J A ' '3 "'32“; - :3 -.-:. : .3 .13." $130.3 ‘*-""'"3'*" ' " 55.33:. ’3'" 33-3-33 3M?" 5.433 3‘L.¢'3.‘;/ ' '.' ' 'I "" 3.3} "it"; '39:" #4:“ VI}? fift’éfiféfflfl.‘ ‘ j. Y I)" \O '1' 'u- . n - » l . _. fi‘I I'f .L."I_1 '3 ”.317?! w; "f r fileI '/ ’3'??ch '3'?" I: my 3:3 :rr: ~:‘ '7 q ..5 lac-V. I ' ' "2;“. 1’. ‘3 | I . 33% 35’3‘3‘3 2'3} 3 - . ,‘E: '. 3‘16" . I. 6 .._\ 69‘!” .‘ 5;:‘- 5:, . {‘3‘ h" :..m.mm~m' ' . - O ‘ ‘0 , .- .. ..e; . _ , .... 3 . 2-. -- ' ‘7 ~55 w} 3 iii: -'Q. a: . .,._ “ " "1 r ' I F p 0 I : *‘t" (a. '~‘l NF... . - v: . fl. . o ’ ‘ u 'v k‘ ‘ --. ~ ,9 ‘ ‘. )0. “ ..-.‘.J ‘ H ’ - 9 'J't- .Kt' 3. l A 'u i ‘ hk "q f.'o_ q ,1 . 81y: ' ‘3 " ‘ ‘ 1 r " o ' .. .' 1 . ' ‘Q -'~'.3ét:.".};~'.‘-..‘J:;3:” - 1 : ~~ f; ‘4‘)"; :3,an ; . '. 3‘; n .. ..‘i. J. \ .4 t 53“,. ‘ - ‘3" ..h ‘1". ..f. u. '. - ‘_ v0. | .‘I "‘6’ "" Q . ‘no: | -?.~ I0.1:;?¢_~j “‘1‘ '. \ '3'.‘.o I III. | ~:‘6':351 3 = '14 | ‘3’ ' I" t 3‘ '03. ..- . 3" ‘.c‘.‘ ‘:{(I- . “a II“. .‘IU“. . I Q a "1.’ w . ... 6 ..J ~ ~ 5).;- ~‘ -.~ . . . a ' I . 3' "1“339. v.‘s_ ', .}I6 I, ‘ _ '0‘ ‘. v: 6 4 0“ I J ‘ Io' .0 ‘I .. I. 4%"? ~41». I 2: ' I .\ I O . z. . n - ‘.' I. . . s— t.’ .‘I ‘ 3:553:33 343' x... x a. . :3 ‘0 k v ‘ .. t. " . rt .- a p." o \‘l 3“..w.'. I: ~T‘:¢... "l 0 fl "' L“ ' .i.‘ ' «"6. 7: L'WJ‘L ', h i "' " " J“ ' L -.r- ,. -w.¢w~~. . . , Q Fflr,o.ll :3, 1" {3 ~ , 5-,, W.- .. . . r Q )- .u‘ 0 -~"..‘ .0. A 1 . I. . ,' . g ‘ 4" ‘ " 3' U- 59 h | l ,‘. i g 1 {gitwgkwaz-fig . .3 . 3. . .. 5 i . I b H. h «I, .I' , ~\'- 3333333-'3:"-‘:4 ~..- 3 3.. ' - ~£ 3 “'- 3.33 3: '45 r33 .3 . .. - ’ ‘ 'v‘fl‘h .? 3s. 3.31 ..3; . ‘ I ..- O..:. ‘ ’ a ’. a a I O ‘0. -.' h I o :‘b‘ ‘ ' ‘ 'J o w... _ Isa I_.. " \j‘ ‘ .‘ - y s i ‘0 "5V ‘ ‘ , n :3... C D 4 o I t i . . , I u. .z\. .1: :01: 1:11 .:‘ .(I 1 . v2.3 . . I. ~ .. .3 ' 't '1) ‘ ‘.\' v J'IQ.."QU"": "'((‘g .. 35. l r'.‘ ..‘z; c‘.;..“ "K"; till , I \ o ..,'« 4 "3 ' ."'."...'. I.‘ . u '- ‘ 3“}? *. I-III x"';‘r."-"1"'.°'\' 3..., .. .3 v . M .o - ‘;‘f.1.o'-i~‘.‘.:‘- ‘Z’J’J‘ I'LJ’Z""I-I;.I:IQD". - . - .. : ‘ 5 .3 -' . g":- 3“ ' r 2'3” o a" '3- 3' _., "" V .V nftr.‘ ‘ J ‘. IZ-l ':"":I ’" J..I‘..' 'Q"|'..- . ‘N‘i‘ 3! -4. Un I @.wm3 fivwwmeu.u . 1 ‘7‘- :g .. ‘o‘ uli‘at o.. ‘6' - n__\ ’ o , . ., ( .po{' 3‘?; &,Q (’91,. ”#352“; dig-i .A‘HEWX . .' l,‘-1.I..(v'—:. .. ., ‘4’. .. _d. x VEP‘I,’ I. 94;; 933$ "*° 'C\’-IJ-‘J‘b"-'a‘ ‘1 -'::'o‘a ' 6‘: ' o ‘1 f .b ‘. .3 ’I' '.- .IOU'IuI ' . '3 ‘9 ‘ . I . ‘ _ ' O ‘.'J’- -'..'. :r‘ ‘."""‘-'..’. .- .r I. 0"]..lu".\ A" f. 1 "V" a)?” ‘1" '3'?“.'.!, .Hr'iv‘ "1'5"" '9‘" ‘7 ' . i ‘ 1.." K.“ .xf 0‘ "J.'-I-" . .'. _ ' 41 I - '4 JI '- . I. . . - '- 33333333339; 5-23.: «.3 3.. -.: . 3. 9".' _I' '~.-1Q1 ‘ ,;.I'.' ‘0 I .a;;‘: .H .J,/??- '0' 'l‘.' ,1; . a ,, .0, . I ilgbl 'I‘I-‘Itf I} I‘. ,IJ(:-.:: “_ I“J.;»" . J: 2“) I. A 'I l . I ' ‘ ’l - . If'fl“. . )V :0." ’0 ”:'. lfxr:r 'r l"(" ...‘¢..:'O '3 . . ‘ t | O 5" fly, .11 ..r J , 3 3 3..I' 51"". .. - . . 93:. 22-:‘#-’33.'.:_ big." : : 3:32.. :.: 3.. .' 4- .4' .- . 9.33.53: $.13}? , .3 2‘? rh'igafj o J Eff 3.62:; ". :‘ 2"”."3 7 ‘f ’ '_' ' 3 -- . ‘ . 3 J I- . .' '. L' ." ' l - . I ._‘:O oJ-t':l if ‘i' ’3. t'.” *1” . ..'.:“:I ..I'Ir:.".'r(.‘ :l'llI H. :‘i‘I/I-‘t'o- I JV- ‘3 t'l‘." 0. :r‘.0'_. .' I. I,'. _ ‘ I n o I" >.',,'?1'3'}"'J '9' . - .' é ..flhflo" (‘3‘; :‘VMV" ,2" 2'53, “71?. '5'? 3"“? $5,; 'r ’51:}: If, 'H r "3'32"!“ .3 .' 'l "3 ”WI" . "7"; '3'")? 3.63:. {AV 30’ ”‘9’; .1?‘”3’_rr’j:’r; fly: {it} ”If, “If-"'8‘." l'ff’t\f§'»"f: ['3'an o". ..' . ”If." . J ; . 3 ' "I H 3.‘ . ‘ .ID. q. :;?_'.. C u if!” 0 _ . ;. ..._. ‘ g .2 “r" ". ,‘I.‘; O '. I'I..." I «p -I_ c I I ”I. ' “3’(/'3 ‘313'%}’§'.9'LI3L3;'I. .‘3 .-"r :' .3. .' 1333')! {.5‘4V":.3133;€VW w 'i" "'6 in": VJ“, '( ' 9'1" €"".3’3fl3"]/h';1) If!" ' .I in”: 3". _. '"r' " ' .335 -' $.31 j.‘ 32" .9.”va '3‘ - "3"".- ' ’ {3 ’33 -“- ,,.t;s..-.3.;.J.-.,.3.,-,3/3329-1;“ .. 3 3 - . . / ..... .1 3:. 3 .4 .39. x: 9:3 ' «3 . .33 =- -:-'3.'3%:-r:z3'-3:::..;: 3'3: »:.3::zf3:7:< .-. ?éru 1:333 . , «3'5 (fir! p/" '- ,Zt‘flnif. r) ‘ . u" - '. 3-".P'f'. ‘3 ’ L. '.',J ‘ff'J’uhu-JWT- 3‘3'. ’ ".‘. 3‘ "fl". .. ‘_ ' " Alt; 1'" "A 2‘23; 5" rigé {5% In: in o 5 5/ f‘fi 3': v ”":- '3‘, ’3);( 31f ”¢ 'T ' '3' o [a "4" ‘ r~ .' .. ' - " ' I ‘ . .‘ ' Q "0.5”" ”1""WW ( I" his ‘I' . enr"' . ' 'J} ‘ ‘ c"'- ".“‘¥u' ‘- 5' ' " ’ . i ‘ 'h 0 ,' '.. u" ' ,- far,“ .".') . r-II . li' r'~' ‘ ’1'; 4-1;??? . -.r' 'l , (65" 3 if: i; '0’ 4‘1)?" }.,/ I b '1 "‘éf’fi;r;.¢"-‘J?'9J$l'I‘litg'r‘fitfj.i'a"¢“‘ '0‘;:',£‘.4. gait .5 ‘5" :‘r’. F": :I..’0II‘:I IIII II .3 .III : m: .. 3:33.. . 3, r , 3 .3’.::.:;3:3.-;=-:3;-,;3§33?; :. g -. - ~ » . . , ‘J'rb. ' t.‘ 'r .' J‘ I. .0 " .' -'. I. a. n f ' ' -'. ' '0' H‘ ' . .06 .l r . .1 1'3 3‘" 3'." 'EJ’W' " '.'“..’ 13r-!-..3v’}"-‘-»r’3-r 53.3%.: '4'. m7 fl‘flr 33.3.33 *4/33: '3 .1 V 23' 3, .13 . r33 . 3 .’ , ‘ .; ’3 -r3”33 3"‘(3 33' ' 33:33:33.3: 333323.33“. 7’33"- 3.313%«4'-;3.3~:3-u-'--: '55:? 7' 3.: ’ '3 ' 3 : . ' ,a’; _ _' ,' 0"...“ J_(v :fi' . ""3 13". . '. rhr'126' ". Y -Po., I V ‘.- . ‘ . 0'. 0’1'.':. 1-. . . - .’ o o 0 ’0 End), . - ' . 1,... o ' . . .. .. 1. I f ‘1 4’“ .,.., . . . . I fif\ 3"}. ‘ It": l:f. ’ifia‘tI {III '1'???) ‘. ’. 25:]. ;"'J'"{8':fififfl Inn)- ,Ikfil‘if y' @2531“. psi“; 1H,J*.,!E,I .’:’.'2‘: #16:" (f '1 g"! t.’_ . :‘ é)! i l 4'. 4‘ ¢ '. ‘ ‘. . ' "' -‘-‘-3 E'S'FV'Q» -"' ’13! iaJ'W'fl'. 3 'Vh“ ’ '3‘3 '3 '7'3 '93::3 ”'53:! ' “Kn/W ”1 2‘ r'." . 3 ' Yf‘f’l‘; 33‘3"", " I""-//r «UN? "1 4 ’..}on".‘-“. 3." 3’- . .y'm 'f ' "I: 1.1 3'92." ‘ ft! $5,?“ -#? "33-0“ ‘ ’ ::'9‘) U ' j o .. Ifll‘p’.‘ “I ' fl, ‘3: u' I ‘ 1'. lg. ' 2.)"! :3". .j- {I}. C :‘IE? "(1‘ 16...: a”. 4...,” (Li’fi 153?“.11" 0",” :.“‘I Jru'; if"... ‘ f H." 1:. 'IO ‘ "I . I. . II.‘ ' ”V 1' -"i""‘ ’11'?§L¥"..'r..rg 1.3““ 3.33: 1353'. .- ' o [I .1 ' .' ' . D J ..4..'-3o (as. 3:3 .33.»... ffrc’r- . 35.1.53353 n . ..' 1, . '.¢"v"'-':.' .' 'I, '--.‘ '1 '3" "'- 31»- {air}? 3:4 "V: WEI; .5. t .Iu 13313.55} 3“! ’3 'l; V33; (f3: 9 TX? 9 ‘5'” "’I ' ‘3 J " “'3' {3‘1" 5" .3": '5"?r";.'1"‘r"'3-£'vtf'f',rJ.3 '1'}.."' it! ’.'('.O ' 0' 1:1" - ‘ ‘1}:I ' "'5‘. .‘3 '.-?.'.‘r."‘ 1;)9AIC” .. I. 55 '(‘vflfo :5‘..'- ($5.. I" ; III- '1 .f \‘ ‘7. gr .l .‘ ; r' C,“ r .‘I 5%.. 6": Vflff‘t _". a "W '6' 'I' ;I.":“ 3. f 22:,‘2'. ‘1‘!)- i;‘;;:. _”j:; '2 2%,, .le‘ '97,- I [l’v‘fi'jn’f .. . ‘»l P, ‘Y'- “-_v-'1‘. 12105.3“ ‘1'-“ ‘ 6 I ' ' u —. “a , " I 0 , A; ffV'i- WE: “‘3 :6: 15 1"... '5’.) 1, P4}? 3' "'13" n '3'”. 35:,- :73 34"; 45" . ‘1': 3.: ~' ,. .. 0 I. O J‘ I 1,? U. “'1’; ‘ ’.":'/"1}’ :3 3’ «6 Ci: .3 l a" ‘ ' I.» ‘ "..u o '3' g (ls/D9“ iérlro‘é?‘ “‘ffimit r2 rtdi" ";’ fi:l'h. gt...) ..I‘I’ I 0 t. .I. "'f' I' " . /J’r"4f fl/Ir: If.:f"’ v. 16% tf‘,,(, . 'gfil‘t‘L." J’:':0 c’dvil :Qg’. é}: '- .r: ' . I .r . . I I. 01" l 1' 3‘ ""‘L‘.’ 3 JV}: IT“ ,' 4'3 :V'l"o.9tl.' " 5:”11' ' " A:'f.1u'~r5;; M, I. 5.. "f '3 {rah}! ta‘n’u‘ o '- ' 3" 3" '. ' f 3:"... . . is"! 'l. 1-; ((I' ,_—_.'{’r3 ‘0‘ r'rl9 "".”. {hi/:71} N... ""r‘fJ/‘Wlfil “rift; "Sty... " ’c” .' .1Jl” 1’0. . -'o" ‘0 " ' ‘ If. . 'a fik"’,..’ f O‘{.’o‘ "OIOr:...-.‘"lo:(.' J‘J. ..J .' 0’"? ; KI, ’-.,4U ¢ ‘1'" [r . _ ’ _" .td riflf’.’:€”‘:, lef’ 1""... .o'j“.oouv ' "I. . , ' 'f/ 1) 0.1 I ' 5339“, '.'.."‘v'»|‘-'-"” 9’? )3? f'.""' .’l "c.” .1131"! 1,3: 9" '1" 1' . ' ulna-1L1; :Uf .. .1"? 60' '5- "t3 o' '3 .1 "f..J!J!'-g:’.' _' “‘9i .rfg. Sir/“$3 jl.‘r'/I.' I”: I,” {'12,}! v d'PV::”(f./"E¢f:&l{, tn" I: f,‘.iho';:w1.":”'r *‘ ' .' ‘ ' ,5.“ v3.1 :g'z. '1”; .3 rh‘. lz‘h affil- .51:‘I.:.. I .. .3. ' u,” ‘. ’, ... "’_ ‘ .tn ‘ ”Hf. '3’", or}; ""J- .. 1"“; I?! 'J‘ I. "' $10!): 33.1.) j,_!_3,‘_:',' ' 0 ,J‘. .3 .6 ~ ,. 7.6.1... . '0' y ‘9: .W‘Ofl’ '4— 54,“. "Eff-'21! if}? if} ' . :15?“ H. 'v‘u '31’::::(;;,"’Efrk}?’)"ql}'£ll.‘ZIf; b Wtfixfiql. A I‘ 4.36 ""."“~}L I" 6 TI, .£f’.{”" ;'{." ..rIo/II I . I I ' ,M'f , L'.‘ I); (a! 3.. (LOJUNJ', ‘IW' 'I"-I'.' _ . " ('93): '71.}‘353’5563'QI .471'... ' 7 ' A'flb‘Ef-Vm’l 1" 56ft"! '1 " """" I "' "-3 ' '°' W "I“ " j”! ‘ 1/2: “'2‘, IV“. é’ég.’ I R729; A“?- .'::'Y ‘ f I“? “I ' ' A? 3""? ' Q (I. '1." g ' O:".'{ 5.! 3’) ”Info”: 9:2,: I. *3 3 ‘20:” I I' .'l f.‘ f. "T.- 66' 3r.” 'EI'I‘. m OJ'.‘ I 1’”)! r f 'o ' W' 6% J71”: ’ 'I. ' :Zr- ”1’. ”If" I 3 6 .i‘é’f.‘ ." :71? .; 5-1.5; 'l’loi‘J-i: ‘55.; .15"; {551.1% 9’! J .' ”lg": “ L'o'o-.v‘.ll 1'..qu '4' -'g)’l .zfi‘l‘gr'u' r;f_;:v (P‘s'rv .. c"f'6 . f‘ o a ::.: riff. _ .II - ... ‘ ‘ 'I' f .-o .1" ' L" 9' ’o I.’1{II.;.}bo,\.'Ih- (’0: r. {4’ v3' ‘ 95’: {-31,}. 9 :' 6 ' 'I‘ ”fit“ Viv/'3! 5?. (ff-has; m: L!JJ;:§3".' ‘9?! .3 o - - 1 ' '- ' I. ' " . '3 ..H" ("l r ' . ' -r ‘ r'l‘l . -" . t. '43.. r-ir "‘ ? ,;;;::; ".0 3 '31."; 'Il‘. '. I: : JV " .- I, 0-: 600 ‘,o' o" "I‘ll , ‘ u ‘_ . 91 v.“ I-"' 3.“ ' o. ’00,. 2' ?..{.:;1,0“ 3-.’33.' ..‘v-u‘v’ ' ;I‘I,y:;:1 V 5::2-‘3! _":\l.'1,'I’f '-"3’ I’JI :ég-{rr"".'f1,; :'_.'J:?.:v' '1! '35:“; to 5;} ' v '0 fkfi‘qf'ééfl':p$r'}:<{fjf{it‘f“I. ' .10.. J5}; 1%,}?‘59 42";:J../u; ’!;f. ..lzérfrv é“ 3. A! Lab’s” JOIII If; u'lIo .‘.U_ _|‘ {I #f’ 3'”. ’v‘o'ol' ..-. Io.". I .rh..({ Ilv. . "' - u: 0‘ _ 'f/' ' _.-Ir.¢ " V. '_‘_I," O} ,3" .ao"r. 1| a:;,;.* liar?“ "flj‘r'J J's". :h' L-Ocryl )/[.”‘{ Jéi’ “I ;;jj::v-‘r‘ L;$:;;{{él'g~311f‘giygfl‘ {gfi'ZJflI 2%.;Ee ‘z.’, I. 'f!/’.:'!}ullsl’;: .:://é');l; 1"? '2 ‘f. 1’: ff’:& “3;:n"$/’;h u." :fi:f.!.:u .‘. . a. . ..ff‘ .1sz " J ’1'!" H ' " 3. " '( 4 'l ' (3.. gr c, "v 'z. ' 0r ".3 a 4:7: 36- '3' f3" 3 _ - ....3,‘z.~./ .3] ' ~33 9,-3.4 :33 "3' 3' "3’; 333’ 33 531.;1‘3’5’7'1’ ”'3 3.553.; .o-soéfirz-v‘ .. 33.333 . m, '3’: ; 3.3. .’ :3:v.-. www'wwmwflwwfl ."Y%Mw 3 ww3:x" M ~JWf% . I ' ."' ‘ o '3’ t 0H" . '|O ' FY l“.JJ ’15). 3/. {"..’.‘Iv'. .' ;. h 2%: 316:!” {4.) i‘é'a'c '3'.. 'Ifif; 5335‘ 63:1,?” I ifgigzp'f' .';’/u‘;o:.’f:o' (’gé:: I 3:. y k.“ in”: 6-7:]! lief”? rd}. 5- yryr- #- (LII : {2513' ffi'lx/ 11:! 7.1.3": ':’. (3/, £591! .’ . ’, 7’ ‘9“, ,5 W .4, ttif’I-‘h VI; '1') :J!.;'.' 2.9,. /;"I‘I-J_":: 17W 3:31,... i‘ . .3 . '{ff '1'?! o .£;;I.:,§ pg.) .2"; . ."'}'¢”* '. . “5"?!" :- o.r<">'_o[ ; 3': i ’ .33" 3:1' Mrs vi' 337:. . 3.3;? ’m 3.: 7-" ”3 .. .4 . ' 33333.53» 1.3"? {32.3 , .3" " {-1333 m3; . 1333.3. 5:23:33 .33. . : . . 3 f ') 'C... c.‘\ .3 f'f‘.._'!," ' ’ ..',' f.“ . o. I 4.56‘fl'..r. ‘5 I" ,fiflJ/Jfi ”'1?!" :&'1 3'7. ‘54” '32:.“ ,.”f'£;'/-{.' J8 I'I' "'3‘;§(g;a :. o’rotOP'lr‘I’v$,3:3.-'f} 'o'Iv'v' v ' .J r4l'&zh.";'é}n:.;fl 3%,“: ’1..':-" f;f_'{!; ;: '?;l;:r 'I ‘I‘ 31%“; l':'l.r.:{ff:2£‘$9{e.rv;‘gik,é:'r {/le f/OO.“ O‘: fl’larz'?e:.2’;:y ..:.' J; ' ~- .3.-.._.::,..,.. 3 1.:- v'» if '3'. "» 13;,» ‘-3 3:74.! .3 “/c 3 3&1' 313 31933.3: ~43. .. ,7 I I.“ l. J . I , I , 3, ,. 1H,!) ,0 l ..y :"M I: “I!" "I . ,3; ,f,‘.;:, ' ., - ’ gi.‘:'?:;‘.l lit: ‘h'qu-J :t' I" Illrnll'lofid I”: "5.)"; 1'59. If} ’VA'I“ .7 “’ Wad" ..‘le‘ I‘JP": I-w 3 .:;;:1’::0’.f I? '1 "l’ {fern} ”if::'),[' ‘:'-.< if '{ - I fr 6. I l r . . ‘ . q '0 a . q . g. I.‘ zip, ‘ (.1 .9 Jar.” ' ‘T ‘0'." I” a", at O " -. ',3. £3.17" ’3)" £2335}: ‘33:???" '6'???" 1'. "33.3'3‘33‘3': 33"...3 ”.y‘rIWEIWEI’ii'flf “335' £32”; ..332 ’3'" ‘ ‘d"‘r.I| {,1‘ .,‘r I I 'I ' ’ ' #37555; '3' 53.31.54, 33': M31. 5:723; "3331;. I559. 3/,'3'33~33.33:$p’§:‘-;«. . y 6 3 'l 111 W” 3".” ”3.153 :«m:: . . 3 . ,J.4.34 H. , . . '. agggl"); 5%."; J' 36:1},71 1,?” W3 3""; "$122178" it, If ’3’); MICHI NST .INI JERSITY LIBRAR RIE ES gm 0 o I 7 7 3 III III III IIIIIII III III II T ’ 0077 5969 LIBRARY Michigan State University This is to certify that the dissertation entitled PARENTAL MEDIATION OF ADOLESCENT TV VIEWING ACROSS TRADITIONAL AND NEW VIDEO HOUSEHOLDS presented by David John Atkin has been accepted towards fulfillment of the requirements for Ph.D. degree in mass media ~ qajoWofessor O I Date MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 1V1ESI.J RETURNING MATERIALS: Place in book drop to LIBRARIES remove this checkout from .—:—. your record. FINES will be charged if book is returned after the date stamped below. m’fi’m PEI? 9343338 I {06 0 4 0 1 MAR 07 ‘99! i NUIVJILO I’M-2' ,V f": . -. PARENTAL MEDIATION OF ADOLESCENT TV VIEWING ACROSS TRADITIONAL AND NEW VIDEO HOUSEHOLDS By David John Atkin A DISSERTATION Submitted to Michigan State University ‘ in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Mass Media Ph.D. Program College of Communication Arts and Sciences 1988 ABSTRACT PARENTAL MEDIATION OF ADOLESCENT TELEVISION VIEWING ACROSS TRADITIONAL AND NEW TECHNOLOGY HOUSEHOLDS By David John Atkin This study addresses the influence of pay cable, basic cable and broadcast viewing environments on family viewing and mediation patterns. A set of family, sociodemographic and media—environment variables are incorporated into Bron— fenbrenner’s typology of ecological niches. The theoreti- cal model advanced suggests that cable status (an exosystem variable) plays a determining role in family mediation patterns, alongside traditional demographic locators. Hypotheses suggest that children from pay homes will spend the most time with TV and View the highest number of R-rated movies. They will be followed by basic and finally broadcast homes. Pay homes are also expected to report the greatest degree of restrictive and nonrestrictive media- tion, followed by broadcast and finally basic homes. Data were collected from an in-class survey of 425 fifth graders and 401 tenth graders from a suburban Midwes- tern community during November of 1984. Parent instruments were also sent home with students, as 359 fifth grade and 172 tenth grade parent cohorts were returned. The survey instrument addressed attitudes and reported behaviors with rulemaking (e. g. on viewing times and content) and media- tion (or guidance) with child TV viewing. Anova contrasts compared pay, basic and broadcast homes in terms of the criterion measures. Regressions were then run to examine the degree to which media use and sociodemographic factors predict viewing time, content, mediation and rulemaking. The data reveal few differences in viewing time by cable status. There was partial support for expectation of differences in R and PG-movie viewing, with basic cable homes viewing more such fare than their broadcast counter— parts. In terms of restrictive mediation, there were no consistent differences with the student contrasts, although parents from pay homes reported more viewing rules and concern over excessive child viewing. Similarly, there were few differences in terms of restrictive mediation, though 10th grade parent and student reports suggest that broadcast homes engage in coviewing to the greatest degree. Regression results confirm the fact that TV use variables play only a minor role in explaining variance in mediation. ACKNOWLEDGEMENTS The author is greatly indebted to a number of people during his stay at M.S.U. Special thanks are directed to Dr. Bradley Greenberg, who helped me develop research skills and served as my dissertation director. Gratitude is also due Dr. Thomas Baldwin, my committe chair, who generously shared with me his knowledge on cable TV. Thanks also to Dr. Thomas Muth, who enlightened me with multidisciplinary perspectives on telecommunication policy and research. Finally, Dr. Martin Block is appreciated for his methodological suggestions for my dissertation. The writer wishes to extend appreciation to his wife, Carolyn Lin. Without her continuous support, this disser- tation would not have been completed. Thanks also to my classmates--Ed Cohen, Rob and Jan Wicks, Gary Pizante, Allen Harris, Carrie Heeter, Renato Linsangan, Gloria Viscassillas, Keith Kenney, Larry Collette, Indra DeSilva, Jean-Luc Renaud, Megumi Komiya and Kwawdo Anokwa--a fine group of colleagues. I am grateful to Dr. Charles Atkin and Janet Atkin, who gave us support and advice over the years. My parents--Dr. Kenward Atkin and Jane Atkin--as well as Thomas and Penny Atkin, are appreciated for their love and encouragement. Finally, I want to thank the wonderful secretaries--Ann Alchin and Brenda Sprite. TABLE OF CONTENTS LIST OF TABLES. LIST OF FIGURES CHAPTER 1. PART I: INTRODUCTION Background . Definition of major variables. An ecological perspective. PART 11: REVIEW OF LITERATURE. Overview Evolution of research on television and development The changing exosystem: a profile of program environments. The microsystem and parental interaction during viewing . . . . . . Nonrestrictive mediation: microsystem responses Parental restrictions on TV viewing: the last defense. . . . Macrosystem influences . . . . New video technologies and consumption: macrosystem elements New video and microsystem elements Endnotes . . . . . . A PARENTAL MEDIATION MODEL FOR NEW VIDEO HOUSEHOLDS . . Interlinkages among the nested environments Microsystem Macrosystem influences Exosystem. . Time spent with television . . . . . Determinants of mediation: an interface among niches Content patronage and expected benefits from mediation Effort required to intervene Restrictive mediation. Traditional television 11' viii H (Date 10 10 11 13 17 21 23 28 30 31 34 37 37 40 41 43 44 45 48 51 51 52 Basic cable Pay channels . Nonrestrictive mediation Endnotes METHODOLOGY Testing a model of parental mediation. Hypothesis testing The survey instrument. Student questionnaire. Media use variables Content availability. Restrictive mediation Nonrestrictive mediation. Background and media- -use variables Parent questionnaire Restrictive mediation Nonrestrictive mediation. Data collection Data analysis. Anova. . Multiple regression analysis . Predictors of TV viewing and mediation Overview RESULTS Viewing time Content availability Restrictive mediation. Parent and child groups Nonrestrictive mediation Parent and child comparisons . Index means and comparisons for student groups Regression analysis Predictors of viewing time Predictors of content availability Predictors of restrictive mediation Predictors of nonrestrictive mediation Endnotes SUMMARY AND DISCUSSION Viewing time . . Predictors of viewing time Content availability . Predictors of content availability Restrictive mediation. Predictors of restrictive mediation Nonrestrictive mediation Predictors of nonrestrictive mediation iii '0 D"... 54 56 61 63 63 64 65 66 66 66 67 67 67 68 69 70 70 72 73 75 77 78 84 84 88 90 92 99 101 105 107 108 110 111 114 116 117 120 122 125 127 128 130 132 134 Comparisons among group findings . . . The exosystem and mediation: an evaluation of the model Measurement limitations . Implications of study findings Further applications Conclusion Appendix 1: Survey Instrument Appendix 2: Correlations Among Dependent Variables Bibliography iv 134. 136 139 144 146 149 153 162 163 Table 10: 11: 12: 13: 14: 15: 16: LIST OF TABLES Student background information Parent background information Hours spent with TV before school (5th children) Hours spent with TV before school (10th children) Hours spent with TV before supper (5th children) Hours spent with TV before supper (10th children) Hours spent with TV after supper (5th children) Hours spent with TV after supper (10th children) Viewership of R-rated movies (5th children) Viewership of R-rated movies (10th Viewership of PG-rated movies (5th How late child allowed to watch TV (5th children) How late child allowed to watch TV (10th children) . . PreSence of rules in the house for viewing (5th children) Presence of rules in the house for viewing (5th parents) Presence of rules in the house for viewing (10th children) children). children). TV TV TV 80 82 85 86 86 87 87 88 89 89 90 91 93 17: 18: 19: 26: 27: 28: 29: 30: 31: 32: 33: 34: 35: Rules regarding viewing should be enforced (10th parents) Parental disapproval for certain shows (5th children) Prevent viewing of particular channels or show (5th parents) Parental disapproval for certain shows (10th children) Prevent viewing of particular channels (10th parents) Parent expressions over excessive TV viewing (5th children) Concern with viewing over a set number of hours (5th parents) Parent expressions over excessive TV viewing (10th children) Concern with viewing over a set number of hours (10th parents) Parent recommends show to watch (5th children) Parent recommends show to watch (10th children) Parents change channels while coviewing (5th children) Parents change channels while coviewing (10th children) Parent views TV with child (5th children) Parent views TV with child (5th parents) Parent views TV with child (10th children) Parent views TV with child (10th parents) Parent discussion while viewing (5th children) Parent discussion while viewing' (5th parents) vi? 94 94 95 96 96 97 97 98 98 99 100 100 101 101 102 102 103 104 104 36: 37: 38: 39: 40: 41: 43: 44: 45: 46: Parent discussion while viewing (10th children) Parent discussion while viewing (10th parents) Index means and comparisons for student groups Weekday TV viewing Viewership of R-rated movies Viewership of PG-rated movies Restrictive mediation Parents tell you to stop switching channels Late night TV viewing I Nonrestrictive mediation Summary of findings for mediation & viewing v11 104 105 106 109 110 111 112 113 113 114 119 Figure [0 N) N) LIST OF FIGURES Bronfenbrenner’s typology of ecological components applied to home media use. The ecology of parental mediation Krugman’s continuum of new telecommunication technology Predicted comparative mediation across services viii Chapter 1 PART 1: INTRODUCTION AND OVERVIEW Recent developments in video technology have greatly altered the content availability and viewing conditions for American families. In particular, the prolific growth of cable TV and VCRs has been propelled by the availability of "uncut" adult programming (Baldwin and McVoy, 1988). Many have since expressed concern regarding the impact this material might have on children (Surgeon General’s Commis— sion on Pornography, 1986). Such apprehension is not entirely unprecedented in conjunction with new media. As Wartella and Reeves (1985) note, the introduction of radio, film and broadcast TV was marked by concerns regarding their impact on children.1 It seems logical, then, that cable and VCRs--with their range of potentially offensive program alternatives--represent the latest focus for concern (Meyerson, 1985). This dissertation assesses the extent to which parental guidance of TV viewing varies in accordance with different home environment factors (e. g. family size) and content settings. Specifically, Bronfenbrenner’s (1979) model of hierarchically embedded niches will be applied to 2 predict relationships among traditional locators of parent- al mediation.' That model should also aid in further theory building and testing with unexplored mediation influences across TV service archetypes (including pay and VCR as well as basic and noncable settings). Background: Few areas in the realm of social science have generated as much concern and controversy as the effects of television on children. Twenty years of research have failed to conclusively develop theory and conclusions that explain the conditions for this influence. Researchers are, however, in agreement that TV can play at least a contributory role in conveying information, attitudes and behaviors (Jeffres, 1986). As Klapper (1960) notes, this impact is rarely direct; it operates instead through a number of intervening variables. For children, parental mediation of TV viewing represents one such factor (Brown and Linne, 1976). Subsequent research has firmly established the study of parental intervention behaviors as a distinct subdiscip— 1ine in mass communication (Messaris and Sarrett, 1981). When parents engage in such activity, they are taking part in a process by which the family ”screens, interprets, criticizes, reinforces, complements, counteracts, refracts and transforms" (Leichter, 1984). Despite the recent interest in this area, researchers have yet to assess how 3 VCRs and cable technologies influence family interaction. The recent proliferation of nonbroadcast TV alterna- tives, combined with regulatory accommodation for nonob- scene adult programming (HBO v. FCC, 1977) underscores the need for parental awareness of their child’s viewing. As Roe (1983) notes, perhaps no group of children is more involved in the process of new media adoption than early to mid-adolescents (aged 12-16). Despite that fact, Brown, Bauman, Lentz and Koch (1987) maintain that studies of this group have been overlooked in favor of those addressing media effects on young children: ...studies of early adolescent media use have been relatively rare. Yet dramatic changes in the media available to adolescents and in the families in which they live suggest that current descriptions of adolescent media use are in order (:2). Since the present study focuses upon parent and child concerns or attitudes rather than actual effects, it is in- tructive to examine children approaching or in adolesence.2 For, as Bandura (1969) notes, that is the time at which social system (e. g. media) influences begin to supersede those of the family. Studies in the parallel area of popular music use among teens suggest as much, as such content apparently enables them to strengthen peer bonds while moving away from parental influences (Lull, 1985; Brown et a1., 1987).3 One might, then, assess whether changes in media fare associated with pay TV and VCRs might motivate changes in 4 family viewing and guidance patterns. Interest in the general issue of parental mediation stems from the fact that children spend more time with TV than with any other waking activity (Bandura, 1969; Timmer, Eccles and O’Brien, 1985).4 Critics have long cited this pervasive TV presence as a basis for more strictly regulat— ing the types of programming that TV programmers may be forced to carry, as well as that which must be censored (e. g. FCC v. Pacifica, 1977; Act v. FCC, 1977). But it may well be that functions concerning mediation of TV viewing by children are being taken up by parents. Clearly, it would be in the interest of programmers, regulators and consumers to know how TV is being mediated, in all of its different forms. Toward that end, Bronfenbrenner’s conception of hierarchically embedded familial, social and cultural contexts should prove instructive, as past work in this area is largely uncoordinated. From this base, pro- jections regarding present and future adaptations to new video program environments can be presented and tested. Definition of major variables; Parental influence mechanisms might include rules and regulations5 on viewing times and fare (restrictive media- tion) as well as coviewing and discussions about TV (non- restrictive mediation). This study will investigate the interaction between family media use, documented mediation 5 patterns by subgroups and any changes associated with child age. The specific theoretical framework advanced suggests that the type of TV service a family receives can, in conjunction with certain environmental factors, play a determining role in specific parental mediation behaviors. Webster (1986) describes the two contextual media variables as follow: ..new media include all those video technologies that have the net effect of opening the distribution system to potentially unlimited channel capacities. As a practical matter, these are video recorders, cable television and other high capacity distribution networks. Conversely, old mgdig are traditional broadcast television systems that use a relatively small number of channels to deliver content on a fixed timetable (1986: 78). Of interest to the present analysis is any family accommodation to the differential content areas presented by each. Webster’s definition is helpful in that regard, although not complete.6 The most important dimension of these technologies is not their ”newness”, in and of itself. Neither the technologies nor the adult fare they " " to marketplace. Rather, the newness carry are truly new of these technologies is a function of their recent expan- sion into home marketplace (with hardware) and importation of adult content to that realm (in terms of software). One can then begin by characterizing these "traditional" and "nontraditional" areas in terms of their relative potential for offensive programming. Once those settings have been conceptually distinguished, the viewing audience for each will be contrasted. This dichotomization is in accordance 6 with earlier studies where TV service types are cast in extreme opposition along content dimensions "so as to make polar the factors affecting audience behavior and their relevance to social theory" (Webster, 1986: 78). In order to better understand mediation across view- ing environments, it is necessary to look in the related field of psychology to see how similar factors influence development. An ecological perspective Bronfenbrenner (1979) provides an appropriate topology by which researchers can conceptually distinguish inputs external to homes and responses to those ecological phenomena. One might, for instance, see certain viewing patterns result from certain viewing environments. As such, the framework provides a strong background for investigating family accommodations across different video viewing environments. Components of Bronfenbrenner’s topology of internal and external environmental niches can be summarized as follow: (a) A mesosystem, involving interrelations among two or more settings in which the developing person actively participates . (e. g. the child’s relations among home media environments, school and peer groups), (b) an exosystem, involving one or more settings that do not involve the developing person as an active participant, but in which events occur that affect, or are affected by, what happens in the setting containing the developing person, - (c) a microsystem, or pattern of activities, roles and interpersonal relations experienced by the developing person in a given setting with particular physical and material characteristics (e. g. interpersonal interaction in the home), (d) a macrosystem, involving consistencies, in the form and content of lower-order systems (micro-, meso- and exo-) that exist, or could exist, at the level of the subculture or the culture as a whole, along with any belief systems or ideology underlying such consistencies, and (e) an ontogenic system, which comprises the child’s individual psychological competencies for development (1979: pp. 222-226).7 While researchers have examined television viewing as it relates to each of the first four elements, Bronfenbrenner (1979) focuses his analysis on TV’s exosystem impact:8 Since the television program enters the home from an external source, it constitutes part of the child’s exosystem. To the extent that this powerful medium exerts its influence not directly but through its effect on the parents and their interaction with their children, it represents another instance of a second-order effect, in this case operating not completely within a microsystem but rather across ecological borders as an exosystem phenomenon. Thus once again we see the isomorphism of relationships at the different levels of ecological structure (:242). Thus, television represents an exosystem force whose influences transcends a wide range of ecological settings. This multi-faceted approach is due, in part, to the limited scope of Bronfenbrenner’s discussion of TV. His review of parental mediation is dated, and constrained to traditional viewing environments. Bronfrenbrenner (1979) is, nevertheless, certain with regard to TV’s impact on family interaction. Like the sorcerer of old, the television set casts its magic spell, freezing speech and action and turning the living into silent statues so long as the enchantment lasts. The primary danger of the television screen lies not so much in the behavior it produces as the behavior it prevents--the talks, the games, the family festivities and arguments through which much of the child’s learning takes place and his character is formed (:170). It could be argued that many of those activities that are "lost" to TV may, in part, be replaced by parental sup— ervision of TV viewing. As recently as 1979, Bronfenbrenner rather hastily concluded that no further research on the problem had been undertaken since Maccoby’s (1951) study (showing little mediation) was published over a quarter of a century ago. What follows is an attempt to broaden Bron- fenbrenner’s topology with subsequent findings and perspec- tives on nonrestrictive as well as restrictive mediation. Of particular interest, Bronfenbrenner’s concept of a domestically based "microsystem" aptly describes the interpersonal interaction which accompanies TV viewing. TV, itself, represents an external or "exosystem" component yielding indirect effects; the greater potential for adult content available with pay cable and VCRs present a basis for differentiation across that dimension. Specific parental responses to this external stimulus vary in accordance with cultural, or "macrosystem" influences. Additionally, parental reactions to new video environments might evoke different concerns regarding TV’s presumed effect on the child’s competencies for development (onto— genic system). Interactions across settings outside of the home (mesosystem) will also not be addressed, because TV viewing is essentially a domestically based activity. Each of these elements is pictured in Figure 1.1 below: 9 FIGURE 1.1: BRONFENBRENNER’S TYPOLOGY OF ECOLOGICAL COMPONENTS AS APPLIED TO HOME TV VIEWING / \ I MACROSYSTEM / \ I (Social Setting) / \ ' / \ I / \ I / \ I I I MICROSYSTEM I I I (Home) I I EXOSYSTEM I I (Media Interface) ONTOGENIC NICHE ' j ::::::::::::::= (Individual) : : MESOSYSTEM —- ------------------------ (Border area) . The remaining concepts should, however, obtain clues in the study of mediation. These settings, combined with those related to child age,media use and family background, comprise a confluence of influences in which the process of mediation is determined. Rather than focus on actual effects, this study investigates potential differences in parent and child reports of mediation across different TV service environ- ments and age groups. Thus, as Feldman, Wolf and Warmouth (1977) note, this type of activity can arise from parental concerns that TV might present "value orientations...oppos- ing those of the parents” (p. 131). It will be interesting to see how program environments with new video technologies figure into the process of mediation. PART 2: REVIEW OF LITERATURE Overview Studies addressing parental mediation have, by and large, ignored the possible influence of new program environments afforded by such sources as VCRs and cable TV. Much of the existing work, focusing on traditional (i. e. over-the-air) TV environments, has generated a good deal of controversy within the social sciences. The bulk of this research concerns the parallel area of TV and aggression (e. g. Liebert, Sprafkin and David- son, 1982). Clearly, as Klapper (1960) argues, such media influences rarely present a necessary and sufficient cause for behavioral effects. Rather, mass communication effects are mediated by a range of internal filters and external environmental influences. The family ecology of TV viewing plays a key role in that regard (e. g. Rubin, 1982). Much of this work can be expressed in terms of Bronfenbrenner’s topology of ecological settings. Schramm, Lyle and Parker (1961) were among the first to investigate the role of parental intervention with family viewing. They found that fewer than 20% of parents intervene in family viewing processes. These findings have been widely replicated since that time (Mohr, 1979; Haefner, Hunter and 10' ll Wartella, 1986). It would seem that things have changed little since Maccoby (1951) first noted that the TV dominates all other family activities when it is on.9 As Neuendorf (1980) notes, "the term ’parental mediation’ does not enjoy standardized usage," taking on dimensions of rules regarding viewing (restrictive) and coviewing or discussion behavior during/after viewing (proactive). For that reason, a cautionary note should be made regarding comparisons to be made across the rather N " diverse literature on new and "old" viewing environments. The limited data base relevant to the various studies, along with the complex topology into which they are inte- grated, preclude definitive comparisons between settings. Nevertheless, there is little reason to expect that mediation practices remain fixed across different home video contexts. Researchers have found that macrosystem variables such as ethnicity are predictive of viewing habits (e. g. Greenberg and Dervin, 1970; Beville, 1985). It is instructive, then, to consider these documented media use patterns in light of past research on home viewing environments. Evolgtion of Research on Television and Development Early studies regarding children’s media use behaviors focused upon family consumption patterns. Garbarino’s (1972) summary suggests that a major reason for initial purchase of a TV set was to bring families together in the 12 home. From a conceptual standpoint, studies have focused on parent-child TV interactions with the hope that mediation could promote desired changes in the child’s behavior and development (Williams, Smart and Epstein, 1979). As Bronfenbrenner (1972) notes, TV provides an ideal stimulating activity, furnishing an opportunity to study parental intervention on a scale not commonly found in other disciplines. Despite that potential, much of the work in this realm suggests that TV viewing is a largely noninter- active process (Maccoby, 1951, Mohr, 1979, Haefner, Hunter and Wartella, 1986). Few Canadian parents even knew what programs their children were watching (TV and our Children, 1976). Upon finding that 78% of the respondents from one study indicated no conversation during program viewing, Maccoby (1951) concluded: The nature of the family social life during a program could be described as "parallel" rather than interactive, and the set does seem quite clearly to dominate family life when it is on (:428). Much of the subsequent work in this area focuses upon the impact of televised aggression on children, a summary of which is provided by Liebert et a1. (1982). With this research there evolved better general understanding of parental influences in the media use process. Chaffee, McLeod and Atkin (1971) and Bandura (1963) were among the first to examine children’s media use from a developmental learning perspective. Bandura’s social 13 learning theory suggests that children’s personalities may be shaped through interaction with environmental stimuli including the media (in addition to family and peer influences). While the Bandura group and other mass communication researchers have generally taken account of these factors, they do not incorporate the full range of variables found in Bronfenbrenner’s earlier mentioned topology. It is useful now to consider the first of those elements. The changing exosystem: a profile of proggggAenvironments As Bronfenbrenner (1979) notes, TV represents one of the broad social forms of influence which comprise the exosystem. For the purposes of this discussion, the exosystem encompasses the first three elements of Berlo’s (1960) oft cited "source-message-channel-receiver" topology. Again, the emphasis is on different available program environments, as opposed to transmission modality. More than any other setting, the exosystem influences and is influenced by public and private policy decisions, especially where the media are concerned.10 While these elements are not a customary dimension of developmental research, Bronfenbrenner (1979) notes the critical need to integrate such policy: Knowledge and analysis of social policy are essential for progress in developmental research because they alert the investigator to those aspects of the environment,both imme- diate and more remote, that are most critical for the cog- nitive, emotional and social development of the person(:8). l4 Returning to Webster’s definition of "new" and "old" media, one finds a series of social, economic and regula- tory factors germane to each. As Klapper (1960) and others have noted, audience maximization concerns constrain commercial broadcasters to a relatively narrow range of discourse.11 When they exceed the bounds of normative appropriateness, the FCC and judiciary impose limitations on broadcast communication, including stringent indecency standards. In so doing, the Supreme Court has recognized that children in the audience must be protected (See FCC v. Pacifica, 1978). So, for reasons of economics as well as public policy, the programming of traditional broadcast television is widely characterized as inoffensive, uniform, uncorrelated with channels, and universally available (Webster, 1986). New technology environments, on the other hand, have been allowed to feature "offensive" (adult—oriented) mate- rials because they do have the same degree of broadcast- ing’s public trustee nature (See HBO v. FCC, 1977). But, with services such as cable, reception modes are virtually identical (1. e. all programming comes through the same set). For that reason, several litigants have tried to apply broadcast indecency standards to cable; the Supreme Court recently rejected that argument (e. g. Community Television of Utah v. Wilkinggn, 1987).12 Given this laissez-faire approach to content regulation, the last line 15 of defense lies in the home, with parental mediation of programming. Feldman, Wolf and Warmouth (1977) note the following with regard to parental mediation: Parents often feel uniquely responsible for promoting certain values in their children. In a heterogeneous and changing society, however, other value orientations--some opposing those of the parents--are often communicated (:131). Perhaps nowhere is this offensive program potential greater than in cable and VCR viewing environments. Uncut R-rated movies are a staple for both media. Indeed, at no time in history has such a wide range of mass video material been available to such a large number of people. Roughly 50% of all U.S. TV households subscribe to cable; half of them (or 25% of the total) subscribe to at least one pay channel (Nielsen, 1987). Additionally, cable subscription encourages ownership of VCRs, which are now available to over 50% of TV households (TV information office, 1987). This pervasity complicates the process of parental control of child viewing, as parents are confront- ed with "uncut" movies not typically available on broadcast TV (U.S. News, 1985). In terms of the more widely available "basic" cable programming, Baldwin and McVoy (1988) maintain that adult themes on music videos are a source of great concern and criticism. Additionally, Schmidt (1976) notes that portrayals of offensive speech and sexual conduct are 16 common to another basic cable service--public access. New technologies also alter interrelations between elements that comprise the mesosystem (i. e. across family and peer settings). For instance, VCRs offer a greater potential for "peer" viewing (Roe, 1987), allowing teens to explore anti-authority identities and a greater degree of overall viewing freedom in the U.S. (Levy, 1984) as well as Europe (Johnsson-Smargdi, Roe 1986). These concerns can be added to those accompanying a traditional TV industry which has, itself, attracted criticism. Groups such as ACT have long tried to effect changes in programming at principal sources--pressuring the networks to eliminate violent content (Atkin and Lin, 1988). Many parents share this apprehension, voicing additional concern over sex (e.g. Wartella and Reeves, 1985) and advertising (Ward, Wackman and Wartella, 1977, Reid, 1979). Thus, as Mander (1983) notes, this adult- oriented programming constitutes a form of "developmental eavesdropping" which can elicit a range of mediational behaviors from parents: ...first, they debrief the child, or interpret the event in such a way that helps the child make sense of it in his own terms; second, they attempt to instill general modes of responding so that the child will be able to interpret in the future similar or analogous events; and third, parents will try to censor the environment that the child is not exposed, to the extent possible, to avoid confusing or distressing events (:245). The question is, how might the discontinuous world of viewing options delivered by newer cable and VCR services 17 alter parental concerns and mediation patterns? Are cable parents likely to show more or less interest in mediation relative to their broadcast counterparts? The next section outlines the work on restrictive mediation for traditional TV and the nascent literature on new video environments. The microsystem and parental interaction during viewing Where TV programming entering the home represents an exosystem component, it influences interactive processes in the home, or microsystem. Epstein and Bozler (1976) note that 60% of the children studied had been allowed to watch TV before they had reached the age of one. As the child approaches adolescence, TV viewing typically subsumes an ever larger proportion of their daily activity, surpassed only by sleep in terms of time allocation (e. g. Liebert, et a1. 1982; Timmer et a1., 1985). Work in this area stems, in large part, from earlier mentioned concerns that children may imitate antisocial behaviors from the visual media use (Bandura, Ross and Ross, 1961, Bandura and Walters, 1963; Bandura, 1969). It was logical, then, for scholars to investigate the parent’s role in minimizing TV’s negative consequences (Schramm et a1., 1961; Atkin and Greenberg, 1977) and maximizing positive impacts (e.g. Mander, 1983). Much of the early research on broadcast TV use, stems from the assumptions of social learning theory, suggesting that children will model their parent’s viewing behavior 18 (Himmelweit, Oppenheim and Vince, 1958; Schramm et a1., 1961). However, Bronfenbrenner (1979) suggests, this influence is not static: The further one goes in school, the more likely one would be to experience freedom from close supervision...and opportunity for self direction(:248). Applying this to the microsystem, one finds children may also originate their own content preferences and opportuni- ties for viewing as they develop (Bandura, 1969; McDonald, 1986). There may, of course, be a reciprocal influence for children with regard to their parents. Some have found that childrens’ media use influences that of their parents (Clarke, 1963; Chaffee, McLeod and Atkin, 1971). Chaffee et a1. (1971) term this a "reverse modeling" effect. They further note that such modeling influences are dependent upon individual family characteristics: it appears that families with similar parent-child communication structures indirectly produce characteristic media use patterns that are shared by parent and adolescent...The parent child communication milieu perhaps operates as a separate factor that independently leads parent and adolescent to behave similarly in other communication situations involving mass media (: 336). Indeed, as the authors note, "...one can assume that these individual differences grow out of background experiences that are more pervasive than simply observing the use their N parents make of the mass media. (p. 330). Examining the components of this family configura- tion, a number of studies have found that it is the 19 influence of the mother that prevails (Hess and Goldman, 1962; Clarke, 1963; Fonts, 1976; McLeod and Brown, 1976). This finding stems logically from the fact that mothers are more likely to act in the role as caregiver within the microsystem (Bronfenbrenner, 1979). In terms of other family configuration variables, children watch more if they have a working-mother or absent father (Medrich, Rozen, Rubin and Buckley, 1982). Brown et a1. (1987) likewise find that maternal employment is negatively related to TV content rulemaking, while having rules is negatively related to the amount of time spent with TV; the presence of fathers bore no influence on either viewing time or mediation, however. While the role of family size is less clear, studies have found that larger families are not as restrictive in terms of time spent with TV (Webster, 1983). There can be little doubt that families exert the primary force in a child’s development. Over time, though, that presence may be challenged by peer and media influences (Bandura, 1969). This, combined with the fact that younger children are seen to be more impressionable, might motivate different parental mediation patterns as the child develops. More recently, Brown et a1. (1987) confirmed those trends, noting that the presence of cable and VCRs are predictive of the amount of time a child spends with the media, but not of parental rules for viewing. They further 20 note that new video households are more likely to have a second TV. Over half of all children now have their own set, and over one-third of all homes have more than two sets (Broadcasting, 1987); This growth in the number of household sets is likely to reduce the latitude for parental supervision of viewing. More general examinations of family viewing note that parental actions to control their children’s viewing behavior decline from early (12-13) years to middle (15-16 years) adolescence (Greenberg and Dominick, 1969; Lyle and Hoffman, 1972; McLeod et. a1, 1972). Chaffee and McLeod (1972) find that the time at which TV viewing decreases among developing children—-after early adolescence--is also the time in which they become progressively less influenced by their parents. Hence, as Bandura (1969) notes, social systems influences (e.g. peers, media) may increase over time relative to parental influence. In terms of Bronfen- brenner’s topology, this means microsystem influences are gradually replaced by mesosystem or exosystem influences. This is certainly true of TV. Several studies have found that children approaching their teens (ages 11 and 12) watch more TV than any other age group (Petersen and Zill, 1980; Timmer et a1. 1985). In terms of specific mediation behaviors, Mohr (1979) found that parents of younger children (elementary school age) indicated more positive and negative guidance than parents of seventh 21 through ninth graders. While viewing limitations or rules may be less frequent as the child matures, direct mediation might continue in different forms. In a landmark project, Brown and Linne (1976) forwarded the now classic model whereby parental influence was treated as an intervening variable to the direct effects of television. They found evidence of parental control in terms of viewing restrictions, coviewing and discussion regarding content, and the assertion of a general "family ethos." Specifically, positive parental attitudes toward programming were corre- lative of higher degrees of viewing among children. Each of these elements will be separately addressed in terms of the hierarchically embedded niches in which they occur. Nonrestrictive mediation: microsystem responses Bronfenbrenner casts his discussion of verbal media- tion patterns as part of the exosystem, beyond the microsystem. The TV viewing room, however, represents an ecological border where any parental control is likely to be motivated by concerns about their child’s psychological (ontogenic) development. It is useful to examine verbal mediation patterns defining that boundary. Neuendorf (1980) provides a summary of studies which shows that many parents interpret TV programs for their children. They often do so with formal planning, in the hopes of aiding their child’s learning from TV (Messaris 22 and Sarett 1981). Such comments have been found to aid child recall of central and incidental plot materials from children’s programs (Watkins et a1., 1980) and adult-orien- ted adventure programming (Collins, Sobol and Westby,1981). Fewer than 25% of all parents engage in such mediation, however (Mohr, 1979). Some work suggests that the presence of an adult coviewer can facilitate understanding (Atkin and Greenberg, 1977) while others stress the need for interaction (Mohr, 1979). Still others maintain that interaction per se is not enough; adults must impose viewing rules to aid their child’s comprehension of TV (Desmond, Singer, Singer, Calam and Colimore, 1985). Much of this work does, however, show that parental interpretation enables children to learn significantly more from TV (Walling, 1976; O’Bryant and Corder-Bolz, 1978). Other researchers have investigated specific dimensions of microsystem activities designed to further the child’s psychological competencies. Singer and Singer (1983) found that parents who foster imaginative and cognitive activities for their children’s viewing prompted them to watch fewer variety/ game shows and use signi- ficantly fewer TV references in their language. The importance of such interaction in averting potential child trauma is underscored by Mander (1983): Just as a child may be upset by the most trivial quarrel between his parents, so he may be upset by a verbal or 23 physical conflict on television that an adult would read in a "proper perspective" (:246). So one can add preventive concerns to the range of motivations for nonrestrictive mediation at the microsystem level. Here again, family ecology researchers are not so much concerned with the actual outcomes as they are with parental perceptions that they can mediate TV’s influence. Aside from minimizing possible (or perceived) child traumas related to viewing, parents may wish to ameliorate the antisocial effects of aggressive portrayals. Several researchers have found that verbal mediation can accomplish that end (McLeod, Chaffee and Atkin, 1972; Korzenny, Greenberg and Atkin, 1979). On balance, then, there is much evidence that coview— ing, discussions and other forms of nonrestrictive media— tion shapes children’s responses to a range of specific exosystem influences. Such interaction reduces the impact of advertising and entertainment, facilitates the acquisition of prosocial behaviors and inhibits the acquisition of antisocial behaviors from television. Should parents operate on this assumption in the more sophisticated "new-tech” video environment, it could well influence their mediation behaviors. At thevery least, parental perceptions of potential harm or benefit could well be altered by these exosystem changes. Parental restrictions on TV viewing: the last defense Recent concerns regarding TV violence and indecency13 24 have underscored the importance of parental mediation as a home control device. Aside from the nonrestrictive strategies mentioned earlier, parents may simply opt for restrictive mediation. In terms of microsystem interactions, one can look to the rules perspective of communication as a means by which to conceptualize family regulation of TV viewing (Lull, 1982). Interpersonal communication researchers imported the rules perspective to communication theory, which Lull (1982) then adopted to the study of media use behaviors. As Meyer and White observe, "this theory assumes that human communication is a purposive, goal-oriented activity and that people participate in and make sense out of their situations by employing normative rules" (1982 :814). Shimanoff further notes that these rules ...may be explicitly stated, or they may be known tacitly...A rule is a followable prescription that indicates what behavior is obligated, preferred or prohibited in certain contexts (1980:71 cited in Wolf,1982). According to Lull (1982), one specific type of rule-- habits--is useful for studying parental restrictions of their children’s TV viewing. Habitual rules have been defined as ...firmly established, regularly repeated, nonnegotiable patterns of action which are not likely to be questioned. These habits are often defined by authority figures as in cases where parents coviewing behaviors of their children (Wolf, 1982: 817). Whether conducted from a rules perspective or not, 25 the research in this area suggests that such mediation is not widely practiced. For example, a Canadian study found that one-quarter of the parents thought that children below the age of seven should be allowed to make their own program choices (TV and our children, 1976). Differences in social customs notwithstanding, results from work in the U.S. reveal similar conclusions. Studying the parents of adolescent children, Musgrave (1969) found that most made no effort to prohibit the viewing of any program. Lyle and Hoffman (1972) found that preschoolers frequently made their own program choice decisions, dropping off even further after children enter grade school. Comparing restrictive and nonrestrictive mediation practices, Barcus (1969) drew a further conceptualization distinguishing between positive (recommended) and negative (prohibited) program controls. He found that negative formal rules and regulations were the most commonly reported control. In another study addressing disciplinary practices, Korzenny, Greenberg, and Atkin (1979) applied Aronfreed’s (1969) concepts of induction (attempts to use reasoning or explanation) and sensitization (disciplinary assertions utilizing power, physical and verbal threats). They found that children from high sensitizing/low inductive families were most affected by aggressive TV fare. Those from low sensitizing/high inductive families were least affected by 26 physically and verbally aggressive TV content. So it would seem that the "soft sell" works better as a microsystem screen against external influences. To be sure, a range of other factors influence these aggressive tendencies (for a summary, see Liebert et a1., 1982). Very few are directly relevant to the process of parental mediation. Suffice to say that the earlier summa- rized media-effects findings are consistent with parental perceptions. In terms of content, studies have found that parents are more likely to intervene out of concern for program content rather than advertising (Feldman, Wolf and Warmouth, 1977). Specifically, TV violence has been found to be a primary motivator of parental intervention. Child viewing has also been found to vary in accordance with specific household predispositions. For instance, Barnes, Kelloway and Russel (1978) found that positive parental attitudes toward television are a predictor of child viewing. This finding is consistent with work addressing the new video technologies, which shows that families that are more interested in TV tend to watch more, across all age levels (e. g. Krugman and Eckrich, 1982). Roberts (1981) added several measures of rulemaking (restrictive) and coviewing (nonrestrictive) mediation behaviors into a regression equation predicting viewing time for children. ’Changes explained by this model were modest, however, accounting for 20% of the variance in 27 viewing. When parents do express an intent to engage in mediation, they report more guidance than their children (Greenberg, Ericson and Vlahos, 1972; Rossiter and Robertson, 1975). Here again, mediation was lower among older children. Robertson (1976) suggests a social desirability rationale for this difference, stating that parents seek to appear in control while children minimize the presence of those rules in order to gain peer status. Contrary to those expectations, however, Mohr (1979) found that parents reported even less guidance than did their children. Specifically, 85% of parents and 84% of children reported that no guidance was given with family viewing. This difference was not statistically signifi- cant, though, and it did not hold true for all programs. So it would seem that the earlier studies, predicting greater parental control, present a more convincing picture of media interaction. In the related area of VCR uses, Levy (1984) suggests that different types of TV entertainment technologies may be associated with different types of mediational activity. Researchers (Greenberg and Heeter, 1987; Lin and Atkin, 1988) have found that VCR homes are no more likely than their non-VCR counterparts to engage in parental mediation. More generally, Haefner et a1. (1986) found only minimal levels of parental discussions regarding programming in 28 "new video" households (1. e. those with cable, computers and video games). In a preliminary examination of the present data base,14 Greenberg (1985) found that fifth graders watched an average of five hours on cable (the previous day), while their broadcast counterparts watched for three hours; tenth grade subscribers watched an average of four hours, compared to two hours among broadcast homes. Kerkman, Wright, Huston and Bremer (1983) investigated the differences across cable and noncable households, finding cable children scored lower on Peabody 1.0. tests. Aside from that, only a handful of studies have addressed family interaction patterns across TV service archetypes, as much of the remaining research emphasizes traditional viewing environments. It is useful now to review that work as it relates to (1) background influences on mediation and (2) new video consumption patterns. Macrosystem inflpences As Bronfenbrenner (1979) notes, the macrosystem is the broadest component of the ecological framework. In the present case, it would involve cultural beliefs, values and attitudes that influence the TV viewing environment. While these and other parental constructs are not directly applicable to children, they provide researchers with a proxy measure of a child’s background. As Bronfenbrenner (1979) notes, such intrasocietal contrasts play a key developmental role in that 29 .systems blueprints differ for various socioeconomic, ethnic, religious, and other subcultural groups, reflecting contrasting belief systems and lifestyles, which in turn help to perpetuate the ecological environments specific to each group (:26). Here again, one can readily integrate much of the work on parental mediation in terms of macrosystem influences that vary across classes. As Chaffee, McLeod and Atkin (1971) note, the importance of class influences was noted in studies during TV’s early decades: ..a... serious challenge to any modeling inference is the likelihood that parent and adolescent are independently led to use the mass media in similar ways by other factors which on persons in the same family are identical or nearly so--socio-economic status, residential locale, intelligence, physical capacity, family conflict or tension, and so forth. Factors such as these have often proven to be useful predictors of individual television use (: 337 citing Schramm et a1., 1961). Ethnicity has emerged as a powerful macrosystem determinant of viewing. Specifically, black house- holds have been found to view more TV (Greenberg and Dervin,1970; Beville, 1985) while showing no less interest in mediation (Greenberg and Dervin, 1970). Hispanic households also have fewer limitations on viewing (Greenberg, Burgoon, Burgoon and Korzenny, 1983). Most recently, though, Brown et a1. (1987) and Blosser and Heintz (1988) did not find any significant differences in either viewing time or restrictions based on ethnicity. These traditional factors can be joined by another transition occurring outside of the home-—the explosion of single—parent households (Kurdek, 1981). While this 30 influence has yet to be fully investigated, it seems logical that supervision of viewing might be less in single-parent or mother-employed households. Early work failed to find any support for that contention, however (Brown, Bauman, Lentz and Koch, 1987). In any case, it would seem that children’s time with parents has been increasingly supplanted by time spent with the mass media (Lyle and Hoffman, 1972, Timmer et al. 1985). Moreover, with increases in the number of TV sets per household, family viewing has become further fragmented-~thereby reducing the latitude for parental mediation and family interaction (Rubin, 1982; Haefner et a1., 1986), The following section examines factors related to TV consumption that may influence the mediation process. New video technologig§ and consumptionzmacrosystem elements Much has been written concerning the general viewing patterns of cable and VCR patrons versus those of broadcast-only media users. To begin, broadcast viewers tend to be older, have fewer children, less education and lower incomes (Webster, 1983, Krugman, 1985). Barnes et a1. cite reports suggesting that, before cable was available, families with lower income and education tended to view TV more extensively than higher status families. Clearly, cable and VCRs have reversed that trend, confounding some of the earlier noted S.E.S. configurations for TV viewing. Johnsson-Smargdi and Roe (1986) found that TV viewing 31 was negatively correlated with VCR use among Swedish teenagers. In terms of domestic media use, Greenberg and Heeter (1987) found that teens from VCR homes watch more programming in all of its various forms (including R~ movies). These teens generally hail from higher S.E.S. households. New video and microsystem elements In addition to the macrosystem predictors of new media use, one can add a list of more specific motivations for viewing. Chief among them is added diversity among program options for cable households of varying economic backgrounds (Baldwin and McVoy, 1983). Clearly, cable viewers are not as satisfied with TV in its traditional form, and seek to make much greater use of the medium (Metzger, 1978; Ducey et a1. 1983). These viewing segments vary in accordance with their consumption attitudes toward broadcast TV as well. Appro- ximating a relatively more innovative consumer, cable subscribers are more willing to take risks with purchases. Nonsubscribers are not as receptive to new ideas (Banks and Gagnard, 1984), seeming content with traditional TV. VCR owners are likewise open to new ideas (Levy, 1981). With regard to specific cable household types, basic subscribers seek greater program variety than traditional viewers, but not as much as pay subscribers (Metzger,1978; Becker, Dunwoody and Rafaeli, 1983). They also reside 32 between pay and non-subscribers in terms of their receptiveness to new ideas (Krugman and Eckrich, 1982). Basic subscribers, as a whole, are characterized by younger, wealthier, and larger families than broadcast TV households (Webster 1983; Krugman, 1985). And finally, at a relatively "higher" end of the socio-economic and media use spectrum one finds pay households. Pay families tend to be somewhat larger and younger than either their traditional or basic counterparts (Krugman and Eckrich, 1982, Baldwin and McVoy, 1983; Webster, 1983). They also watch slightly more television and see TV as a more important source of entertainment (Baldwin and McVoy 1983). Kerkman et. al (1983) evaluated the effects of various TV viewing environments upon preschoolers. The researchers found, among other things, that children in basic cable and pay environments watched more TV than in broadcast only homes. Moreover, preschoolers in families with two-pay channels had substantially lower Peabody verbal IQ’s than any other group, independent of their parent’s education level. These studies, taken together, present a picture of differential exposure levels (microsystem), living con- ditions (macrosystem), and academic performances (meso- system) associated with different media exosystems. While the latter will not be investigated here, this study 33 examines whether mediation varies across media settings. The following section incorporates Bronfenbrenner’s topology into a model of parental mediation across pay cable, basic cable and broadcast TV settings. 34 ENDNOTES 1. Radio is one of several services that was initially broadcast. Early versions of the phone were also operated on a broadcast model in Eastern Europe. This paper focuses upon mass video technologies (e. g. cable), and is not intended to be exhaustive in its treatment of telecommuni- cation innovations. 2. Parental mediation research focuses upon parental concerns that might motivate guidance in child viewing. While some research examines the impact of these mediation patterns on learning of prosocial and antisocial behaviors, the present study does not address itself to such effects. 3. Popular music useahas traditionally been overlooked in past work with parental mediation. As cable services such as MTV provide video versions of popular music, it will be important to consider how TV and VCRs enable teenagers to strengthen peer bonds. For a discussion, see.Roe,(1983). 4. It will be important to revisit the question of time spent with TV across different decades, to examine the degree to which changing media environments alter viewing patterns. 5. The terms "rules" and "regulations" have been used interchangeably in much of the parental mediation work. The present study will treat them as synonymous. 6. The term "new" is not linked to the age of the technology itself, as many carriers such as cable and microwave dis- tribution are as old as broadcast TV. Instead, the criti- cal distinction with new technologies concerns their recent diffusion into the video marketplace. VCRs also fit the definition of an expanding new technology, though they will not be analyzed here, as penetration was too low at the time of this study. 1 Bronfenbrenner’s concept of an ontogenic system lends itself more appropriately to the study of potentially traumatic situations for children, and their capacities for dealing with them. Such cases are not likely to arise with TV, and will not be separately considered. They are, 35 however, relevant in the context of parental concerns which may motivate mediation of TV content. 8. In this sense, the term "home" or "house" refers to the primary place of residence in which the child develops. Events occurring within the home are commonly associated with the microsystem, although homes actually represent an intersection point for the different ecological settings. 9. This observation is based on the assumption . that the family watches one primary TV. The recent proliferation of secondary TV sets obviously complicates this picture. Implications for this and other exosystem changes are fully discussed in Chapter 2. 10. In a very general sense, public policy refers to planning decisions undertaken by public officials who are charged with regulating various activities. An example might include the FCC’s promulgationof decency restrictions on broadcast programming. Private policy refers to the planning activities undertaken by businesses. While not formally guided by public officials, these private deci- sions are often shaped by "public relations" concerns regarding the image of a company (e. g. a cable system that chooses not to carry R-rated programming where such fare is likely to alienate a large segment of the viewing audience) 11. It should be noted that this policy also covers noncommercial mass telecommunications. But audience maximization concerns are most prominent with broad— casting. 12. 61] F. Supp. 1099 (D. Utah, 1985), cert. denied, 108 S. Ct 985. This and a series of cases in Utah and Miami, Florida established that cable is a distinct medium from broadcast in that it enters the home though a private or closed circuit, is a subscriber service, and does not utilize ubiquitous over-air delivery modes. See Cruz v. Ferre, 755 F. 2d. 1415 (11th Cir. 1985); 9 Media Law Rptr. 2050 (S.D. Fla. 1983); Hgge Box Office v. Wilkinggp, 8 Media Law Rptr. 1108 (1982); Community Televigion of Utah, Inc. v. Roy City, 555 F. Supp. 1164 (E. Utah, 1982). 13. See U.S.CLA, 1464. 14. Greenberg’s (1985) report was based upon an analysis of different viewing measures than those analyzed in the present study. As a preliminary analysis of the data, those findings did not yet control for the influences of individual and family characteristics on the decision to subscribe for cable. Further distinctions with that 36 approach and the one employed here will be discussed in Chapter 5. CHAPTER 2 A PARENTAL MEDIATION MODEL FOR NEW VIDEO HOUSEHOLDS Interlinkages among the nested environments As Chapter 1 suggests, parental mediation is likely to vary across different micro- and macrosystem archetypes. Researchers have yet to assess this impact as it relates to different exosystem configurations. It is clear, though, that new communication technologies such as cable feature unconventional types of programming which attract audiences that are different from those of broadcast TV. They also present new opportunities for parental mediation, which may or may not result in action. In an attempt to better understand those processes, this chapter provides a taxonomy of subject areas related to parental mediation of viewing across different environ- ments. Dimensions of family viewing include viewing time, content patronage, restrictive mediation and nonrestrictive mediation (i. e. parental recommendations or coviewing). Based upon the previously cited literature, this section advances a model containing several assumptions regarding media use and interactions between the hier- archically embedded developmental niches. They can be summarized as follow. 37 38 Exosystem: 1) Children from new video households will have access to a wider range of programming than their traditional TV counterparts. Microsystem: 1) Children, if left unregulated, will view a greater variety and duration of programming across a wider range of times. 2) Parents will more closely mediate viewing in households with younger (i. e. preteen) children. Macrosystem: 1) In traditional video settings, lower S.E.S. households typically watch more TV and engage less frequently in proactive and restrictive mediation. 2) New video households, generally higher in S.E.S., tend to watch more TV than traditional video households; parental mediation is largely unidentified in these settings. Specific dimensions of parental mediation, and the variables that should help predict it, are included in Figure 2.1. FIGURE 2.1 ABOUT HERE Each of the elements included in Figure 2.1 will certainly vary in their availability and degree of in~ fluence across different households. The typology to follow will explore the role each of these ecological niches plays in determining viewing time, content patron- age, restrictive mediation and coviewing or recommendation behaviors (i. e. nonrestrictive mediation). Specific hypothetic expectations will then be presented for the home ecology variables of primary interest—~those addressing the 39 FIGURE 2.1: THE ECOLOGY OF PARENTAL MEDIATION Home media Parental Predictors environment mediation (macro- and (exosystem) (microsystem) microsystem) _______ Content_______ General Adult Restrictive MACRO —--—Traditional ------- \ I .daypart and night- Race / c \5 .time limitations __S.E.S. / h \3 .duration limitations Marital i ' .program prohibitions status 1 Basic Cable Parent age d Nonrestrictive Sibling ages .coviewing MICRO \ 5 .discussions \ e Pay Cable .recommendations -——t —————————————— / ! POSSIBLE EFFECT (ontogenic) 4O exosystem. Other relationships focusing upon the macrosystem and microsystem will be included as part of an overall model of parental mediation; they will not be addressed separately, as their interaction with the exosystem is of primary interest. In order to better understand these interrelationships, it is important to first examine microsystem and macrosystem influences. l. Microsystem The previously cited findings on family viewing influences, taken together, can be easily integrated into Bronfenbrenner’s ecological perspective. Within the micro- system, family size and age dimensions are likely to affect viewing time and mediation practices. In terms of the for— mer, it is likely that TV would be patronized to a greater extent in larger families, all other factors held constant. Viewing would also be greater for children approaching ado- lescence (age 10-13)--a time at which parental influences are waning and peer activities do not yet predominate. These family age and size dimensions are also likely to influence mediation practices. In each case, the pre- sence of younger children would motivate intervention. This stems logically from perceived needs to protect younger viewers from harsher aspects of adult-oriented fare. Additionally, demographic characteristics influencing the mediational resources available to parents may also affect mediation. Specifically, families with more children are 41 likely to display less mediation, as they have a greater number of prospects to "police." The presence of this type of parental concern is, itself, likely to influence viewing time. While dimensions of parental mediation will be discussed more fully in con- junction with other ecological niches, it is possible to now observe that higher levels of parental interest in restricting viewing hours and times will likely reduce the amount of time children spend with TV. Likewise, parental restrictions with activities such as channel switching or doing homework while watching TV are likely to have the same influence. Other nonrestrictive practices should also place a downward force on viewing time, as they’re sympto- matic of the same underlying parental mediation dynamic. These interrelationships between viewing time and mediation were demonstrated by Roberts (1981). And finally, in terms of child gender, recall that studies have found parents to be more protective of daughters than sons when it comes to TV viewing (Barnes, Kelloway and Russel, 1978). This may stem from the same concerns that might obtain for younger children--namely, that their delicate state of mind warrants a degree of special parental protection. 2. Macrosystem influences: These microsystem factors, no doubt, play a role in certain mediational activities. Yet, as the literature 42 suggests, few have emerged as consistent predictors of mediation. Such findings may stem from the difficulty of categorizing different permutations of those factors occurring from family to family. While parental employment status ultimately affects TV viewing at home, compensated work typically occurs outside of the home--as part of the macrosystem. Within this realm, the growth of single-parent households may reduce parent-child activity (Kurdek, 1981). That is, with fewer parents available to supervise viewing, less guidance might be expected. It would seem that children’s time with parents has been increasingly supplanted by time spent with the mass media (Lyle and Hoffman, 1972, Timmer et al. 1985). This trend can only be exacerbated by increasing numbers of women in the US work outside of the home, who now represent 46.8% of the workforce. In terms of mediation, single-parent families would have fewer resources with which to police their children, so lower degrees of mediation can be expected. As Figure 2.1 indicates, race is also likely to predict mediation, with black households displaying lower degrees of nonrestrictive and restrictive mediation. This may be a function of the greater degree of affinity that blacks feel toward TV, as witnessed by generally higher use rates (e. g. Greenberg and Dominick, 1969). The section to follow examines how these macro- and 43 microsystem factors interact with the exosystem in the process of parental mediation. 3. Exosystem: Overview: The complex nature of microsystem, macrosystem and exosystem interaction precludes a strict accounting of how each affects prospects for mediation here. This is because of the difficulty in partialling out influences of social status and new video use, since the two are positively related. Similar interactions occur among microsystem settings. Thus, while the added exosystem dimension of adult programming might motivate different parent intervention behaviors at the microsystem level, it is not easy to establish a directional force in new video settings. What follows is an attempt to conceptually distinguish those countervailing ecological influences. To begin, these new forms represent an unfamiliar exosystem dimension which might motivate a greater degree of parental mediation. As the earlier discussion on policy suggests, changes in regulation and delivery afforded by cable imply a larger range of "restrictable" material available to children. That potential is somewhat greater for pay cable, where adult pay channels typically feature "R" rated fare. Given this increased latitude for programming, the following section outlines likely implications for overall TV consumption and its relation to mediation patterns across TV service types. 44 Time spent with television The link between viewing opportunities and viewing time or mediation is often a function of equipment avail- ability within the home. Few could dispute that the child’s latitude for evading parental control in these and other settings is influenced by access to secondary TV sets. According to a recent study, 45% of children surveyed (ages 6-15) have their ,own TV (Broadcasting, 1987). These multi-set households are more likely to own other new video products (Krugman, 1985). As noted earlier, secondary TV sets are widely available across cable and noncable homes. For each case, the availability of secondary outlets makes it more difficult for parents to mediate child viewing in most any household. In addition to investments in additional TV sets, it seems logical that those who pay for and receive a more diversified channel menu are likely to spend more time with the medium (Krugman, 1985). This heightened TV use tendency among TV pay homes is likely to play a role in determining mediation in that environment. One might, then, hypo- thesize as follows: H 1: Pay homes will spend the most time with TV, followed by basic and finally broadcast homes. Specific measures of time with TV are expressed in the following subhypotheses. 45 Hl.l: Students from pay homes will spend the most time with TV before school, followed by basic and finally broadcast homes. H 1.2: Students from pay homes will spend the most time with TV after school and before supper, followed by basic and finally broadcast homes. H 1.3: Students from pay homes will spend the most time with time with TV after supper, followed by basic and finally broadcast homes. The section to follow examines how this affinity for TV might influence parental mediation of TV viewing across pay, basic and broadcast households. Determinants ofgmgdiation: an interface among niches If one were to focus strictly on the macrosystem dimension of class, a clear direction of TV influence would emerge. Recall that, .in broadcast viewing environments, lower S.E.S. households tended to be more laissez-faire in their approach to viewing regulation. Higher S.E.S. families were more likely to engage in mediation. So, to the extent that basic and pay households are also typically "upscale," macrosystem influences would imply a higher degree of mediation in new video settings. The confluence of variables favoring mediation, involving class and content influences is, nevertheless, contradicted by heightened TV consumption patterns with pay and basic cable. That is, viewers who patronize new media (i. e. basic and pay cable) are likely to be comfortable with the TV in all of its different forms. These different class and TV use tendencies can be 46 reconciled by observing that new technologies bring in multiple program environments, only some of which are "objectionable." Even though parents may seek to control the latter element, they obviously don’t find that concern sufficient to discontinue TV use altogether. Clearly, as Meyrowitz (1986) notes, the difficulties accompanying direct intervention force parents to take the good with the bad. This is especially true when a child has their own set. The child’s television set...is like a new doorway to the home. Through it come many welcome and unwelcome visitors: schoolteachers, Presidents, salesmen, police officers, prostitutes, murderers, friends and strangers ...Those parents who tackle the monumental task of censoring their children’s television viewing are faced with at least two significant dilemmas. First, controlling television involves a conflict in values: protecting children vs. allowing them to learn as much as they can...A good deal of the protection from adult information in books was taken care of automatically by the inherent features of print. Now parents find themselves in the uncomfortable position of actively intervening in the "learning process" of their youngsters. Second, it is difficult for parents to control their children’s viewing...without limiting their own viewing as well (:245). More to the point, the act of mediating a child’s TV viewing may well prove ”painful" to parents in certain respects. As Meyrowitz (1986) goes on to note, ..even if parents act to censor programs, parental control is nevertheless weakened because the control becomes overt and therefore often unpalatable to both children and adults. For these reasons, perhaps,...parents exercise surprisingly little control over what children watch on television (:247). So it would seem that the consistent lack of me- diational interest revealed by past studies might be 47 attributable to the "unpalatability" of parental supervision and confrontation. In ecological terms, then, new video technologies may complicate parental mediation of exosystem influences at the microsystem level. This is not to suggest that all new video households would adopt a "laissez faire" approach. In addition to the many individual viewer tendencies, one might expect that mediation may change in accordance with the specific types of technologies and services offered. Simply put, there is no reason to expect this level of mediational concern would remain constant across all TV service archetypes. Mediation can instead be expressed as a function of rewards and efforts. A more ”adult-like" pay environment might thus provide new incentives for restrict- ing content, as well as enhanced mechanical means for con- trol. In terms of the former, efforts required to mediate must be justified by associated benefits (in the case of proactive effort) or harms avoided (in the case of restric- tive intervention). This relation is expressed as follows. Likelihood of action = Expected benefit from intervention Effort required to intervene Hence, mediation would most likely ensue when the benefits of intervention are greatest and effort required to intervene is least. If, for instance, parents receiving adult programming sought to "censor” that content, then the numerator would be relatively large. One could designate 48 that benefit of avoiding indecency to be, say, plus 10 on an arbitrary scale. If the effort required to censor content was lower (a. g. plus 5 on that same scale), than mediation would likely follow; such would be the case when this ratio exceeded the value of one. Conversely, if that ratio was less than one (it e. when required effort to in— tervene exceeds 10, in this case) then mediation would not ensue. This might occur when parents don’t think child learning is worth the trouble of discussing TV content. The greater the relative disparity in these terms, the greater our ability to predict a family’s impetus for mediation. Higher values represent a stronger impetus. Lower values represent a weaker impetus. Both the numerator and denomi- nator terms are affected by new video. It is important now to examine the role content plays in viewing and mediation. Content patronageggnd expected benefits from_medigtion As Krugman (1985) notes, pay and basic services represent relatively more discontinuous media formats relative to broadcast TV, in terms of content environment and delivery modes. His relative ranking of those video environments is included on the following page (1985: 24). This stems from Robertson’s (1971) typology of innovations, which includes products which are continuous (slightly modified) relative to existing technologies, dynamically continuous (extensively modified, in the case of adult channels) and discontinuous (entirely different, 49 FIGURE 2.2: KRUGMAN’S CONTINUUM OF NEW TELECOMMUNICA- ATION TECHNOLOGY Standard Continuous Discontinuous Dynamically consumption Discontinuous consumption Broadcast TV Basic cable Pay cable VCRs Source: Krugman, D. (1985). "Evaluating the audiences of the new media." Journal of Advertising, 1985, l4, 4: 21-27. requiring new consumption patterns). It would seem, then, that these different "exosystem" (content) influences could well motivate different patterns of mediation. This typology of content availability stems from the differential "exosystem" environments mentioned earlier. Specifically, it is logical to expect that homes with the capacity to receive "R" and ”PG” movies through pay channels will indicate seeing more adult fare. Such channels show movies around the clock. And, while the bulk of these are rated "PG", "R" films represent a popular component after 10 p.m. (Baldwin and McVoy, 1988). Such is not the case with broadcast and basic cable settings, even though the latter has elements which might prove too "adult" for broadcasters (e.g. sex therapy programs). Because basic cable is marketed as a general 50 family service, it is still highly similar to the broadcast exosystem (Webster, 1986). The greater latitude for indecent programming for basic would still render that service more "discontinuous" than broadcast TV. Relative to pay TV, though, basic cable and broadcast homes would present families with a relatively lower "need" to censor programming; the expected benefit from intervention (numerator term) would hence be lower. In terms of assessing this influence, recall that re- searchers (Roberts, 1981; Blosser and Heintz, 1988) have noted an inverse relationship between parental allowance of viewing and mediation. It is possible, then, to obtain clues regarding mediation by examining exposure to adult TV. While exposure to such fare is likely to be higher as a function its increased availability in pay settings, mediation could also play a role in viewership. In fact, both R-movie viewing and viewing prohibitions could be greater in pay homes, as there are more opportunities for restraint. Even so, a perspective on different content types should help establish that those settings do actually prompt different viewing patterns, even if not ultimately linked to mediation. Applying this perspective to different program environments, one might expect that those less predisposed toward mediation might allow their children to view PG and R-movies to a greater extent. 51 H 2: Pay homes will report watching R and PG-rated movies to the greatest extent, followed by basic and finally broadcast homes. Specific sub~measures include the following: H 2.1: Students from pay homes will report watching R—rated movies to the greatest extent, followed by basic and finally broadcast homes. H 2.2: Pay homes will report watching PG-rated movies to the greatest extent, followed by basic and finally broadcast homes. To better understand this influence, it is helpful to examine specific actions involved in mediational efforts. Effort required to intervene: Further differences in mediation across TV service groups might arise from differential mechanical controls within those exosystems. Here again, pay cable patrons possess a limited means of system control. In these settings, electronic channel "lockboxes" allow parents a measure of control for individual pay channels. This microsystem (or domestic) control over exosystem (media) content is less comprehensive with basic cable, as not all channels can be ”locked-out.” Broadcast TV offers an even lower relative degree of system control; the only filter for content in that setting is the "on-off" switch. The following section explores the degree to which this lack of control might inhibit restrictive mediation across an otherwise "traditional" broadcast viewing environment. Restrictive mediation: Both terms of the fraction of determination are 52 likely to be affected by new technologies. For, in addition to exosystem concerns over benefits for avoiding or emphasizing programming, electronic devices (1. e. channel lockboxes) also influence the effort required to intervene at the microsystem level. In order to gain a better understanding of these factors, it is necessary to examine their implications for specific home media environments. Traditional: As the preceding review indicates, traditional viewing environments invite a low degree of mediation. One might summarize, in terms of the model, that this is chiefly attributable to the relative constancy of broadcast fare. On the other side of the equation, the lack of any convenient "lockbox" control might make mediation more difficult, and hence less palatable for broadcast homes. This logistical inconvenience should, however, be out- weighed by the same concerns that often prevent one from subscribing to cable--concern over "sexy" content. This overriding content consideration will prompt broadcast homes to restrict viewing to a greater extent than basic homes--but not to the degree displayed by pay homes. Basic Cable: When confronted with the task of mediating the more plentiful cable environment, basic-TV households would likely "give-up" in the face of_TV’s new diversity. That is, both logistical and dispositional factors would 53 mitigate against strict supervision. In terms of the latter, a cable household’s affinity for TV should make them more receptive to non-adult fare. That content dimension should also play a role with mediation in basic homes. Otherwise unconventional formats (e. g.. public access) are increasingly coming under local scrutiny. So it would seem that basic is moving back in the direction of broadcast TV in terms of content obnoxiousness; that is, cable programmers are increasingly aware of the need for a non-indecent service alternative. But, even where viewers are concerned about such fare, the logistical difficulty of knowing exactly what is appearing on each of the non-pay channels undermines control opportunities. Basic cable households thus face a lower fraction of determination for intervention relative to their broadcast counterparts. According to recent federal legislation, cable systems have had to make channel lockout devices available by sale or lease.1 These devices are typically a premium beyond existing service, and have proven attractive to a minority of subscribers (Baldwin and McVoy, 1988). They may become an increasingly important factor in restrictive mediation, with their ability to screen-out unwanted channels. Here again, though, control may be undermined by the fact (or perception) that "secret" codes for various channels might be learned by a child. More importantly, 54 the fixed memory of lockout devices precludes comprehensive electronic censorship across a typical 36 channel environment. So, short of actually hiding the cable keypad, a parent cannot fully prevent exposure to all channels all of the time.2 This would prove especially true where auxiliary TV sets are wired to cable feeds. Even when a parent is present, the child may—-at least momentarily—— happen across an "adult" discussion on one of the basic channels. Despite this potential uncertainty, parents would remain reasonably certain that their children would not see "R"-rated adult programming. In the absence of that countervailing concern, basic households are less likely to regulate viewing than either pay or basic viewers. Pay channels: Pay households, of course, share many of the macro- and microsystem characteristics of their basic counterparts. However, where a pay channel is entirely or partially devoted to adult programming, it represents another "step-up" from basic in terms of sophistication. This added exosystem dimension should affect concerns regarding TV’s impact on children, thereby increasing the likelihood for intervention. Even where the actual change in programming is only slight, parental perceptions of TV’s ."new obnoxiousness" might be sufficient to motivate further intervention. 55 Thus, concerns regarding adult fare should override the heightened TV consumption motivation characteristic of other cable households. The second element, involving the logistical difficulty of mediating TV use, is also different in pay households. Here again, lockboxes greatly facilitate the process of restrictive intervention, as little effort is required to make 1-2 channels "off—limits" to children. Figure 2.3 depicts the relationship between TV viewing and the threshold for parental interaction. FIGURE 2.3: PREDICTED COMPARATIVE MEDIATION ACROSS SERVICES HI I Level of ! ---- media- 2 tion ! ---- LO 2 --—- ifiXfifi‘iéfiXi """"" ERIE ------ 133?"— Lowest level of Highest level of offensive content offensive content In light of this discussion, a hypothesis on restrictive mediation across viewing environments can be stated as follows. 56 H 3: Pay homes will report the greatest degree of restrictive mediation, followed by broadcast and finally basic cable homes. Subhypotheses on restrictive mediation are as follow following: H 3.1: Pay homes will report having to stop watching TV earliest on school nights, followed by broadcast and then by basic cable homes. H 3.2: Pay homes will report the greatest number of rules in the house about TV viewing, followed by broadcast and finally basic homes. H 3.3: Pay homes will report the greatest degree of parental disapproval for specific shows, followed by broadcast and finally basic homes. H 3.4: Pay homes will most often report that parents tell their children that they watch too much TV, followed by broadcast and finally basic homes. Nonrestrictive mediation: As the preceding discussion indicates, parental restriction of viewing will be greatest in homes where adult content is greatest and the prospects for control are least. In terms of the proposed model, program (exosystem) environments are relatively more discontinuous in pay cable homes. The fraction would hence be greater in those cases, and even overcome the degree of affinity (an anti- mediational force) those homes might otherwise feel for TV. In the absence of adult fare, that affinity would motivate lower mediation in basic homes relative to broadcast homes. Assuming that adult pay cable fare invites a greater degree of concern, several of the factors that encourage restrictive intervention may also prompt nonrestrictive 57 mediation. For instance, it is likely that parents may wish to clarify any "developmental eavesdropping" their children do with cable. Again, the basis for expecting greater concern regarding pay cable homes stems from their more disruptive program environment. Thus, one could hypothesize that the arrangement for TV service archetypes with proactive mediation would approximate that of restrictive mediation. H 4: Pay homes will most often report engaging in nonres— trictive mediation, followed by broadcast and finally basic homes. Subhypotheses can be stated as follow: H 4.1: Pay homes will most often report parental recommen- dations for certain shows, followed by broadcast and finally basic homes. H 4.2: Pay homes will most often report that parents change channels while viewing, followed by broadcast and finally basic homes. H 4.3: Pay homes will most often report that at least one parent watches with them, followed by broadcast and finally basic homes. H 4.4: Pay homes will most often report that a parent talks about shows when viewing, followed by broad— cast and finally basic homes. In providing a rationale for these rankings, one can note that similar content concerns again play a determining role in nonrestrictive mediation. In terms of pay homes, the exosystem defined by their programming should prompt parents to actively "decipher” TV’s meaning. This may be facilitated via coviewing, discussion about, or recommendations for programs. 58 While comparable group viewing situations may be found in basic and broadcast homes, they would not likely operate to the same extent. As the previous review of literature suggests, nonrestrictive mediation is practiced only in a minority of broadcast TV households (e. g. Mohr, 1979). Basic cable homes have been found to be even less concerned about the impact of TV (Haefner et a1., 1986) and more "at ease" with various contents (Baldwin and McVoy, 1983). In the absence of adult program settings, these viewers could be expected to actively mediate to a lesser extent than pay homes. 3 One can further posit that parental guidance would be greatest when the technology first arrives, as has been the case with other communications innovations (e. g. Wartella and Reeves, 1985). These concerns should dissipate as the family becomes more familiar with its channel environment. Of course, some restrictions would likely remain where premium adult services are present. But, as families become more familiar with basic or non-adult rental formats, mediation should approach or even drop below levels found in traditional households within similar strata. In this regard, TV-use motivations can overpower class or content-related factors prompting mediation in non-adult video environments. As mentioned, this should hold true despite the fact that many may-~by virtue of their high S.E.S. orientation-- 59 face social desirability norms favoring mediation. Those most concerned about the impact of adult content are, no doubt, screened out in the cable purchase decision. For them, real or imagined fears about new video fare figure in the decision not to purchase those technologies. With or without that restrictive sub-element, S.E.S. influences should predict both types of mediation across most every media exosystem.4 The typology in Figure 2.3 represents only relative levels of mediation--not absolute terms. Even though both forms of mediation vary across these settings, it is expected that parents will attach more absolute importance to restrictive practices, especially where adult programming is present. In other words, while some parents might not worry about being on hand to recommend or discuss certain programs, they may worry about not being on hand to prohibit the viewing of adult shows. Here again, these relative attributions of importance are not reflected on the above chart, but would prove most germane to pay cable homes. This is because the concerns that usually motivate restrictive mediation (e. g. for adult shows) are more pressing than those related to recommendations or coviewing. Of course, both methods may still be used when parents allow their children to view such content; nonrestrictive attempts to clarify pay cable programming could hence ameliorate any harms associated 60 with "developmental eavesdropping" in pay households. For reasons of convenience as well as moral concern, though, restrictive intervention should provide a focus for parental actions. Thus, as the earlier discussion indicates, pay cable presents an extreme case in terms of content and parent control. The general model suggests that, absent an overriding content concern, TV-use motivations will eventually overpower social desirability norms in other settings. There should not be as much nonrestrictive mediation among basic cable households, all things considered equal. 61 ENDNOTES 1. See P.L. 98-543, The Communications Act of 1984. For a discussion, see Klieman (1986). 2. It should be noted that encryption techniques for lockout devices are not, in most cases, likely to be learned by children. In cases where they are, codes can be changed by parent viewers. 3. The role of macrosystem influences, however, is not as clear. In general, one would not expect to find many differences with S.E.S. in terms of viewing restrictions, for reasons described earlier. There should, however, be a positive relationship with cable subscribership and nonrestrictive mediation. This stems from the fact that higher S.E.S. homes may, in many cases, subscribe to basic in order to receive the greatly enhanced offerings (e. g. Nickelodeon, U.S.A. cartoons, etc.). Thus, where viewers go to the trouble of paying money for such fare, they can be expected to recommend programs, discuss content, and view with their children. Relatively speaking, though, this should not exceed the level of nonrestrictive mediation present in broadcast homes. Again, absent any overriding concern over ”R" movies, one would expect that content motivations encouraging cable subscription could overpower any extensive mediational initiatives. In more general terms, higher SES households face a stronger class norm for restricting TV use, as mentioned earlier. They may also, as better educated individuals, be more aware of the importance of nonrestrictive mediation. Even so, these mediational tendencies run counter to the heightened consumption trends of new video users, who are also higher in SE8 (Krugman, 1985). While it would be instructive to isolate the effects of SES in the process of cable mediation, the homogeneity of most cable environments precludes such control; that is, cable users are typically ”self-selected" from among higher SES homes (Krugman and Eckrich, 1983). In the case of cable, lower income areas are rarely wired (Baldwin and McVoy, 1983), so we may never be able to fully examine cable use in urban environments, as researchers (Greenberg and Dervin, 1970) have done with broadcast TV. 52 4. As with other macro- and macrosystem variables, S.E.S. will be entered into a regression equation in line with theoretical expectations discussed in this section. They will not be the basis of specific hypothesis testing (described in Chapter 3). Since the focus of this research is on exosystem (i. e. pay v. basic v. broadcast) variables, hypotheses will only address those elements. Chapter 3: Methodology This chapter describes data collection and analysis procedures used to test the hypotheses outlined in Chapter 2. Those hypotheses reflect the goal of this study, which is to assess parental mediation behaviors across pay, basic and broadcast homes. In order to assure adequate statisti- cal power across the various TV service categories, it was necessary to collect data on a large scale. Since the topic of interest involved parent child viewing behaviors, a classroom survey emerged as the most appropriate data collection technique. And, while resource limitations favored the selection of a single community, it was felt that classroom surveys will allow for a degree of generalizability across other cabled communities. Such a technique should also provide consistency with past study designs addressing the area of parental mediation. Testingga model of parentalgmediation As Chapter 2 suggests, this study seeks to examine the role that viewing environments might play in mediation practices for children in early and late adolescence. The bulk of past work reveals that fewer than a quarter of all parents are likely to engage in mediation (e. g. Mohr, 1979). There were, nevertheless, interesting differences 63 64 across different home and social settings. In terms of the macrosystem, restrictive mediation was found to be higher in wealthier homes (Himmelweit and Swift, 1976), working mother or father-absent homes (Medrich et a1., 1982). Associated microsystem characteristics included smaller families (Webster, 1983) as well as having younger or female children (Barnes et a1., 1978). These are just a few examples; the specific mediation elements addressed in the research hypotheses included restrictive and nonrestrictive mediation. Additionally, viewing time and household TV use measures were introduced, as they have also been related to mediation practices in past work. Hypothesis testing The formal hypotheses addressed relations within the exosystem. As Hypothesis 1 suggests, pay cable viewers are likely to view the most TV, followed by basic and broadcast viewers. This much is apparent from past work; it stems from pay viewers’ greater affinity (and willingness to pay) for cable services prompts greater use. According to Hypothesis 2, consumption of "R" and "PC" TV movies would also be highest in pay homes, as a function of greater content availability in that environment. In terms of restrictive (Hypothesis 3) and non- restrictive (Hypothesis 4) mediation, it is expected that pay homes might express greater degrees of interest in mediating their child’s viewing. This activity would arise 65 from concern regarding the developmental impact of "adult— oriented" fare associated with that content environment. Lacking that concern, however, one might expect to find a different situation in basic homes. For, within that exosystem, the cable viewer’s greater affinity (and time allocated) toward the medium will result in less proactive and restrictive mediation than might be expected in broadcast homes. Further exosystem variables, as well as macro— and microsystem influences discussed in Chapter 2, will be discussed in conjunction with the data analysis. The following section describes how these variables will be operationalized. The survey instrument The questionnaire measures a host of media use behaviors which might potentially vary across different demographic and TV service archetypes. As mentioned, the key dependent variable encompasses restrictive and nonres- trictive mediation behaviors. The larger questionnaire also assessed items related to viewer knowledge of services available on cable, channel usage, innovativeness and ownership of other electronic media devices (e. g. "walkman" radios, personal computers and video games). Only those pertaining to TV consumption, patronage or mediation will be analyzed. As mentioned, the survey was administered to fifth 66 and tenth graders, as well as their parents. Except for the addition of an item on ethnicity for tenth graders, the two student survey instruments were identical. Opera- tionalizations of media use, mediation and background variables for that survey are included in Appendix 1. Those measures can be summarized as follow. Smmdent qmgstionnsire 1. Media use vsriables: Time spent with television was measured by series of dif- ferent daypart viewing measures. 1. Number of hours before supper a child spent watching TV (ranging by half hour intervals with a scale from 0=none to 12=more than 5.5 hours). 2. Number of hours a child reported spending with TV after supper, before bed (scaled by half hour intervals from 0=none to 12=more than 5.5 hour intervals). 3. Number of hours a child reported spending with TV before school (ranging by half hour intervals from 0=none to ll=more than 4.5 hours). II. Content availability 1. Frequency of R-rated TV movie viewership (ranging from 0=very often to 3=not at all). 2. Frequency of PG-rated movie viewership on TV (ranging from 0=very often to 3=not at all). This item was asked only of fifth graders. 67 III. Restrictiveimediation Restrictive mediation practices were measured by the following variables. 1. How late the child is allowed to watch TV on a school night (ranging by hour intervals from 9:00 p.m. to midnight on a four point Likert-type scale). 2. Number of TV rules in the house regarding viewing (ran- ging on a four-point scale (ranging from 0=yes, lots to 3=no, not really). 3. A measure of whether a parent says not to watch a given program (scaled along four points from 0=very-often to 3=not at all). IV. Nonrestrictive mediation 1. Frequency with which parent recommends show to watch (ranging from 0=very often to 3=not at all). 2. At least one parent watches with me (ranging from 0=very often to 3=not at all). 3. Talk about show with parent when viewing together (ranging from 0=very often to 3=not at all). 4. Frequency with which parents change channels (ranging from 0=very often to 3=not at all). V. Backgmomnd and media-use variables Non-content variables, mentioned in the hypotheses, reflect background factors. While not all are directly related to mediation behaviors, these factors should help obtain clues across different family and media settings. 68 They were operationalized as follow: Gender (0=male; l=female), age (open), number of younger siblings, number of older siblings (for each variable ranging from 0=none to 6:5 or more), number of people living at home (ranging from 1:1 to 10:10), ,number of parents living at home (0=none; 1:1; 2:2), father’s employment, mother’s employment (0=l; l=no, for each), cable TV subscribership status (0=yes, l=no, 3=former), VCR ownership (0=yes, l=no), whether the child has their own set (0=yes, l=no) and ethnicity (for tenth graders only, l=white, 2=black, 3=spanish american, 4=indian, 5=other). Rationales behind the inclusion of these and other variables are outlined in Chapter 2. Parent questionnaire Parent background and media use factors included cohort measures of cable subscription, VCR ownership and secondary TV ownership (0=yes, l=no). Pay cable subscription was also measured among adults. In terms of adult demographics, the following measures were used: sex (0=male; l=female), age (open ended), race (0=white, l=black, 2=oriental, 3=hispanic, 4=other), number of people living at home (open ended), number of children living at home (ranging in number from 0=none to 6=more than 5), occupation of working adults (open ended), household income (0=below $10,000 by $10,000 intervals to 7=$70,000 or over), last year completed in 69 school (ranging by one year intervals from 0=none to 12:12 and on up to some college [13], college graduate [14], graduate degree [15]). A comparable set of questions was asked of adults. For the purposes of this analysis, however, only those items that are redundant with child variables will be reported. These items can be summarized as follow. I. Restrictive mediation l. A measure of how often parents watch R~rated TV movies in (scaled from [0]"very often” to [3]"not at all"). 2. A measure of how often the parent works while watching TV (scaled from [0]"very often" to [3]”not at all"). 3. The number of rules about TV viewing the parent has for their child ("ranging from [0]"yes, lots" to [3]”no, not really”). 4. Whether actions were taken to prevent the viewing of particular channels or programs. 5. Whether children were allowed to watch R—rated movies on TV ([0]=yes and [l]=no). 6. Whether parents allow their child to watch PG-rated movies on TV ([0]=yes and [l]=no). 7. Whether the parents ever try to prevent the viewing of particular channels or programs ([0]=yes and [l]=no) 8. How late on a school night a child is allowed to watch TV (ranging by hour intervals from 9:00 p.m. [0] to "later”[4]). 70 II. Nonrestrictive mediation 1. When watching with your child, how often do you talk with him/her about the show (ranging from 0=very often to 3=not at all). 2. When you watch TV, how often is at least one teenager watching with you (ranging from 0=very often to 3=not at all). Datsgcollection The data were collected from a classroom survey which the researcher helped plan, design and implement as part of a multi-faceted project addressing children and cable. The larger study examined program content and consumption patterns across pay, basic, broadcast homes, and was directed by Dr. Bradley Greenberg and Dr. Thomas Baldwin, through the Department of Telecommunication at Michigan State University (MSU). After having been approved by university and school district officials, the survey instrument was pretested during the Spring of 1984. The pretest sample consisted of four classes of fourth graders from an elementary school in a medium-sized Midwestern community. As in the final data collection, research assistants from M.S.U. were available to help students complete the questionnaire. To ensure that children answered questions relevant only to their household-type, pages appropriate to each group were color coded. 71 On balance, the pretest group had little difficulty understanding the questionnaire. Items pertaining to home TV service-type created a degree of confusion among some of the respondents. These questions were taken into consi- deration in redesigning the survey. In particular, items pertaining to cable subscribership status were represented through the use of illustrations for various cable channels and converter boxes. These and other questions were incorporated into the final instrument, which was presented during November of the 1984—85 school year. The survey was administered in class to groups of fifth and tenth graders in a medium— sized Midwestern community. Research assistants from MSU were available to help students complete the question~ naire. To ensure that children answered questions relevant only to their household-type, pages appropriate to each group were color coded. A companion parent instrument was sent home with consenting students. Students were asked to have a custodial parent complete the cohort. To facilitate the prompt return of questionnaires, students were given pencils for returning it within three days. Some 421 fifth graders (aged 10-13) and 399 tenth graders (aged 14-16) completed the survey (along with 359 and 172 parent cohorts, respectively). Each respondent group filled out a similar survey, roughly 45 minutes in length, measuring 72 several aspects of TV viewing. In terms of respondent composition, the fifth grade, tenth grade and parent data sets revealed a fairly upscale picture. The majority of parents attended college and had a yearly household income in excess of $40,000. Penetra— tion levels for pay cable, basic cable, secondary TV sets and other video items were also above the national average. Specific student profiles are presented in Table l, and adult information is in Table 2. In terms of ethnicity, asked only of parents, fewer than 5% among fifth and tenth grade parents were nonwhite. This percentage was too low to warrant any meaningful statistical results. As a consequence, it will be dropped from any further analysis. Similarly, respondent frequencies were too low in the basic cable service category to warrant a sufficient degree of statistical power for tests involving parents. Data analysis The size and distribution of these data enable comparisons across pay, basic, and broadcast TV user archetypes. Interval and ratio-level response scales were incorporated to facilitate the use of multivariate statistics. Statistical breakdowns were done for each service type (i. e. pay, basic or broadcast) as well as for aggregate measures of sample data. Upon completion of the survey, data were keypunched 73 and stored on computer tape. Data were processed through the Statistical Package for Social Scientists program, version 10, on the Southern Illinois University I.B.M. computer system (3270 emulation program). Overview of data analysis Frequencies were run to gain an overall profile of the respondent groups. Information on cable services was derived from a subscription measure (cable or not) and-- among subscribers--a measure of pay subscription. Owing to the low response rate among male adults, the adult sex variable was weighted among tenth graders. Specifically, the total number of male responses was doubled (tenth grade parents) or tripled (fifth grade parents), bringing the total to a point where 50% of responses were from male respondents. MEN—8 Oneway analysis of variance (Anova) tests were then computed to compare reports of mediation across the various TV service groups. The test was used in is the most appropriate method in which to provide mean compari-sons when three or more contrasts are made. Specific contrasts for the viewing time and content availability hypotheses involved pay v. basic and basic v. broadcast only households. Hypotheses on restrictive and nonres- trictive mediation were facilitated with pay v. broadcast and broadcast v. basic contrasts. Companion parent 74 responses were subjected to contrasts between pay and broadcast homes. Contrasts involving basic homes among fifth and tenth grade parents were not run, owing to difficulties associated with low response rates among those groups. Original group sizes for pay, basic and broadcast groups were 118, 27, and 214 for fifth grade parents, and 67, 7 and 98 among tenth grade parents. In addition to the cell size concerns, wording among adult and child items threaten the validity of 2-way comparisons (with children and parents). Items where parent data are scaled differently will also not be directly compared. Means for both groups will hence be reported separately in those cases. Other media use tendencies where adult responses are available are also displayed. While different scales and wording preclude statistical comparison, it is possible to discern overall patterns among pay, basic and broadcast homes. Such information helps to establish the degree to which children approximate parental media use tendencies across different home video settings. Parental mediation behaviors may, of course, change as the child develops. So responses of fifth and tenth graders were further compared to check for different group means. Moreover, this comparative format was designed to provide information on how modeling tendencies for other media use might change over time. Specific mean compari- 75 sons across pay, basic and broadcast homes for fifth versus tenth graders were conducted with anova mean contrasts. It should be noted that differences in respondent group composition represent a threat to validity. The simultan- eous analysis of two groups is not as valid as a longitudi- nal analysis of one group over time. These date should, however, provide an acceptable measure of age group differences. Multiple regpession analysis After assessing this media use information, specific predictors of media use (including S.E.S., TV service and the like) were entered into a regression equation. This method will not be employed for hypothesis testing. It should, however, provide an added measure of robustness regarding the role of TV status in determining mediation and viewing. Most importantly, it represents the best method by which to gauge the relative contribution of cable status and other exosystem variables in comparison with traditional micro- and macrosystem locators. Placed in terms of the regression analysis, the four dependent variable areas would be viewing time, R and PG- movie viewership, restrictive mediation and nonrestrictive mediation. Though not a formal sub-category of mediation, the media use measure was included because of its close relationships with mediation (see Roberts, 1981). In order to break the data down into more manageable analytical 76 blocks, indices were created for like-scaled measures with viewing time, restrictive and nonrestrictive media‘ tion. Reliability tests were conducted for each item, using Cronbach’s alpha coefficients. The viewing time measure (alpha = .64) is a composite of the before school, after school and after- supper viewing measures. Restrictive mediation (alpha = .53) is assessed through an index based upon the measures of viewing rules, excessive viewing and the prevention of viewing. Similarly, nonrestrictive mediation (alpha=.50) is a composite of the items concerning parental recommendations, coviewing discussions and channel changing while coviewing. Measures of R and PG-rated movie viewing were not combined, as only the latter was assessed across both student groups. While the alphas measuring scale reliability could be stronger, they were judged sufficient for use in regression. Many of the other predictors represent the "usual suspects" with regard to parental mediation, as outlined in Chapter 2. In terms of the Bronfenbrenner continuum, documented relationships inhabit the macrosystem and microsystem domains. To this, one can also add exosystem variables, such as home cable status and VCR ownership. Prior to the initiation of regression analyses, Pearson Correlation tests were run. They were used to assess the degree of association for dependent as well as 77 independent variables. Multicollinearity was not deemed a concern among the independent variables, there were no correlations approaching .8. Correlation results for tests involving the dependent variables are included in Appendix 2. In testing the components of Bronfenbrenner’s larger model, the following variables (outlined in Figure 2.3) were entered into regression runs: child age, size, numbers of younger siblings, older siblings, parents living at home as well as people living at home, parental employment, income, education, VCR ownership, secondary set ownership, level TV service and viewing time. Owing to the exploratory nature of mediational research within the context of new media, no a priori hierarchical orderings were forwarded. The following section reviews the likely direction of influence of those variables within the framework of multiple regression and other analyses. Predictors of TV viewing and medistion This section reviews the rationales outlined in Chapter 2 in conjunction with the regression model. Recall that viewing time is likely to be positively predicted by family size and number of younger children. Similarly, the presence of younger siblings should predict greater restrictive and nonrestrictive mediation. Family size however, should emerge as a negative predictor in the regression equation. 78 The prediction model also suggests that parent access to children will be a determinant of TV viewing time. Single-parent and mother-absent homes will, thus, be predictive of less mediation. While the earlier mentioned cell-strength concerns prohibit inclusion of ethnicity into the regression equation, there other proxy measures of background status are worthy of consideration. It is likely that parental employment, for instance, would prove a positive predictor of restrictive and nonrestrictive mediation. Where anova contrasts provide the primary measure for Hypotheses 1—4, the level of TV service will also be entered into the regression equation. Here one might expect to see that level of TV service positively predicts viewing time (Hypothesis 1) and exposure to "adult” movie fare (Hypothesis 2) as well as restrictive (Hypothesis 3) and nonrestrictive mediation (Hypothesis 4). In terms of the relationships among the measures of viewing time and mediation, past work illustrates the value of examining each outside of their conventional roles as dependent variables. It is likely that restrictive and nonrestrictive mediation will emerge as inverse predictors of viewing time. Overview In sum, oneway Anovas should help ascertain differences across the different media environments. 79 Regression analysis should aid in the exploration of causal relationships. Taken together, analysis derived from these tests should help uncover differences in TV use and mediation behaviors across different parent, child and TV service groups. 80 TABLE 1: STUDENT BACKGROUND INFORMATION Attribute 1. Student characteristic Sex (male) Ethnic background White Black Spanish-American American Indian Other Proportion 5th .53 na na D8 na na II. Family characteristics Older siblings No older siblings One older sibling Two older siblings Three older siblings Four older siblings Five older siblings Six + older siblings Younger siblings ' No younger siblings One younger siblings Two younger siblings Three younger siblings Four younger siblings Five younger siblings Six + younger siblings Parents at home No parents at home One parent at home Both parents at home Parental employment Father employed Mother employed na: not available .38 .35 .07 .03 .02 .01 .37 .39 .16 .04 .01 .01 .01 .01 .13 .87 .97 .66 Proportion 10th .54 .92 .02 .01 .01 .04 .41 .16 .07 .03 .03 .02 .37 .39 .16 .05 .01 .01 .01 .01 .18 .81 .95 .75 (continued) 81 TABLE 1: STUDENT BACKGROUND INFORMATION (continued) Proportion Proportion Attribute 5th (n=421) 10th (n=399) III. Home video characteristics TV service Broadcast only .33 .36 Basic-only .15 .40 Pay .52 .24 Home computer .43 .37 Video recorder .32 .27 Video game .82 .62 Cordless phone .18 .22 IV. Media owned by student TV set .47 .56 Record or tape player .83 .87 Calculator .66 .85 Phone .27 .44 Walkman .65 .64 82 Proportion 10th .26 .89 .05 .01 .04 .01 .00 .07 .48 .29 .14 .01 .01 .03 .05 .14 .19 .25 .13 .08 .13 .04 .35 .27 .21 .13 (n=172) 10th parent mean 4.4 2.4 TABLE 2: PARENT BACKGROUND INFORMATION Proportion Attribute 5th (n=359) I. Individual characteristic Sex (male) .34 Ethnic background White .96 Black .01 Hispanic .01 Oriental .01 Other .01 II. Family characteristics Children at home None .01 One child .17 Two children .46 Three children .26 Four children .07 Five children .02 Six or more children .02 Household income Below $10,000 .02 $10,000 to $19,999 .04 $20,000 to $29,000 .14 $30,000 to $39,999 .25 $40,000 to $49,999 .28 $50,000 to $59,999 .18 $60,000 to $69,999 .05 $70,000 or over .04 Parent education Failed to finish H.S. .02 Up to high school .28 Some college .36 College degree .21 Graduate school .13 5th parent mean Family size 4.3 Children at home 2.3 Age 37 40 (continued) 83 TABLE 2: PARENT BACKGROUND INFORMATION (continued) Proportion Proportion Attribute 5th (n=359) 10th (n=172) III. Home video characteristics TV service Broadcast only .33 .39 Cable .07 .05 Pay .60 .56 Home computer .39 .36 Video recorder .25 .28 Video game .83 .65 Cordless phone .14 .22 Chapter 4: Results This chapter presents statistical results of hypo- thesis testing, along with further tests concerning inter- relationships among the dependent variables. For each contrast, a minimum significance was set at the .05 level.1 In terms of overall significance, results that are not statistically significant will be so designated by the sign "n.s." within statistical tables. Further information, including obtained levels of significance, degrees of freedom and other relevant data will also be included. Viewing,timg H 1: Pay homes will spend the most time with TV, followed by basic and finally broadcast homes. The first phase of testing for this general hypo— thesis involves one-way anova tests. Specific mean contrasts for adults involved pay and broadcast homes. Student contrasts for involved broadcast v. basic and basic v. pay homes (for viewing time and content availability measures). Analyses with restrictive and nonrestrictive mediation involved pay v. broadcast and broadcast v. pay contrasts. Significant paired contrasts will be indicated by superscripts (with "Q" designating statistically signi— ficant contrasts at the .05 level). 84 85 Each table will further report information on the overall F probabilities and degrees of freedom (df) involv- ed with specific contrasts. The df term will be divided into ”between group" and ”within group" terms. As mentioned in Chapter 3, the total number of subjects included 421 fifth graders, 359 fifth parents, 399 tenth graders and 172 tenth parents. Once again, the cell fre- quencies associated with each TV service group (pay,basic, broadcast) were as follow: fifth grade students (141,63, 217), fifth grade parents (118/27/214), tenth grade students (145/159/95) and tenth grade parents (67,7,98). H 1.1: Students from pay homes will spend the most time with TV before school, followed by basic and finally broadcast homes. Table 3 presents mean comparisons for measures of hours spent with TV before school among fifth graders. Obtained responses ranged from 0 to "more than 4.5 hours" among both student groups Table 3: Hours spent with TV before school Group pay basic brosdcast (F,,df) F prob. 5th children 1.04 1.02 1.18 (.58, 2/420) ns Absolute mean values suggest that viewing is actually highest among broadcast homes. None of these differences is significant, however. Hypothesis 1.1 is hence without 86 support among fifth graders. Table 4 details the situation for tenth graders. Table 4: Hours spent with TV before school Group pay, basic broadcast (F, df) F prob. 10th children .19 .22 .18 (.10, 2/397) us While absolute values suggest that TV viewing among cable homes is marginally greater than that among broadcast homes, with a peak among basic homes, these differences are not significant. H 1.1 is thus not supported among tenth grade students. H 1.2: Students from pay homes will spend the most time with TV after school and before supper, followed by basic and finally broadcast homes. A similarly constructed measure assessed the amount of time students spend with TV after school and before supper. Student responses on this item ranged from 0 to "more than 5.5 hours." The results of this test among fifth graders are included in Table 5. Table 5: Hours spent with TV before supper Group pay basic broadcast (F, df) F prob. 5th children 3.06 2.97 2.72 (.75, 2/417) ns While the absolute differences are in line with those predicted by Hypothesis 1.2, the differences among TV service groups are not statistically significant. The 87 hypothesis is not supported among fifth graders. Results among tenth graders are shown in Table 6. ——--\-—-‘_-----—-_——-------_-—-------_----—-_—_—-—-———--—- Table 6: Hours spent with TV before supper Group psy basic broadcast (F, df) F prob. 10th children 2.31 2.62 2.40 (.76, 2/398) us As with the "before school" time measure, viewing is greater among basic homes. Pay homes view the least TV. This, combined with the lack of any statistical signi- ficance, leaves H 1.2 without support. H 1.3: Students from pay homes will spend the most time with TV after supper, followed by basic and finally broadcast homes. As with the "before dinner" viewing item, responses ranged from 0 to ”more than 5.5 hours." Table 7 shows the mean distributions among fifth graders. Table 7: Hours spent with TV after supper Group pay bssjc broadcast (F. df) F prob. 5th children 4.68 5.03 4.89 (.50, 2/418) us As the data suggest, basic viewers are likely to view to the greatest extent. Since these mean differences are not statistically significant, H 1.3 is not supported among fifth graders. 88 Table 8 summarizes this relationship among tenth Table 8: Hours spent with TV after supper Group pay bssic broadcast (FLAdf) F prob. 10th children 3.78 4.02 3.45 (2.0, 2/396) us Here it appears that there are no significant differences in viewing time between basic viewers and their pay or broadcast counterparts and broadcast viewers. Absolute values suggest that basic homes view a quarter-hour more than pay homes and a half-hour more than broadcast homes. significant. The lack of any significant differences, however, leaves Hypothesis 1.4 without support among tenth graders. Content availability H 2.1: Pay homes will report watching R-rated movies to the greatest extent, followed by basic and finally broadcast homes. Responses to the item measuring R-rated movie viewer- ship ranged the full scale from 0 (very often) to 3 (not at all) for fifth and tenth graders.2 Table 9 provides TV service group means among fifth graders. Table 9: Viewership of R-rated movies Group pay, basic broadcast (F, df) F prob. 5th children 1.44 1.54 1.99 (12.3,2/418) .000* * Broadcast v. basic contrast significant at pg.05. As the data show, cable homes watch R-rated movies to a greater extent than broadcast homes. Absolute mean values are in line with theoretical expectations; only broadcast and basic homes differ significantly, however. Pay homes do not differ significantly from basic homes, though absolute mean values suggest that viewership of R-movies is higher with pay. Hypothesis 2.1 is partially supported. Table 10 presents this information for a parallel measure among tenth graders. Here again, it appears Table 10: Viewership of R-rated movies Group pay basic broadcast (F, df) F peob. 10th children .57 .53 .98 (ll.80,2/396).000* * Broadcast v. basic contrast significant at p$.05. that basic viewers are able to watch a significantly greater degree of R-rated movies than their broadcast counterparts. The lack of any such difference between pay and basic homes leaves this hypothesis with only limited 90 support. Further comparisons with R-movie viewing across fifth and tenth grade student groups will be reviewed subsequently. H 2.2: Pay homes will report watching PG-rated movies to the greatest extent, followed by basic and finally broadcast homes. The item assessing PG-movie viewership was asked only of fifth graders. Student responses ranged from 0 (very often) to 3 (not at all). Results are listed below in Table 11. Table 11: Viewership of PG-rated movies Group pay basic broadcast (F, df) F prob. 5th children .27 .24 .48 (5.1, 2/415) .006* * Broadcast v. basic contrast significant at pi.05. As is the case with R~movies, viewership of PG~movies is significantly greater among basic homes than broadcast homes. Pay homes, though, do not significantly differ from other groups-~50 Hypothesis 2.2 is partially supported. Restrictive mediation H 3 : Pay cable homes will report the greatest degree of restrictive mediation, followed by broadcast and finally basic homes. Specific operationalizations of this hypothesis ranged in scope from limitations on late-night viewing to 91 prohibitions on the viewing of specific programs. With the exception of one measure addressing late-night TV viewing, restrictive mediation concepts were also included in adult cohorts. Not all of these can be directly compared, however, owing to different measurement scales for each respondent group. In terms of analyses involving child- ren, table displays are re-ordered to reflect pay v. broad- cast and broadcast v. pay contrasts. H 3.1: Students from pay cable homes will report having to stop watching TV earliest on school nights, follow— ed by broadcast and finally basic homes. Obtained responses on this item ranged the full spectrum from 0 (8:00 p.m.) to 4 (midnight or later) among fifth and tenth graders. Fifth grade means are represented in Table 12 below. Table 12: How late child allowed to watch TV Group pmy broadcast pmsic (F, df) F prob. 5th children 1.67 1.74 1.83 (.77, 2/407) n.s. —_-—~-------—--_-n--—-——_—-—~——_——--—n—-~-—-———--~-a—‘-—_~—*—---- Absolute values are in line with Hypothesis 3.1, with basic homes exercising the least restraint. Since none of these differences is significant, the hypothesis is not supported. Responses among tenth graders are summarized in Table 13 below. Table 13: How late child allowed to watch TV Group pay broadcast,bssic 4(F, df). F prob. 10th children 2.96 2.72 2.93 (2.54,2/388) us As with the previous measure, there are no significant differences in late-night viewing, so Hypothesis 3.1 is unsupported. In terms of absolute values, basic homes stay up with TV later than their broadcast counterparts, watching a quarter-hour later on average. Emrentpsnd child groups: H 3.2: Pay homes will report the greatest number of rules in the house about TV viewing, followed by broad- cast and finally basic homes. Table 14 presents mean comparisons for measures of household rules for TV viewing. Responses ranged from 0 ("yes") for the presence of rules on up to 3, indicating that rules are ”not really” present for fifth and tenth grade students as well as parents. Here again, the adult contrast involved pay v. broadcast homes, owing to the low cell frequencies in the basic group. Degrees of freedom for adults represent a sample weighting in which male cases were duplicated to assure an even representation of sexes. Table 14: Presence of rules in the house for TV viewing Group psy brosdcast basic (F, df) F prob. 5th children 2.23 2.19 2.03 (1.2,2/411) ns As the above table suggests, student mean values are not in the direction hypothesized, and differences are not statis- tically significant. Table 15 shows parent means. Table 15: Rules regarding viewing should be enforced Grpmp ,psy broadcast basic (FLydf) F prob. 5th parents 1.22 1.41 1.05 (4.2,2/526) .015* ‘~‘~_--‘—*~-~*-‘-~~-*--”_-*_-_--~-_~-_*—~_-_--~_~_—_—_-—- Here it appears that pay homes are significantly more concerned about enforcing viewing rules than their broad- cast counterparts. This finding is in line with the expectations of H 3.2 for pay and broadcast homes. That support is qualified, however, as basic homes actually express the greatest concern over rules (though weak cell strength precluded contrasts with basic homes). The hypothesis is thus partially supported. Table 16 presents means among 10th graders. Table 16: Presence of rules in the house for TV viewing Group pmy broadcast basic (F, df) F prob. 10th children 2.71 2.61 2.73 (1.4,2/393) ns -———-__—-_-_-_‘_—_—-_-——————-_—..---—-_—--_‘-q————n——_¢———__‘-— 94 Here absolute mean values suggest tenth graders from broadcast homes are more likely to report rules than their cable counterparts. The lack of any significant differences leaves Hypothesis 3.2 without support. Tenth grade parent responses are in Table 17. Table 17: Rules regarding viewing should be enforced Group pay broadcast bssic (F, df) F prob. 10th parents 2.06 2.05 2.13 (.03,2/227) ns Here again, means are not significantly differentiated. Hypothesis 3.2 is thus lacking support from any of the fifth or tenth grade data groups. H 3.3: Pay homes will report the greatest degree of parental disapproval for specific shows, followed by broadcast and finally basic homes. As with the previous item, measures of parental dis— approval for specific programs were scaled differently for student and parent groups. Among students, responses ranged from 0 (very often) to 3 (not at all). Table 18 presents mean comparisons for fifth grade children. Table 18: Parental disapproval for certain shows Group psy broadcsst basic (F, df) F prob. 5th children 1.70 1.77 1.52 (l.94,2/418) ns As the data suggest, basic homes are somewhat less likely 95 to report prohibitions on programs, relative to broadcast homes. The significance (p=.058) of this difference is marginal, however. That, combined with the nonsignificant difference with pay homes leaves H 3.3 weakly supported. Parents, by comparison, were asked if they ever try to prevent viewing of particular channels or programs. Responses ranged from 0 (yes) to 1 (no). Table 19: Prevent viewing of particular channels or show Group pay broadcast basic (F, df) F prob. 5th parents .26 .34 .00 (4.3,2/423) .014 Here it seems that there is a difference approaching significance in the only mean contrast that was conducted (pay v. broadcast). Recall that mean contrasts involving basic homes were not run for parent data, owing to the low number of basic homes. The absolute value for basic homes is, however, high--contrary to the expectations of Hypo- thesis 3.3. That extreme basic value accounts for the significant F probability in spite of the nonsignificant pay/broadcast contrast. As expected, pay homes prohibit the viewing of specific content to a greater extent than broadcast homes. Table 20 shows measures of parental disapproval for certain shows among tenth graders (using scales identical to those for fifth graders). Table 20: Parental disapproval for certain shows Group pay broadcast basic (F, df) F prob. 10th children 2.44 2.37 2.35 (.54,2/398) ns Among tenth grade children, the group means for pay, basic and broadcast homes were not significantly different. Table 21 details data with tenth grade parents. Table 21: Prevent viewing of particular channels or shows Group pay broadcast basic (F, df) F prob. 10th parents .46 .61 .20 (2.0,2/200) ns As Table 21 shows, there are no significant differences across the TV service groups. Absolute mean values suggest that, as with fifth graders, pay homes are more likely to prevent the viewing of specific shows, relative to broad- cast viewers. The lack of significance with this relation- 1eaves Hypothesis 3.3 without support. H 3.4: Pay homes will most often report that parents tell their children that they watch too much TV, fol- lowed by broadcast and finally basic homes. This hypothesis is assessed by measures of parental expressions of concern regarding TV viewing time. Here again, the parent item differed somewhat in wording, though each was scaled on a four-point scale--with 0=very 97 often (student) or strongly agree (parent) through 3=not at all (student) or strongly disagree (parent). Obtained responses ranged across each of those categories. Table 22 shows mean comparisons for reported parental limitations on viewing time. Table 22: Parent expressions over excessive TV viewing Group I psy brosdcsst basic (F, df) F prob. 5th children 1.80 1.75 1.63 (.60,2/416) n.s. As with prohibitions on viewing, the lack of any signifi- cant differences leaves the hypothesis without support among fifth graders. Absolute mean values reveal a ten~ dency for basic homes to report more expressions of parental concerns then their counterparts. Table 23 details this irelationship among parents of fifth graders. Table 23: Concern with viewing over set number of hours Group psy broadcast basic (F, df) F prob. 5th parents 1.31 1.46 '1.09 (6.6,2/526) .016* Here the contrast involving pay and broadcast homes is statistically significant, with the former expressing a greater degree of concern about viewing hours. While the low number of basic respondents precluded a valid contrast 98 involving that group, overall mean values suggest that they are even more concerned than pay homes. The ordering of TV service archetypes leaves Hypothesis 3.4 unsupported among fifth graders’ parents. In terms of tenth graders, Table 24 details the responses concerning reported parental disapproval for certain shows. Table 24: Parent expressions over excessive TV viewing Group pay broadcsst basic (F, df) F prob. 10th children 2.42 2.19 2.23 (2.40.2/397) ns Here there are no significant differences, though it seems pay homes are less likely to express concern over excessive TV viewing, relative to basic and broadcast homes. Hypothe- sis 3.4 is not supported by the tenth grade student data. Table 25 describes parental concerns that their child should not be allowed to watch TV more than a set number of hours per week. Table 25: Concern with viewing over set number of hours Group pay broadcast basic (F, df) F prob. 10th parents 1.65 1.87 1.49 (2.77.2/222) .06* * Pay v. broadcast contrast significant at pg.06. From the table, broadcast homes are significantly more likely than pay homes to express concern over the time 99 their child spends with TV. This is not in line with model predictions. Hypothesis 3.4 is therefore without support among these parents. Nonrestrictive,mediation H 4 : Pay homes will most often report engaging in nonres- trictive mediation, followed by broadcast and finally basic homes. Specific measures of nonrestrictive mediation range from parental recommendations to having discussions of content while viewing. Two of the items have parent cohorts--measures of coviewing and discussions about shows. Among these items, Anova runs for students and parents are similar to those for Hypotheses 3.1-3.3. H 4.1: Pay homes will most often report parental recommen- dations for certain shows, followed by broadcast and finally basic homes. The item measuring parental recommendations, asked only of students, was scaled from 0 (very often) to 3 (not at all). Obtained responses span each of those categories. Table 26 details means across TV service categories. Table 26: Parent recommends show to watch ngmp psy broadcast basic (F, df) F prob. 5th children 1.86 1.83 1.73 (.65,2/418) n.s. 100 By looking at absolute values, it would appear that basic homes are most likely to engage in parental recommenda- tions. This, combined with the lack of any statistically significant differences, leaves Hypothesis 4.1 unsupported. Table 27 presents the means among tenth graders. Table 27: Parent recommends show to watch Group pay broadcast basic (F, df) F prob. 10th children ‘ 2.12 1.99 2.15 (2.28,2/398) ns The tenth grade data reveal no significant differences in parental recommendations about what shows to watch. This leaves Hypothesis 4.1 without support. H 4.2: Pay homes will most often report that parents change channels while viewing, followed by broadcast and finally basic homes. Measures of parental channel changing while coviewing were included only in student questionnaires. Obtained responses spanned the full range from 0 (very often) to 3 (not at all). Table 28 details mean differences among fifth grade children. Table 28: Parents change channels while coviewing Group pay broadcast basic (F, df) F prob. 5th children 1.51 1.61 1.52 (.58,2/418) n.s. 101 Here there are no significant differences in reported parental channel changing behaviors among fifth grade subscriber groups. Hypothesis 4.2 is hence not supported. Table 29 details tenth grade responses. —-—_-----_-—_—---—--—-—-_——--————_--——--——-—-—_—--—-_-—-u Table 29: Parents change channels while viewing Group pay broadcast basic (FL df) F prob. 10th children 1.71 1.40 1.58 (5.86,2/398) .004* * Broadcast v. basic contrast significant at pi.05. From the table, it appears that broadcast viewers are significantly more likely than basic viewers to have their parents control channel changing behaviors. Though not evidence of a linear trend, this reduced control in broad~ cast homes provides partial support for Hypothesis 4.2. Parent and child comparisons H 4.3: Pay homes will most often report that at least one parent watches with them, followed by broadcast and finally basic homes. The concept of parental coviewing is measured by student and parent items, each of which has identical scales (0=very often to 3=not at all). Answers for both groups ranged across each of the response categories. Fifth grade student means are in Table 30. Table 30: Parent views TV with child Qmpmp pay broadcast basic (F,,df) F prob. 5th children 1.70 1.77 1.52 (1.9,2/418) n.s. 102 As the data suggest, there are no significant differences in 5th grade reports of coviewing. Absolute values suggest broadcast homes are less likely to view with their children. As with the previous measure, this contradicts the expectations of Hypothesis 4.3. Table 31 details this relationship among parents of fifth graders. Table 31: Parent views TV with child Group pay broadcast basic (F, df) F,prob. 5th parents 1.02 1.14 1.23 (3.14,2/530) .044* * Pay v. broadcast contrast significant at p i .05. Here it appears that pay parents view with their children to a significantly greater extent than broadcast viewers. Absolute values suggest that basic viewers are least likely to view with children--though cell strength is not sufficient to. provide a valid contrast with that group. Hypothesis 4.3 is hence supported with reservation among parents of fifth graders. Table 32 details mean differences by TV service type among tenth graders. Table 32: Parent views TV with child Group pay broadcast basic (F, df) F prob. 10th children 1.71 1.40 1.58 (5.86,2/398) .000* * Pay v. broadcast v. basic contrast significant at pi.05. 103 The data indicate a significant relationship, with broad- cast homes coviewing to the greatest extent, followed by basic and finally broadcast homes. On the whole, though, this lends partial support to Hypothesis 4.3’s prediction of less coviewing activity in basic homes, relative to broadcast homes. In terms of parents of tenth graders, Table 33 presents parent reports of coviewing. Table 33: Parent views TV with child Group pay broadcsst basic (F, df) F prob. 10th parents 1.37 1.16 .87 (3.57,2/227) .03* * Broadcast v. pay contrast significant at pg.05. The data show that broadcast homes coview to a greater extent than their pay counterparts. And, while the cell strength is not sufficient to warrant testing with basic homes, absolute values suggest that they coview to an even greater degree. In each case, the tenth grade parent data do not support the relationship posited in Hypothesis 4.3. H 4.4: Pay homes will most often report that a parent talks about shows when viewing, followed by broadcast and finally basic homes. Measures of parental discussion while viewing were included in both the parent and student questionnaires. In each case, obtained responses spanned the full range from 0 (very often) to 3 (not at all). Table 34 details mean 104 differences among fifth grade children. Table 34: Parent discussion while viewing Group psy broadcsst basic (F, df) F prob. 5th children 1.64 1.52 1.60 (.84,2/417) n.s. Absolute values suggest broadcast homes are more likely to engage in discussions while viewing. There aren’t any significant differences between pay, basic and broadcast homes in terms of discussions while viewing, however. Hypothesis 4.4 is hence not supported among fifth graders. Table 35 shows data from fifth grade parents. Table 35: Parent discussion while viewing Group pay broadcast basic (F, df) F prob. 5th parents 1.21 1.21 1.20 (.80, 2/533) n.s. As is the case among fifth grade students, there are no significant differences among means. Hypothesis 4.4 is hence not supported with this group. Table 36 details mean differences by TV service type among tenth graders. Table 36: Parent discussion while viewing Gmpmp, pay broadcast basic (F, df) F prob. 10th children 1.57 1.41 1.58 (2.13,2/397) ns —-_--————-————-—_-_——-———————-——_————_—————-——————_-‘-_-— 105 As with the coviewing measure, the data indicate that broadcast homes are significantly more likely than pay homes to discuss program content with their children. Using a slightly expanded significance range, the data contradict Hypothesis 4.4. In terms of parents of tenth graders, Table 37 presents mean contrasts for cable and broadcast homes. Table 37 Parent discussion while viewing Group pay broadcast basic (F, df) F prob. 10th parents 1.41 1.15 1.00 (3.94.2/225) .02* * Broadcast v. pay contrast significant at p$.05. Here, as with tenth grade children, the data reveal significant mean differences. Yet, in this case, it appears as if pay homes are significantly less likely to engage in program discussions, relative to their broadcast counterparts. Hypothesis 4.4 is thus not supported among tenth grade parents. Index means and comparisons for for student groups Significance tests for differences between fifth and tenth graders by TV service group are included in Table 38. In order to simplify those comparisons, a set of indicators was constructed from the measures of viewing time, restrictive mediation and nonrestrictive mediation. The viewing time measure is a composite of the before school, after school and after supper viewing measures reviewed in 106 Tables 3 to 8. Theses indexes are described in Chapter 3. Recall that PG and R-movie viewing were not combined, however, as only the latter was assessed across both student groups. Fifth and tenth grade comparisons are thus based upon the R-movie viewership measure. Table 38: Index means and comparisons for student groups Variable Pay Basic Broadcast 1. Viewing time 5th grade 2.92 2.97 2.91 10th grade 2.09 2.29 2.01 (p<.001; df=1/308) (p<.01 ;df=l/218) (p<.0 1;df=1/281) II. R-rated exposure \ 5th grade 1.44 1.54 1.99 10th grade ' .57 .53 .98 (p<.001; df=l/309) (p=. 00;df=l/220) (p<.0 1;df=1/280) III. Restrictive mediation 5th grade 1.90 1.74 1.90 10th grade 2.53 2.44 2.39 (p=.000; df=l/303) (p=.000;df=l/213) (p<.000;df=1/279) IV. Nonrestrictive mediation 5th grade 1.64 1.59 1.60 10th grade 1.76 1.73 1.67 (p=.034; df=l/302) (p=.05; df=1/220) (p=.15; df=l/279) p= F probability; significant values are underlined. —_————————-—————————-———-————————-o—e.——————--——_—--——-—————_———.~— 107 With the exception of nonrestrictive mediation (broadcast homes), all of the measures were significantly differentiated between fifth and tenth graders. As expected, fifth graders spend more time with TV, though watch fewer R-rated movies. That latter finding may be a consequence of the enhanced restrictive mediation which fifth graders report. The fact that both restrictive and nonrestrictive mediation were generally higher among young- er students confirms expectations. It would seem that the same dynamic motivating restrictive mediation is operative with nonrestrictive mediation in pay and basic homes. Regression anslysis Differences between TV service archetypes aside, remaining exosystem variables were entered into a regres- sion equation alongside microsystem and macrosystem measures. As outlined in the preceding chapter, specific microsystem predictors included child age, sex, family size and number as well as age of siblings. Macrosystem pre- dictors included number of parents, their marital status, income, education, age and employment. In terms of addi- tional exosystem variables, cable subscribership status was entered alongside secondary TV set ownership, use of child TV sets, and VCR ownership. Owing to the lack of correspondence between parent and child background mea- sures, only the child variables will be analyzed.3 108 Criterion variables in the regression included compo- site indexes constructed from the measures of viewing time, restrictive mediation and nonrestrictive mediation (as described in the previous section). Restrictive and nonrestrictive mediation measures were. entered into the equations predicting viewing time. This was based upon the expectation that practices of mediation are likely to negatively influence time spent with TV. When interpreting these results, it should be borne in mind that content and mediation variables in this construct have reverse scales --unless otherwise noted (with 0=very often and 3=not at all). To avoid counter-intuitive beta weight displays, signs were inverted where a reverse-scaled predictor was used on a positively scaled criterion variable, or visa versa. Once again, those measures that are scaled differently or otherwise conceptually distinct from index measures will be presented separately. Predictorsmof viewing tims On the whole, predictor variables were only able to account for a modest degree of variance. As in Table 38, the three viewing measures (before school, after school/be- fore supper, and after supper) were combined into a single viewing time index. Table 39 reports standardized beta weights for each of the significant predictors in the regression equation for viewing time among fifth and tenth graders. Table 39: Weekday TV viewing 5th grade: R Square =.277, p = .000 Predictors: Standardised Beta p R-movie viewing .23 .000 Excessive viewing concern .30 .000 Late night viewing .21 .004 TV rules -.15 .008 Female child .14 .011 10th grade: R Square =.191, p = .000 Predictors: Standardised Beta Excessive viewing concern .33 .000 Homework with TV .25 .002 Among fifth graders, R-movie viewing and late-night viewing are predictors of TV viewing before school, as is concern over excessive viewing (both were scaled in reverse fashion). In terms of the latter, families that are more concerned about excessive viewing are (paradoxically) likely to report greater child viewing. TV rules are, however, an inverse predictor of viewing time; that is, families with more rules are likely to view less TV. And finally, female sex is a predictor of viewing time (i. e. females watch more). All told, the prediction equation accounted for 28% of the variance in viewing time. For tenth graders, concern over excessive viewing is, again, a predictor of TV viewing among tenth graders. Doing homework while watching TV is also a predictor. Thus, viewing is likely to be greater in homes where parents are concerned about viewing and children are allowed to do homework while viewing. The prediction 110 equation accounted for 19% of the variance in viewing time among tenth graders. Predictors of content availability The same set of background variables was regressed against measures of content availability. Only background variables were entered into the prediction equations. Table 40 outlines viewership of R-rated movies among fifth graders. Table 40: Viewership of R-rated movies 5th: R Square =.25, p = .000 Predictors: Standardised Beta p TV after supper .27 .000 Cable subscription .29 .000 Child age .14 .002 Female child —.14 .002 Child set .10 .040 Older siblings .11 .036 VCR .09 .000 10th graders: R Square = .07, p = .000 Predictors: Standardized Beta p Cable status .17 .000 Child set .15 .000 Child age .12 .02 ———————--—————u—-——-————.————_————————————_—-——-¢.———-————~n_—_— For fifth graders, several variables emerge as predictors of R—movie viewership, collectively accounting for 25% of the variance. From the table, having broadcast TV is a negative predictor (or subscription to cable predicts) R- movie viewership. Time spent with TV (after supper) is also a powerful predictor, along with age, male status, child TV ownership, having more older siblings and family lll VCR ownership. Three of those fifth grade variables are also pre- dictors among tenth graders. Specifically, R-movie viewing was predicted by subscribership to cable services, child set ownership, and child age. Relative to fifth graders, a comparatively modest proportidn of variance (7%) is explained by this predictor set. Table 41 outlines regression results for PG—movie viewership, which was assessed only among fifth graders. Table 41: Viewership of PG-rated movies 5th: R Square = .037, p = .001 Predictors: Standardised Beta p TV after supper .20 .000 TV before school -.10 .04 —————————_——————.——-—_——ul——-—-———-———-—————————————-.————-—— As the results indicate, time spent with TV after supper is a predictor of PG movie viewership. Time spent with TV before school, however, is an inverse predictor. Only a modest degree of variance in PG-movie viewing is explained by the predictors. Predictors of restrictive mediation Background predictors entered for restrictive media- tion were identical to those used in the viewing time and content patronage equations. Measures of viewing time were not included in the prediction equation, as they were 112 deemed more a result than a cause of parental concerns motivating such action (for a discussion, see Chapter 2). Regression results for parental requests not to watch a particular program are described in Table 42. Table 42: Restrictive mediation 5th grade: R Square =.042, p = .006 (continued) Predictorsp Standardised Beta p Child TV set —.15 .015 Parent education .13 .026 10th grade: R Square =.055, p = .000 Predictor: Standardised Bets p Female child sex .23 .007 Among fifth graders, children with their own TV are less likely to report restrictive mediation (which was scaled in a reverse direction). Also, those with better educated parents report more such mediation. The amount of variance explained in these cases, however, is a marginal 4.2%. For tenth graders, female child sex emerges as the sole significant predictor of restrictive mediation. This parallels findings with viewing time (Table 39), suggesting that child sex exerts a similar influence on viewing and mediation measures. Aside from the index findings, Table 43 outlines child reports of parental intervention in the item addres- sing channel switching. Data on this item are available only from tenth graders. Table 43: Parents tell you to stop switching channels 10th grade (only): R Square = .045, p = .000 Predictors: Standardised Beta p Cable subscribership .16 .00 Child age -.11 .02 Parents -.11 .04 Here it is apparent that subscribership to cable services predicts parental intervention in channel switching. Nega- tive predictors include child age and number of parents. 4.5% of the variance is explained by these predictors. Late night viewing is scaled in a positive fashion, with higher values indicative of later viewing. Table 44 details responses among fifth and tenth graders. Table 44: Late night TV viewing 5th grade: R Square =.086, p = .000 Predictors: Standardized Beta p Child set - .17 .000 Child age .15 .000 Young siblings -.10 .04 Old siblings .09 .048 10th grade: R Square = .046, p = .000 Predictors: Standardised Beta p Child Age .17 00 Child set .15 00 ———-————-—-_-_——————_———-——_———‘———__-———_--—_— Among fifth graders, late night TV viewing is child set ownership, child age and having siblings. Having more younger siblings is as predictor of viewing. explained by the predictors. predicted by more older a negative_ All told, 8.6% of the variance is 114 For tenth graders, age and child set ownership are positive predictors of late night viewing. A modest 4.6% of the variance is explained by age and set ownership. Predictors of nonrestrictive medistion The same set of predictors for restrictive mediation were entered into the regression equation for nonrestric- tive mediation. To summarize, the criterion index is- comprised of measures for parental recommendations, coview- ing, discussions and channel changing while coviewing. Table 45 presents results for the index measure of nonrestrictive mediation among both student groups. Table 45: Predictors of nonrestrictive mediation 5th grade: R Square = .032, p = .003 Predictors: Standardised Beta p Income .18 .003 10th grade: R Square = .053, p = .008 Predictors: Standardised Beta p Child set .23 .008 Among fifth graders, parent income is the sole significant predictor of mediation. Thus, it seems that wealthier homes are more likely to engage in nonrestrictive mediation. Only 3% of the variance is explained by this equation, however. A slightly higher proportion of variance is explained among tenth graders (5%). In this case, child set owner- ship is a negative predictor of nonrestrictive mediation. 115 That is, children with their own set report more such mediation. On balance, the regression results reveal a mixed bag, from the perspective of data explanation. Variance explained by the various prediction equations is greatest among the measures of viewing time, even after the varying number of predictors is taken into account. In no cases, however, was more than 30% of the variance explained. The implications of these findings, in light of the model advanced in Chapter 2 will be reviewed in the chapter to follow. 116 ENDNOTES 1. This suggests that 5 times out of 100 one could expect to see differences occurring as a function of chance. 2. It should be noted that these measures are a remote proxy measure of movie viewership. When reading the findings, the observer should note that respondent self reports suffer from two threats to validity: (1) some many not have been certain that they had basic or pay service, hence movie descriptions are confused, and, more important- 1y, (2) respondents may think that the ”edited down" R- movie that they see on basic or broadcast TV are the same as the "uncut" versions on pay TV. The implications of those validity threats are further discussed in Chapter 5. 3. A table of correlations among dependent variables is included in Appendix 2 Chapter 5: Summary and Discussion This chapter discusses the implications of hypothe- sis testing from the preceding chapter. After reviewing those findings, reasons for the apparent strengths and weaknesses of the proposed models are discussed. Results will then be examined in light of those from the further regression analysis. Where the findings are not as expect- ed, shortcomings in the conceptual model, internal validity and external validity are noted. Based upon this discus- sion, social implications of findings will be explored, along with avenues for later work. It is important to consider past work when reviewing the implications of this dissertation. Parental mediation research is now a fairly mature subdiscipline within the media effects area. Previous studies have aptly outlined the role of household (microsystem), socio-economic (macro- system), viewing time and parental mediation variables in child development. Yet none of the published literature has assessed the impact of pay, basic and broadcast viewing environments on parent-child viewing and interaction pat- terns. In examining that impact, this study imported a model of ecological niches first forwarded by Bronfenbren- ner (1979). Pay cable, basic cable and broadcast settings 117 118 ner (1979). Pay cable, basic cable and broadcast settings were each cast as different exosystem environments. It was felt that this new exosystem construct cbuld explain mediation and viewing patterns alongside the previously documented macro— and microsystem predictors. General model propositions suggest that pay viewers would view TV to the greatest extent, followed by basic and broadcast homes. This prediction is based on past work with adult viewing patterns, and stems logically from cable viewers’ greater affinity toward the TV medium. The same order was expected with regard to R-movie viewership, owing to the content availability associated with each exosystem (Krugman, 1985). In response to that more "discontinuous" new video exosystem, pay homes were expected to be most active in terms of restrictive and nonrestrictive media- tion. Basic homes, though, were expected to mediate less than their broadcast counterparts; that is, content con- cerns among basic homes were not deemed adequate to override the greater use tendencies associated with cable homes. Table 46 below summarizes the directional tendencies apparent throughout hypothesis testing for viewing time, R and PG movie viewership, restrictive mediation and nonrestrictive mediation. Following that summary, results from anova tests and supplementary regression analyses will be discussed for each criterion variable area. 119 Table 46: Summary of findings for mediation and viewing Variable 5th child 5th Parent 10th Child 10th Parent 1- Viewing time (hypothesis: P>BA>BR) ——————————————————— Before school n.s. n.a. n.s. n.a. Before supper n.s. n.a. n.s. n.a. After supper n.s. n.a. n.s. n.a. II. R-rated exposure (Hypothesis: P>BA>BR) ————————————— R—movie BA>BR n.a. BA>BR n.a. PG-movie BA>BR n.a. n.a. n.a. III. Restrictive mediation (hypothesis: P>BR>BA) ——————— Late TV n.s. n.a. n.s. n.a. TV rules n.s. P>BR n.s. n.s. Prevent viewing n.s. n.s. n.s. n.s. Excessive view n.s. P>BR n.s. P>BR IV. Nonrestrictive mediation (hypothesis: P>BR>BA) ------ Recommendation n.s n.a. n.s. n.a. Channel change n.s. n.a. BRBR BA