Portuguese validation of the Positive Youth Development Short Form (PYD-SF) : a shorten version

PSICOLOGIA, SAÚDE & DOENÇAS, 2018, 19(3), 477-489 ISSN - 2182-8407 Sociedade Po uguesa de Psicologia da Saúde - SPPS - www.sp-ps.p DOI: h p://dx.doi.o g/10.15309/18psd190301 www.sp-ps.p 477 PORTUGUESE VALIDATION OF THE POSITIVE YOUTH DEVELOPMENT SHORT FORM (PYD-SF): A SHORTEN VERSION Ma ga ida Gaspa de Ma os1, Te esa San os1, Ma a Reis2, Adilson Ma ques2, Gina Tomé2, Celes e Simões2, & No a Wiium3 1 Faculdade de Mo icidade Humana (Equipa A en u a Social), Uni e sidade de Lisboa, Po ugal; Ins i u o de Saúde Ambien al, Faculdade de Medicina, Uni e sidade de Lisboa, Po ugal; William James Cen e o Resea ch, ISPA - Ins i u o Uni e si á io, Lisboa, Po ugal, [email p o ec ed]a.p , [email p o ec ed] 2Faculdade de Mo icidade Humana (Equipa A en u a Social), Uni e sidade de Lisboa, Po ugal; Ins i u o de Saúde Ambien al, Faculdade de Medicina, Uni e sidade de Lisboa, Po ugal, [email p o ec ed], ama ques@ mh.ulisboa.p , [email p o ec ed], [email p o ec ed] 3Depa men o Psychosocial Science, Facul y o Psychology, Uni e si y o Be gen, No way, [email p o ec ed]o ___________________________________________________________________________ ABSTRACT: The Posi i e You h De elopmen -PYD concep highligh ed he impo ance o s eng hening in e nal and ex e nal de elopmen al asse s in he you h’s con ex s. This pape examines he psychome ic p ope ies o PYD-Sho Fo m (PYD-SF) and i s alida ion o Po uguese you hs. The sample included 2700 college s uden s (73.3% gi ls), wi h a mean age o 21.3±2.79 yea s old. Explo a o y-EFA and Con i ma o y Fac o ial Analysis-CFA we e conduc ed on he 34 i ems o he PYD-SF, a inal sho en ins umen (20 i ems) was achie ed, and in he inal model, CFA displayed an accep able model i (χ2 (174)=1349.20; GFI=0.95; NFI=0.95; CFI=0.95; and RMSEA=0.05, CI(RMSEA)=0.05-0.06). The de eloped measu e showed adequa e psychome ic p ope ies ( alidi y and eliabili y) o es ima e PYD in Po uguese you hs. Such ins umen can be an ad an age o i s adminis a ion in la ge popula ion samples and a suppo o decision-making in p o essional p ac ices and policies, encou aging a shi in you h ca e se ices app oach: om a pe spec i e ocused on de ici s/di icul ies, owa ds a iewpoin ocalized on possibili ies/oppo uni ies. Keywo ds: adolescen s, de elopmen al asse s, i e cs, po ugal, posi i e you h de elopmen sho o m (PYD-SF), you h ___________________________________________________________________________ VALIDAÇÃO PORTUGUESA DO POSITIVE YOUTH DEVELOPMENT SHORT FORM (PYD-SF): UMA VERSÃO REDUZIDA RESUMO: O concei o de Desen ol imen o Posi i o em Jo ens (Posi i e You h De elopmen -PYD) des acou a impo ância de o alece as “ o ças” in e nas e ex e nas do desen ol imen o nos con ex os da ju en ude. Es e a igo examina as p op iedades psicomé icas de PYD-Sho Fo m (PYD-SF) e sua alidação pa a jo ens Po ugueses. A amos a incluiu 2700 es udan es uni e si á ios (73,3% apa igas), com idade média e des io- pad ão de 21,3±2,79 anos. Aos 34 i ems do PYD-SF oi aplicada a Análise Fac o ial Explo a ó ia (AFE) e Análise Fac o ial Con i ma ó ia (AFC), ob endo-se um ins umen o inal mais eduzido (com 20 i ens). No modelo inal, a AFC ap esen ou um modelo ajus ado acei á el (χ2 (174) = 1349,20, GFI = 0,95, NFI = 0,95, CFI = 0,95 e RMSEA = 0,05, CI Es ada da Cos a 1495-688 C uz Queb ada, Po ugal. e-mail: gaudi_ @ho mail.com Ma ga ida Gaspa de Ma os, Te esa San os, Ma a Reis, Adilson Ma ques, Gina Tomé, Celes e Simões, & No a Wiium www.sp-ps.p 478 (RMSEA) = 0,05-0,06). O ins umen o desen ol ido demons ou p op iedades psicomé icas adequadas ( alidade e con iabilidade) pa a es ima o PYD em jo ens Po ugueses. Tal ins umen o pode cons i ui -se uma an agem pa a a sua adminis ação em amos as populacionais maio es e um supo e pa a a omada de decisões nas polí icas e p á icas p o issionais, incen i ando uma mudança na abo dagem dos se iços de p es ação de cuidados à ju en ude: de um pon o de is a ocado em dé ici s/di iculdades, pa a uma pe spec i a cen ada em possibilidades/opo unidades. Pala as-cha e: adolescen es, “ o ças” do desen ol imen o, Cinco Cs, Po ugal, Desen ol imen o Posi i o, Posi i e You h De elopmen Sho Fo m (PYD-SF), Jo ens ___________________________________________________________________________ Recebido em 23 de Janei o de 2018/ Acei e em 25 de Se emb o de 2018 In he pas cen u y he s udy o young people has been essen ially ou lined by a “de ici pe spec i e”, and wi hin his app oach, he concep o posi i e de elopmen was simply cha ac e ized by he absence o he dec ease o p oblems (Geldho e al., 2014). This pe spec i e in luenced policies, esea ch and p ac ice ha we e in o med by measu es used by p og ams and se ice o ganiza ions o assess you h unc ioning ha was mainly based on isk and p oblem beha iou s. Howe e , o e he pas wo decades, he s udy o posi i e indica o s o child well-being ha e signi ican ly inc eased (Lippman, Moo e, & McIn osh, 2011). Addi ionally, i was la gely ecognized in he li e a u e ha you h p og ams and in e en ions would ha e limi ed impac i hey mainly ocus on isks and ulne abili ies. Thus, s eng h-based app oaches ha e been ound o be empowe ing and e ec i e in a ious con ex s (Benson, Scales, Hamil on, & Sesma, 2006) and he in e es in p e en ing you h p oblems and p omo ing heal hy you h de elopmen has led p ac i ione s, policy make s and esea che s o de elop a wide ange o such app oaches elying on se e al heo e ical amewo ks (Small & Memmo, 2004). Posi i e You h De elopmen -PYD has highligh ed he impo ance o s eng hening in e nal and ex e nal de elopmen al asse s ha a e embedded in he social ecology o an you h’s ne wo ks and oppo uni ies (Ca alano, Be glund, Ryan, Lonczak, & Hawkins, 2004; Small & Memmo, 2004). Wi hin he PYD app oach, posi i e indica o s such as he Sea ch Ins i u e’s De elopmen al Asse s model’s (Benson, Le e , Scales, & Bly h, 1998) and he Fi e Cs model o PYD (Le ne e al., 2013; Le ne e al., 2011) ha e been widely s udied. The Sea ch Ins i u e’s De elopmen al Asse s amewo k iden i ied and comp ised 40 ele an de elopmen al asse s (Benson e al., 1998), ha e lec wo majo asse blocks, each consis ing o 4 asse ca ego ies, espec i ely: 1) ex e nal asse s (ca ego ies: suppo ; empowe men ; bounda ies and expec a ions; cons uc i e use o ime); and 2) in e nal asse s (ca ego ies: commi men o lea ning; posi i e alues; social compe encies; posi i e iden i y). In line he empi ical e idence (Busse i & Rose-K asno , 2009; Le ne e al., 2005; Le ne e al., 2011), he Fi e Cs (Le ne e al., 2013; Le ne e al., 2011) ha e been ope a ionalized h ough he assessmen o di e en Cs: 1) Compe ence, Con idence, Cha ac e , Connec ion and Ca ing. Compe ence is ela ed o he posi i e iew o one’s ac ions in specific a eas, such as social (in e pe sonal skills), academic (school g ades, a endance and es sco es), cogni i e (cogni i e abili ies) and oca ional (wo king habi s, he sea ch o ca ee op ions and en ep eneu ship). Con idence e e s o an in e nal sense o o e all posi i e sel -wo h, sel -e ficacy and iden i y. Cha ac e conce ns he espec o socie al and cul u al alues, he exis ence o s anda d ideas o co ec beha iou s, a gene al sense o mo ali y ( igh /w ong) and in eg i y. Connec ion is ela ed o ela ionships o sel and o he s, posi i e bidi ec ional exchanges wi h he indi idual and o he people, ins i u ions, pee s, amily, school, and communi y. Ca ing e e s o empa hy and compassion o o he s (Le ne e al., 2005). When young people expe ience de elopmen al asse s ha e lec an alignmen o you h s eng hs PORTUGUESE VALIDATION OF THE POSITIVE YOUTH www.sp-ps.p 479 wi h human and s uc u al esou ces in you h con ex , posi i e de elopmen is acili a ed. Acco dingly, an adap i e de elopmen al egula ion p ocess akes place whe e young people a e able o con ibu e o he de elopmen o bo h sel and socie y. The Fi e Cs model o PYD has been used as a measu e o posi i e ou comes, mainly in US samples, bu also in non-US samples o some ex end (Scales, 2011). Se e al e sions o he PYD scale ( o example he 78-i em scale and he 34-i em scale) ha e been de eloped and es ed, and hei psychome ic p ope ies ha e been la gely e i ied (Geldho e al., 2014; Le ne e al., 2014). The sho e e sion o he PYD scale was de eloped o p omo e among o he s, ease o use by esea che s and p ac i ione s (Geldho e al., 2014). Mo e s udies looking a he psychome ic p ope ies o he sho e e sion a e needed o asce ain i s psychome ic p ope ies and applicabili y, especially in non-US con ex s. This pape aims o examine he psychome ic p ope ies o he 34-i em scale o PYD (PYD Sho Fo m, PYD-SF) and i s alida ion o Po uguese you hs. METHOD S udy design The p esen alida ion was conduc ed wi hin he p ojec BePosi i e, a c oss-sec ional s udy and a pa o he local na ional su ey Heal h Beha iou in School-aged Child en (HBSC/WHO) ex ended o Po uguese Uni e si ies (HBSC/JUnP). The HBSC/JUnP ollowed all he ules o esea ch ou lined in he Decla a ion o Helsinki (WMA, 2008) and was app o ed by he E hics Commission o he Medicine Academic Cen e o Lisbon (Cen o Académico de Medicina de Lisboa), CHLN/FMUL/IMM (Cen o Hospi ala Lisboa No e/Faculdade de Medicina de Lisboa/Ins i u o de Medicina Molecula ). This Uni e si y s udy is a esul o he Po uguese pa ne ship wi hin he Posi i e You h De elopmen c oss- na ional p ojec (h p://www.uib.no/en/ g/sipa/pydc ossna ional), which in ol es mo e han wen y coun ies ac oss Eu ope, A ica, Asia, he US and La in Ame ica. The c oss-na ional p ojec aims o: 1) examine he ex en o which de elopmen al (in e nal and ex e nal) asse s a e accessible o young people in di e en na ional con ex s; 2) unde s and how hese asse s can be ela ed o posi i e you h ou comes, such as he “5 Cs” o PYD (i.e., con idence, compe ence, cha ac e , ca ing and connec ion) and h i ing indic o s (e.g., school success, alues di e si y, esis s dange and exhibi leade ship); 3) highligh how posi i e ou comes a e associa ed o young people’s con ibu ion o he de elopmen o he sel and o he in ol ing socie y. Pa icipan s A o al o 2700 uni e si y s uden s (1979 women), aged om 16 o 29 yea s old (21.3±2.8) pa icipa ed in he s udy. Mos o he pa icipan s had Po uguese na ionali y (n=2614), we e bo n in Lisbon (n=902), we e li ing in an u ban a ea (n=1253), we e g adua e s uden s (n=1715) and had a middle socioeconomic s a us (SES) le el (n=1828). P ocedu e Da a collec ion was pe o med h oughou an online su ey, using he Limesu ey pla o m. P io o da a collec ion, he s uden s we e in o med abou he pu pose o he s udy and in o med consen s we e ob ained. In o med consen om pa en s was sough as well whe e necessa y. S uden s we e assu ed Ma ga ida Gaspa de Ma os, Te esa San os, Ma a Reis, Adilson Ma ques, Gina Tomé, Celes e Simões, & No a Wiium www.sp-ps.p 480 con iden iali y o hei esponses o he ques ionnai e and du ing he handling o he da a. The su ey ollowed all he ules o esea ch ou lined in he Decla a ion o Helsinki o he Wo ld Medical Associa ion and was app o ed by he E hics Commission o he Medicine Academic Cen e o Lisbon o he Facul y o Medicine, Uni e si y o Lisbon. Ins umen The o iginal e sion o PYD scale was de eloped using da a om he 4-H S udy in i s di e en wa es (Bowe s e al., 2010; Jelicic, Bobek, Phelps, Le ne , & Le ne , 2007; Le ne e al., 2005) ha p oposed and es ed a highe -o de measu e o PYD, consis ing o a fi e fi s -o de la en cons uc s, each ep esen ing one o he Fi e Cs o PYD, published elsewhe e (Geldho e al., 2014). The s uc u al componen o PYD has been confi med wi hin he 4-H S udy da a se wi h ea ly and mid-adolescence samples (i.e., om app oxima ely 10–16 yea s) (Geldho e al., 2014). The 78 i ems om he o iginal scale we e d awn and adap ed om P o ile o S uden s Li e: A i udes and Beha iou s (PSL) (Ins i u e, 2006), Sel -Pe cep ion P o ile o Adolescen s (SPPA) (Ha e , 1988), Teen Assessmen P ojec (TAP) Su ey Ques ion Bank (Small & Rodge s, 1995), Eisenbe g Sympa hy Scale (ESS) (Eisenbe g e al., 1996), and Empa hic Conce n Subscale o he In e pe sonal Reac i i y Index (IRI) (Da is, 1980). The i ems o he “compe ence” we e e ie ed om he SPPA, he i ems o he “connec ion” om he PSL and TAP, he i ems o he “con idence” and “cha ac e ” om he PSL and SPPA, and he i ems o he “ca ing” a e a composi e o modi ied i ems om he ESS and IRI. Mo e ecen ly, a sho e e sion o his scale was de eloped – he Posi i e You h De elopmen Sho Fo m (PYD-SF). The conduc ed analysis examined he measu e ac oss mo e wa es o measu emen , educed he leng h o he scale om 78 o 34 i ems, scanned i ems a he han pa cels and empi ically es ed he enabili y o a highe -o de ac o s uc u e. I was also emphasized a pa simonious ep esen a ion o he Fi e Cs and c ea ed sepa a e o ms o ea ly s. middle/la e adolescen s (Geldho e al., 2014). The PYD-SF was ansla ed om he o iginal English e sion in o Po uguese language (and back ansla ion), and was hen e ised by a g oup o specialized expe s wi hin he a ea. A e wa ds he psychome ics p ope ies o he ins umen we e examined in e ms o he equency, in e nal consis ency, eliabili y, co ela ion analysis, explo a o y ac o analysis and con i ma o y ac o analysis. MEASURES De ailed in o ma ion conce ning he measu es o he PYD-SF is desc ibed in Table 1. Table 1. O iginal i ems o he PYD-SF (34 i ems) Fac o (Fi e Cs) O iginal I ems Sco ing* Con idence1 All in all, I am glad I am he e. 5-poin s Like scale When I am an adul , I'm su e I will ha e a good li e. I eally like he way I look I am good looking o a ac i e. I am happy wi h mysel mos o he ime. I am e y happy being he way I am. Compe ence2 I am jus as sma as o he s my age. 5-poin s Like scale I could do well a jus abou any new ou doo ac i i y I ha e no ied PORTUGUESE VALIDATION OF THE POSITIVE YOUTH www.sp-ps.p 481 be o e. I ha e a lo o iends. I do e y well in my class wo k a school. I am popula wi h o he s my age. I am be e han o he s my age a spo s. Connec ion3 Teache s a school push me o be he bes I can be. 5-poin s Like scale Adul s in my own o ci y lis en wha I ha e o say. I eel my iends a e good iends. My iends ca e abou me. I ha e lo s o good con e sa ions wi h my pa en s. In my amily I eel use ul and impo an . I ge a lo o encou agemen a my school. Adul s in my own o ci y make me eel impo an . Ca ing4 I eel so y o o he people who don’ ha e wha I ha e. 5-poin s Like scale When I see someone being picked on, I eel so y o hem. When I see someone being aken ad an age o , I wan o help hem. I bo he s me when bad hings happen o any pe son. I makes me sad o see a pe son who doesn’ ha e iends. When I see ano he pe son who is hu o upse , I eel so y o hem. Cha ac e 1 Knowing a lo abou people o o he aces. 5-poin s Like scale Enjoying being wi h people who a e o a di e en ace han I am. I ha dly e e do hings I know I shouldn’ do. I usually ac he way I am supposed o. Helping o make he wo ld a be e place o li e in. Gi ing ime and money o make li e be e o o he people. Doing wha I belie e is igh e en i my iends make un o me. Accep ing esponsibili y o my ac ions when I make a mis ake o ge in ouble. 1 The i ems om Con idence and Cha ac e we e d awn and adap ed om he P o ile o S uden s Li e: A i udes and Beha iou s (PSL) (Ins i u e, 2006) and Sel -Pe cep ion P o ile o Adolescen s (SPPA) (Ha e , 1988). 2 The i ems om Compe ence we e d awn and adap ed om he Sel -Pe cep ion P o ile o Adolescen s (SPPA) (Ha e , 1988). 3 The i ems om Connec ion we e d awn and adap ed om he P o ile o S uden s Li e: A i udes and Beha iou s (PSL) (Ins i u e, 2006) and Teen Assessmen P ojec (TAP) Su ey Ques ion Bank (Small & Rodge s, 1995). 4 The i ems om Ca ing a e a composi e o modi ied i ems om he Eisenbe g Sympa hy Scale (ESS) (Eisenbe g e al., 1996), and Empa hic Conce n Subscale o he In e pe sonal Reac i i y Index (IRI) (Da is, 1983). * The sco ing sys em can be: 1=S ongly disag ee, 5=S ongly ag ee; 1=Almos ne e ue o ne e ue, 5=Always ue; 1=No all like me, 5=Ve y much like me; 1=No impo an , 5=Ex emely impo an . Da a analysis Da a om Limesu ey was ans e ed o an elec onic da a ile. All a iables we e checked o da a inaccu acy by unning SPSS equencies, and a e wa ds, an analysis on missing alues was conduc ed. Desc ip i e analysis (means, s anda d de ia ion and pe cen age) we e used o cha ac e ize he sample. All da a we e es ed o no mali y p io o any analyses using Kolmogo o -Smi no es s, as well as Le ene's es o he homogenei y o he a iance. To examine he ac o s uc u e o he PYD-SF, Explo a o y Fac o Analysis (EFA) wi h Maximum Likelihood Es ima ion and Oblimin o a ion was used. The Ma ga ida Gaspa de Ma os, Te esa San os, Ma a Reis, Adilson Ma ques, Gina Tomé, Celes e Simões, & No a Wiium www.sp-ps.p 482 Kaise -Gu man ule was used o e ain ac o s wi h eigen alues o ≥1.00. C onbach’s alpha was compu ed o examine in e nal consis ency o he PYD-SF dimensions (Fi e Cs). A Con i ma o y Fac o ial Analysis (CFA) was pe o med o con i m he Fi e Cs s uc u e o he PYD- SF. The Maximum Es ima ion p ocedu e was used o de i e pa ame e es ima es in PYD-SF model. The analysis we e pe o med on he co a iance ma ix. Fo model iden i ica ion pu poses, a scale was assigned o each la en a iable by se ing one o i s ac o loadings o uni y. Mul iple i indexes we e used o e alua e goodness o i o models es ed in he s udy: chi-squa e s a is ics (χ2), compa a i e i index (CFI), non-no med i index (NNFI), oo mean squa e e o o app oxima ion (RMSEA), and s anda dized oo mean squa ed esidual (SRMR). The c oss alidi y o he model was o mally es ed o measu e and s uc u al in a iance be ween he wo spli hal es o he sample. The analyses we e conduc ed using he SPSS 24.0 (S a is ical Package o Social Sciences) and he S uc u al Equa ion Modelling wi h so wa e AMOS 24.0. The signi icance le el was se a p<0.05. RESULTS Explo a o y ac o analysis Da a om he 34 i ems we e subjec ed o an Explo a o y Fac o Analysis (EFA). The Maximum Likelihood ex ac ion wi h Oblimin Ro a ion p oduced a nine- ac o solu ion based on eigen alues g ea e han 1, and accoun ed o 68.08% o he explained a iance. The Kaise -Meye -Olkin (KMO) measu e o sampling adequacy was sa is ac o y (KMO=0.84). Ba le ’s es o sphe ici y (χ2(561)=47618.85, p<0.001) indica ed ha i ems co ela ion we e su icien ly la ge o ac o analysis (Field, 2013). Al hough he EFA esul ed in a nine- ac o solu ion, Le ne ’s heo y de ined he exis ence o 5 dimensions, conside ed he Fi e Cs (Le ne e al., 2013; Le ne e al., 2011). The ob ained solu ion in he p esen s udy showed ha he i ems we e no well ep esen ed acco ding o Le ne ’s heo y and we e sp ead be ween nine- ac o s. Thus, an e o was made o main ain he Fi e Cs de ined in he heo y. Th ee i ems wi h communali ies lowe han 0.50 we e subsequen ly emo ed, and an analysis o assess he in e nal consis ency o each dimension (Ch onbach’s Alpha) was un. The analysis showed ha he dele ion o he i ems wi h a wo se ep esen a ion would inc ease he Alpha alues, in each dimension. The e o e, en i ems we e dele ed. Wi h he emaining 21 i ems, ano he EFA was pe o med. The Maximum Likelihood ex ac ion wi h Oblimin Ro a ion p oduced a ou - ac o solu ion based on eigen alues g ea e han 1, and accoun ed o 60.57% o he explained a iance. The Kaise -Meye -Olkin (KMO) measu e o sampling adequacy was good (KMO=0.86). Ba le ’s es o sphe ici y was (χ2(210)=27568.67, p<0.001). I ems we e conside ed o ha e loaded a ac o i alues exceeded 0.50 and we e conside ed uniquely loaded i c oss-loadings on o he ac o s we e less han 0.50. The i s componen was named “Con idence-Compe ence” and comp ised nine i ems. The o he componen s we e named “Cha ac e ”, “Connec ion” and “Ca ing”, each one consis ing o ou i ems. The explo a o y ac o analysis loadings o PYD a e p esen ed in Table 2. PORTUGUESE VALIDATION OF THE POSITIVE YOUTH www.sp-ps.p 483 Table 2. Explo a o y ac o analysis loadings o PYD Componen s CF and CP CH CN CA I am be e han o he s my age a spo s. 0.95 I am happy wi h mysel mos o he ime. 0.82 I am good looking o a ac i e. 0.78 I am e y happy being he way I am. 0.77 I do e y well in my class wo k a school. 0.75 I am popula wi h o he s my age. 0.75 I ha e a lo o iends. 0.75 I could do well a jus abou any new a hle ic ac i i y. 0.71 I eally like he way I look. 0.70 Doing wha I belie e is igh e en i my iends make un o me. 0.85 Helping o make he wo ld a be e place o li e in. 0.81 Accep ing esponsibili y o my ac ions when I make a mis ake o ge in ouble. 0.81 Gi ing ime and money o make li e be e o o he people. 0.73 In my amily I eel use ul and impo an . 0.84 I ha e lo s o good con e sa ions wi h my pa en s. 0.84 Adul s in my own o ci y make me eel impo an . 0.66 I ge a lo o encou agemen a my school. 0.63 I bo he s me when bad hings happen o any pe son. 0.79 When I see someone being aken ad an age o , I wan o help hem. 0.75 When I see ano he pe son who is hu o upse , I eel so y o hem. 0.73 I makes me sad o see a pe son who doesn’ ha e iends. 0.67 Abb e ia ions: CF, con idence, CP, compe ence; CH, cha ac e ; CN, connec ion; CA, ca ing Ex ac ion Me hod: P incipal Componen Analysis. Ro a ion Me hod: Oblimin wi h Kaise No maliza ion. In e nal Consis ency To u he examine he eliabili y o he i ems e lec ing he PYD componen s, C onbach’s Alpha was compu ed o each componen . Fo such analysis and despi e he ou - ac o solu ion, ha has been ob ained in p e ious analysis, he au ho s decided o ha e a conse a i e app oach, main aining he Fi e Cs s uc u e based on Le ne ’s heo y (Le ne , Phelps, Fo man, & Bowe s, 2009; Le ne e al., 2013; Le ne e al., 2011). This means ha he ac o “Con idence-Compe ence” was di ided in o wo sepa a ed componen s, espec i ely “Con idence” and “Compe ence”. “Con idence” had an in e nal consis ency o α=0.83, “Compe ence” α=0.87, “Cha ac e ” α=0.82, “Connec ion” α=0.74, “Ca ing” α=0.74. Conside ing ha each componen included ou and one o hem included i e i ems, he in e nal consis encies can be conside ed as adequa e. Con i ma o y ac o analysis In a sepa a e analysis, a CFA was conduc ed on da a om he 2700 pa icipan s. Mul iple i indexes we e used as guides o e alua e goodness o i o models es ed in he s udy (χ2, GFI, NFI, CFI and Ma ga ida Gaspa de Ma os, Te esa San os, Ma a Reis, Adilson Ma ques, Gina Tomé, Celes e Simões, & No a Wiium www.sp-ps.p 484 RMSEA). The i indices o bo h he o iginal and modi ied models a e showed in able 3. The i indices o he o iginal model we e indica i e o a poo i ing model (χ2(179)=3486.10; GFI=0.89; NFI=0.87; CFI=0.88; and RMSEA=0.08, CI(RMSEA)=0.08-0.09). Modi ica ion indices we e examined, and model i was imp o ed by adding a ows be ween he highes e o ’s co a iance in each dimension. Al hough he Chi-Squa e alues emained poo , he model i gene ally imp o ed (χ2(174)=2370.72; GFI=0.92; NFI=0.91; CFI=0.92; and RMSEA=0.07, CI(RMSEA)=0.06-0.07). Based on modi ica ion indices, he e o ’s co a iance among i ems we e inspec ed again. The modi ica ion indices sugges ed ha one i em om Compe ence was also sa u a ed in Con idence and also showed a highe e o co a iance wi h o he i ems om he Con idence and Compe ence. Thus, his i em was dele ed and he esul ing CFA displayed accep able model i (χ2(174)=1349.20; GFI=0.95; NFI=0.95; CFI=0.95; and RMSEA=0.05, CI(RMSEA)=0.05-0.06). The χ2/d alues emained poo , howe e , χ2/d is no conside ed o be a use ul i index (Hayduk, Cummings, Boadu, Pazde ka-Robinson, & Boulianne, 2007), because i is a ec ed by he sample size and he dis ibu ion o a iables. The indices o e alua ing goodness o i o he models in he di e en phases a e p esen ed in Table 3. Table 3. Indices o e alua ing goodness o i o models in di e en phases Models χ² (gl) GFI NFI CFI RMSEA (90% IC) S ep 1 3486.10 (179) 0.89 0.87 0.88 0.08 (0.08-0.09) S ep 2 2370.72 (174) 0.92 0.91 0.92 0.07 (0.06-0.07) S ep 3 1349.20 (174) 0.95 0.95 0.95 0.05 (0.05-0.06) Ab e ia ions: GFI, goodness o i index; NFI, no mal i index; CFI, compa a i e i index; RMSEA, oo mean squa e e o o app oxima ion The composi e eliabili y o he ac o s was high o he ollowing C componen s: Con idence (0.83), Compe ence (0.85), Cha ac e (0.82); and adequa e o he ollowing C componen s: Connec ion (0.74) and Ca ing (0.74). The disc iminan alidi y o he ac o s was es ed by he compa ing he a e age a iance ex ac ed wi h he squa ed co ela ion among ac o s. The ac o s showed disc iminan alidi y be ween each o he , wi h he excep ion o he C’s Con idence and Compe ence. The ela ion be ween Con idence and Compe ence had been al eady obse ed in he EFA. In he cu en analysis, hese wo Cs we e no ound o be sepa a e componen s o ac o s, al hough heo e ically wo cons uc s su aced in Le ne ’s heo y. Aiming o main ain heo e ical cohe ence and o p ese e he s uc u e as close as possible o he o iginal, Con idence and Compe ence we e e ained as wo cons uc s, e en hough hey had no disc imina ing alidi y. The inal model is shown in Figu e 1. PORTUGUESE VALIDATION OF THE POSITIVE YOUTH www.sp-ps.p 485 Figu e 1. Pa h diag am o he model o PYD wi h s anda dized alues o coe icien s and co a ia es In o de o c oss alida e he esul s, measu emen and s uc u al in a iance we e es ed on he wo spli hal es o he sample. To es o measu emen in a iance, he o mal s uc u e om he calib a ion o he inal s anda dized solu ion was examined. Fi indices o his mul i-g oup model we e good, hus p o ing i s alidi y o in e p e ing he PYD model (χ2(310)=2079.08; GFI=0.93; NFI=0.92; CFI=0.93; and RMSEA=0.05, CI(RMSEA)=0.04-0.05). A e wa ds, a psychome ic analysis o in e nal consis ency was pe o med and he ollowing C onbach’s Alphas we e achie ed: α=0.83 o Con idence, α=0.85 o Compe ence, α=0.74 o