BUSINESS STUDIES ipmi The Impact of Country of Origin Image and Consumer Ethnocentrism Toward Purchase Decision of Casual Apparel in Young Adult

The objective of the study is to analyse the effect of Leverage, Pro�itability, and Market Value on Stock Prices. The study used quantitative approach in which the population were hotel, restaurant, and tourism companies listed on the Indonesia Stock Exchange (IDX). Non-probability sampling method is used to collect 41 hotel, restaurant, and tourism companies. Panel data regression analysis method use Eviews 11.0 program to test the hypothesis. This study showed that Leverage measured by Debt to Equity (DER) has no effect on Stock Prices, Pro�itability is measured by Return on Equity (ROE) has a positive effect on Stock Prices, and Market Value is measured by Price Earnings Ratio ( PER) has no effect on shares prices.


INTRODUCTION
According to the Ministry of Tourism and Creative Economy (KEMENPAREKRAF) in 2020, Indonesian fashion products are still facing challenges in competing against imported products. The market still prioritizes its space for imported products, leaving not enough room for local fashion to grow rapidly. Alongside that, local fashion also needs to keep the synergy of the supply chain from upstream to downstream, starting from the textile/garment factories, fashion designers to market affairs. Public mindset also needs to be shifted from thinking that local products are not comparable to foreign brands. Indonesian textile and apparel industry is very pronounced globally. Indonesia contributes a large amount of export around the world. According to Textile Focus (2019), Indonesia is among the top 10 textile and apparel producers in the world and with that being the 12th of the leading textile and apparel exporters in ASEAN region, contributing up to $13.8 billion export earning in 2019. The number is optimistically expected to increase and reach 5% global market share of textile and apparel product's in 2030. Data showed that the industry has become more competitive in both domestic and international markets, referring to 2018's growth rate of 8.75%. Indonesian Government has also been making tremendous effort in boosting local product purchases with the aim of making domestic consumption as the main pillar of Indonesian economy (Al Hikam, 2020). According to Riani (2020), local products seem to no longer be the underdog due the rise of information technology. The ever so growing easiness to access information has been strategically utilized by local brands to promote their products through social media in�luencers, actresses and actors who are painting new pictures for local products, creating a sense of pride and a sense of contribution towards the domestic economic chain, especially during the current pandemic situation. Despite the pandemic, Indonesian designers claim that demand for apparel remains. Thus, the use of social media and e-commerce are optimized. However, as seen on data in 2019, ninety percent of Rp. 144 trillion in Indonesian e-commerce transactions are imported products (Andriani, 2020;Gareta, 2019) with 19.57 million packages counted by Indonesian Custom in Import Document Records. This �igure has increased sharply in 2019, reaching up to 49.69 million packages with 254 percent nominal growth compared to 2018 (Hidayat, 2019). This data was taken before the Ministry of Finance, Sri Mulyani, revised the threshold value for import shipments from US$ 75 per shipment to US$ 3 per shipment, in the effort to protect Indonesian SMEs (Hidayat, 2019). According to Utami (2020), four reasons why Indonesian consumers may still prefer imported products include perceived low quality over price, less attractive packaging, lack of product innovation and non-strategic store locations. Additionally, Lembaga Ilmu Pengetahuan Indonesia (LIPI), revealed that consumers buying more of imported products are often due the scarcity of the products and/or have lower prices for the same type of products. They added that the trend of buying imported products is also a result of availability and easy access to purchase imported products online and even having giant local e-commerce to accommodate and facilitate foreign sellers (Sa�itri, 2019). Local brand appeal public awareness about local products and its geographical indication (IG) for Indonesian consumers is still considered low (Yulistara, 2018;dgip, 2020) Directorate General of Intellectual Property (DGIP) even went as far as claiming that Indonesian consumers would worry about the authenticity of Indonesian products when other countries start to claim them as theirs (dgip, 2020). The Head Section of Geographical Property (DJKI) and Ministry of Law and Human Rights (KEMENKUMHAM) believe that Indonesian people need to increase the effort of appreciating local products, look after them, and protect their authenticity (dgip, 2020). Preliminary research was conducted through a social media platform on June 2020 (using a polling system) to see whether DKI Jakarta consumers would prefer Indonesian or non-Indonesian apparel brands when compared to one another within the same clothing segment. From the result, it can be seen that the majority of the people participating in

LITERATURE REVIEW Country of Origin
Country-of-Origin has been widely used by organizations through their product's labels with indicators such "made in", "country of assembly", "country of manufacture", "designed in" or "produced in" as a cue of product quality to consumers (Amine et al., 2005). According to Azadi et al. (2015), Country-of-Origin in�luences consumers evaluation through the effect of original manufacturer's expertise, which is considered an intangible barrier for products as they are entering a market. This is based on the consumers' country stereotyping upon the perceived image of the country. This notion is backed by the result from Pappu et al. (2006) which indicated that when consumers value a brand, it is signi�icantly associated with the images of the country-origin of the brand. In other words, consumers evaluate from which country the brand is coming from and associate it with their personal scheme of the country. Hence, Country-of-Origin can be said to in�luence consumers' perception of quality (Darley and Lim, 1994). It becomes a salient attribute within consumer evaluation that could affect consumer interest, behavioral intention, feelings, and decision making (Azadi et al., 2015). With differences on the level of country development across the world, some countries own the advantages of technological advances and competitive nature of the markets. On the other hand, developing countries are considered to have less of these bene�its. With that, consumers around the world are conferred to a tighter evaluation towards product quality (Kalicharan, 2014), because paying attention to the origin of the products become a part of their evaluation and purchasing decision process (Bandyopadhyay and Banerjee, 2002;Bhuian, 1997;Cordell, 1991;Ettenson and Klien, 1998;Katsanis and Thakor 1997). Another study further supported this notion that Country-of-Origin directly affects consumers' perception of product quality, attitudes, behavior, purchase intention, and decision making (Teas and Agrawal, 2000;Baker and Ballington, 2002). Based on this founding, it can be assumed that Indonesia, as a developing country, has less of the bene�it of owning the advantage of competitiveness in the market. However, several recent studies argued contradictory with these �indings. According to Diamantopolous (2011), Country-of-Origin is not a salient cue in consumer decision making and that it was not found to contribute as a determinant of consumer's purchase intention. Several empirical studies also showed that it is common for consumers to not know the (correct) Country-of-Origin of many well-known brands, which suggests that consumers might �ind Country-of-Origin information unimportant and thus unworthy of memory retention or consumer have limited recognition of Country-of-Origin (Samiee et al., 2005;Diamantopoulos, 2009). Furthermore, Liefeld (2004) found in surveys on the use of Country-of-Origin as an informational cue that Country-of-Origin counts for only a small portion of consumers' product choice and that for the majority of the respondents considered Country-of-Origin as "is not a relevant attribute for making choices between alternatives". Agrawal and Kamakura (1999) suggested that as consumers get closer to the actual choice behavior, the effect of Country-of-Origin is likely to decline. This is because, in actual day-to-day setting, Country-of-Origin has to compete with other informational cues presented by the product, which consequently making the relevance of Country-of-Origin effect on actual consumer decision making behaviour likely to be small. These contradicting �indings become one of the interests of this presented paper to explore, particularly in Indonesia.
International Journal of Business Studies Vol. 6 No. 2 (February 2022)

Consumer Ethnocentrism
Another variable that has been tested many times to affect consumers' purchase decision is Consumer Ethnocentrism. A study by Kalicharan (2014) revealed that countries with high ethnocentrism prefer products that are produced locally. Although, it is notable that with globalization, these biases are becoming less and less apparent, however country-speci�ic factors are still commonly found and recognized (Ahmed & D' Astous, 1996;Bhuian, 1997). However, a study in Indonesia reported that Country-of-Origin was found to be more important than price in respect to purchase intention (Hamin, 2006). This was proven true for both goods as well as services and it was tested to consumers with both low and high Consumer Ethnocentrism. Results of the study showed that Consumer Ethnocentrism might not seem to in�luence their purchase decision, especially when Country-of-Origin information was available. The study suggested that even when Indonesian consumers scored high in Consumer Ethnocentrism, they evaluated products from MDC's higher than others from LDCs. This is contradicting other theories which prove consumers in countries with high ethnocentrism prefer local products. Thus, it will be explored more in this paper.

METHODOLOGY
The population of this study includes young adult consumers in DKI Jakarta, ranging from 16-30 years old (BPS, 2019). Non-probability, purposive sampling was utilized with 100 participants collected because the targeted samples have to suit the purpose of this study. other characteristics include living in DKI Jakarta with valid Kartu Tanda Penduduk (KTP), and minimum yearly expense for apparel Rp. 500.000, referring to the average of clothing, footwear, and headgear expenditure per capita per month in DKI Jakarta (BPS, 2020). All measurements were taken from items which had been used and tested in previous studies. The presented study adopts the method used in Reinaldo (2020) to test Purchase Decision. The items are measured using 5-Likert scale (1=strongly disagree, 5=strongly agree). For the independent variables, this study considers the components stated by Costa (2016) to measure Indonesian Country-of-Origin Image with 6-items, rated with 5-Likert scale (1=strongly disagree, 5=strongly agree). Finally, to test Indonesian Consumer Ethnocentrism level, the study is using 17-items CETSCALE with 5-Likert scales (1=strongly disagree, 5=strongly agree) developed by Shimp and Sharma (1987) to measure consumers' ethnocentric tendencies, which the �irst 9-items has been adopted by Syabanita & Hanfan (2019) to �it Indonesian samples. A selfadministered questionnaires were distributed through Google From by the Author to the targeted samples. There are sections within the questionnaire that participants must complete. The �irst section consists of participants demographic information, which includes sex, age, Indonesian ID (KTP) ownership, location, and annual spending. The second section intends to measure Indonesian Country-of-Origin Image towards Indonesia. The third section intends to measure the level of Indonesian Consumer Ethnocentrism. Lastly, the fourth section intends to measure Indonesian willingness to purchase Indonesian-made casual apparel.

Research Variables
There are two types of variables and a total of three used in this study, which includes independent variables (Country-of-Origin Image of Indonesia and Indonesian Consumer Ethnocentrism) and a dependent variable (Indonesian Purchase Decision when it comes to buying Indonesian-made apparel). Dependent variable refers to the main variable that the author is interested to measure, whereas independent variables are the variables will-positively or negatively--impact the dependent variables (Sekaran & Bougie, 2016)

Preliminary Test Validity and Reliability Test
The validity test would include assessment of Pearson correlation (r) to measure how well the observed distribution �its the expected distribution, t-test (level of signi�icance), and direction. The expected Pearson correlation would be -1 < r < 1 and the standardized level Purchase Decision of Casual Apparel in Young Adult / 101 -111 of signi�icance would be 5% (-1.96 < t < 1.96) (Pallant, 2020;Karras, 1997). Reliability would be tested with Cronbach's Alpha. The general rule of thumb for reliability test is when the value is at 0.6 to 0.7 to be considered reliable. Anything below 0.6 will be considered insuf�icient and a value higher than 0.7 will be considered suf�icient.

Classical Assumption Tests
Assumptions are needed to be made before testing for multiple regression.

Linearity Test
Linearity must be assumed; thus, the model should be linear. Scatterplot is used to test whether the independent variable has a linear relationship with the dependent variable or not. When independent does not appear to have linear relationship with the dependent relationship, then the variable cannot be tested with multiple regression. Further, p-value is checked to further con�irm the correlation of the variables.

Normality Test
Normality must be assumed; thus, the data should have normal distribution. Normality would be tested with Kolmogorov-Smirnov (p-value > 0.05). When p-value is greater than 0.05, the data is considered normally distributed and null hypothesis is accepted. Conversely, when p-value is less than 0.05, the data is considered not normally distributed and null hypothesis is rejected (Ghozali, 2011).

Multicollinearity Test
Multicollinearity must not be assumed; thus, the independent variable should not have high correlation with each other. Variance In�lation Factor (VIF) is used. To determine a linear regression, the value of VIF has to be between 2-5. A value larger than 5 may indicate that multicollinearity may be present. When the value is above 10, multicollinearity certainly presents.

Testing (Multiple Linear Regression Analysis)
The hypothesis would be tested quantitatively using multiple regression in SPSS version 25.
Multiple regression is used to predict whether the independent(s) variables affect the dependent variable. Within this study, the level of Indonesian Country-of-Origin Image toward Indonesia and Consumer Ethnocentrism are assessed through the questionnaire based on the operation de�inition stated above. The formula of multiple regression is as follow (Priyatno, 2014): Y = dependent variable X = independent variable a = intercept b = slope e = regression residual (error)

Simultaneous Test (F-Test)
The F-test was conducted to test the model using the f-value of the variables, which are Country-of-Origin Image of Indonesia (X₁) and Indonesian Consumer Ethnocentrism (X₂), and whether they are simultaneously having impact toward the dependent variable, that is Indonesian Purchase Decision when it comes to buying Indonesianmade apparel (Y).

Partial Test (T-Test)
T-test was used to test whether there is signi�icant difference of means between the independent variables, which are Country-of-Origin Image of Indonesia (X₁) and Indonesian Consumer Ethnocentrism (X₂) partially toward the dependent variable, that is Indonesian Purchase Decision when it comes to buying local apparel (Y).

Coef�icient of Determination (Adjusted R²)
An increase of independent variable will consequently increase the value of R². On the other hand, the value of adjusted R² can either raise or fall depending on the independent variable added (Ghozali, 2005). Thus, it is more appropriate to use adjusted R² for the purpose of this study. A correlation of 0 indicates no relationship at all and 1 indicates a perfect correlation (range of value: -1 < r < 1) (Pallant, 2020).

RESULTS AND DISCUSSION Descriptive Statistics
The questionnaire was distributed online with 141 responses. However, only 70.9% (100 responses) are eligible to use for further testing based on the requirements needed for the respondents. Seventy two per cent (72%) female were recorded to participate and the rest were male. One possible explanation on this is due to the limitations of this study, given the nature of the data collection technique that was heavily reliant on online method, the reality was that the data collected unintentionally consisting of mostly female participants. All of the participants ful�illed the age, nationality, domicile, minimum yearly expanse and frequency of purchasing local apparel products.

Preliminary Test Validity and Reliability Test
Pre-test (n=30) and post-test (n=100) for validity was run on all variables. The pre-test results showed all variables namely Country of Origin Image (COO Image), Consumer Ethnocentrism Tendencies Scale (CETSCALE) and Purchase Decision have positive correlations and are signi�icant with p < 0.05 level. Thus, all statements used within the variables are considered valid for further testing with larger sample size. All of the variables also indicated positive correlation and are signi�icant with p < 0.05 level and thus are considered valid for further testing. Pre-test (n=30) and post-test (n=100) for were also run for reliability with Cronbach's Alpha. All of the scores for Cronbach's Alpha in the pre-test showed reliable values (>.60) that are suf�icient for further testing with larger sample size. The post-test scores of Country of Origin Image (.760), Consumer Ethnocentrism Tendencies Scales (.919), and Purchase Decision (.687) also indicate reliable and suf�icient values (>.60), which indicates that all of variables can be further tested.

Classical Assumption Tests Linearity Test
Each independent variable was tested independently against the dependent variable for linearity. As seen on the data, the independent variables appear to have linear relationship with the dependent variable with upward trends suggesting positive relationships. Thus, the assumption is true and �it for testing.

Normality Test
Kolmogorov-Smirnov was run to test the normality of the data. The p-value showed .051 indicating that the data is considered to be normally distributed (p-value > 0.05) and the null hypothesis is accepted. Thus, the assumption is true and �it for testing. Additionally, Normal P-P Plot of Regression Standardized Residual for Dependent Variable was drawn. The points formed a relatively straight diagonal line suggesting no major deviations from normality. Chairiena, Heteroscedasticity Test Scatterplot was used to test and check the pattern which indicates heteroscedasticity. According to the result, the data seem to spread in a random and irregular pattern above and below zero on the Regression Standardized Predicted Value. This indicates that heteroscedasticity does not exist and that the residuals have constant variance (homoscedasticity). Additionally, as seen on the approximation of Breusch-Pagan statistic using SPSS, the result shows that the signi�icant value is well above 0.05. Thus, the assumption is true and �it for testing.

Multicollinearity Test
To check multicollinearity, Variance In�lation Factor (VIF) is used. For each independent variable, the tolerance showed .802 which is well above the cut-off points of <.10. Additionally, the VIF showed 1.246 and therefore does not violate the multicollinearity assumption for multiple regression analysis. Thus, the assumption is true and �it for testing.

Hypotheses Testing Simultaneous Test (F-Test)
The F-test was conducted to test the model and see if the independent variables simultaneously have impact toward the dependent variable.  (2017) in Colombia about Consumer Ethnocentrism, Country Image and Local Brand Preference in textile, apparel and leather industry found that when consumers have positive image toward their own country, they will likely to develop more ethnocentric behavior patterns. These �indings support the �inding of this study. Majority of the participants scored relatively high on consumer ethnocentrism and responded positively toward the image of Indonesia.
Although the model has been correct to support that the independent variables in predicting the dependent variable, the adjusted R2 for this model is rather small. One possible explanation for this is because there were only two variables used to predict the dependent variable. Referring back to Agrawal and Kamakura (1999), in an actual day-to-day setting, Country-of-Origin has to compete with other informational cues presented by the product, which consequently making the relevance of Country-of-Origin effect on actual consumer decision making behavior likely to be small. Hence, more independent variables can be added to better predict the dependent variable. Companies can continue to improve embracing and utilizing local culture (kearifan lokal) as part of their sales and marketing strategies to maximize their competitive advantage in the DKI Jakarta market.