Validação de uma escala sobre o comportamento do consumidor de se autopresentear

Abstract

Due to the highlight in the Administration area of studies related to consumer behavior, by means of which one intends to identify some influences, stimuli and stages in the consumption, it was decided to carry out a research in this field (ONO et al., 2012). Analyzing how each type of behavior may influence the consumer's purchasing decision is of great importance (FUTRELL, 2003). Among the types of consumer behavior observed is the behavior of giving, which, by chance, is rooted in our society. The gifting behavior can happen of 2 (two) ways: to another person; or for yourself. It is worth mentioning that many researchers have focused on the first, but few, on the second, which can be termed self-gifting consumer behaviour (SGCB). In this respect, Mick and DeMoss (1990a, 1990b) were the pioneers in the investigation of this type of behavior. Thus, in order to fill a theoretical gap, referring to the self-gifting consumer behaviour, the present study has as general objective to validate the scale on the SGCB, proposed by Mortimer et al. (2015), for the Brazilian context and, as specific objectives: describe the profile of the respondents; analyze the constructs and respective variables of the scale on the SGCB, proposed by Mortimer et al. (2015); verify the differences in the perception of the respondents in relation to the variables of the scale on the SGCB, proposed by Mortimer et al. (2015); perform the depuration and validation for the Brazilian context of the scale on the SGCB, proposed by Mortimer et al. (2015); verify the differences of perception of the respondents in relation to the constructs of the scale on the SGCB, proposed by Mortimer et al. (2015), after being validated for the Brazilian context. The survey was applied to a sample of 1.315 (one thousand, three hundred and fifteen) respondents (web users, via e-mail, Facebook, Whastapp and Twitter). Among the results found, it was noted that, at the time of the Exploratory Factorial Analysis (EFA), there was a need to exclude only two variables of the model (CO_1 - “I would not buy gifts for myself to remind me of special times in my life” and MT_3 - “I do not purchase gifts for myself to feel motivated to achieve my goals”), because they presented values of communality and/or factorial load less than 0.50 (zero point fifty). In addition, at the time of the Confirmatory Factor Analysis (CFA), all adjustment indices of the model were satisfactory (χ²/gl, RMSR, RMSEA, GFI, IFC, NFI and TLI). Furthermore, all other measures were also adequate: reliability, through the results of Cronbach's Alpha; the convergent validity, through the analysis of the standardized loads, the extracted variance and the composite reliability; and the discriminant validity, by means of the analysis of the correlations between the constructs and of the comparison between the explained variance and the square of the estimates of the correlations between the constructs. Thus, after the exploratory factorial analysis and the confirmatory factorial analysis, we reached indices within the expected limits. In this context, the validity of the model is accepted as a research tool for the Brazilian context.