FORECASTING MODEL OF VIETNAMESE CONSUMERS’ PURCHASE DECISION OF DOMESTIC APPAREL

The study of the determinants of consumer purchase decision of domestic goods is necessary in the context of Vietnams’ international integration in order to support domestic firms to improve their competitiveness. This study aims to analyze the factors influencing Vietnamese consumers’ purchase decision of domestic apparel based on the Binary Logistic regression model. This study uses survey data gathered from 240 consumers in Vietnam in 2019. The research results showed that i) Perceived price, ii) Perceived quality, iii) Consumer ethnocentrism, and iv) Demographic variables (age; the family with a child; the level of education; income; and sex) have significant influences on the consumer purchase decision of domestic apparel in Vietnam.


INTRODUCTION
Vietnam's economy continues its deep international integration process with the continued signing of the Comprehensive and Progressive Partnership Trans-Pacific Partnership (CPTPP), the world's largest free-trade area with a combined market of 600 million people, and together with 16 signed linkage mechanism  promises to continue to increase growth opportunities for many economic sectors. At the same time, competitive pressure on the domestic market will also increase strongly in Vietnamese firms. The relaxation of trade policies has provided consumers with more foreign manufactured product choices than ever before. Consequently, their attitudes toward products originating from foreign countries have been of interest to domestic business and consumer behavior researchers for decades (Lu Wang & Xiong Chen, 2004). Therefore, it is important to understand both the theoretical and practical aspects of the factors affecting the behavior of the Vietnamese consumer purchase decision.
Consumer behavior and consumer purchase behavior of domestic goods have received significant attention from researchers (Knight, 1999;Watson & Wright, 2000;Vida & Damjan, 2001;Lu Wang & Xiong Chen, 2004;Nguyen, Nguyen & Barrett, 2008;Dmitrovic, Vida, & Reardon, 2009;Renko, Crnjak Karanović, & Matić, 2012;Ha-Brookshire, 2012;Purwanto, 2014;He & Wang, 2015;Asshidin, Abidin, & Borhan, 2016;Mansi & Pandey, 2016;Kacprzak & Pawłowska, 2017;Puška et al., 2018). Previous studies used a microeconomic theory approach, the behavioral economics theory to explain consumers' domestic purchasing behavior and test for multiple product groups. However, there is no full research model in all cases, and many studies, rather than the use of dependent variables, are the real decisions of consumers made using other variables such as a willingness to buy or intention to purchase. Thus, the predictability power of these research models is not really significant.
The present study aims to determine factors affecting Vietnamese consumers' purchase decision of domestic apparel based on the binary logistic regression model. The structure of the article consists of five parts: i) Introduction, ii) Literature review, iii) Methodology, iv) Results and discussions, and v) Conclusions. Kotler and Armstrong (2009) argue that consumers' decision-making processes were influenced by the factors including cultural, social, personal, and psychological. Consumer behavior patterns are used to describe the relationship between the three factors, including the stimulus, black box consciousness, and responses responding to consumer stimuli that affect purchasing decisions of consumers. The Theory of Reasoned Action (TRA) developed by Fishbein and Ajzen (1975) shows that the intention of an individual's behavior is influenced by two factors, including behavioral attitude, and subjective norm. These two factors directly affect the behavioral intent and then affect the actual behavior of an individual. The Theory of Planned Behavior (TPB) states that the intention of an individual is influenced by three factors: behavioral attitudes, subjective norms, and perceived behavior control. Perceived Behavioral Control (PBC) refers to the ease or difficulty of performing a behavior and whether or not the behavior is controlled (Ajzen, 1991). This extension involves explanations for when people intend to perform an activity that is impeded by their lack of confidence or lack of the right to conduct behavior (Ajzen, 1991).

LITERATURE REVIEW
There are many previous kinds of research that have identified the factors that explain consumers' purchase behavior towards domestic goods. Key elements proposed in the previous studies include consumer ethnocentrism, perceived price, perceived quality, and demographic variables.

Consumers ethnocentrism
Ethnocentrism refers to the tendency of individuals to see their cultural group as proving the norms for acceptable behaviors and preferences. Highly ethnocentric individuals are intolerant and judgmental with respect to cultures different from their own (Luque-Martinez, Ibanez-Zapata, & Barrio-Garcia, 2000). The tendency of consumers to be ethnocentric represents their beliefs about the appropriateness and moral legitimacy of purchasing foreign products (Shimp & Sharma, 1987). Ethnocentric consumers prefer domestic goods because they believe that products from their own country are the best (Klein, Ettenson, & Morris, 1998). Moreover, a concern for morality leads consumers to purchase domestic products, even though the quality is poorer than that of imports (Wall & Heslop, 1986). Consumer ethnocentrism may play a significant role when people believe that their personal or national wellbeing is under threat from imports (Sharma, Shimp, & Shin, 1995;Shimp & Sharma, 1987). The more importance a consumer places on whether or not a product is made in his/her home country, the higher his/her ethnocentric tendency (Huddleston, Good, & Stoel, 2001). Sharma, Shimp, and Shin (1995) show the effect of consumer preference leads to a preference for domestic goods, low prices for foreign goods and unwillingness to buy foreign goods while always favoring domestic goods. According to Netemeyer, Durvasula, and Lichtenstein (1991), the level of consumer preference indicates the degree of influence of buyers' beliefs, attitudes and behavior toward domestic products. In a study that examined the relationship between consumer ethnocentrism and evaluations of foreign-sourced products. Lantz and Loeb (1996) found that highly ethnocentric consumers have more favorable attitudes toward products from culturally similar countries. The higher the purchasing power, the higher the domestic buying behavior (Shimp & Sharma, 1987;Watson & Wright, 2000;Lu Wang & Xiong Chen, 2004;Nguyen, Nguyen, & Barrett, 2008;Zafer Erdogan & Cevahir Uzkurt, 2010). In addition, individuals with high levels of consumer ethnocentrism would have more favorable attitudes toward products from culturally similar countries in comparison to products from culturally dissimilar countries, consumers' ethnocentric tendencies play a significant role in predicting purchase intentions towards domestically produced goods (Vida & Damjan, 2001;Dmitrovic, Vida, & Reardon, 2009;Renko, Crnjak Karanović, & Matić, 2012;Purwanto, 2014;He & Wang, 2015).
H 1 -Consumers' ethnocentrism has a positive impact on purchase decision of domestic apparel in Vietnam.

Perceived price
The price represents an extrinsic sign and provides one of the most important forms of information available to consumers when making a purchase decision (Jin & Sternquist, 2003). Diehl, Kornish, and Lynch (2003) argue that consumers are price sensitive because they tend to look for lower-priced products and they want satisfaction through price comparisons between different products. Teas and Agarwal (2000) argue that the price offered was positively related to the perception of product quality and sacrificed by consumers. Zeithaml (1988) distinguished prices into two categories: objective price is the actual price of the product or service and price is perceived by the consumer. Price perception is a comparison between the actual objective price and the price at which consumers refer to, reflecting the buyer's subjective feelings or feelings about the price objective. On the consumer side, they often do not remember or know the actual price of the product, but instead feel the price suitability based on the perceived usefulness of the merchandise. This approach refers to the suitability of the price of the product or service that the user perceives. When consumers feel positive about price stability, and compare it to the usefulness they receive, they will shape their buying intentions and behavior (Ha-Brookshire, 2012;Winit & Gregory, 2009;Prasad, 2014). Higher price is associated with higher likelihood of purchase (Sternquist, Byun, & Jin, 2004).
H 2 -The perceived price of domestic apparel has a positive influence on consumer purchase decision in Vietnam.

Perceived quality
Consumers' intention to purchase domestic products will be influenced by perceived quality. Perceived quality is a cognitive response to a product which influences product purchase (Kumar, Lee, & Kim, 2009). In literature, quality perception is treated as a multi-dimensional concept, including appearance, color and design, durability, fashion, functionality, prestige, reliability, technical advancement, value for money, and workmanship (Darling & Wood, 1990;Klein, Ettenson, & Morris, 1998;Scott, Powers, & Swan, 1997;Watson & Wright, 2000). Thelen, Ford, and Honeycutt (2006) concluded that product characteristics may influence product preference for domestic versus imported more than consumer ethnocentrism levels. Consumers of the developing countries will go for non-local products because they are generally deemed to be of high quality (Khattak, Saeed, & Shah, 2011). One of the ways in which consumers form perceptions about a brand is based upon the quality (Dooley & Fryxell, 1999). A positive perception of quality is the source of the consumer's purchase decision (Aaker, 1991). Quality perception influences consumers' buying intentions (Dodds, Monroe, & Grewal, 1991;Asshidin, Abidin, & Borhan, 2016).
H 3 -The perceived quality of domestic apparel has a positive influence on consumer purchase decision in Vietnam.
H 4 -There is a significant relationship between demographic variables and consumer purchase decision of domestic apparel in Vietnam.
The logit model was estimated to explain and predict consumer's purchase decision of domestic products. The logit model was chosen for this study because of its mathematical simplicity and because of its asymptotic characteristics that constrain the predicted probabilities to a range between zero and one (Maddala, 1983). The maximum likelihood (ML) estimation procedure was used to obtain the model parameters. The TIEN LUU D.  FORECASTING MODEL OF VIETNAMESE CONSUMERS' PURCHASE DECISION OF DOMESTIC APPAREL model, selected to analyze the dependence of purchase decision (Y) on the main factors, was specified as: Y = β1 + β2CONSUMER ETHNOCENTRISM + β3PERCEIVED QUALITY + β4 PERCEIVED PRICE + β5AGE + β6FAMILY WITH CHILD + β7LEVEL OF EDUCATION + β8INCOME + β9SEX + µ.
In which: -Y: Purchasing decision.
-β1 to β9: Regression coefficients of the independent variables (Consumer ethnocentrism, Perceived quality, Perceived price, Age, Family with child, Level of education, Income, and Sex, respectively.) -µ: Residual (other factors not in the model).

METHODOLOGY
Binary Logistic Model was used to estimate parameters in the model. There are 14 items to measure Perceived price, Perceived quality, and Consumer ethnocentrism in the model (see Table 2). All items are measured by 5-point Likert scales, which were 5 -strongly agree, 4 -agree, 3 -not sure, 2 -disagree and 1 -strongly disagree. We adopted previously validated measures with additional testing of reliability for indicator scale in this study.
The logit model is generally specified as follows: β is a scalar parameter to be estimated in equation (1).
The logarithm of the Odds ratio is a linear function with independent variables Xi (Cox, 1970). Agresti (2007) indicated the prediction function of the binary logit model:  (2) According to Hair, Black, Babin, and Anderson (2010), the sample size used in the exploratory factor analysis was determined by the minimum (min = 50) and the number of variables included in the model. The proportion of samples compared to an analytical variable was 5/1 or 10/1, which included 14 observations and therefore the sample size is at least 14*5 = 70 observations. Green (1991) suggested that in the regression model the minimum sample size was determined by the empirical formula 50 + 8*independent variables in the model. This study has 08 independent variables so the sample size is at least 50 + 8*8 = 114 observations. Statistical analysis of the data obtained in this study was performed using the SPSS 22.0 software tool.
The sample for the study was drawn conveniently from 240 consumers in Ho Chi Minh city which represent different geographic, cultural, and commercial backgrounds of Vietnam. The stratified sampling plan was followed based on the population distribution in the districts of the city in order to ensure the representation of the research sample. Personal interviews were conducted at the large supermarkets by university students recruited from the city. The raw data set was publicly available at https://data.mendeley.com/datasets/dwj7hg3dt2/1 The research required that we examine a product category in which a domestic alternative was available. Apparel products were chosen as the domestic product category in this study.
The result of the survey showed that 121 cases (50.4 percent) purchased domestic apparel regularly while 119 cases (49.6 percent) did not. Demographic variables were described in Table 1

Reliability and Validity
The reliability and validity of indicators in the model are tested by the system of criteria. As can be seen from Table 2, the lowest Cronbach's alpha value is 0.712, exceeding the cut value of 0.70 recommended (Hair, Black, Babin, & Anderson, 2010). Convergent validity was estimated by factor loading. The value of Kaiser-Mayer-Olkin (KMO) was 0.833 (between 0.5 and 1.0) which means that data is significant for conducting a factor analysis. All loadings of variables are higher than the 0.50 (see Table 2). According to Hair, Black, Babin, and Anderson (2010), loadings ± 0.50 or greater are considered practically significant.

PP1
The price of domestic apparel is more acceptable than foreign manufactured products. 0.626

PP2
Compared to the quality, the price of domestic apparel is cheaper. 0.744

PP3
The amount of money to buy domestic apparel is perfectly suited to me personally. 0.793

PP4
The amount of money to buy domestic apparel compared with foreign manufactured products is reasonable. 0.549  The probabilities presented in Table 3 show the effects of changes in the individual explanatory variables on the likelihood of consumer purchase decision, assuming that all other explanatory variables are set to zero. The likelihood of the consumer purchase decision increases by 18.52 percent if consumers get one more unit in their ethnocentrism. The likelihood of the consumer purchase decision increases by 17.88 percent if consumers have one more unit in their perceived quality of domestic apparel. If consumers have one more unit in their perceived price of domestic apparel, there is a 19.68 percent increase in the likelihood of purchasing decision.

Hypothesis Testing
Among the demographic variables, the coefficients related to the families with a child, with a high school education or above, younger consumer, female consumer, and lower income have impacts on consumer purchase decision and are significant at the 10%, 5%, and 1% level. Probabilistically, those families with a child were 19.85% more likely to buy domestic apparel regularly than others. Those with a high school education or higher were 33.53% more likely to buy domestic products regularly than were those with less than high school education. Similarly, those who are younger consumers, female consumers, and lower income consumers were 9.37%, 5.93%, and 4.30% more likely to buy domestic apparel, respectively.
Facing the context of Vietnam's extensive international integration, competition between domestic and foreign manufactured goods is fierce. Consumer purchase decisions for domestic products play an im-portant role in the competitiveness of domestic enterprises. This study suggests that the research model is quite adequate when assessing the impact of behavioral, psychological, and demographic factors on consumer purchase decision of domestic product. The research results contribute significantly to the current theoretical framework in clarifying consumer purchase behavior for domestic goods in one of the emerging markets. The research results are similar to the results of previous studies cited in the literature review of this article.
Based on the research results, the forecasting model of domestic apparel purchase by theVietnamese consumer could be shown as

CONCLUSIONS
The study uses the Binary logistic econometric model to explain the factors affecting the Vietnamese consumer purchase decision of domestic apparel based on the data obtained from 240 consumers in Vietnam. The results show that consumer purchased decision was significant influences by i) the perception of price, ii) the perception of quality, iii) consumer ethnocentrism, and iv) demographic variables (age; families with a child; the level of education; income; and sex). This study provides valuable implications for business in the domestic market in the context of Vietnam's international integration. Domestic firms must recognize the significant relationship between consumers' perceived price, perceived quality of domestic products, consumer ethnocentrism, and their demographic variables with their purchase decision. Positive perceptions of quality and price are important for the long-term success of a brand. Therefore, it is recommended that domestic firms employ aggressive strategies to improve consumer perception of their brands in terms of quality and price appeal among those who have a higher interest in domestic products. This study has its limitations which can be resolved in future studies. Firstly, this study only conducted a product survey in one area of Vietnam, and therefore it is possible that the sample is not representative. In future studies, a wide range of products should be examined, and in more areas too, such as urban and rural ones. In addition, a comparison should be made with as many countries as possible so as to have more multidimensional assessments of consumers' domestic purchases; Secondly, this study uses the Binary Logistic model in accordance with the nature of the dependent variable. However, consumers' buying behavior also involves buying intentions, willingness to pay, satisfaction and loyalty after purchase. Therefore, further research should consider including these variables in the research model which are based on other quantitative models such as the structural equation model.