UM INVENTÁRIO DAS CARACTERÍSTICAS DAS ESCALAS EM MARKETING PROPOSTAS E VALIDADES NO BRASIL

VIEIRA, VALTER AFONSO; SLONGO, LUIZ ANTONIO

doi:10.1590/1678-69712007/administracao.v8n4p11-34

Acessibilidade / Reportar erro

Brasil

RAM. Revista de Administração Mackenzie

Español English

Brasil

Español English

sumário « anterior atual seguinte »

Sumário

Artigos • RAM, Rev. Adm. Mackenzie 8 (4) • May-Jun 2007 • https://doi.org/10.1590/1678-69712007/administracao.v8n4p11-34 copiar

UM INVENTÁRIO DAS CARACTERÍSTICAS DAS ESCALAS EM MARKETING PROPOSTAS E VALIDADES NO BRASIL

Autoria SCIMAGO INSTITUTIONS RANKINGS

RESUMO

As escalas utilizadas na pesquisa em marketing visam à obtenção fiel e confiável dos dados. Entretanto, diversos problemas surgem quando o assunto é a utilização de medidas válidas. Nesse contexto, este artigo tem a intenção de analisar as escalas propostas e validadas em âmbito nacional. Para tal fim, discute-se a teoria sugerida para validar e/ou criar escalas (sugerindo uma figura-resumo) e em seguida verifica-se tal teoria empiricamente. Foram analisadas escalas exclusivamente no campo de marketing, obtendo-se um total de 26 instrumentos. Os resultados sugerem que é necessário utilizar mais os métodos alternativos sugeridos pela teoria, tais como análise fatorial confirmatória e validade nomológica.

PALAVRAS-CHAVE
Escalas; Marketing; Medição; Validade; Instrumento

BRAZILIAN AUTHOR	SCALE NAME	DOMAIN OF CONSTRUCT	SCALE GENERAL DESCRIPTION			SAMPLE	PRE-TEST	RELIABILITY				VALIDITY			CFA	OBSERVATIONS
			SCALE DIMENSIONS	SCALE POINTS AND ITEMS	ALPHAS α	SAMPLE FEATURES		TEST- RETEST	ALTERNATIVE FORM	INTERNAL CONSISTENCE	CONTENT	CONSTRUCT
			SCALE DIMENSIONS	SCALE POINTS AND ITEMS	ALPHAS α	SAMPLE FEATURES		TEST- RETEST	ALTERNATIVE FORM	INTERNAL CONSISTENCE	CONTENT	CONVERGENT	DISCRIMINANT	NOMOLOGICAL
Hor-Meyll (2004)	Perceived Risk in On-Line Environment	X	Performance and Financial; Psychological; Time; and Total	7 points and 15 items	0,82; 0,89; 0,75; 0,80	n=15; n=10 and n=2 for items generation; n=32; n=75 and n=84 for collection I, and n=752 for collection II				X						The study did not obtain the construct validity. Each dimension presented high reliability. VE=0,69% KMO=0,80
Almeida and Nique (2004)	Consumer Delight in the Post-Consumption Evaluation	X	Affect Aspects; Cognitive Aspects; Positive Surprise; Personalize; and Repurchase	5 points and 23 items	0,91; 0,84; 0,77; 0,78; and 0,77	n=30 for items generation; and for data collection I,II and III were used n=146; 129 and 240	X			X	X	X	X		X	The results showed that there is content validity, reliability, and unidimensionality. The results also presented a compose validity, discriminant and convergent. VE=68% and KMO=0,88
Santos and Muniz (2004)	Delivery in Service Quality	X	Reliability; Flexibility; Price; Accuracy; and Readiness	7 points and 44 items	0,96; 0,90; 0,72; 0,95; and 0,95	n=38 for items generation=227 for data collection	X	X		X	X	X	X			The authors achieve Internal consistence (>,70 all cases), external validity, convergent validity (33% >,50) and discriminant validity
Paiva (2004)	Value Structure for retailing bank consumers	X	Empathy; Reliability; Price; Benevolence; and Competence	6 points and 25 items; a point for nonopinion	0,62; 0,72; 0,56; 0,63; and 0,64	n=38 for item generation, n=227 for data collection I and n=111 for data collection II	X	X		X	X	X	X			0,5<KMO<1,0. The scale complements SERVQUAL in price dimension and was submitted to 2 convergent and discriminant tests
Ayrosa (2003)	Enduring Involvement	X	Product Importance/Interest; Product Symbolic Value; and Consumer Knowledge and Information	13 items	0,91; 0,82; and 0,81	Items were based on 5 dimensions of McQuarrie and Munson (1992); data collection used n=501	X			X					X	The scale is a first step towards a fully reliable enduring involvement scale in Brazilian Portuguese. Since the final configuration of the scale is quite different from McQuarrie and Munson’s (1987, 1992)
Nicolau and Rossi (2003)	Consumer Regret	X	Two dimensions. Affective and Cognitive Aspects of Regret	5 points	Car 0,92 ; 0,86 and Shoes 0,86 ; 0,80	n=500 for data collection	X			X		X	X		X	Albeit it was tested the discriminant validity; it was not found
Prado (1997)	Affective Consumer Response in supermarket		Happiness; Interest; Unfounded; Anger; Fear; Anguish; Blame; and Fatigue	7 points and 36 items	0,92; 0,83; 0,75; 0,76; 0,70; 0,60; 0,79 and 0,61	theory for items; n=100 and n=282 for data collection I and II				X					X	The author also verified the impact of Affective Response on Consumer Satisfaction, where the results were that two factor were positive, and two were negative, albeit both pair were significant, and the model R²=,317
Prado and Marchetti (1996)	Consumer satisfaction in supermarket	X	Cashier Service; Supermarket. Environment; Bakery; Attendance; and Meat Department	7 points and 19 items	0,90; 0,87; 0,81; 0,75; 0,70	n=25 for item generation; n=60 for data collection I; n=120 data collection II				X	X	X				The scale points out the importance of aspects such as cashier, environment, politeness and convenience to consumer
Ayrosa (1998)	Country Image		General Aspects of the Country; Emotional Response; Attitude toward Arts; and Marketing Aspects	7 points and 17 items	0,96; 0,88; 0,71; and 0,83	Items were generated from Pisharodi and Parameswaran (1992) and n=9; n=124 data collection I and n=330 data collection II	X			X					X	The dimensional structure of the scale is consistent with the theory.
Wolff (2002)	Symbolism X		Brand; Self-Concept/Prestige; Self-Concept/Security; and Experience	7 points and 21 items	0,62; 0,55; 0,67; and 0,68	Items were based on Kleine III and Kernan (1998), n=32 and literature. Data collectin n=600	X			X					X
Rodrigues (2000, (2001)	Service Quality		Supplier Responsibility; Client and Tangible Item	7 points and 22 items	Between 0,95 and 0,97	Focus group + Servqual for items generation; n=331 for data collection	X			X						A total of 27 items were generated and 22 were keeped in the final version.

BRAZILIAN AUTHOR(S)	SCALE IN EVALUATION	ORIGINAL SCALE AUTHOR(S)	SCALE GENERAL DESCRIPTION			SAMPLE	PRE-TEST	RELIABILITY				VALIDITY			CFA	MAIN RESULTS
			SCALE DIMENSIONS	DIMENSIONS AFTER THE STUDY	SC ALE POINTS	SAMPLE FEATURES		TEST RETEST	ALTERNATIVE FORM	INTERNAL CONSISTENCE	CONTENT	CONSTRUCT
			SCALE DIMENSIONS	DIMENSIONS AFTER THE STUDY	SC ALE POINTS	SAMPLE FEATURES		TEST RETEST	ALTERNATIVE FORM	INTERNAL CONSISTENCE	CONTENT	CONVERGENT	DISCRIMINANT	NOMONOLOGICAL
Sampaio and Perin (2001)	MARKOR	Kohli et al.(1993)⁵	3	1	5 points for electronic industry and 6 points for retailing	n=192 for Retailing and n=198 for Electronic business firms	X			X	X		X		X	The authors found that the three dimensions of MARKOR should be one. It means that the market orientation construct (multidimensional) should be one, instead of three. The authors did not support the discriminant validity.
Strehlau and Aranha (2002)	Status Consumption Scale - SCS	Eastman (1999)	1	1	7 points	n=385 undergraduate students				X			X		X	The authors, similar to the original study, also supported the undimensionality of the construct
Lages and Neto (2002)	Ecological/Environmental Awareness Scale	Straughan and Roberts (1999)	6		5 points	n=400 consumers	X			X						The authors found that the relation between the demographic variables and ecological awareness was comproved, indicating a strong validity and reliability of such scale. The level of ecological awareness was defined by the sum of the six dimensions of the scale.
Fonseca and Rossi (1999)	New Involvement Profile (NIP)	Jain and Srinivasan (1999)	5		5 points	n=420 consumers	X			X	X	X	X			The tests of reliability supported the robustness of reliability, validity of content and construct. The multidimensionality of the scale was also supported and two products had validity problems in two dimensions.
Viana et al. (1999)	Relationship Marketing	Wilson and Vlosky (1997)	6		5 points	n=359 business company	X			X	X	X	X		X	Content Validity was also supported by in depth interview
Siqueira (2004)	SERVPERF	Cronin and Taylor (1992)	4		7 points	n=213 undergraduate business students	X			X	X	X	X		X	The results indicated that new variables need to be included in the model, that nine variables were excluded and that sixteen variables were maintained
Espartel et al. (2004)	Consumer Self-Confidence	Beardenet al. (2001)	6		5 points	n=186 postgraduate students	X			X		X			X	The results suggest that the scale has discriminant the convergent validity and Internal consistence.
Urdan (2001); Urdan and Zuñiga (2003)	QEP – Experienced by the Patient	Brown and Swartz (1989)	9	5	5 points	n=344 patient from a hospital	X			X		X	X			There is not discriminant validity between the scale and a global measure of satisfaction.
Souza and Luce (2003)	TRI - Technology Readiness Index	Parasuraman (2000), Parasuraman and Colby (2001)	4	5	5 points	n=731 consumers				X		X	X		X	One dimension (Optimism) did not present satisfactory results (unidimensionality and discriminant validity). The unidimensional analysis indicated that three dimensions should be reduce to one.
Johnston et al. (1996)⁶	SERVQUAL and SERVPERF	Parasuraman et al. (1988) and Cronin and Taylor (1992)		7 points	n=332 clients					X						The results indicated that the bank service is formed by three factors. The SERVQUAL validity is not confirmed and a version based on SERVPERF presented better performance
Arruda et al. (2003)	CES - Consumption Emotion Set	Marsha Richins (1997	17	12	6 points	n=417 people	X			X	X	X	X			Just three dimensions of twelve presented both convergent and discriminant validity. The other ones (nine) presented either convergent validity or discriminant validity or none.
Mello and Rocha (2003)	Market Orientation	Day (1999)	5	5	5 points and 6 points	n=114 companies				X		X	X			Six points scale is better to measure the, instead of 5 as the original scale suggest. One factor of the scale (distribution) had problems in its variables.
Gosling and Gonçalves (2003)	Relationship Marketing between banks and consumers	Morgan and Hunt (1994)	8	8	7 points	n=238 undergraduate business students				X					X	The simples reliability and compose reliability indicated that the instrument is good and might be used in future research
Mello and Collings (2001)	Perceived risk in business-to- business field	The authors tested two models, which one of them is based on Peter and Tapey (1975)			7 points	n=162 business firms		X		X		X	X			The findings indicate that all three types of scale (riskness-scale; likelihood of loss/seriousness and likelihood of loss in details) had construct validity but the riskiness scale seemed to provoke some fuzzy effects when compared to other methods. Moreover, it did not perform as well as the other methods in distinguishing among traits. The results from the MTMM method failed to provide strong evidence about which of these methods is better in distinguishing among traits.
Lages and Brasil (2003)	Experienced by the Consumer in on-line environment	Novak et al. (2000)	17	12	7 points	n=400 on-line consumers	X			X	X	X	X		X	From the original structure of 13 constructs, just 2 had serious problems. Other conclusion is the unidimensionlity of the constructs. In addition, all construct have better results compared to the original study (in terms of compose reliability).

Editora Mackenzie; Universidade Presbiteriana Mackenzie Rua da Consolação, 896, Edifício Rev. Modesto Carvalhosa, Térreo - Coordenação da RAM, Consolação - São Paulo - SP - Brasil - cep 01302-907 - São Paulo - SP - Brazil
E-mail: revista.adm@mackenzie.br

Acompanhe os números deste periódico no seu leitor de RSS

[TFN1] CFA: Confirmatory Factor Analysis; VE: Variance explained; letter “x” means that it was verified in the research.