristics. Whether the conflict is influenced through beliefs or evaluations would be important information for channel man— agement in trying to reduce the intensity of conflict. As will be shown in the chapter six section, Implications for Channel Management, the various conflict reduction strategies available to management tend to work through either the belief or evaluation component. The implications of this will be further discussed in chapter six. Hypotheses Since the integrative model has potential implications for channel management, empirical verification of several posited associations critical to its overall structure is warranted. Five sets of hypotheses will be developed to test key relationships in the model. Each hypothesis set will be stated initially in theoretical terms in order to emphasize the constructs under investigation. Each will then be restated in operational terms using the variables which will be measured in testing the theoretical hypothesis. 65 Hypothesis Set I Theoretical: Latent conflict variables are associated with channel members' level of affective conflict. The hypothesis will test whether variables Stern and Heskett theorize as underlying conflict are systematically related to the degree of affective conflict. To the extent such a rela— tionship exists, the goals, domains, and perceptions of parti— cipants at one level of a channel should be related to attitudes held toward firms at a different level. Operational: A Channel member's goals, domains, and perceptions are associated with attitudes held toward members at other levels of the channel. Hypothesis Set II Theoretical: Latent conflict variables are associated with the belief component of attitude. This set is designed to examine whether the underlying conflict variables are systematically related to the cognitive component of attitude. The three causal variables should be related to the characteristics attributed to channel partici— pants occupying other levels of the channel, if the hypothesized association exists. As noted in Chapter II's section, Components of Fishbein's Model, the belief construct is defined as the probability or improbability that a particular relationship exists between the object of belief and some attribute or characteristic. Therefore, to operationalize the belief component the position 66 on a probability dimension an individual ascribes to statement linking a reciprocal Channel member to characteristics will be measured. Operational: A channel member's goals, domains, and perceptions are associated with the posi— tion on a probability dimension ascribed to relational statements, linking charac- teristics to other channel members. Hypotehsis Set III Theoretical: Latent conflict variables are associated with the evaluation component of attitude. Similar to Set II, the hypothesis' purpose is to establish whether the variables underlying conflict are systematically related to a component of attitude. If so, the three causal variables should be associated with the evaluation of charac- teristics in a good-bad sense. Fishbein intended the evalua— tion component as a measure of the desirability each attribute possesses for respondents. Therefore, the latent conditions should be associated with the value channel participants attri- bute to characteristics of reciprocal Channel members. Operational: A channel member's goals, domains, and perceptions are associated with the de— sirability of reciprocal channel member characteristics. Hypotehsis Set IV Theoretical: Belief component of attitude is associated with channel members' level of affective conflict. The hypothesis establishes whether participants' attitude 67 toward members at other channel levels is related to a specific component in Fishbein's attitude model. To the extent this is the case, the position on a probability dimension that a par— ticipant ascribes to relational statements should be related to attitudes held toward reciprocal channel members. Operational: Position on a probability dimension ascribed to relational statements linking characteristics to other channel members, is associated with attitudes toward those members. Hypothesis Set V Theoretical: Evaluation component of attitude is associated with channel member's level of affective conflict. Similar to Set IV, the hypothesis will test whether atti- tude toward reciprocal channel members is related to a particu— lar attitude model component. If true, the value channel par— ticipants attribute to characteristics of other channel members should be associated with attitudes held toward the channel members. Operational: Desirability of reciprocal channel member characteristics is associated with attitudes toward those members. Overall, the five hypotheses test the critical internal relationships of the integrative model. To the extent associa— tions posited are confirmed for a channel system, insights into the functioning of the conflict process offered by the model would be available for channel management. 68 Recapitulation In developing an integrative model of the conflict process, delineating its managerial significance, and speci— fying testable hypotheses of its structure, this chapter has presented the conceptual foundation of the dissertation. The following chapter will detail the research strategy necessary to empirically verify the hypotheses developed in this chapter. 69 CHAPTER III FOOTNOTES 1Larry J. Rosenberg, An Empirical Examination of the Causes, Level, and Consequences of Conflict in a High Stake Distribution Channel, Dissertation, Ohio State University, 1969. 2Ibid., p. 124. 31bid., p. 142. 4Louis W. Stern and Adel I. El-Ansary, Marketing Channels (Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 1977), p. 283. 5lhid., p. 293. CHAPTER IV METHODOLOGY Before the research hypotheses can be empirically veri— fied, important decisions must be made regarding the methods utilized to collect and analyze the data. This chapter will discuss the procedures and criteria employed: 1) in select— ing the channel system to serve as the empirical basis for the dissertation, 2) in developing operational measures of the research variables, 3) in designing the specific data collec- tion format, and 4) in statistically analyzing the data. The chapter will conclude_by assessing the limitations these pro— cedures impose on the ability to draw conclusions from the study. Selection of the Channel Candidate Channel Systems As mentioned in the Chapter I discussion on the scope of the channel system, administrative and franchise systems pro- vide channel participants with a balance between autonomy and commitment. In order for conflict to develop in distribution relations, members must be able to pursue their own objectives but still have a high stake in the particular channel and be functionally interdependent with other members. Initially, the institutional food distribution channel was investigated as a candidate system to empirically test the dissertation's hypotheses. However, the affective conflict 70 71 restaurant owners hold toward their food wholesalers or dis— tributors was found to be low. Separate personal interviews with several restauranteurs and their distributors revealed that the institutional food wholesale market is highly compe— titive resulting in restaurant owners having a low stake in particular channel systems. Restauranteurs tended to have a multiple supplier policy, purchasing from several distributors and frequently adding or dropping suppliers. In addition, the deveIOpment and pre-testing of an instrument to measure the channel's underlying potential for conflict revealed that the limited stake and functional interdependence lead to an ab— sence of conditions necessary for the study. Similarly, the beauty supply channel was investigated and rejected as a potential empirical system for the disserta- tion. .Again, the low stake and interdependence of barber and beauty shop operators with their suppliers lead to low levels of overt conflict. Personal interviews revealed operators terminate relationships with disagreeable suppliers. For ex— clusive distributors of sought—after branded products, func— tional interdependence was found to be limited to billing and physical distribution functions. Operators reported few prob— lems in these areas. In an attempt to identify a channel fitting the theoreti- cal criteria mentioned above, the franchised channel for the distribution of new automobiles was explored. Personal inter- views with dealers and manufacturer field personnel revealed a high amount of affective conflict between the parties. 74 divergence in member's goals is typical in distribution chan- nels:2 Goals between and among vertically linked firms are often different, and may be in— compatible and even mutually exclusive... one firm's goals may comprise another firm's constraints so that conflict results. Of course, as March and Simon point out, the reward structure of money payments between participants is developed especially to overcome differences in individual goals.3 However, they note the reward structure itself can stimulate goal conflict if it places members in competition for scarce resources. This appears to be the case in automobile distri— bution since manufacturer's production requirements for a large and steady output of all models Clashes with dealer sales needs for large supplies of a few popular models. Although dealers are rewarded by the gross margin received on cars sold, manufacturers faced with a shortage of popular models place dealers on allocations, thus limiting their ac- cess to monetary rewards and cause tensions. In order to examine the impact goals have on dealer atti— tudes toward their manufacturer, the goal construct should be carefully defined. To Perrow, goals define the ends de- sired through operating policies.4 The goals are formed by the unique problems and tasks the organization must emphasize since these tasks determine the organization's success. Furthermore, Perrow notes goals are important because they reflect choices among competing values in selecting and 75 defining the specific content of general business objectives such as growth and profitability. Since decisions must be made among many alternative ways of achieving these general objectives, dealers are likely to have different sets of ac- tive goals or at least have different preference orderings for the same goals. This possibility is critical for the analysis of conflict since, as Simon states, goals are the, "...value premises that can serve as inputs to decisions.H5 If dealers for a manu— facturer hold a variety of value premises, it seems likely that operating decisions made by the manufacturer would not satisfy the goals of all dealers to the same degree. Similarly, to the extent this condition existed, the various operating policies adopted by dealers would be unlikely to uniformly satisfy the value premises of the manufacturer. Biddle offers a formal definition of the goal construct. He states a goal,6 ...is a cognition consisting of a value held by a person which maps an end trace for the behavior of an actor in a context. According to this View, a dealer goal could be considered a valued end to which daily operating activities appear to tend. They reflect the dealer's value structure. In View of these approaches to the goal construct, the dissertation will adopt the following operational definition of goal: ,Goal is a value premise underlying task elements or Operating ends. 76 Goal Statements In order to translate the goal definition into a measure— able variable, questionnaire statements reflecting the under— lying value premises for important task elements had to be developed. Personal interviews were conducted with the follow- ing parties in an attempt to identify critical task elements and their associated value premises. 1. Marketing and Finance Editors of Automotive News (trade paper) 2. Officers of Michigan Automobile Dealers Association (trade association) 3. Manufacturer headquarter and field personnel 4. Dealers, i.e., individuals holding franchise agreements Figure 3 lists the results of these discussions. The five task elements represent critical features in the manufact- urer-dealer relationship. During the interviews, the parties mentioned these elements as being necessary for the efficient distribution of new automobiles. For each element a specific questionnaire statement was constructed that reflected the major value premise associated with it. For example, in dis— cussing selling effort, the interviewed parties recognized not all dealers shared the same viewpoint. It was noted that some dealers feel their time and effort is best spent on selling only models with high degrees of consumer acceptance, where as other dealers believe higher overall sales result from aggressively selling the entire line. Shown in Figure 3, a sentence explicitly stating a value premise for this task Goal Topics Selling Effort Profit Reinvestment. National Policies Profitability Trade—Off. Dealer Dependence. 77 Questionnaire Statements .When sales are slow for a model that is difficult to sell, a Dealer should devote more time and effort to selling it. .A Dealer should prefer to reinvest the majority of his profits back into the Dealership. .Ford adoption of nation-wide policies (as planned by Ford personnel and outside experts) is to be preferred to local flexibility of Dealers. .A Dealer should always be willing to satisfy customer demand even if doing so reduces Dealership profitability. .Having a Ford franchise is well worth the accompanying restric— tions on Dealer's independence in operating his business. Figure 3 Five Goal Topics and Related Questionnaire Statements 78 element was developed. Similarly, the underlying goal or value premise relating to the other task elements were constructed from the inter— views.7 Of course, whether these goals systematically in- fluence a dealer's attitude toward his manufacturer remains an empirical issue and will be explored in the next chapter. Domains A firm's domain refers to the role it occupies in the channel of distribution. To Stern and Heskett, ”Roles con- stitute the functions and the degree of functional performance expected of a channel member by other system members.”8 Roles represent families of norms that are maintained by the common expectations of other channel members. The behavior of any position occupant in a channel system is prescribed by the role expectations others hold for the firm. These role prescriptions, by specifying expected beha- viors, allow for the coordination of activities among channel participants. When a firm considers an activity as falling in its domain and the role expectations of others also attri— bute the activity to the firm, channel members are able to predict his behavior facilitating cooperation and enabling the entire channel to function as an integrated system.9 However, as El—Ansary and Robicheaux observe, role performance is a function not only of role prescriptions but personalized expectations, situational factors such as dependence and power tolerances, and organizational objectives.lO 79 These latter factors could cause role performance to deviate from role prescriptions. When this occurs, the domain of a firm would be defined in a way which is incongruent with the expectations of the other channel members. Such domain dis— agreements serve as a major cause of conflict creating insta— bility in the Channel resulting in inefficient resource allo- cation and lowering performance. Stern and Heskett note, ”When roles are unclear or not performed as expected, conflict is inevitable."11 In order to develop an operational definition of domain, the construct must be examined more formally. Stern and Gorman state,12 Each role in the channel represents a normative model of conduct which defines the expected contribution of the unit occupying the position to the achievement of the system's task. According to Sarbin, the model of conduct consists of ”rights” each member has in terms of expected behaviors from others and ”obligations” which are expectations of performance by others.13 Therefore, the dissertation will utilize the following operational definition of domain: Domains are rights and obligations held toward members at other channel levels. Domain Statements Questionnaire statements reflecting rights and obligations of channel participants were developed in a manner similar to goal statements. During interviews, the parties described Domain Topics Performance Pressure. Discretion Exercised. Control Exercised Inventory Requirements. Additional Dealerships. 80 Questionnaire Statements In order to insure that all Dealers perform adequately, Ford is justi— fied in setting performance stan— dards and insuring Dealers meet them. In developing marketing policies, Ford has the right to determine what is in the best interest of the overall franchise system. Because of the need for uniformity throughout its franchise system, Ford is entitled to control some important Dealer operating deci- sions. In order to maintain production of all models at an economic level, Ford is justified in requiring Dealers to take slow selling as well as fast selling models. Ford is entitled to determine the number of Dealerships necessary to satisfy consumer demand in a mar— ket area. Figure 4 Five Domain Topics and Related Questionnaire Statements 81 several important domain elements operating at the manufac- turer-dealer interface during the automobile distribution process. Figure 4 lists the elements and the associated questionnaire statement. Sentence construction utilized information obtained in the interviews expressing the rights and obligations of channel members relative to domain ele— ments. For example, the discussions revealed performance pressure was accepted by some dealers as a legitimate function of the manufacturer in managing its distribution system. On the other hand, other dealers were much less willing to attri— bute this activity as a right. Perceptions Perceptions of reality refer to channel member's subjec- tive interpretation of channel phenomena. Since information is a critical ingredient in a firm's decision making process, conflict between members can occur if information or an abstraction of reality is interpreted in varying subjective ways. In this regard, Stern and Gorman state,l4 Perception is the process by which an individual selects and interprets environmental stimuli. The interpretation need not, however, be con— gruent with reality. Perceptual errors are common occurrences...Thus, two decision makers within the same system may observe the same phenomenon in exactly reversed terms. Differences in perception indicate that there will be differing bases of action in response to the same situation. Divergent perceptions can have an important impact on the relationships between channel participants. March and Simon 82 observe that perceptions account for: 1. Criteria used to determine preferred courses of action 2. Expectations of consequences from each course of action 3. Knowledge of the existing alternative courses of action If members hold different knowledge, expectations, and criteria for courses of action, they are unlikely to closely coordinate their activities. The following operational definition of perception will be adopted: Perception is the subjective interpretation of a situation. Perception Statements Interviews revealed several critical perceptual topics in the manufacturer—dealer dyad. Figure 5 lists these ele— ments and their associated situational statements. Sentence construction involved expressing situations which would elicit respondent's subjective interpretation of the percep- tual element under consideration. Beliefs In the Chapter II discussion on the components of Fish— bein's model, belief was defined as the position a respondent ascribes to a relational statement on a probability dimension. A relational statement specifies an associated link between the object and some attribute. The probability dimension Perception Topics Business Guidance Manufacturer Support. Car Design. Advantageous Programs Dealer Density. 83 Questionnaire Statements .The guidance and advice Dealer receives from Ford is extremely useful in operating a Dealership profitably. .Ford provides Dealer with adequate support to enable Dealer to be competitive with other brands in his market area. .Ford designs cars exactly the way consumers want them. .Ford has nearly always proved trustworthy in introducing poli- cies and programs which are of mutual advantage to Dealer and Ford. .The number of Ford Dealers in most market areas is few enough to allow each to earn a satisfac- tory profit. Figure 5 Five Perception Topics and Related Questionnaire Statements 84 refers to the respondent‘s subjective probability that the particular relationship expressed in the statement exists. Of course, the selection of attributes to be included in the study is critical. The objective is to include only those characteristics of attitude objects which actually de— termine the subject's attitude. Relational statements linking the object to attributes subjects actually use in the forma- tion of their attitudes are referred to as salient beliefs. Fishbein and Ajzen have stated,16 ...a person's attitude toward an object is primarily determined by no more than five to nine beliefs about the object; these are the beliefs that are salient at a given point in time. Since only a relatively small number of beliefs serve as determinants of attitude, the attribute selection procedure must identify the few critical characteristics of attitude objects from the large number of beliefs respondents hold. Furthermore, the authors caution,l7 When belief statements are selected in an intuitive fashion, many beliefs will tend to be nonsalient, and they may also be unrelated to the underlying attitude. Belief Statements Fortunately, prior research on automotive channel systems provides an empirical basis for the selection of manufacturer attributes. Lusch in a study of automotive manufacturer's sources of power analyzed twenty-four manufacturer attributes.l8 85 Through discussions with channel participants, Lusch ident- ified sixteen noncoercive and six coercive characteristics which functioned as the important manufacturer attributes in manufacturer—dealer interactions. Since Lusch was interested in examining the impact sources of power have on intrachannel conflict, he obtained from dealers measures of manufacturer characteristics and the frequency of dealer disagreement with the manufacturer. Through a national mail survey involving domestic and foreign new car dealers, Lusch established a relationship between each attribute and the level of perceived conflict. Since the Fishbein model requires attributes that are related to the respondent's attitude, the attributes Lusch identified as having the greatest systematic association to conflict will be used in the present study. Thus, manufacturer characteristics bearing little or no relationship to the way dealers perceive their manufacturer will be excluded. The coercive and noncoercive attributes which will be utilized in the study are: Coercive Slow payment on warranty work Unfair vehicle allocation Unnecessary bureaucratic red tape Slow delivery of ordered vehicles Threats of termination to get Dealer cooperation 86 Noncoercive Effective district sales representative Excellent quality product warranty Effective year-end inventory rebate plan Worthwhile Dealer incentive programs Excellent quality mechanic training To incorporate these attributes into belief statements, it is necessary to form an associative link with the attitude object. To accomplish this, sentences were developed in which each attribute was preceded by the noun, Ford. For example, ”Ford has effective district sales representatives.” Respondent's would register their belief strength for this attribute on a probability scale. Scaling will be dis— cussed in the section on questionnaire development. Evaluations The second component in Fishbein's model refers to the evaluation of the attributes in a good—bad sense. The compo- nent is intended as a measure of the desirability each attri- bute holds for the subject. The potential importance of the model's value component in the analysis of Channel conflict is demonstrated in an ob- servation by McCammon and Little,19 Research undertaken at all levels of the dis— tribution system indicates that the firm's behavior and decisions are strongly conditioned 87 by the entrepreneur's value hierarchy. Thus, the way a dealer evaluates manufacturer attri- butes could influence the degree of conflict present in the dyad. Evaluation Statements The ten attributes derived from the Lusch study will be used to obtain evaluation measures. For the noncoercive attributes respondents will register their evaluations accord— ing to the degree they are perceived as beneficial. Similar— ly, evaluations of the coercive attributions will be measured in terms of the degree they are perceived as harmful. Data Collection Procedures In order to gather the data necessary for testing the research hypotheses, a questionnaire had to be designed and distributed to Michigan Ford dealers. The specific data collection procedures utilized are described below. Initial Research Activities Officers and members of the Michigan Automobile Dealers Association (MADA) were consulted regarding possible procedures to be used in collecting information from Michigan Ford deal- ers. During those discussions the following procedures were deemed necessary to obtain an acceptable response rate. 88 1. Develop and pre-test questionnaire format by personal interviews with dealers. 2. Prenotify dealer of study by personal telephone call. 3. Personalize mail questionnaire by hand addressing envelope using dealer's first and last name. 4. Include cover letter on Michigan State Uni— versity letterhead signed by researcher and endorsed by MADA. (see appendix A) Using these procedures, it was felt an adequate response could be obtained by mailing a questionnaire to dealers rather than attempting to personally interview dealers throughout the state of Michigan. A stamped pre—addressed return en— velope was included in the mailing. Names, addresses, and phone numbers of Michigan Ford dealers were obtained from the 1978 Yearbook and Directory of the MADA, and from a list of dealers not currently members of the trade association. Regarding the use of telephone prenotification, Jolson identifies several important advantages over conventional mail—only approaches:20 Differentiates the mailing piece from ”junk mail” by alerting the respondent to its impending arrival. Stresses the importance of the respondent's role and provides a medium for answering his questions. Introduces a dimension of intimacy and elicits his promise to cooperate. Furthermore, Jolson observes,21 89 The use of a well—administered initial phone contact in advance of the mailing may offer the potential for a real breakthrough — a doubling or tripling of return rates...However, the effectiveness of the method is highly dependent on the design of the phone message and the skills of the personnel who deliver it. Adopting Jolson's telephone message format to the present study resulted in the standard phone presentation shown in Figure 6. The MADA generously agreed to support the research by permitting the researcher to use the organization's phone system to contact a sample of dealers. To enlarge the sample and to call back dealers not initially available, additional calls were made from the researcher's residence. Since the phone calls were made during the last weeks of June, 1978, several dealers were on vacation or otherwise unavailable. Therefore, it was decided to include as many dealers as pos— sible in the sample. The following section discusses the specific content of the mail questionnaire. 90 Hello. I'm Dan Bello with the College of Business at Michigan State University. I'm studying the problems owners of car dealerships in Michigan have with their manufacturers. Are you the owner? I've randomly selected dealerships in Michigan including yours, and will be sending a brief ques- tionnaire to get dealer opinions about their manufacturer. The questionnaire is endorsed by the Michigan Automobile Dealers Association. You wouldn't mind filling out the questionnaire would you, Mr. ? Your name is not on the questionnaire and you are asked to give your opinion to some questions by placing an X. I'll mail the questionnaire today and it will arrive in a Michigan State University enve— lope. I've called you in advance since we've found if we don't call first, some dealers fail to return the questionnaire and our results suffer. Your reply is very important. Will you try to return the question— naire soon after you receive it? Thank you, Mr. . I appreciate your help. Figure 6 Standard Presentation Used In Telephone Pre—notification 91 Questionnaire Development The statements associated with each measured variable were organized into a three—page questionnaire shown in Appendix B. Part I consists of ten belief statements linking Ford to five noncoercive and five coercive manufacturer characteristics. To encourage respondent to consider each statement individually and avoid possible response patterns, the statements are alternated beginning with a noncoercive characteristic. A seven-point response scale, anchored by the descriptions ”strongly agree" and ”strongly disagree,” was used to register dealer!s subjective probability for each of the ten Ford—attribute associations. 1. Belief Statement Strongly Strongly Agree : : : : : : Disagree Before this scale was selected, two other scales were developed and pre—tested. However, dealers expressed uncertainty in their interpretation and found them generally confusing. 1. Belief Statement Very Very Unlikely —3 -2 —1 0 +1 +2 +3 Likely 92 1. Belief Statement Extremely Extremely Improbable : : : : : : Probable Parts II and III were designed to measure dealer evaluations for the noncoercive and coercive attributes, respectively. In each section, the five attributes were listed and dealers were instructed to indicate their evalua— tions on a seven—point scale. The end point descriptors for the noncoercive statements were ”moderately beneficial” and ”extremely beneficial.” For the coercive statements, the end points were ”moderately harmful” and ”extremely harmful.” 1. Noncoercive attribute Moderately Extremely Beneficial : : : : : : Beneficial 2. Coercive attribute Moderately Extremely Beneficial : : : : : : Harmful 93 Again, prior to the selection of these scales, alterna— tive versions were developed and pre—tested. For example, the folloiwng noncoercive scale was judged to be too confusing by respondents. (Coercive scales were similar in construction.) 1. Noncoercive Attribute Least Most Desirable —3 —2 —1 0 +1 +2 +3 Desirable Part IV listed fifteen sentences which consisted of alternated goal, domain, and perception statements. Dealers were instructed to respond on a seven—point anchored scale similar to the scale used in Part I. Part V consists of a set of standard semantic differen— tial scales. Following the technique pioneered by Osgood et. al., a series of bipolar adjective scales were used to mea— sure overall affect toward the attitude object, Ford.22 The original studies by these authors on the nature of semantic space identified the three major dimensions of evaluation, potency, and activity. Since Fishbein considers the attitude construct, ”...to refer solely to a person's location on a bipolar evaluative or affective dimension with respect to some object...,“ the evaluative dimension will be of primary interest.23 Eleven sets of adjective pairs appear in Part V and represent the following semantic dimensions. In the question— naire the adjective pairs are placed in random order and are occasionally reversed in direction to avoid possible pattern- ing of responses. 94 Evaluation Dimension Foolish - Wise Awful — Nice Unpleasant — Pleasant Wrong — Right Bad - Good Cruel - Kind Punishing - Rewarding Potency Dimension Weak - Strong Activity Dimension Passive — Active In addition to these adjective pairs which were taken from Osgood et. al.'s original Thesaurus study, two conflict related semantic adjective pairs were developed and included in Part V. These additional pairs appeared on a priori basis to be related to the channel conflict construct. Conflict Related Descriptors Frustrating — Helpful Adversary - Partner Following the usual semantic differential procedure, all eleven adjective pairs were submitted to a factor analysis procedure to determine the relevant dimensions for the speci— fic context of the research. The results of the factor analysis for the current distribution channel context are discussed in the following chapter. Additional questions were included in the questionnaire to obtain the following descriptive information about the 95 respondent. 1. Number of years holding Ford franchise 2. Additional automotive brands represented 3. Percentage of time spent operating Ford dealership 4. Ford new car and truck unit volume in 1977 5. Comparative 1977 dealership profit perfor— mance General degree of Ford control over deal- ership O) 7. General degree of satisfaction with Ford 8. Frequency of disagreements with Ford Statistical Analysis of Data To empirically test the research hypotheses constructed in Chapter III, the operational measures just discussed will be statistically analyzed in the next chapter. The data collected on the automotive channel system will be investi— gated in order to determine if the posited theoretical rela— tions can be verified. Multipleregression will be utilized to determine if the associations proposed in the dissertation's conceptual model and formalized in the research hypotheses exist for the channel system under study. As Kerlinger and Pedhazner note, multiple regression is an appropriate statistical technique to investigate such theoretical relationships.21 Multiple regression is a method of analyzing the collective and separate contributions of two or more independent variables, X, to the 96 variation of a dependent variable, Y. The fundamental task of science is to explain phenomena. As Braithwaite (1953) says, its basic aim is to discover or invent general explanations of natural events. The purpose of science, then, is theory. A theory is an interrelated set of constructs or variables “that presents a systematic View of phenomena by specifying relations among variables, with the purpose of explaining...the phenomena.” But this view of science is close to the de— finition of multiple regression. Therefore, the interrelated set of constructs comprising the theoretical basis for the dissertation will be evaluated by testing the strength and statistical significance of the association between criterion and predictor variables. By evaluating and measuring the overall dependence of a criterion on a set of predictor variables, multiple regression can be used to assess the strength of hypothesized associations. However, beyond regression's ability to describe and summar— ize existing relationships in the data, the question of statistical inference can be addressed. The question of whether the results of the sample observation can be general- ized to the population is critical in drawing any conclusions from the study. Both the F and t tests will be utilized in testing the statistical significance of the regression analy— sis. The specific manner in which multiple regression will be used to test the individual research hypotheses will be discussed in detail in the next chapter. 97 Limitations of the Study Although the project attempts to examine hypothesized relationships among selected variables of distribution systems in general, the evidence is based on a single channel of distribution. Consequently, the generalizability of any study conclusions are limited to channel systems similar in nature to the system used as the empirical basis for the study. However, as discussed in Chapter I, research indi— cates conflict intensity is fairly similar across various types of channel systems which suggests insights gained from the automobile franchise channel may have broader application. Nevertheless, several important aspects of the study's channel system serve as limitations. Since data was collected in a restricted geographical area for a single manufacturer, the study may not represent the complete automotive channel. Furthermore, non—responding dealers reduce the thoroughness with which even the one manufacturer's system was studied. However, a follow up survey using an abbreviated questionnaire suggests this is not a serious problem. In a general sense, the conclusions may be influenced somewhat by the peculiari— ties of the particular manufacturer and region. In addition, the study was conducted at a point in time, the last weeks in June, 1978, and may not be representative of other econo— mic, political, and social environments. Unlike a laboratory experiment where variables exogenous to the process under examination can be controlled, field research sacrifices such control for realistic settings. As 98 a result, a controlled examination of the conflict process and a complete understanding of the causal connections in the dissertation's conceptual model is not possible with the survey type field research utilized. Of course, this limitation is offset by the fact artificial settings associated with the experimental approach introduce biases which are not operating when the study is conducted in the environment natural to the process under examination. Another limitation of the study is based on the statis— tical tests used to verify the hypotheses. The regression and correlation analysis utilized measures the degree of association among variables but does not imply causation. Furthermore, the techniques have certain mathematical assump- tions which if not met can distort the data analysis. Kerlinger and Pedhazur note that in order for regression analysis to yield a meaningful test of statistical signifi— cance the dependent variable should be normally distributed and have equal variances.25 Also, the distribution of the deviations from the regression should be random and normal. Furthermore, the analysis assumes an interval level of data. The scales utilized in the research were assumed to yield interval data because of the way respondents perceive equally spaced anchored scales. However, Kerlinger and Pedhazur conclude the statistics used to test significance in the re- gression model are fairly robust to deviations from the underlying mathematical assumptions and, consequently, are not likely to seriously distort the study's conclusions.26 99 CHAPTER IV FOOTNOTES 1James R. Brown, ”Toward Improved Measures of Distri— bution Channel Conflict,” in B. A. Greenberg and D. N. Bellenger (eds.), Contemporary Marketing Thought (1977 American Marketing Association Proceedings), p. 385. 2Larry J. Rosenberg and Louis W. Stern, ”Toward the Analysis of Conflict in Distribution Channels: A Descrip— tive Model,” Journal of Marketing, 34:40-46 (October, 1970). 3James G. March and Herbert A. Simon, Organizations (New York: John Wiley and Sons, Inc., 1958), p. 124. 4Charles Perrow, ”The Analysis of Goals in Complex Organizations,” American Sociological Review, 26:855 (December, 1961). 5Herbert A. Simon, ”On the Concept of Organizational Goal,” Administrative Science Quarterly, 9:3 (June, 1964). 6Bruce J. Biddle, ”Roles, Goals, and Value Structures in Organizations,” in W. W. Cooper (ed.), New Perspectives 1L Organization Research (New York: John Wiley and Sons, Inc., 1964), p. 163. 7Questionnaire item format similar to Larry J. Rosen— berg, "An Empirical Examination of the Causes, Level, and Consequences of Conflict in a High Stake Distribution Channel,” Dissertation, Ohio State University, 1969. 8Louis W. Stern and James L. Heskett, ”Conflict Manage— ment in Interorganization Relations: A Conceptual Framework,” in L. W. Stern (ed.), Distribution Channels: Behavioral Dimensions, p. 294. 9A. I. El—Ansary and R. A. Robicheaux, ”A General Model for Understanding Channel Member Behavior,” Journal of Retailing, 52:13—30 (Winter, 1976). lOIbid. 11Stern and Heskett, p. 294. 12Louis W. Stern and Ronald H. Gorman, ”Conflict in Distribution Channels: An Exploration,” in Stern (ed.), op.cit., p. 157. 13Ibid. 14Ibid., p. 159. 100 15March and Simon, p. 11. 16Martin Fishbein and I. Ajzen, Belief, Attitude, Inten- tion and Behavior: An Introduction to Theory and Research (Reading, Mass.: Addison-Wesley Publishing Co., 1975), p. 218. 17Ibid., p. 228. 18Robert F. Lusch, "Sources of Power: Their Impact on Intrachannel Conflict,” Journal of Marketing Research, 13: 382—390 (November, 1976). 19Bert C. McCammon and Robert W. Little, "Marketing Channels: Analytical Systems and Approaches," in G. Schwartz (ed.) Science in Marketing (New York: John Wiley and Sons, 1965), p. 344. 20Marvin A. Jolson, ”How to Double or Triple Mail-Survey Response Rates,” Journal of Marketing, 41:78 (October, 1977). 211bid., p. 81. 22Charles E. Osgood, George J. Suci, and Percy H. Tannen- baum, The Measurement of Meaning (Urbana, Illinois: University of Illinois Press, 1957). 23Fishbein and Ajzen, p. 216. 24Fred N. Kerlinger and Elazar J. Pedhazur, Multiple Regression in Behavioral Research (New York: Holt, Rinehart and Winston, Inc., 1973), p. 3. 25Ibid., p. 47. 26Ihid. CHAPTER V ANALYSIS OF RESULTS In this Chapter the data generated by implementing the previously described methodology will be analyzed. The chapter begins with an assessment of how representative the responding sample of dealers are of the dealer population by evaluating information gathered on non—respondents. Next, a profile of the responding dealers will be developed detail— ing demographic and attitudinal dimensions of the sample. The balance of the chapter will be devoted to a detailed statistical analysis of the research hypotheses. Analysis of the Sample Non—Respondents In order to investigate whether the dealers participating in the study were representative of all Ford dealers in the state, a shortened version of the original three—page ques- tionnaire was mailed three weeks after the initial mailing. The accompanying cover letter (Appendix C) reminded the dealer of the academic and confidential nature of the project and restated its endorsement by the trade association. The letter further urged the dealer to complete and return the original long questionnaire but asked, if that was not possible, for the return of the shortened version. Although containing only eight questions, the abbrevi— ated questionnaire (Appendix D) requested both descriptive and attitudinal information about the non—responding dealers. 101 102 Questions were identical to ones asked in the long version. Dealers were asked the length of time they had been Ford dealers, the percentage of time spent running their dealer- ships, their annual unit sales volume, and their relative profitability. Perhaps more important in a study of Channel conflict, however, is whether respondents and non-respondents hold similar attitudes and perceptions toward the manufacturer. To assess these dimensions non—respondents were asked the degree of control Ford possesses over the operation of their dealership, their degree of satisfaction with Ford, and the frequency with which they disagree with the manufacturer. Table 1 compares the responses for the two groups of dealers on the common descriptive and attitudinal questions. A t-test was conducted to determine whether the groups dif— 'fered significantly on the items. Since none of the mean responses differed for the two groups, it was concluded the dealers responding to the original long questionnaire were representative of the total Ford dealer population in Michi— gan. 103 pom s on 30H How H Eoum manCmH mamom m do Hom. mom. smo. mam. Nwo. mqm. mom. NN.H N.q mo.m ©.m ww.o H.¢ 4N.H N.¢ mm.H 0.6 ma.o n.0m wo.ma m.ma .Q.m and: cauz mamamm Hmamom ucmvcoamCMIcoz .Hm>oa mo. can um uamummmwe kaucmoflmwawwm mum mance can mo oaoz n wowmzmuw muoB msowummsv mafia wo wmwuamouma mam mute» How uacoxm sm.a m.q wm.a H.m om.H o.q mm.H ©.q aq.a m.m H©.oa q.sm Ho.qa m.ma Iamnm. sees mqauz oaaamm seamen wafivaoammm .ann w .mumawmw nmfiuo cu commaaoo %uwaflnmuwmoum aflnmumammm .masao> HHS: Momma can umo 3mm Hmsaaw mo huowmumu .Cuom suHB quECCHmmme mo mouwon .QHSmaoHuwHou @Hom SuwB coauommmwumm mo common .QHSmuonme mcflumumao cw mmmwoumxm whom Houucoo mo wmuwmo .mflnmumamou wcwumummo ucoam mawu mo mwmucmupmm .umawow wuom m mm memoh mo Hmnasz mmamuH oufimccowummmd HmofiuaopH osflmccofimeSG copwfl>osn9¢ cw mafipofiaaoo mponoQ wcflecoamomucoz we ®HQEdm m can mpnocqogmom yo mfimzads< H oases 104 Profile of Respondents Table 2 compares the geographic location of dealers who returned the questionnaire to the location of all Ford dealerships in Michigan. The 149 responding dealers are located in each of the major geographic subdivisions of the state and represent 67% of the total number of dealerships. Since the questionnaire was anonymous the postmark on the return envelope was used to identify the dealers' location. Although the originating location for twelve questionnaires could not be identified, the sample of dealers appears to be representative of the actual geographic distribution of dealers. Table 2 Geographic Location of Dealerships and Responding Dealers in State of Michigan Total Number Number of Respondents as of Responding Percentage of DealerShipsa Dealers Dealerships Upper Peninsula 18 . ll 61% Lower Peninsula Upper Half 40 17 43% Lower Half 165 109 66% Unknown Postmark —— 12 -— Total: 223 149 67% a From Michigan Automobile Dealers Association 1978 Directory and non—member list. b Location based on return envelope postmark. C A few dealers operate more than one dealership either through partner— ships or special franchise agreements. 105 The average length of time dealers in the sample had held their Ford franchise was 15.7 years. Table 3 shows the distribution of dealers according to the number of years they had represented the manufacturer. Response ranged from several months to over fifty years. However, dealers' atti- tude toward Ford was not related to their tenure with the company. The distribution of unit sales volume for the dealer sample is shown in Table 4. Approximately half the dealers sold between 200 and 550 new cars and trucks during the pre- vious calendar year. Although the table indicates a broad range of sales levels is contained in the sample, a dealer's unit sales volume is independent of his attitude toward Ford. Table 3 Number of Years Responding Dealers have been Ford Dealers Years as Number of Percentage of Ford Dealer Respondents Respondents less 3 17 11.4 3—5 22 14.8 6—10 23 15.4 11-15 24 16.1 16-20 20 13.4 21—30 24 16.1 31—50 8 5.4 over 50 7 4.7 no response _4 2.7 Total 149 100.0 Mean Years 15.7 Correlation with attitude .04a a not significant at .05 level 106 Table 4 Annual New Car and Truck Unit Sales Volume of Responding Dealers lumber of Percentage of Unit Volume Respondents Respondents 1 to 199 25 16.8 200 to 349 36 24.2 350 to 549 36 24.2 550 to 999 17 11.4 1,000 to 1,700 19 12.7 over 1,700 16 10.7 Total: 149 100.0 Correlation with attitude .04a a not significant at .05 level Table 5 displays dealer responses to a question asking them to compare their profitability to all Michigan Ford dealers. Approximately 24% of the dealers view themselves Has earning below average profits, 28% as earning average profits, and 48% as earning above average profits. However, when profit responses were correlated with attitude toward Ford only a relatively small correlation coefficient (r = .16) was obtained. Since a dealer's profit- ability accounts for less than 3% (r2 = .026) of the varia— tion in attitude toward the manufacturer, the question arises as to the nature of the forces which account for the unex— plained variation in dealer's attitudes. Later in this chapter, hypotheses derived from the dissertation's concep— tual model will be empirically tested. Table 6 shows responses to questions in three areas of 107 Table 5 Responding Dealers' Perception of Their Profitability Compared to All Michigan Ford Dealers Responsea Number of Percentage of Measure Respondents Respondents 1 6 4.1 2 10 6.8 3 19 12.8 4 42 28.4 5 33 22.3 6 18 12.1 7 20 13.5 no response 1 — Total: 149 100.0 Correlation with attitude .16b 3Responses were measured on scale ranging from ”Greatly Below Average“ = 1 to ”Greatly Above Average” = 7. bsignificant at .05 level a dealer's relationship with the manufacturer. In terms of the degree of control dealers feel Ford has over the opera— tion of their dealership, 46% feel the manufacturer has little or no control, 22% some control, and 32% attribute a high degree of control. The perception of control was not signi— ficantly related to dealers' attitude toward the manufacturer. When asked how satisfied they were with their relation— ship with Ford, 36% stated they were dissatisfied, 19% were neutral, and 57% were satisfied. Since only a little over half of the dealers were satisfied with their manufacturer relationship, the impression obtained during the initial Responding Dealers' 108 Table 6 Perception of Their Relationship With Ford Questions 1. In general how much control would you say Ford has over the operation of your Dealership? 2. How satisfied are you in general with your relationship with Ford? 3. How frequently do you disagree with Ford? Response Measure 1 2 Total: Mean Correlation with attitude asignificant at the .001 level Control (Question 1) Freg. _j§_ 5 3.3 33 22.1 30 20.1 33 22.1 33 22.1 13 9.0 __i; 1.3 149 100.0 3.7 -0.13 Satisfaction (Question 2) Freq. _;§_ 3 2.1 13 8.7 20 13.4 28 18.8 33 22.1 38 25.5 _14_ 9.4 149 100.0 4.6 -o.4aa No Control Complete Control Very Dissatisfied = Very Satisfied Very Infrequently = Very Frequently Disagreement (Question 3) Freg. _ZL_ 8 5.4 20 13.4 28 18.8 33 22.1 31 20.8 22 14.8 __1_ 4.7 149 100.0 4.0 -0.318 109 research phase that the channel exhibited conflict was con— firmed. Similarly, the responses to a question on how frequently they disagreed with the manufacturer, dealer responses indicate a substantial amount of channel conflict. Only 38% report infrequent disagreement with Ford, 22% were neutral, while 40% report they frequently disagree. Furthermore, both the disagreement and satisfaction questions were significantly correlated with dealer attitudes toward Ford indicating a close association with the operational measure of affective conflict used in the study. Testing of Hypotheses Although attitude toward Ford was independent of the number of years the franchise was held, annual unit sales volume, and the perceived degree of control Ford has over dealership operations, the relationships posited between latent causal variables, evaluations, beliefs, and affective conflict may be demonstrable. The theoretical relations identified in the literature, described in the dissertation's conceptual model, and formalized as research hypotheses will be empirically analyzed in this section. Hypothesis I The first hypothesis states goals, domains, and percep— tions of participants at one level of a channel should be related to attitudes held toward firms at a different level. 110 Before the association can be tested the degree of intercorrelation among the fifteen latent variables should be assessed. Table 7 shows the matrix of correlation co— efficients for these variables. The intercorrelations range from a high of .47 for dealer dependence and advantageous programs to a low of .004 for national policies and sell- ing effort. In order to assess the pattern of the intercorrelations a factor analysis of the fifteen independent variables was conducted. Four underlying latent dimensions were revealed. Table 8 shows the latent variables having the highest load— ings on each of the factors. An analysis of the question-3 naire statements for the set of latent variables associated with each factor suggests the following factor interpretations: Factor Interpretation Policy Domain Dealership Perception Profit Goal Support Perception ewmw Each of the latent variables with high loadings on the first factor refer to the manufacturer's right to establish various policies in operating its franchise dealer system. The highest loading variable, national policies, expresses the value premise that manufacturer adoption of nation—wide policies is to be preferred to local dealer flexibility. NNH. ovo. NHH. mMH. mom. mmH. mHo. HHm. ham. Ohm. ova. mom. HoN. oMH. mom. mam. NoN. hmm. NBN. omm. oHo. ohm. oom. vom. vmm. omN. Mbo. mNH. me. mmH. oNN. GUN. mom. omN. omH. NON. HNN. mow. mNN. CNN. 11 OOO.H vmo. mmo. moo. OON. 11 11 emo. ooo.H omm. mma. osa. mmo. mom. OOO.H mmm. fiHm. moo. mmH. mmm. ooo.H on. OON. ohH. va. on. OOO.H moo. mmH. mvm. 5mm. mNN. va. OOH. OBH. woo. mod. ucoEUmo>cflom uuommsm whommohm .mmmmmmmm .mmmmmmmm uflmoum Housuommscmz oocmEuomuom apnoeumz mmoCHmsm moa. flaw. wwm. ohm. How. wow. ova. mmm. moo. mma. mom. hmm. mNN. ooo.H OHH. ommflouoxm coflaouomflo mNH. bra. mvm. mom. oho. mom. mom. mha. vow. OOH. 05H. woo. me. OHH. ooo.H uuommm ocaaaom mufimcoo Hmawmo uaaounoaooo Hoeoanaooa mocoocomoa Hoamom mficumoum msoommucm>om mpcmEoHflswom muoucm>cH uuonooene soaaaoooauona daemon HMO oomflouoxm Houpcoo ucoeumo>CHdm uflmoum unommdm MoH5u06mdcmz ousmmonm moccauomuom mofloflaom HMGOHumz mocmoflso mmocflmsm Comflouoxm coauouomfla unomum moaaaom moanwflsw> pcopmq ma how mucoHOAHwooo noflpaaoppoo we Xenon: b mHQmB 112 mNm. «mm. vmv. ooo.H vom. vma. mom. mam. mam. mom. 5mm. New. ohm. ohm. mom. mEonoum msoommucm>©¢ wha. mmm. mom. wom. ooo.H ova. boa. th. oHo. ohm. oom. VOM. vow. How. one. quwEoHHSwom muoucm>cH oma. moa. new. oma. smm. mew. ema. mom. osa. sea. ooo.a How. How. . ooo.H and. mam. oom. 6mm. meo. oem. mma. mom. mma. 6mm. mma. ooa. eom. oea. mom. mom. Geolooons .mmmmmm snaaaoonauona noo A©.uGOUO h OHQMB who. NmN. vow. mam. th. wha. mom. ooo.H mom. ANN. mow. mNN. Ohm. mNm. th. oomflonmxm Houucou unflmcmo seamen moannuoaooo Hoeoauaooa oocowcomoa Hmamoa mamumoum msoommucm>om mucofimuflswom muoucm>cH nuonoooue snaaaoonauona smflmoo HMO oomflouoxm Houpcoo pcoEumo>aHmm uflmoum uuommsm HmHSHOMMDCHE ouummonm moccauomuom mofloflaom HMGOHumz mocoofldw mmocflmdm oomfloumxm cofluouomfla uuowmm mcaaaom 113 ooo.H mow. mom. mmm. NBA. oNH. moH. who. NNH. mvo. baa. mma. mom. moH. mNH. soanooo uonoQ ave. mmm. ooo.H Nam. Nam. ooo.H vmm. vhv. mmm. mom. wvm. hmm. omH. mom. mom. vow. mma. ova. mao. mom. Ham. How. ham. oma. Ohm. mom. vav. vow. boa. mom. mmflnmuoamoo oozeocommo HMCOHpfloom Hmamma Ao.ncooo e moans mafimcmo seamen umarnuoaooo aoeoanaopa ooqmocomoo Hoamoo mamumoum msoommucm>om mpcoEouflswom mnouao>cH uuouooone snaaaoonauoua cmwwmo moo oomflouoxm Houucoo ucofipmo>cflom uflwoum whommdm HmH5u06mscmz ousmmonm oocmsuomuom wofloflaom Hmcowumz oosmoflsw mmocflmsm oomflouoxm coflumuomfla uuommm ocaaaom 114 Table 8 Factor Analysis of Latent Variables Varimax Rotated Factor Matrix a Latent Variables Selling Effort Discretion Exercised Business Guidance National Policies Performance Pressure Manufacturer Support Profit Reinvestment Control Exercised Car Design Profitability Trade—off Inventory Requirements Advantageous Programs Dealer Dependence Additional Dealerships Dealer Density Eigenvalue Percentage of total Factor Variance aLoadings above .5 are in parentheses. -0 GOO 000 29. Factor 1 .06493 (O. 0. 70292) 35054 .70544) .63228) .16420 .00215 .57370) .18883 .30321 .44284 .22115 .09022 .42261 .00395 .36 Factor 2 0. 0.23192 0. GOO 000 (O. (O. (O. 9. 17277 38915 .04524 .09163 .01127 .07232 .06792 .07638 .09638 .34975 .54482) 65357) 68667) 80584) .47 8 A 000 A COD (0. —0. .24160 0 Factor 3 .64748) .00700 .19921 .04306 .08476 .01688 .75935) .22984 .47723 63586) 07240 .16787 .10971 .06712 .27 7 Factor 4 .22261 .04352 .18055 .07610 .36251 .78323) .02979 .29418 .43593 .14848 .37080 .43188 .38304 —0. —0. 17360 06166 .05 .0 115 .eoonaa one m. o>ono awoaeooa hogs noaooano> sauce womfiouoxm Hosuaoo wooloowwe kuflaflomuwwoum msmuwopm msoowmuam>o< ousmmoum mommauomuom uuowwm wGHHHom mesmeaoaoo Hoamom womwouoxm deflumuomwm uuoaasm mafinmuoamom umuauommscmz ucoEumo>aHmm uflwoum HwGOHuwoo< mowowaom HmGOfiumz “moanmflum> cowumouuom uuoaasm .q Hwoo bemoum .m sowuaooumm awnmhoamom .N aflwaom %uaaom .H ”mucuowm MWMOPOdnH GO MQHUflOQ mmfifldfihdxw UCd WOEGZ .HOPOdnH o manmb 116 Similarly, discretion exercised, performance pressure, and control exercised express the legitimacy of the manufacturer developing policies to control dealers. Since the Chapter IV discussion of the domain construct concluded expressions of a channel member's rights constitute domain statements, the first factor can be interpreted as measuring dealer's views on the manufacturer's policy domain. The variables loading the highest on the second factor refer to situations critical to the operation of a dealer- ship. The highest loading variable, dealer density, asks whether a respondent perceives there are too many dealers in most market areas to allow each to earn a satisfactory pro- fit. The other variable associated with the second factor, additional dealerships, dealer dependence, and advantageous programs require the respondent to make a judgement on mat- .ters directly affecting his ability to successfully operate his dealership. Since the prior discussion of perception defined the construct as a subjective interpretation of a situation, the second factor can be described as measuring dealership perceptions. The interpretation of the third factor as measuring the dealer's profit goal is based on an analysis of the highest loading variables. The variable, profit reinvestment, states a dealer should prefer to reinvest the majority of his profits back into the dealership. Similarly, selling effort and profitability trade—off refer to statements identifying what a dealer should do in situations affecting his 117 profitability. Since the goal construct was operationally defined as a value premise underlying task elements or op— erating ends, the third factor can be labeled profit goal. The perception variable, manufacturer support, loaded the highest on the fourth factor. Since the variable asks whether the manufacturer provides dealers with adequate support to enable them to be competitive with other brands, the fourth factor was identified as manufacturer support. The four factors were treated as predictor variables and regressedcnlthe attitude criterion as a test of the hypothesized association between the latent variables and dealer attitude toward Ford. As shown in Table 10, the regression supports the hypothesis since the multiple corre— lation coefficient is a significant .64. Furthermore, each of the domain, perception, and goal factors were signifi- cantly related to attitude. The results support the hypothesis that the variables Stern and Heskett proposed as causing conflict are systema— tically related to an affective measure of expressed conflict. The data analysis supports this conclusion since twelve of the fifteen latent variables loaded highly on the factors and all the factors significantly contributed to the associa— tion with attitude. Hypothesis II According to Fishbein's attitude model, the various forces which influence attitude work through either cognitive 118 Table 10 Regression Equation: Attitude (Criterion) Latent Factors (Predictors) Latent Regression Significance Factors Coefficient t value Level Policy Domain 0.41 5.76 .001 Dealership Perception 0.45 6.33 .001 Profit Goal 0.19 2.73 .007 Support Perception 0.34 4.71 .001 Intercept 4.52 Multiple R = .64, (32:.41, significant beyond .001 level) beliefs about the object or affective evaluations of object attributes. Since the preceding results indicate the level of affective conflict a channel member holds toward another is related to the member's goals, domains, and perceptions, the question arises as to whether these latent forces influ— ence attitudes by affecting beliefs or evaluations. Hypoth— esis II posits the latent variables are systematically related to the belief component while Hypothesis III states the latent variables are associated with the evaluation component. In order to test the second hypothesis ten separate regressions were performed. Each regression, in turn, utilized one of the ten belief variables as the criterion upon which the fifteen latent variables were regressed. This procedure yielded an analysis of the degree of association 119 between each belief and the latent variables. Table 11 summarizes the results of the analysis. The results support Hypothesis II since nine of ten beliefs show a moderate association, significant beyond .05 level, with a channel member's goals, domains, and percep- tions. The strength of the associations ranged from an R of .56 for the coercive belief, unfair vehicle allocation, to an R of .41 for the non-coercive belief, excellent product warranty. Only the coercive belief, slow vehicle delivery, was not significantly associated with the latent variables. However, the strength and significance of the associations for all the non—coercive beliefs and for four of the coer— cive beliefs support the posited relationship. An examination of the ten regression equations identi— fies the specific latent variables significantly associated with each belief. Table 12 indicates on an individual basis only a few of the fifteen latent variables are so related. The number of latent variables significantly contributing to the association with a belief ranges from three for the non— coercive belief, threats of termination, to only one for the non-coercive beliefs, effective inventory rebate plan, excellent mechanic training, and excellent product warrnaty, plus the coercive belief, slow vehicle delivery. However, as previously discussed the fifteen latent variables are reducible to four meaningful latent factors. Further evidence regarding Hypothesis II was obtained by performing ten additional regressions using the four latent 120 Table 11 Summary of Multiple Correlation Coefficients: Belief (Criterion) and Latent Variables (l5 Predictors) Signi— 2 ficance Beliefs R R Level Non—Coercive 1. Effective district sales reps .50 .25 .001 2. Effective inventory rebate plan .47 .22 .004 3. Worthwhile incentive programs .55 .31 .000 4. Excellent mechanic training .48 .23 .001 5. Excellent product warranty .41 .17 .049 Coercive 1. Slow warranty payment .45 .20 .009 2. Unfair vehicle allocation .56 .31 .000 3. Bureaucratic red tape .49 .24 .001 4. Slow vehicle delivery .37 .14 .148 5. Threats of termination .50 .25 .001 Note: Each regression, in turn, utilized a different belief as the criterion upon which the fifteen latent variable predictors were regressed. This procedure yielded a multiple R for each belief, indicating its degree of association with the predictors. 121 .womoquopma an oopmowocfl one Ho>oH oo. one ASH.0 60.m 00.0 w0.0 no.0 A0m.00 NH.0- 00.0 No.0! No.0 oo.0 00.0I o0.0I oo.o oo.0I 00.01 No.0 seaweeds posooem Amm.v 00.H NH.o NH.o HH.0| 0m.0 40.0 Awfi.00 00.0 0H.01 00.0 No.0! oo.0I oo.o ho.o wo.o vo.o mcficfidpe oficmnooz Aam.v 00.0 MH.0 No.0 oH.0 mo.0I o0.OI 80.0 SH.0 wo.0I No.01 AHN.00 m0.0 00.0- Amm.00 00.0 90.0 mEmAwOHa o>Hpcoqu Amm.v ob.m oo.OI HH.o 0H.o AHN.00 00.01 00.0 00.0! No.0 No.0 va.o oo.OI 0H.OI NH.OI 0H.o oH.o opwnom zsopco>cH mwoflaom o>Hosoooqoz p0 pcdoflwflcwflm mucoHOwaooo ”ouoz Aom.v oposvmnm wo.H pmoosoch AwH.00 >pwmqmm season 00.0: mmflsmpoasom Homoepfloom 0H.0 mocoocomom Reason 0H.0I mawstHm msoowwpam>o¢ vH.0| mucothfisvom asepco>cH 00.0 amouooeae aphaaoopaeona no.0 cmemom H00 No.0: oomflosoxm Hoquoo H0.0| psoEpmm>cfimm pfiwoum H0.0| pnomasm thDpOdwsqnz w0.0 ossmmopm oommEpoHHom o0.0| mofiowaom chofipoz 0H.0 oocmoflso mmocflmsm ASN.00 oomflosoxm :oHpoHomHQ NH.0I epochs woafiaom mmom moanmflhw> moawm pumped Annoooaeoaa 0H0 nofiooann> pooped one Acofinopflpov mwoflfiom o>Hosooocoz ”mcoflpmsvm scammohwom oa 00. one A00.0 00.0 0H.0 00.0: 00.0 A00.0uv Aem.00 0H.0 0H.0: A00.0:0 0H.0 50.0I No.0 HH.0I 0H.0I 0H.0I No.0! mesoHcB :oHp0:HEsoB 0H. 00.0 00.0 HH.0 00.0I 00.0! 00.0 v0.0 00.0! 00.0I wo.0l 00.0- H0.0 0H.0| ASH.0-0 H0.0- 80.0- Npo>HHoQ ofioaeo> Aem.v H0.0 00.01 00.0 No.0I vfl.0l 0H.0 HH.0! 00.0 00.0 mN.0I 00.0! 00.01 HH.0I 00.0-. 0H.0I 00.0 @009 com AH0.0 00.0 H0.0I H0.0 AOH.o|0 hH.0I 00.0 00.0 00.0I oH.0I oo.0l H0.0 A00.0IO 0H.0I H0.0I H0.0 00.0 eoaonooafia oHoHeo> p0 pcoofiwflcwflm mpqofloflwwooo ”opoz A00.0 oswswmlm 00.0 poochoqu 00.01 mpflmcom season 00.0- wagonaofiooe Hoooaoaoo< 00.01 mosooaogoo season HN.0I medprHQ msoo00p20>0¢ 0H.0 mecothflswom 0s0pco>aH 0H.0- heosoeoae spafiaoooaeona ASN.00 004000 one 0H.0I oomfloeoxm Hospcoo 00.0: pcosumo>qwom pagOpm 00.0: psoamsm Roadwowwscos 00.0: ossmwomm oomoEnowpom 00.0: wofloeaom Hocofipaz Amm.00 oocoofizo mmoaflwsm 0H.0| oomflosoxm cofiposomfio 00.0 snowmm wcflaaom pamEuom moHQ0HH0> seaweeds psopaq mummflfimm ®>HOH®OU Acoflpopfipov mwofiaom o>flopooo Anaoboaeoaa 0H0 nofioeane> bemoan 000 mNH mandB ”mcofip0som :onmoswom 123 factors as predictors and each of the ten beliefs, in turn, as the criterion. The regression equations are shown in Table 13. All of the regressions are significant beyond the .05 level indicating the latent factors are associated with each belief. Furthermore, for the noncoercive belief, worthwhile incentive programs, and the coercive belief, unfair vehicle allocation, all of the latent factors significantly contri- bute to the association. For the balance of beliefs, two of the latent factors significantly contribute to the associa— tion. In addition, Table 13 indicates the extent each latent factor is related to beliefs held toward another channel member. The factor, dealership perception, significantly contributes to eight of the ten associations with the belief criterions. The remaining factors, profit goal, policy domain, and support perception significantly contribute, respectively, to seven, five, and four of the ten associa— tions. The overall analysis of the data, both in terms of the strength and significance of the associations, warrants the acceptance of the second hypothesis. Hypothesis III In contrast to the second hypothesis, Hypothesis III posits latent variables affect a channel member's degree of affective conflict by influencing the member's evaluation 124 Table 13A Regression Equation: Noncoercive Beliefs (Criterion) and Latent Factors (4 Predictors) Noncoercive Beliefs Latent Sales Reps. Inventory Rebate Incentive Programs Factors Coef. t—value Signi. Coef. t-value Signi. Coef. t-value Signi. Policy Domain 0.39 2.98 .003 0.14 1.00 .319 0.29 2.26 .025 Dealership Perception 0.55 4.27 .000 0.45 3.30 .001 0.52 4.05 .000 Profit Goal 0.12 0.89 .375 0.31 2.28 .024 0.32 2.45 .016 Support Perception 0.07 0.54 .592 0.25 1.84 .067 0.42 3.24 .001 Intercept 4.02 5.23 4.34 Rquuare .17 2.65 .000 .12 2.26 .001 .21 3.08 .000 Noncoercive Beliefs Latent Mechanic Training Product Warranty, Factors Coef. t—value Signi. Coef. t-value Signi. Policy Domain 0.33 2.47 .015 0.046 0.39 .695 Dealership Perception 0.56 4.12 .000 0.401 3.51 .001 Profit Goal 0.37 2.76 .007 0.317 2.77 .006 Support Perception -0.03 -0.19 .852 -0.02 —0.20 .841 Intercept 3.93 4.47 Rquuare .18 2.77 .000 .12 2.25 .001 Regression Equation: 125 Table 138 and Latent Factors (4 Predictors) Coercive Beliefs Coercive Beliefs (Criterion) Latent Warranty Payment Vehicle Allocation Red Tape Factors Coef. t-value Signi. Coef. t—value Signi. Coef. t-value Signi. Policy Domain -0.37 —2.60 .010 -0.65 -4.74 .000 -0.28 -l.80 .073 Dealership Perception —0.37 -2.61 .010 -0.41 —3.03 .003 -0.44 -2.90 .004 Profit Goal -0.11 —0.74 .459 —0.33 -2.41 .017 —0.47 -3.10 .002 Support Perception -0.03 -0.20 .845 —0.50 -3.68 .000 —0.25 -l.63 .105 Intercept 4.32 4.86 2.71 R—Square .09 1.88 .009 .26 3.57 .000 .14 2.45 .000 Coercive Beliefs Latent Vehicle Delivery Termination Threats Factors Coef. t-value Signi Coef. t-value Signi. Policy Domain -0.l7 -l.40 .164 -0.46 -3.47 .001 Dealership Perception 0.01 0.09 .930 -0.04 -0.28 .777 Profit Goal ~0.24 —1.98 .050 —0.07 -O.53 .596 Support Perception —0.34 -2.72 .001 -0.33 -2.52 .013 Intercept 5.65 5.17 Rquuare .08 1.82 .012 .12 2.16 .001 126 of manufacturer attributes. The hypothesis was tested by performing ten regressions where the fifteen latent variables were predictors and each of the ten evaluations were, in turn, used as the criterion. Table 14 also shows the results of this procedure when the four latent factors were used as pre- dictors. Table 14 Summary of Multiple Correlation Coefficients: Evaluation (Criterion) and Latent Variables (l5 Predictors), Latent Factors (4 Predictors) Latent variables Latent Factors Signi. Signi. R Level R Level Non-Coercive Effective district sales reps. .30 .626 .21 .181 Effective inventory rebate plan .35 .283 .14 .576 Worthwhile incentive programs .33 .414 .11 .792 Excellent mechanic training .32 .440 .18 .291 Excellent product warranty .38 .144 .15 .537 Coercive Slow warranty payment .45 .012 .19 .263 Unfair vehicle allocation .33 .402 .06 .971 Bureaucratic Red Tape .39 .092 .20 .185 Slow vehicle delivery .28 .769 .21 .154 Threats of termination .24 .940 .18 .333 The data analysis does not support the third hypothesis. Of the ten regressions nine lack any significant association between the latent variables and the evaluations. Only the evaluation of attribute, slow warranty payment, 127 was significantly associated with the latent variables. For this evaluation only one of the four latent factors, dealer— ship perception, contributed significantly to the association. Since the dealer perception factor measures dealer's subjective interpretations of situations critical to the Operation of a dealership, its relationship to dealer's evaluation of the particular manufacturer's attribute is understandable. The association can be explained by recognizing the way a dealer evaluates slow warranty payments is likely to be related to the perception of various matters affecting the Operation of his dealership. However, the general lack of any systematic relationships between latent variables and evalutions leads to the rejection of the third hypothesis. For the Channel system investigated, a dealer's goals, domains, and perceptions are unrelated to evaluations of manufacturer attributes. Hypothesis IV Since Hypothesis II positing an association between the latent variables and beliefs was supported, the question arises as to whether these beliefs, in turn, are systematically re— lated to a channel member's degree of affective conflict. The fourth hypothesis proposes beliefs are associated with dealer attitudes toward the manufacturer. Because both coercive and noncoercive beliefs were in- vestigated the hypothesis can be tested utilizing each type of belief separately, in addition to an overall test of both belief 128 categories. Consequently, three regression equations were calculated each using the same attitude criterion but with different belief predictors. Section one of Table 15 summar- izes the results of the procedure. All three regression equations contain significant asso- ciations, thus supporting the fourth hypothesis. While each of the two separate regressions utilizing the five beliefs from each category as predictors shows a moderate relationship with attitude, a stronger association was found when both categories of beliefs were used in a single regression. The single regression resulted in a multiple correlation coeffi- cient of .63 indicating the belief component of attitude is significantly associated with channel member's level of affec- tive conflict. Additional insight into the nature of the association can be gained by examining the specific beliefs that contribute to the relationship with attitude. Table 16 contains the three regression equations under discussion. In the separate regression of the noncoercive beliefs, four contribute signi— ficantly to the association while in the separate coercive belief regression two are significant. Perhaps more important, however, are the significant beliefs in the single combined regression. Both coercive and noncoercive beliefs are signi— ficant indicating the overall association holds equally for both belief categories. While the two noncoercive beliefs, effective inventory rebate plan and worthwhile incentive pro— grams, are significant the two coercive beliefs, unfair vehicle 129 Table 15 Summary of Multiple Correlation Coefficients: Attitude (Criterion) with Beliefs, Evaluations as Predictors Section 1: Beliefs as Predictors Two Separate Regressions Non—coercive Beliefs (5 Predictors) Coercive Beliefs (5 Predictors) One Regression Non-coercive and coercive Beliefs (10 Predictors) Section 2: Evaluations as Predictors Two Separate Regressions Non—coercive Evaluations (5 Predictors) Coercive Evaluations (5 Predictors) One Regression Non—coercive and coercive Evaluations (10 Predictors) .54 .51 .63 .24 .25 Significance R2 Level .29 .000 .26 .000 .40 .000 .02 .794 .06 .152 .06 .546 000. 00.0 0v. 00.0 00H. 00.HI 00.0: 000. 00.0 00.0 moo. ma.m| NH.0I 000. Nh.ml 0H.0I ~00. vH.0I H0.0I 00 Q0 .1 000. 0o.H 00.0 00v. 00.0 00.0 000. 00.0 0H.o mvo. v0.0 00.0 000. hm.a 00.0 Ho>oa 05Hm>lu ucoHOmemoo wocmoflmwcmflm coflmmoummm 000. 050. >05. 000. 000. Nam. 000. HNO. ohm. 000. 000. 000. Ho>oH mocmoHMAcmwm ha.0 00.HI 00.0 om.vl mo.0l oo.0I 00.0 00.0 00.H 00.N H0.0 0o.N MMflMHM om. oH.o o0.0I No.0 0H.0I 0H.0I 00.0I 0N. N0.H va.o 00.0 NH.0 vH.0 0H.0 ucoflowmmoou cowmmmummm coammoumom oco macammoummm oumummom 039 oH mafldb Anaopoaooaa 040 .neoaaom 000 Aeoaaobaaov oesoappa mumsomnm umooumucH COHDMCHEHmp mo mumwuna .m >Mo>fiaoo oHOH£o> 30am .v 0069 000 OHUMHODMoHsm .0 coeuoooaam oHOH£o> Macon: .N 0:08080 >0cmuum3 30am .H o>wouooo mumsomLm ummououcH mucmunmz uosooum ucoaawoxm .m onwcwmuu caconcofi ucoaamuxm .w maoumoum o>HucoocH oaflzznvuoz .0 scam outflow muouco>cfl o>Huoowmm .m .onu moamm DOHHumHU o>Huommwm .a o>Houooocoz Hmcoflpwsom cOHmmohwom moans 131 allocation and bureaucratic red tape, also contribute signifi— cantly to the association. Because of the strength and significance of the belief and attitude association the fourth hypothesis can be accepted. However, as Fishbein conceptualized the attitude construct, attitudes can also be attributed to respondents'evaluations of object attributes. Examination of the following hypothesis tests whether channel member's evaluation of manufacturer characteristics is systematically related to affective conflict. Hypothesis V Section two of Table 15 shows the results of three regres— sion equations relating to the fifth hypothesis. Regardless of whether the coercive and noncoercive evaluation predictors are used separately or combined in a single regression equation, Hypothesis V remains unsupported. None of the multiple corre— lation coefficients is significantly different from zero. Such results indicate the degree of affective conflict dealers hold toward their manufacturer is independent of dealer evaluation of manufacturer characteristics. CHAPTER VI SUMMARY, CONCLUSIONS, AND IMPLICATIONS The study proposed a theoretical structure of the con— flict process with important managerial implications for channel management. An initial exploratory test of the model was undertaken and evidence compatible with its structure was developed. The study was conducted in the franchised system for the distribution of new automobiles. Overview In the channel system examined, dealers possessed a wide range of affective conflict toward the manufacturer. While some dealers held a positive attitude, others held a very negative attitude toward the manufacturer. Furthermore, such attitudes were associated with dealer satisfaction with the channel relationship and the frequency of disagreements with the manufacturer. However, since these attitudes were only weakly asso— ciated with dealer profitability and were independent of sales volume and tenure with the manufacturer, the explana- tion as to why only certain dealers were in a high state of conflict is unclear. Furthermore, the inability of these descriptive factors to aid in understanding conflict leaves the manufacturer without guidance in developing conflict reduction strategies. 132 133 The dissertation's integrative model posited an explanation of the conflict process which would be useful to management in selecting strategies to reduce conflict. The model integrated Fishbein's two component conceptuali— zation of the attitude construct with Stern and Heskett's proposal that goals, domains, and perceptions are latent conflict variables. Specifically, the model proposed the latent variables influence the affective conflict stage either through Fishbein's cognitive belief or affective evaluation component. Furthermore, it was observed that conflict reduction strategies tend to work either through changing the beliefs or values antagonists possess. The implications of this point for channel management will be developed later in this chapter. Since the relationship of the integrative model to conflict reduction strategies is a major managerial impli— cation of the dissertation, the topic will also be discussed as a direction for future research. 0f more immediate concern is the need to analyze whether the posited relation— ships exist among the latent variables, beliefs, evaluations, and affective conflict. Five hypotheses were developed in order to test the relationships contained in the integrative model. The follow- ing section summarizes the purpose of the hypotheses and the empirical results. 134 Summary of Results Hypothesis I The first hypothesis stated the latent conflict variables are associated with the degree of affective conflict possessed by members. Measures for the latent conflict variables were obtained by having dealers rate their degree of agreement or disagreement to fifteen statements reflecting important goal, domain, and perception issues. For affective conflict, a measure was derived from dealer responses to a semantic differ— ential instrument using the manufacturer as the attitude object. The hypothesis was tested through the use of regression analysis. A regression equation utilizing the attitude mea— sure as the criterion and the fifteen latent variables as predictors indicated a significant association. The multiple correlation coefficient was .71. However, thirteen of the latent variables failed to contribute significantly to the relationship. Further analysis of the fifteen latent variables identi- fied four underlying factors. Twelve of the latent variables had loadings in excess of .5 on the factors. When the four factors were used as the predictors the multiple correlation coefficient was .64. Furthermore, each of the goal, domain, and perception factors contributed significantly to the associa— tion with the attitude criterion. Since most of the fifteen latent variables loaded highly on the factors and all the factors contributed to the 135 association with attitude, the first hypothesis was supported. Hypothesis II An association between the latent conflict variables and the belief component of attitude was stated in this hypothe— sis. Since beliefs and evaluations are the two components of Fishbein's attitude model, the second hypothesis posits the latent conflict variables influence affective conflict through the belief component. The third hypothesis states the alterna— tive scenareo where the latent forces influence affective con— flict through the evaluation component. Since the data analysis for Hypothesis I indicated the latent variables are, in fact, associated with the intensity of affective conflict, the current theoretical question is whether-this association is due to the belief component. Al- though the study was not experimental and explicit causal inferences cannot be made, supportive evidence for such a linkage was derived. Specifically, multiple regression analysis was conducted utilizing beliefs as the criterion and the latent conflict variables as predictors. The posited linkage would be support- ed by the data if the latent conflict variables are associated with dealer beliefs. However, if no relationship is found, then beliefs would be independent of the latent variables and fail to play a direct role in linking the latent to the affec- tive stage of conflict. The results of the multiple regression support the second 136 hypothesis since nine of the ten beliefs showed a significant association with the latent variables. The strength of the associations ranged from a multiple correlation coefficient of .56 to .41, indicating a moderate association. However, only a few of the fifteen latent variables significantly contributed to the association with the beliefs. Since the fifteen latent causal variables are reducible to four meaningful latent factors, the analysis was continued using the latent factors as the predictor variables. Again, the second hypothesis was supported since all the regressions were significant beyond the .05 level. In addition, for two of the beliefs all of the latent factors significantly contri— buted to the association. For the balance of beliefs, two of the latent factors contributed to the association. Hypothesis III In contrast to the second hypothesis, this hypothesis states the latent conflict variables are associated with the evaluation component of attitude. The component refers to the desirability attributes of the manufacturer have for the dealers. Dealers may differ in the value attributed to manu- facturer attributes and such differences may be accounted for by differences in dealer goals, domains, and perceptions. The evaluation component was measured by having dealers rate noncoercive and coercive attributes of manufacturers in terms of how beneficial and harmful they were perceived, re— spectively. In all, evaluation measures were obtained on ten 137 attributes. To test the hypothesis ten separate regressions were performed using each of the ten evaluations, sequentially, as the criterion. The fifteen latent variables were used as the predictors in all the regressions. The results of the regression analysis did not support the third hypothesis. Of the ten regressions nine lacked any significant association between the evaluations and the latent variables. Apparently, for the channel system investigated, the goals, domains, and perceptions held by a dealer were un— related to his evaluations of manufacturer attributes. Hypothesis IV The results of the second hypothesis indicate the beliefs about the manufacturer held by dealers are associated with the latent variables. Whether the beliefs are, in turn, associated with dealer attitudes toward the manufacturer is the subject of Hypothesis IV. The fourth hypothesis states the belief component of attitude is systematically related to a channel member's de- gree of affective conflict. The degree of affective conflict was measured using a semantic differential scale with bipolar descriptive adjectives. The dealers were instructed to indi— cate their feelings toward the manufacturer using the scale. The hypothesis was tested through regression analysis. Three equations were calculated using attitude as the criterion but with noncoercive, coercive, and noncoercive and coercive 138 beliefs together as the respective predictors. All three re— gression equations contained significant associations, thus supporting the fourth hypothesis. Although the data indicates an association between beliefs and attitudes, attitudes can also be based upon respondents' evaluations of object attributes. This is the subject of the next hypothesis. Hypothesis V The fifth hypothesis states the evaluation component of attitude is associated with channel member's level of affective conflict. The hypothesis will test whether the value dealers attribute to characteristics of the manufacturer is systema- tically related to dealer attitudes toward the manufacturer. To test the hypothesis regression equations were calcu- lated using attitude as the criterion and evaluations as the predictors. The data analysis failed to support the hypothe- sis. The absence of a relationship between attitude and evaluation suggests the values dealers assign to manufacturer attributes are independent of their attitudes toward the manu— facturer. Study Conclusions Findings The study was designed to accomplish three research objectives. First, an attempt was made to analyze evidence for a theoretical link between hypothesized latent conflict 139 variables and a later stage of conflict. Second, the nature of the mediating mechanism between the latent and affective stages of channel conflict was investigated. Third, the man— agerial implications of establishing criteria based on the mediating mechanism for evaluating conflict reduction strate— gies was discussed. In light of these objectives and the statistical tests of the hypotheses, several study conclusions can be made. However, in View of the limitations discussed in Chapter IV generalizations are necessarily constrained. The following conclusions will be discuSsed. l. The intensity of affective channel conflict was shown to be related to the latent con— flict variables possessed by channel members. 2. Supportive evidence for the mediating role ' beliefs play in linking the latent and affective stages of conflict was found. 3. For the channel examined, the data indicated the evaluation component of attitude did not serve as a mediator between the latent and affective conflict stages. Discussion The study has provided evidence for the theoretical link between hypothesized latent conflict variables and a later stage of conflict. The variables Stern and Heskett proposed as underlying conflict were found to be related to the affec- tive conflict stage. By empirically demonstrating the rela— tionship between variables posited to cause conflict and a measure of affect toward another channel participant, the study 140 points to the need to consider the covert conflict processes in developing strategies to manage its overt manifestations. Since the goals, domains, and perceptions members pos— sess are related to the intensity of conflict held toward a participant at another channel level, these factors should be explicitly considered in the conflict management process. For the channel studied, the policies developed by the manu— facturer should reflect the differences in dealer goals, domains, and perceptions since these variables are related to dealer attitudes. However, without an understanding of the mechanisms 11 king the latent and affective conflict stages, prescriptions for policy development cannot be made with confidence. Because the attitude construct is complex the manner in which the latent variables influence affect is not clear. Utilizing Fishbein's two component model of attitude, the study investigated the mediating process between the con- flict stages. The results of hypotheses testing associations between the latent variables and components of Fishbein's model revealed a linkage. For the channel investigated, the latent variables were associated with the cognitive belief component of attitude. Although demonstrating the existence of an association between the latent variables and beliefs does not justify the conclusion that latent variables influence the beliefs, it does recognize a relationship between the phenomena. Further- more, the association is compatible with the structure of the - 141 dissertation's integrative model in the sense that indepen— dence among the variables would constitute evidence contra- dictory to the model. Although not sufficient to validate the model, the association would be necessary for it to be valid. Additional evidence compatible with the model is the finding that beliefs, in turn, are associated with attitudes. Thus, the study provides limited supportive evidence of an exploratory nature that the latent variables may influence attitudes toward another Channel member by affecting the belief component. Consistent with this evidence is the finding that latent variables are independent of the evaluation component which, in turn, is independent of attitude. The lack of any associa— tion among these factors, in conjunction with the previous evidence, is compatible with a possible mediating role for the belief component. In View of the empirical evidence, the implications of the model for channel management will be examined next. Implications for Channel Management This section illustrates the potential usefulness of the integrative model in establishing criteria for selecting conflict reduction strategies. First, the various types of conflict reduction techniques available to management will be reviewed. Then the prescriptive uses of the model for strategy selection will be discussed. 142 Conflict Resolution Strategies In stressing the managerial importance of conflict, Stern and El—Ansary state,1 A central task in interorganization manage— ment is to seek ways to manage conflict. In other words, it is essential that ways be found to keep conflict from becoming dysfunctional and to harness the energies in conflict situations to produce innovative resolutions. Conflict theorists have suggested a variety of techni— ques useful to management in resolving conflict. Mack and Snyder identify a long list of potential modes of conflict resolution applicable in a variety of cor-texts.2 They note arbitration, mediation — conciliation, nego- tiation, inquiry, legislation, judicial settlement, informal consensus, the market, force, authoritative command, and voting.are all useful strategies. Boulding considers the principal methods of procedural conflict resolution to be reconciliation, compromise, and award.3 In terms of distribution channels, Stern and Gorman isolate four specific resolution procedures.4 They consider bargaining, implicit mediation, cooperation, and self— encapsulation as a means to restore equilibrium to a disturbed channel system. Assael, in his nine industry study of distributive conflicts, concludes political and self- resolution are the primary means used to resolve conflict.5 Political resolutions refer to all activities directed to power sources outside the Channel system, while self-resolution 143 refers to the many activities internal to the Channel system. However, the prescrnmjwmevalue of offering management a lengthy descriptive list of conflict reduction techniques is questionable. In order to be useful, prescriptions must provide management with criteria to evaluate the appropriate— ness of each strategy to specific conflict situations. The growing body of conceptual and empirical research in the channel conflict area make an initial specification of such criteria possible. As noted in the literature review, the conflict construct has been conceptualized as consisting of an observable overt phase and a covert phase, a phase which Rosenberg describes as conflict's ”invisible period.” However, research on the latent phase has specified variables that create the poten— tial for conflict during this essentially unobservable phase. The integrative model developed in Chapter III posits a mecha— nism whereby these latent variables result in overt conflict. The mechanism links the flare—up phase management observes with its underlying cause. The integrative model suggests the strategies to reduce expressed conflict would be most effective if they operate through the same mechanism that initially generated the overt phase. To do otherwise places management in jeopardy of applying remedies that merely ad— dress conflict symptoms rather than causes. This could result in the chronic recurrence of conflict flare—ups. For as Coser observes, conflict processes, ”...follow a law of 144 social inertia insofar as they continue to operate if no explicit provision for stopping their course is made by the participants."5 To illustrate the managerial implications of selecting strategies using criteria based on the mediating mechanism, several conflict resolution techniques reviewed by Stern and El-Ansary will be discussed. The authors suggest manage- ment could employ either a bargaining, interpenetration, boundary or superorganizational strategy to resolve distri— butive conflicts.7 Based on the integrative model, it is proposed management should select a specific strategy depen— dent on whether the latent variables influence the belief or evaluation component. Criteria for Strategy Selection Bargaining strategies refer to various negotiation tech- niques that attempt to induce parties to accept a compromiSe agreement. Since compromise depends on each party giving up something in order to partially obtain some of its objectives, the strategy will work best when the cognitive belief compo— nent rather than the affective evaluation component is the active conflict agent. In other words, when antagonisms are due to channel participants holding differing beliefs rather than divergent values a bargaining strategy will be most ef~ fective. In such situations conflict reduction occurs as the different belief sets of the parties converge, for as 145 Stern and El—Ansary note:8 The willingness to negotiate a compromise depends, of course, on correct assessment of the conflict situation; such assessment and the following accommodation are possible only if each party is aware of the relative strength of others. On the other hand, bargaining techniques would be inef— fective in situations where conflict is due to channel members holding incongruent ideological values. The authors express this View by observing that ”One major limitation on the scope of negotiation in channel relationships is the difficulty of settling ideological differences through bar— gaining."9 Bargaining is ineffective because the basis for compromise, a shared system of values, is absent and addi— tional time spent in negotiation merely serves to clarify the incOnsistencies in each party's position. Interpenetration refers to a different category of strategies that depend on a variety of direct methods for increasing the number of meaningful, constructive interactions between parties as a means to reduce conflict. Cooptation, an important interpenetration strategy, is the technique of absorbing new elements into the policy determining structure of an organization as a means to lessen antagonisms. The technique is employed in many channel systems that have es— tablished dealer and distributor advisory councils. Typi- cally, a manufacturer in a state of conflict with other channel members employs cooptation in attempting to reduce 146 tensions through establishing interorganizational councils that share decision making responsibility. However, cooptation would be most effective in situations where the evaluation component is the active conflict agent. When latent variables result in overt conflict by influencing member's values, this interpenetration strategy should be selected. Thompson and McEwen state,10 By providing overlapping memberships, coopta— tion is an important social device for increasing the likelihood that organizations related to one another in complicated ways will in fact find compatible goals. The reason cooptation works through the evaluation compo— nent rather than beliefs is due to its tendency to converge the values or goals of the parties. Stern and El-Ansary empha— size this by observing, ”...cooptation of channel members encourages their ideological transformation, so that they subsequently tend to carry the ideology of the coopting unit into their other membership groups.11 Similar to cooptation, the criteria for selecting other interpenetration techniques should be based on the dominant mediating component. This view is reflected in Stern and El—Ansary's listing of specific aims for informational and propaganda conflict management techniques.12 To enhance knowledge and understanding. To shape attitudes among the personnel of another firm so as to influence its management to follow or not to follow a certain course of action. 147 Clearly, informational techniques aimed at enhancing understanding work through the cognitive belief component, while propaganda techniques focused on shaping attitudes have their impact on the affective evaluation component. Uncertainty absorption is an example of the informational approach. The technique refers to one organization drawing inferences from a body of evidence and communicating just the inferences to other members of the channel. Both Ford and General Motors utilize the technique by developing statistical averages on dealer operating data, including sales and pro- fits, which serve as performance norms for dealers. This leads to a uniform set of beliefs regarding operating expec— tations for all channel members. This is significant since it serves, "...to achieve at least a relatively uniform per— ception of the environment in which channel members operate. Once perceptual consensus is established, one can expect at least some reduction in conflict that was based on incongruent views."13 In an opposite manner, propaganda activities work through the evaluation component rather than beliefs. Stern and El-Ansary note, “What the channel propagandist (or educator) may be seeking is some sort of ideological 'conversion'."14 Such an approach has been implemented by manufacturers and wholesalers when attempting to alter retailer evaluations by urging them to adopt return-on-investment rather than gross profit performance measures. Often occurring over several years, such attempts focus on persuading retailers that one 148 measure is better, in an evaluative sense, rather than merely providing informationzuxnn:the measures. Similarly, other conflict reduction strategies can be shown to work primarily through either cognitive beliefs or affective evaluations. Consequently, the integrative model has a prescriptive dimension since it offers management strategy selection criteria. Through a new conceptualization of the conflict process, the model provides additional insight into the management of distributive conflicts. Implications for Future Research Both the theoretical and empirical contributions of the study suggest a new line of research in the analysis of dis— tribution channels. Future research should focus on estab— lishing the implied causal linkages of the model and the prescriptive value of selecting conflict reduction strategies based upon it. As noted in the Chapter IV discussion of the study's limitations, the survey type field research utilized does not control for the effects of all the variables affecting the phenomena under examination. The lack of control makes it difficult to isolate the contribution a few variables of interest have on a process. Experimental laboratory research, on the other hand, does provide the opportunity to control for the effects of extraneous variables so that conclusions can be drawn about the causal effects of the variables under study. 149 Future research might attempt to isolate the causal effects of the latent conflict variables on the belief and evaluation components of affective conflict. By using exper— mental techniques, researchers may be able to provide addi- tional evidence for the mediating role played by beliefs and evaluations in the channel conflict process. As this occurs, additional research will move beyond merely establishing associations among the variables in the integrative model and establish their causal connections. Another implication for future research resulting from the present study is the need to develop supportive evidence for selecting conflict reduction strategies based upon the mediating mechanism between the latent and later stages of conflict. The prescriptive value of linking strategy selec— tion to the components of affective conflict should be empir— ically determined. Since an objective of the present study was to illustrate the managerial implications of establishing criteria for strategy selection, it remains for future researchers to test the applicability of the integrative model in choosing con— flict reduction'techniques. The prior discussion of utiliz— ing bargaining and interpenetration strategies in particular channel circumstances could be examined in future studies. Future research could determine whether bargaining strategies are, in fact, more effective in conflict situations where the cognitive belief component is the active mediator in the conflict process. Similarly, it could be determined whether 150 interpenetration strategies are more effective when the affective evaluation component is the active mediator between the conflict stages. By continuing research along the lines suggested by this study, additional insight can be gained into the functioning of the conflict process. An improved understanding of con- flict in channels of distribution would be an important con— tribution to our knowledge of the marketing process. 151 CHAPTER VI FOOTNOTES 1Louis W. Stern and Adel I. El—Ansary, Marketing Channels (Englewood Cliffs, New Jersey: Prentice—Hall, Inc., 1977), p. 14. 2R. W. Mack and R. C. Snyder, "The Analysis of Social Conflict-Toward an Overview and Synthesis,” Journal of Conflict Resolution, 1:238 (June, 1957). 3K. E. Boulding, Conflict and Defense (New York, Harper and Row, Inc., 1963), p. 310. 4Louis W. Stern and R. H. Gorman, "Conflict in Distri- bution Channels: An Exploration,” in L. W. Stern (ed.), Distribution Channels: Behavioral Dimensions (Boston, Houghton Mifflin Co., 1969), pp. 163-170. 5Henry Assael, ”The Political Role of Trade Associa— tions in Distributive Conflict Resolutions,” Journal of Marketing, 32:21—28 (April, 1968). 6L. A. Coser, ”The Termination of Conflict,” Journal of Conflict Resolution, 5:347 (June, 1962). ' 7Stern and El—Ansary, pp. 292—309. 8Ibid., p. 295. 91bid., p. 296. 10James D. Thompson and William J. McEwen, ”Organiza— tional Goals and Environment,” in Amitai Etzioni (ed.), Complex Organizations: A Sociological Reader, 2nd. ed. (New York: Holt, Rinehart and Winston, Inc., 1969), p. 195. 11Stern and El—Ansary, p. 302. 12Ibid., p. 300. l31bid., p. 301. 14Ibid., p. 300. APPENDICES APPENDIX A Initial Cover Letter 152 Initial Cover Letter MICHIGAN STATE UNIVERSITY GRADUATE SCHOOL OF BL‘SINFSS ADMINISI RAMON I'AST LANSING ' \IICHIGAN ' 438.2% Dl-ZPARTHENT OF \IARK[-.TI.’\C AN!) TRANSPORTATION ADMINISTRAI |()\ June 20, l978 Dear Sir: As a Ph.D. candidate in the College of Business at Michigan State University, I would appreciate your assistance on my dissertation research project. Your deal- ership was selected according to a statistical sampling technique and your reply is important to the validity of my results. This project, which is endorsed by the Michigan Automobile Dealers Association, is concerned with dealer attitudes toward their car manufacturers. Although trade publications such as Automotive News have highlighted the problems dealers have with their manufacturers, little is known about the business practices which either strengthen or weaken relations between dealers and manufacturers. Your experience and judgement in these matters can contribute greatly to an improved understanding of the dealer—manufacturer relationship. The enclosed questionnaire for either GM, Ford, or Chrysler takes only a few min- utes to fill out since most of the questions can be answered by simply placing an X in the appropriate response. Your name does not appear on the questionnaire since your identity is not important to the study. Won‘t you please fill it out and return it to me? Your cooperation will be most appreciated. Sincerely, Endorsement, Daniel C. Bello Gilbert L. Haley Executive Vice President Michigan Automobile Dealers Association DCB/mw APPENDIX B Initial Questionnaire 153 Initial Questionnaire CONFIDENTIAL QUESTIONNAIRE MICHIGAN STATE UNIVERSITY Thank you for agreeing to participate in this study! 1. How many years have you been a Dealer for Ford ? 2. Place an X by any additional automotive brands you currently represent. Mercury or Lincoln Import DEALER OPINION SURVEY years. Other Domestic 3. Since Dealers sometime have additional businesses or represent several automotive brands. tell us what percentage of your PART I. 1. Ford 2. Ford 3. Ford 4. Ford 5. Ford 6. Ford 7. Ford 8. Ford 9. Ford 10. Ford PART II. time is spent operating your Ford Dealership. DOES EACH STATEMENT ACCURATELY DESCRIBE FORD ? Place an X to represent how strongly you agree or disagree with each description of Ford. If you feel a statement is true place an X in a space close to "Strongly Agree" How— ever, if you feel a statement is false place the X closer to “Strongly Diasgree. " SAMPLE: Ford has consistently high product quality. has effective district sales representatives. is unfair in vehicle allocation. has an effective year-end inventory rebate plan. is slow to pay on warranty work. has worthwhile Dealer incentive programs. uses threats of termination to get Dealer cooperation. has excellent quality mechanic training. is slow in delivery of ordered vehicles. has an excellent quality product warranty. creates unnecessary bureaucratic Red Tape. Below are five supposedly “good" things an automotive manufacturer could provi e its Dealers. beneficial _you feel each item would be to your Dealership if it were provided to you. w0uld be 0. only slightgbenefit. place an X near nModerately Beneficial“. cial" Give yOur opinion for all five items. of great benefit, place an X near 'Extremely Benefi SAMPLE: Effective showroom promotion kits. 1. Effective district sales representative. 2. Excellent quality product warranty. (J . Effective year—end inventory rebate plan. 4. Worthwhile Dealer incentive programs. 5. Excellent Quality mechanic training. Strongly Agree Strongly Agree Answer all ten statements. Strongly Disagree Strongly Disagree Strongly Agree Strongly Disagree Strongly Agree Strongly Disagree Strongly Agree Strongly Disagree Strongly Agree Strongly Disagree Strongly Agree Strongly Disagree Strongly Agree Strongly Disagree Strongly Agree Strongly Disagree StrOngly Agree Strongly Disagree Strongly Agree Strongly Disagree Moderately Beneficial Moderately Beneficial Moderately Beneficial Moderately Beneficial Moderately Beneficial Moderately Beneficial __ However, Tell us how If you feel an 1tem if you feel it would be Extremely Beneficial Extremely Beneficial Extremely Beneficial Extremely Beneficial Extremely Beneficial Extremely Beneficial 154 PART III. automotive manufacturer did them. ately Harmful”. your opinion for all five items. 1. Slow payment on warranty work. 2. Unfair vehicle allocation. 3. Unnecessary bureaucratic Red Tape. 4. Slow delivery of ordered vehicles. 5. Threats of termination to get Dealer cooperation. PART IV. SAMPLE: Ford's incentive programs get a Dealer to work harder than normal. . When sales are slow for a model that is difficult to sell, a Dealer should devote more time and effort to selling it. . In developing marketing policies, Ford has the right to determine what is in the best interest of the overall franchise system. useful in operating a Dealership profitably. . Ford adoption of nation-wide policies (as planned by Ford personnel and outside experts) is to be preferred to local flexibility of Dealers. In order to inSure that all Dealers perform adequately, Ford is justified in setting performance standards and insuring Dealers meet them. . Ford provides Dealer with adequate Support to enable Dealer to be competitive with other brands in his market area. A Dealer sh0uld prefer to reinvest the majority of his profits back into the Dealership. . Because of the need for uniformity throughout its franchise system. Ford is entitled to control some important Dealer operating decisions. 9. Ford designs cars exactly the way conSumers want them. 10. II. 12. 13. A Dealer should always be willing to satisfy customer demand even if doing so reduces Dealership profitability. In order to maintain production of all models at an economic level, Ford is justified in requiring Dealers to take slow selling as well as fast selling models. Ford has nearly always proved trustworthy in irtroducing policies and programs which are of mutual advantage to Dealer and Ford. Having a Ford franchise is well worth the accompanying restrictions on Dealer's independence in operating nis bu51ness. . Ford is entitled to determine the number of Dealerships necessary to satisfy consumer demand in a market area. The number of Ford Dealers in most market areas is few enough to allow each to earn a satisfactory profit. However. if you feel it would be of great harm, place an X near "Extremely Harmful". Now place an X to represent how hanmful these five supposedly "bad" items would be to your Dealership if an If you feel an item would be of only slight harm, place an X near "Moder- Give In this section, place an X to represent how strongly you agree or disagree with each statement. to consider the entire scale for each statement before placing an X. Answer all fifteen statements. . The guidance and advice Dealer receives from Ford is extremely Strongly Moderately Extremely Harmful Harmful Moderately Extremely Harmful Harmful Moderately Extremely Harmful Harmful Moderately Extremely Harmful Harmful Moderately Extremely Hannful Harmful sure Strongly Stongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree _ Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree Strongly Strongly Agree Disagree —___—____—_-———_—_. 155 PART V. 0n the scales below, please indicate how you feel about Ford. There are no right or wrong answers. only your own feelings are important. Place only one X per scale and answer all scales. IN MY OPINION, FORD IS foolish : : : : : : wise nice____: : : z 2 : awful pleasant : : : : : : unpleasant wrong : : : : : : right passive : : : : : : active adversary 2 : : : : : partner good : : : : : : bad kind : : : : : : cruel weak 2 : : : : : strong frustrating : : : : : : helpful rewarding : : : : : : punishing PART VI. Listed below are some questions that will be used for statistical purposes. 1. In general how much control would you say Ford No Complete has over the operation of your Dealership? Control : : : : : : Control 2. How satisfied are you in general with your Very Very relationship with Ford? Dissatisfied : : : : : : Satisfied Very Very 3. How frequently do you disagree with Ford? Infrequently : : : : : : Frequently 4. Place an X by your approximate Ford new 1 to 199 550 to 999 ' ' 7 car and truck unit volume in l9.7. 200 to 349 1.000 to 1,700 350 to 549 more than 1.700 5. Compared to all Michigan Ford Dealers, Greatly Greatly your l977 Dealership Profit Performance was Below Above Average : : : : : : Average Thank you 1 Please immediately place the questionnaire in the return envelope and mail it. Complete confidentiality of individual returns is assured. APPENDIX C Follow-up Cover Letter 156 Follow—up Cover Letter MICHIGAN STATE UNIVERSITY GRADUATE SCHOOL OF BUSINI‘SS ADAIINIS'IRAIION I:r\SI' LANSING ' MICHIGAN ' 4582i DEPAR'IHENT OF NIARKFI'ING AND TRANSPORTATION ADMINIS'I R.\ I'li )\ July 14, 1978 Dear Sir: A few weeks ago you received a questionnaire seeking your opinions about your car manufacturer, as part of my Ph.D. dissertation research. The questionnaire is completely confidential since dealer names do ggt_appear on it. The research is endorsed by the Michigan Automobile Dealers Association. Since your dealership was selected according to a statistical sampling tech- nique, it is important to include some information on your firm in order for the study to be valid. If you have not yet replied, I would appreciate your answer- ing the original long questionnaire. If that isn't possible, I have enclosed a shortened version (only eight questions) which I hope you will fill out and - return. Your cooperation will be most appreciated. Sincerely, Endorsement, Daniel C. Bello Gilbert L. Haley Executive Vice President . . Michigan Automobile Dealers Assoc1at1on DCB/Ifiv'i APPENDIX D Short Form Questionnaire 157 Short Form Questionnaire CONFIDENTIAL QUESTIONNAIRE MICHIGAN STATE UNIVERSITY DEALER OPINION SURVEY I. How many years have you been a Dealer for Ford ? years. 2. Place an X by any additional automotive brands y0u currently represent. Mercury or Lincoln ______Import ~_____Other Domestic 3. Since Dealers sometime have additional businesses or represgnt several automotive brands, tell us what percentage of your time is spent operating your Ford Dealership. PART II. Listed below are some questions that will be used for statistical purposes. Place an X to represent how you feel about Ford. Strongly Strongly SAMPLE: Ford has consistently high product Quality. Agree : : : : z : Disagree 1. In general how much control would you say Ford No Complete has over the operation of y0ur Deaiership? Control : : : : : : Control 2. How satisfied are y0u in general with your Very Very relationship with Ford? Dissatisfied : : : : : : Satisfied Very Very 3. How frequently do you disagree with Ford? Infrequently : : : : : : Frequently 4. Place an X by your approximate Ford new I to 199 550 to 999 i ' ' ..._ ' 7 car and truck unit vc1u e in l9.7. 200 to 349 1.000 to 1.700 350 to 549 more than 1,700 5. Compared to all Michigan Ford Dealers, Greatly Greatly your l977 Dealership Profit Per‘onnance was Below Above Average : t : : : : Average Thank you 1 Please immediately place tre Questionnaire in tte return envelope and mail it. Complete confidentiality of individual returns is aSSured. B I BLI OGRAPHY BIBLIOGRAPHY Alderson, Wroe. Dynamic Marketing Behavior. Homewood, Illinois: R. D. Irwin, Inc., 1965. Allport, Gordon W. "Attitudes.” Handbook of Social Psycho— logy, ed. C. Murchison. Worcester, Massachusetts: Clark University Press, 1935. Assael, Henry. ”Dealer Reactions to Factory Policies and Actions: A Case Study of The Automobile Dealer." Ph.D. Dissertation, Columbia University, 1965. Assael, Henry. "The Political Role of Trade Associations in Distributive Conflict Resolutions." Journal of Marketing Research BE (April, 1968), pp. 21-28. Bartles, Robert. History of Marketing Thought. 2d ed. Columbus: Grid, Inc., 1976, Beir, Frederick J., and Stern, Louis W. ”Power in the Channel of Distribution.” Distribution Channels: Behavioral Dimensions, ed. Louis W. Stern. Boston: Houghton Mifflin Co., 1969. ' Berg, Thomas L. ”Designing the Distribution System.” The Social Responsibilities of Marketing, ed. W. D. Stevens, pp. 481-490. Chicago: American Marketing Association, 1961. Biddle, Bruce J. ‘Roles, Goals, and Value Structures in Organizations.” New Perspectives In Organization Research, ed. W. W. Cooper, Chapter 10. New York: John Wiley and Son, Inc., 1964. Brown, James R. ”Toward Improved Measures of Distribution Channel Conflict.” Contemporary Marketing Thought, eds. B. Greenbery and D. Bellenger, pp. 385-389. Chicago: American Marketing Association, 1977. Boulding, Kenneth E. Conflict and Defense. New York: Harper and Row, 1963. Boulding, Kenneth E. ”Economics of Human Conflict.” The Nature of Human Conflict, ed. Elton B. McNeil. Englewood Cliffs: Prentice—Hall, Inc., 1965. Boulding, Kenneth E. ”Organizations and Conflict.” Journal of Conflict Resolution 1 (June, 1957), pp. 122—134. Bucklin, Louis P. A Theory of Distribution Channel Structure. Institute of Business and Economic Research. Berkeley, Calif.: University of California, 1966. 158 159 Carlson, Bjorn, and Kusoffsky, Bertil. ”Distributor Brands and Conflicts In Distributive Organizations.” Distribution Channels: Behavioral Dimensions, ed. Louis W. Stern. Boston: Houghton Mifflin Co., 1969. Coser, L.A. ”The Termination of Conflict.” Journal of Conflict Resolution 5 (June, 1962), pp. 347—353. Day, George S. ”Theories of Attitude Structure and Change.” Consumer Behavior: Theoretical Sources, eds., Scott Ward and Thomas S. Robertson. Englewood Cliffs: Prentice—Hall, Inc., 1973. El-Ansary, A. I., and Robicheaux, B.A. ”A General Model For Understanding Channel Member Behavior.” Journal of Retailing 52 (Winter, 1976), pp. 13-30. Fink, C. F. ”Some Conceptual Difficulties in the Theory of Social Conflict.” Journal of Conflict Resolution 12 (December, 1968), pp. 412—460. Fishbein, M., and Raven, B. H. ”The AB Scales: An Operational Definition of Belief and Attitude.” Readings in Attitude Theory and Measurement, ed. M. Fishbein. New Yorszohn Wiley and Sons, Inc., 1967. Fishbein, M. "A Behavioral Theory Approach to the Relations between Beliefs about an Object and the Attitude Toward the Object.” Readings in Attitude Theory and Measurement, ed. New York: Joba Wiley and Sons, Inc., 1967. Fishbein, M., and Ajzen, I. Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research. Reading, Mass.: Addison—Wesley Publishing Co., 1975. Fishbein, M. ”A Consideration of Beliefs, and Their Role in Attitude Measurement.” Readings in Attitude Theory and Measurement, ed. New York: John Wiley and Sons, Inc., 1967. Goldman, Ralph M. ”A Theory of Conflict Processes and Organ— izational Offices.” Journal of Conflict Resolution 12 (September, 1966), pp. 328—343. Hall, A. D. and Fagen, R. E. ”Definitions of a System.” General Systems, eds. L. Bertalanffy and A. Rapoport. Ann Arbor: University of Mihigan Press, 1956. Harrell, Gilbert D. ”Modeling Physician Prescribing Beha— vior: Attitudes, Normative Beliefs, Motivation to Comply, Confidence, Behavioral Intention and Behavior.” Ph.D. Disser— tation, The Pennsylvania State University, 1972. 160 Hostuick, K. T., and Kurtz, D.L. ”Alderson's Functionalism and the Development of Marketing Theory.” Journal of Business Research 1 (Fall, 1973), pp. 141—156. Jolson, Marvin A. ”How to Double or Triple Mail—Survey Response Rates.” Journal of Marketing 31 (October, 1977), pp. 78-81. Kelly, J. Steven, and Peters, J. Irwin. ”Vertical Conflict: A Comparative Analysis of Franchise and Distributors.” Contemporary Marketing Thought, eds. B. Greenbery and D. Bellenger, pp. 380-384. Chicago: American Marketing Association, 1977. Kerlinger, Fred N., and Pedhazur, Elazar J. Multiple Regres— sion in Behavioral Research. New York: Holt, Rinehart, and Winston, Inc., 1973. LeVine, R.A. ”Anthropology and the Study of Conflict: An Introduction.” Journal of Conflict Resolution 5 (June, 1962), pp. 3—15. Little, Robert W. ”The Marketing Channel: Who Should Lead this Extracorporate Organization.” Journal of Marketing 33 (January, 1970), pp. 31—38. Lusch, Robert F. ”Channel Conflict: Its Impact On Retailer Operating Performance.” Journal of Retailing 52 (Summer, 1976), pp. 3-12. Lusch, Robert F. ”An Empirical Investigation of Power, Con- flict and Satisfaction in a Franchisor-Franchisee Channel of Distribution." Ph.D. Dissertation, University of Wisconsin, 1975. Lusch, Robert F. ”Sources of Power: Their Impact on Intra— channel Conflict.” Journal of Marketing Research 13 (Novem- ber, 1976), pp. 382—390. Mack, R. W., and Snyder, R.C. ”The Analysis of Social Conflict — toward an Overview and Synthesis.” Journal of Conflict Resolution 1 (June, 1957), pp. 212—248. Mallen, Bruce. ”A Theory of Retailer—Supplier Conflict, Control, and Cooperation.” Journal of Retailing 36 (Summer, 1963), pp. 24-32. March, James G., and Simon, Herbert A. Organizations. New York: John Wiley and Sons, Inc., 1958. McCammon, Bert C. ”The Emergence and Growth of Contractually Integrated Channels in the American Economy.” Marketing and Economic Development, ed. Peter D. Bennett, pp. 496—515. Chicago: American Marketing Association, 1966. 161 McCammon, Bert C., and Little, Robert W. ”Marketing Channels: Analytical Systems and Approaches.” Science In Marketing, ed. George Schwartz, pp. 321—385, New York: John Wiley and Sons, 1965. McVey, Phillip. ”Are Channels of Distribution What the Textbook Say?.” Journal of Marketing gfi (January, 1960), pp. 61—65. Osgood, Charles E.; Suci, George J.; and Tannenbaum, Percy H. The Measurement of Meaning. Urbana, Illinois: University of Illinois Press, 1957. Palamountain, Joseph C. The Politics of Distribution. Cambridge, Mass.: Harvard University Press, 1955. Parsons, Talcott, and Smelser, Neil J. Economy and Society: A Study in the Integration of Economic and Social Theory. New York: The Free Press, 1956. Pearson, Michael M. ”The Conflict—Performance Assumption." Journal of Purchasing 9 (February, 1973), pp. 57—69. Perrow, Charles. ”The Analysis of Goals in Complex Organiza- tions.” American Sociological Review g6 (December, 1961), pp. 854-866. Pondy, Louis R. ”Organizational Conflict: Concepts and Models." Atiministrative Science Quarterly 1g (September, 1967), pp. 296-320. Raven, B., and Kruglanski, A. ”Conflict and Power.” The Structure of Conflict, ed. P. Swingle. New York: The Academic Press, 1970. Reve, Torger. "Interorganizational Relations in Distribution Channels.” Paper Presented at Eight Annual Albert Haring Symposium. Bloomington, Indiana: Indiana University, March, 1978. Rosenberg, Larry J.and Stern, Louis W. ”Conflict Measurement in the Distribution Channel.” Journal of Marketing Research. lg (November, 1971), pp. 437—442. Rosenberg, Larry J. ”An Empirical Examination of the Causes, Level, and Consequences of Conflict in a High Stake Distribu— tion Channel.” Ph.D. Dissertation, Ohio State University, 1969. Rosenberg, Larry J. ”A New Approach to Distribution Conflict Management.” Business Horizons 19 (October, 1974), pp. 67—74. Rosenberg, Larry J., and Stern, Louis W. "Toward the Analysis of Conflict in Distribution Channels: A Descriptive Model.” Journal of Marketing g3 (October, 1970), pp. 40-46. 162 Rosenbloom, Bert. ”Conflict and Channel Efficiency: Some Conceptual Models for the Decision Maker.” Journal of Marketing §§ (July, 1973), pp. 26—30. Schmidt, Stuart M. and Kochan, Thomas A. ”Conflict: Toward Conceptual Clarity.” Administrative Science Quarterly 11 (September, 1972), pp. 3594369. Scott, Jerome E. and Bennett, Peter D. "Cognitive Models of Attitude Structure: 'Value Importance' is Important." Com— bined Proceedings: 1971 Spring and Fall Conference. ed. Fred C. Allvine, pp. 346—350. Chicago: American Marketing Association, 1971. Simon, Herbert A. ”On the Concept of Organizational Goal.” Administrative Science Quarterlyg (June, 1964), pp. 1—22. Stern, Louis W. and Gorman, Ronald H. ”Conflict In Distri- bution Channels: An Exploration.” Distribution Channels: Behavioral Dimensions, ed. L. W. Stern. Boston: Houghton Mifflin Co., 1969. Stern, Louis W. and Heskett, J.L. ”Conflict Management in Interorganization Relations: A Conceptual Framework.” ' Distribution Channels: Behavioral Dimensions, ed. L. W. Stern. Boston: Houghton Mifflin Co., 1969. Stern, Louis W. and Brown, Jay W. ”Distribution Channels: A Social Systems Approach.” Distribution Channels: Behavioral Dimensions, ed. L. W. Stern. Boston: Houghton Mifflin Co., 1969. Stern, Louis W.; Sternthal B.; Craig, C.S. ”Managing Con- flict in Distribution Channels: A Laboratory Study.” Journal of Marketing Research l9 (May, 1973), pp. 169-179. “Stern, Louis W., and El—Ansary, Adel I. Marketing Channels. Englewood Cliffs: Prentice—Hall, Inc., 1977. Thompson, James D. Organizations In Action. New York: McGraw- Hill, Inc., 1967. Thompson, James D., and McEwen, William J. ”Organizational Goals and Environment.” Complex Organizations: A Sociological Reader. ed. Amitar Etzioni. New York: Holt, Rinehart, and Winston, Inc., 1969. Walker, Orville C. ”An Experimental Investigation of Con- flict and Power in Marketing Channels.” Ph.D. Dissertation, University of Wisconsin, 1970. Wilkie, William L., and Pessemier, Edgar A. "Issues In Mar— keting's Use of Multi—Attribute Attitude Models.” Journal of Marketing Research 19 (November, 1973), pp. 428—441. if?“ .§ tin-awn; § ’9 1.} {1.519.392 1‘ o} .1.“ I ..It. #4.. 5|: 1 . ‘ ‘IWMMWI“