The Influence of The Electronic Word of Mouth (E-Wom) Dimension in Instagram on Purchasing Decisions At Kedai Ice Kepal in Mataram

: This study aimed at assessing the impact of Electronic Word of Mouth aspects on purchasing decisions at Kedai Ice Kepal. The study utilized purposive sampling in collecting data. The sample size for this study consisted of 100 participants who were active users of Instagram platform and had previously followed the @Ice_Kepal_Mataram Instagram account and made a purchase at the Kedai Ice Kepal. According to the finding, at a significance level of 5%, the intensity dimensions have a favorable impact on purchase decisions. However, this impact is not statistically significant, as indicated by a significant value of 0.933, which is bigger than the threshold of 0.05. The content dimension has a favorable effect on purchase decisions, although this effect is not statistically significant. This is supported by the significance value of 0.103, which is bigger than the threshold of 0.05. The positive valence dimension exerts a positive and statistically significant impact on purchasing decisions, as indicated by a significant value of 0.000, which is below the threshold of 0.05. The negative valence dimension has a favorable effect on purchase decisions, although this effect is not statistically significant. This is supported by a significant value of 0.813, which is bigger than the threshold of 0.05. The test findings for the coefficient of determination (R2) indicate a value of 0.435, which corresponds to 43.5%. The independent variable has a 43.5% impact on the dependent variable, whereas the remaining portion is influenced by other variables that were not considered in this study.


Introduction
The culinary business, especially in the city of Mataram, is currently experiencing rapid growth.This can be evidenced by the proliferation of culinary establishments such as restaurants, eateries, cafes, food stalls, and others that are commencing their operations.Many of these culinary businesses make use of social media as a means of promotion, product reservations, and various other purposes.Consequently, the level of competition is on the rise, rendering the landscape increasingly complex.With an unrestricted market segmentation, companies have the opportunity to expand their market networks.However, Every consumer who experiences dissatisfaction after consuming a product will express negative opinions about the product, posting negative reviews about the products they have consumed on the content provided by the company.This can adversely affect the company, as such negative reviews can influence other consumers not to try the product.H4: The negative valence dimension has a positive and significant impact on purchase decisions at Kedai Ice Kepal in Mataram City.
The theoretical framework proposed to demonstrate the influence of Electronic Word of Mouth (e-WOM) on purchase decisions at Kedai Ice Kepal can be depicted in the following conceptual framework.The independent variables in this study consist of dimensions within Electronic Word of Mouth, namely Intensity (X1), Content (X2), Positive Valence (X3), and Negative Valence (X4).Meanwhile, the dependent variable in this study is Purchase Decision (Y).

Research Method
The data required for this study are primary data obtained from questionnaires and direct interviews with 100 respondents who have Instagram accounts, follow the Instagram account, and have made purchases at Kedai Ice Kepal.The multicollinearity test aims to examine whether there is a correlation between independent (predictor) variables in the regression model.Variables that cause multicollinearity can be determined by looking at the tolerance value or the variance inflation factor (VIF).If the tolerance value is >0.1 or the VIF is less than 10, then there is no multicollinearity (Ariefianto, 2012).

Hypothesis Testing a. Multiple Linear Regression Analysis
This analysis examines the relationship between two or more independent variables and the dependent variable.This research utilizes multiple regression analysis as the analytical tool because the variables under investigation consist of more than one independent variable, namely, Intensity Dimension (X1), Content (X2), Positive Valence (X3), and Negative Valence (X4) on Purchase Decisions (Y).The regression equation can be formulated as follows: Y = α + b1X1 + b2X2 + b3X3 + b4X4 + e b.F-Test This is the initial step to identify whether the estimated regression model is appropriate or not.If the p-value of the F-statistic (shown in the SPSS output under the "sig."column) is less than the significance level (alpha) of 0.05 (as predetermined), it can be concluded that the estimated regression model is suitable.
Conversely, if the p-value of the F-statistic is greater than 0.05, it can be said that the estimated regression model is not appropriate.

c. t-Test
The hypothesis testing is formulated as follows: d.The influence of Intensity on purchase decisions H01: β1 ≤ 0, meaning there is no positive influence of Intensity on purchase decisions.
Ha1: β1 > 0, meaning there is a positive influence of Intensity on purchase decisions.e.The influence of Content on purchase decisions H02: β2 ≤ 0, meaning there is no positive influence of Content on purchase decisions.Ha2: β2 > 0, meaning there is a positive influence of Content on purchase decisions.f.The influence of Positive Valence on purchase decisions H03: β3 ≤ 0, meaning there is no positive influence of Positive Valence on purchase decisions.
Ha3: β3 > 0, meaning there is a positive influence of Positive Valence on purchase decisions.g.The influence of Negative Valence on purchase decisions H04: β4 ≤ 0, meaning there is no positive influence of Negative Valence on purchase decisions.
Ha4: β4 > 0, meaning there is a positive influence of Negative Valence on purchase decisions.
The overall significance of the regression coefficient is tested using the t-Test by examining the significance level of the SPSS calculation:-H0 is rejected and Ha is accepted if the significance value is below 0.05.-H0 is accepted and Ha is rejected if the significance value is above 0.05.

Coefficient of Determination Test (R^2)
The coefficient of determination (R^2) essentially measures how well the model explains the variation in the independent variables.

Testing Assumptions Validity Test
Table 2 The results of the reliability test from the table show that all Cronbach's Alpha coefficients are >0.60.Therefore, it can be concluded that the instruments are reliable (consistent) for use as research instruments.It can be observed that the data distribution shows a spread around the diagonal lines and follows the direction of the diagonal lines.Therefore, it can be concluded that the regression model produced meets the assumption of normality.Multicollinearity Test Source: Data processed with SPSS 23, 2019 (Appendix 6) By examining the table above in the significance column, it can be explained whether the four independent variables are significant or not as follows:

Classical Assumption Testing Normality Test
1. H1: The dimension of intensity has a positive and significant effect on purchase decisions at Kedai Ice Kepal in the city of Mataram.
From table 4.9, it is evident that the significance value for the intensity variable is 0.933, which is greater than 0.05.Therefore, it can be concluded that the intensity variable has no significant effect on purchase decisions.This means that the first hypothesis in this study is rejected.Statistically, the regression coefficient for the intensity variable (X1) is -0.010.A negative coefficient indicates that the intensity variable (X1) has an opposite relationship with purchase decisions (Y).
Based on the partial testing results, as shown in the table, the significance value for the intensity variable is 0.933, which is greater than the probability value of 0.05, and it is positive.Therefore, Ho is accepted, and Ha is not fully accepted or supported by the analyzed research results.It can be concluded that the intensity variable and the purchase decision variable are not related or do not have a significant impact.This means that providing opinions or comments on the Instagram social media account @ice_kepal_mataram more frequently does not necessarily lead to higher purchase decisions at Kedai Ice Kepal.
The results of this study reject or do not support some findings from previous research conducted by Sari (2012), Mustikasari and Widanisngsih (2016), Yunitasari (2018), Adeliasari, Ivana, and Thio (2014), which stated that intensity has a significantly positive effect on purchase decisions.The data collected by the researcher did not confirm the hypothesis.2. H2: The dimension of content has a positive and significant effect on purchase decisions at Kedai Ice Kepal in the city of Mataram.
From the table, the significance value for the content variable is 0.103, which is greater than 0.05.Therefore, it can be concluded that the content variable has no significant effect on purchase decisions.This means that the second hypothesis in this study is rejected.Statistically, the regression coefficient for the content variable (X2) is 0.193.A positive coefficient indicates that the content variable (X2) has a positive relationship with purchase decisions (Y).
Based on the testing results, as shown in the table, the significance value for the content variable is 0.103, which is greater than 0.05, and it is positive.Therefore, Ho is accepted, and Ha is not fully accepted or supported by the analyzed research results.It can be concluded that the content variable and the purchase decision variable are related and have a positive relationship, but it is not statistically significant.This means that even though the content on the Instagram social media account @ice_kepal_mataram is considered informative by respondents, it does not have a significant impact on purchase decisions.
The results of this study reject or do not support some findings from previous research conducted by Sari (2012), Mustikasari and Widanisngsih (2016), Yunitasari (2018), Adeliasari, Ivana, and Thio (2014), which stated that content has a significantly positive effect on purchase decisions.3. H3: The dimension of positive valence has a positive and significant effect on purchase decisions at Kedai Ice Kepal in the city of Mataram.
From the table, it is evident that the significance value for the positive valence variable is 0.000, which is smaller than 0.05.Therefore, it can be concluded that the positive valence variable has a significant and positive effect on purchase decisions.This means that the third hypothesis in this study is accepted.Statistically, the regression coefficient for the positive valence variable (X3) is 0.522.A positive coefficient indicates that the positive valence variable (X3) has a positive relationship with purchase decisions (Y).
Based on the testing results, as shown in the table, the significant value for the positive valence variable is 0.000, which is smaller than 0.05, and it is positive.Therefore, Ho is rejected, and Ha is accepted.
Thus, it can be concluded that the hypothesis that positive valence has a significantly positive effect on consumer purchase decisions at Kedai Ice Kepal is supported by the analyzed research data.These two variables are related and have a positive relationship, meaning that the more consumers provide positive comments or opinions on social media like Instagram, the higher their purchase decisions.This occurs because many consumers who post positive comments or opinions about Kedai Ice Kepal on Instagram can be read by other Instagram users, which reinforces consumers' decisions to purchase Kedai Ice Kepal products.
The results of this study support previous research findings, such as those conducted by Sari (2012), Mustikasari and Widaningsih (2016), Adeliasari, Ivana, and Thio (2014), which stated that positive valence has a significantly positive effect on purchase decisions.However, this study contradicts the findings of Yunitasari (2018), who stated that positive valence does not have a significantly positive effect on purchase decisions.4. H4: The dimension of negative valence has a positive and significant effect on purchase decisions at Kedai Ice Kepal in the city of Mataram.
From the table, the significant value for the negative valence variable is 0.813, which is greater than 0.05.Therefore, it can be concluded that the negative valence variable has no significant effect on purchase decisions.This means that the fourth hypothesis in this study is rejected.Statistically, the regression coefficient for the negative valence variable (X4) is -0.018.A negative coefficient indicates that the negative valence variable (X4) has an opposite relationship with purchase decisions (Y).
Based on the testing results, as shown in the table, the significant value for the negative valence variable is 0.813, which is greater than 0.05, and it is positive.Therefore, Ho is accepted, and Ha is not fully accepted or supported by the analyzed research results.It can be concluded that the negative valence variable and the purchase decision variable are not related or do not have a significant impact.This means that the frequency of social media users posting negative comments or opinions on Instagram does not significantly determine an increase in consumer purchase decisions for Kedai Ice Kepal products.
The results of this study support and reinforce the findings of Yunitasari ( 2018), who stated that negative valence does not have a significantly positive effect on purchase decisions.This study also contradicts the findings of some previous research, such as Sari (2012), Mustikasari andWidaningsih (2016), andAdeliasari, Ivana, andThio (2014), which stated that negative valence has a significantly positive effect on purchase decisions.

Coefficient of Determination Test Results
Table Based on the table above, we obtained an R-value of 0.659.This result indicates a strong correlation or relationship between the variables Intensity (X1), Content (X2), Positive Valence (X3), and Negative Valence (X4) with the purchase decision variable (Y) of 65.9%, implying a close relationship.The higher the R-value, the stronger the relationship between the independent and dependent variables.With an R-value of 0.659, the coefficient of determination (R Square) is 0.659 x 0.659 = 0.435.This means that the ability of the independent variables, namely Intensity (X1), Content (X2), Positive Valence (X3), and Negative Valence (X4), to explain the variance of the dependent variable, which is the purchase decision, is 43.5%.This implies that 55.6% (100% -43.5%) of the variance in the dependent variable is explained by other unexamined factors.Meanwhile, the Adjusted R Square is 0.411, indicating that the contribution of the independent variables, Intensity (X1), Content (X2), Positive Valence (X3), and Negative Valence (X4), to the purchase decision is 41.1%, while the remaining 58.9% (100% -41.1%) is explained by other unexamined variables.

Conclusion
Intensity is not correlated or linearly related to purchasing decisions and does not have a significant impact on purchasing decisions at Ice Kepal Store in Mataram City.This means that the more respondents access information on Instagram and the more often they provide opinions or comments on the Instagram social media @ice_kepal_mataram, it does not necessarily imply a higher likelihood of making a purchase at Ice Kepal Store.Content is correlated and linearly related to purchasing decisions but does not have a significant impact on purchasing decisions at Ice Kepal Store in Mataram City.This implies that although the content found on the Instagram social media @ice_kepal_mataram is considered informative by respondents, it is not significant in influencing purchasing decisions.Positive Valence is correlated and linearly related to purchasing decisions and significantly influences consumer purchasing decisions at Ice Kepal Store in Mataram City.This means that the more frequently consumers provide positive comments or opinions on Instagram social media about Ice Kepal Store, the higher the likelihood of consumers making a purchase.Negative Valence is not correlated or linearly related to purchasing decisions and does not have a significant impact on purchasing decisions at Ice Kepal Store in Mataram City.This means that fewer Instagram users writing negative comments or opinions does not determine an increase in the number of consumers making purchasing decisions to buy Ice Kepal Store products.

Recommendations
As a company, Ice Kepal Store should pay more attention to the appearance of the Instagram account @ice_kepal_mataram, making it more creative in designing Instagram posts, and optimizing the quality of the products offered in posts.This includes uploading newer and previously unposted content, making the Instagram account more attractive to followers, and increasing engagement from respondents compared to other online businesses.Ice Kepal Store should provide more informative content related to the products to be uploaded, such as providing clearer information about the composition and benefits of the products, rather than just mentioning the product name and price.This will make the products offered by Ice Kepal Store more engaging to respondents, providing them with knowledge and experiences about the products they intend to purchase.Ice Kepal Store should host promotions on Instagram, including using Instastories and direct Instagram posts, to promote product purchases through supporting apps like Go-Jek and Grab.Examples of promotions could include discounts with a minimum purchase of Rp30,000, "buy two get one free" for all product types, and other promotional offers.By doing so, respondents will be more inclined to recommend Ice Kepal Store products to others through supporting apps like Go-Jek and Grab on Instagram.Ice Kepal Store should work on minimizing negative comments from consumers and focus more on the points mentioned above.This will help consumers feel secure and comfortable in making purchasing decisions as they won't encounter negative comments about Ice Kepal Store on Instagram.Steps to minimize negative comments include being more creative in Instagram post design, providing informative content, optimizing the quality of posts by sharing new and previously unposted product images, and hosting promotions more frequently.This will encourage consumers to leave more positive comments about Ice Kepal Store for others to see.Future research should consider using other research theories besides Goyette et al, such as theories by Eunha Jeong and Hennig-Thurau et al.Research on Electronic Word of Mouth (e-WOM) should explore variables that haven't been utilized by researchers, such as "Helping the company" and "Concern for others" from Eunha Jeong's theory or "Platform assistance" and "Economic incentives" from Hennig-Thurau et al's theory and other e-WOM researcher theories.Future research should involve a larger sample size than the current study.Research questions arising from the analysis of respondent characteristics may include inquiries such as "Do women indeed post more reviews about food and beverages on Instagram social media?" Figure 1.Results of Normality Testing for the Research Variable Using Normal P-P Plot

Table 1 Operational Definitions of Variables
is conducted by examining the points on the Normal P-Plot of Regression Standardized Residual of the dependent variable.The requirement for the normality test is that if the data is spread around the diagonal line and follows the direction of the diagonal line, then the regression model satisfies the assumption of normality.b.Multicollinearity Test

. Results of Validity Testing for the Variables: Intensity, Content, Positive Valence, Negative Valence, and Purchase Decision
The correlation coefficients for all statement items of the variable are >0.3, which indicates their validity for use as research instruments.Reliability Test

Table . 4 Results of Multicollinearity Testing Coefficients a
Primary data processed with SPSS 23 in 2019From the table above, it is evident that the VIF (Variance Inflation Factor) values are below 10, and the tolerance values are above 0.1, indicating that each independent variable is free from multicollinearity issues.

Summary of Multiple Linear Regression Analysis Results and Regression Equation Coefficients a
on the calculation results, the explanations are as follows: a.The constant value of 8.266 means that when the values of the intensity (X1), Content (X2), Positive Valence (X3), and Negative Valence (X4) variables are all zero, the purchase decision at Kedai Ice Kepal (Y) has a value of 8.266.b.The regression coefficient for the Intensity variable (b1), which is -0.014, signifies that if the intensity (X1) increases by one unit, the purchase decision will decrease by -0.014, but it is not statistically significant.c.The regression coefficient for the Content variable (b2), which is 0.191, means that if the Content variable (X2) increases by one unit, the purchase decision will increase by 0.191, but it is not statistically significant.d.The regression coefficient for the Positive Valence variable (b3), which is 0.676, indicates that if the Positive Valence variable (X3) increases by one unit, the purchase decision will significantly increase by 0.676.e.The regression coefficient for the Negative Valence variable (b4), which is -0.044, suggests that if the Negative Valence variable (X4) increases by one unit, the purchase decision will decrease by -0.044, but it is not statistically significant.F-Test Results