Data Analysis Results : Trust Level And Purchase Intention

DATA ANALYSIS RESULTS : TRUST LEVEL AND PURCHASE INTENTION Λ Θ KP1 0.594 0.334 KP2 0.594 0.722 Trust Level KP3 0.678 0.433 (TRUST) KP4 0.706 0.471 KP5 0.665 0.421 KP6 0.747 0.341 KP7 0.614 0.546 KP1 0.594 0.334 KP2 0.594 0.722 KP3 0.678 0.433 KP4 0.706 0.471 KP5 0.665 0.421 KP6 0.747 0.341 KP7 0.614 0.546 Variables Purchase Intention (INTEREST) Code (Σ λ)2 Σθ C R 1,713 2,895 0,881 2,419 4,472 0,789 Source: Results of Data Processing with AMOS 20 (2018) Based on the results of reliability testing with construct reliability (CR), it was found that all variables obtained a CR value> 0.7. Thus referring to the results of validity and reliability testing, so that the data in this study can be used for further testing. 3. Data normality test The normality test in this study was carried out using Kolmogorov- Smirnov (K-S) analysis using SPSS 22. The test results are said to be normally distributed if the significance value (sig.) > 0.05, and vice versa the data is said to be not normally distributed if the significance value (sig.) <0.05. The processing results obtained a significance value (sig.) on substructure 1 and substructure 2 of 0.107 and 0.490, respectively. These results indicate that the significance value of the two substructures is above 0.05. Thus it can be concluded that the data in this study are normally distributed. 4. Research Model Test with Structural Equation Model (SEM) After conducting validity tests, reliability tests, and data normality tests, researchers conducted data analysis and analyzed the relationship between variables using SEM analysis techniques with the AMOS version 20 program. Data analysis using SEM requires raw data that has been entered into excel format to be processed using the AMOS program. a. Developing a Theoretical Model The model used in this study is a model adopted from a research model conducted by Ghadikolaei (2016) and Ambolau et.al. (2015) which states that purchase intention is influenced by halal labeling, brand image, product quality and the level of trust from consumers. This research model is as follows: b. Developing a Flowchart The next step after compiling the theoretical model along with the variables and indicators is to create or develop a path diagram. c. Convert Flowchart into Structural Equation Structural equations in SEM are carried out to express the relationship between causality between various constructs. The formula developed is endogenous variable = exogenous variable + endogenous variable + error, this step will automatically be performed by AMOS 20. The flowchart in this study consists of three main independent variables, namely halal labeling, brand image, and vitamin quality, as well as two dependent variables, namely the level of trust and purchase intention. These variables are measured by indicators in the form of questions in a questionnaire in the form of a Likert scale of 1-4. d. Conduct Good of Fit Test Structural model analysis in SEM begins with testing the overall model fit as seen based on the Goodness-of-fit Index (GFI) statistics from AMOS output.21 Overall, a summary of the critical values of the overall model fit test can be seen from the summary in Table 5. MODEL FIT TEST RESULTS : GFI VALUE Degree of Fit Measure Goodness of Fit Indices (GFI) Root Mean Square Error of Approximation (RMSEA) Normed Fit Index (NFI) Adjusted GFI (AGFI) Comparative Fit Index (CFI) Incremental Fit Index (IFI) Relative Fit Index (RFI) The level of fit that can Retrieved Value Ket GFI > 0.9 0,906 Good Fit RMSEA ≤ 0.08 NFI > 0.90 AGFI ≥ 0.09 0,126 Poor Fit 0,966 0,948 Good Fit Good Fit CFI > 0.6 0,653 Marginal Fit IFI > 0.90 RFI > 0.90 0,958 0,930 Good Fit Good Fit From this data, it can be seen that the GFI value of 0.906> 0.90 indicates that the model is a good fit. The RMSEA value of 0.126> 0.08 indicates that the model is poor fit. The RMSEA value which is above the 0.08 requirement indicates that the model fit is weak in terms of the root mean or Chi-Square value. However, it is not entirely that the model in this study can be said to meet the specified level of fit, because from all tests, in other tests such as CFI or in measuring the degree of fit by determining the CFI value, it shows that the model is already Fit which refers to a model said to be good if it has a CFI value close to 1 and 0.653 indicates that the model is fit. Therefore, the research model can still be continued or can be used as modeling in this study.23 Measurement of the degree of fit with NFI of 0.966 indicates that the model fits. The results of the fit model are emphasized by obtaining an IFI value of 0.958, where the IFI value close to 1 indicates that the data is fit. The measurement of the degree of fit by determining the AGFI value of 0.948 indicates that the model is moderate Fit.23 Meanwhile, regarding the Relative Fit Index (RFI), it states that the RFI ranges from 0 to 1 where a value close to 1 indicates a fit model. 23 The model in this study obtained an RFI value of 0.930, so it can be said that the model is close to 1 so it can be said that the model is fit. Based on the results of testing the degree of fit above, it can be concluded that the model in this study is fit with the data. e. Model Interpretation After conducting the results of the research model, structural analysis is also carried out. Structural analysis is an analysis with AMOS that contains a variety of Information. Hypothesis testing in this study was carried out by looking at the critical value (CR) at the 95% confidence level or 5% error, the accepted CR value is 1.96.21 The three information above can be seen in the results of data processing with AMOS in the form of structural equations: