Global Journal of Management and Business Research, E: Marketing, Volume 23 Issue 2

Consumers’ Food Delivery Apps (FDAs) Continuance Intention: An Empirical Investigation using the Extended UTAUT2 Model 9 Global Journal of Management and Business Research Volume XXIII Issue II Version I Year 2023 ( )E © 2023 Global Journals agree” to measure the items for the latent constructs. All of the items were derived from previous research. Table A1 lists the study items. The questionnaire contained 25 items previously developed and designed to assess the perception of online food delivery services applications, such as performance expectancy, effort expectancy, social influence, facilitating conditions, price value, hedonic motivation, and habit (Venkatesh et al., 2012), 11 items were used to assess the information quality (Lee et al., 2019), time-saving (Yeo et al., 2017) and convenience (Chotigo & Kadono, 2021). Three items were used to assess continuance usage intention (Cho et al., 2019). IV. D ata A nalysis The descriptive statistics of a demographic profile were examined using the SPSS 25 packages, and the partial least squares structural equation modeling (PLS-SEM) with SmartPLS (SEM) version 3.2.8 was used to quantify the association between the variables. According to Hair Jr et al. (2017), one of the most prominent advantages of the PLS-SEM approach is that it possesses a strong predictive capacity for endogenous target variables, which is especially useful in highly completed models. a) Common Method Bias Because the data for this study came from a single source (an experienced user), there is a chance of common method bias (CMB), which involves measuring both dependent and independent variables. Also, the study focused on statistical and method- logical fixes before and after data collection to reduce the possibility of variance. Harman’s single-factor test was used to figure out the CMB. In the principal component factor analysis, 11 of the factors had Eigen values that were greater than 1.0. These eleven factors explained 61% of the variance. Also, the first factor did not explain most of the variance (29.43%). Based on the results, we can conclude that the CMB is not an issue for this research (Podsakoff & Organ, 1986). V. R esult a) Measurement Model We examined construct reliability, composite reliability, convergent validity, and discriminant validity using the approaches proposed by Hair et al. (2017). Composite reliability (CR), Cronbach’s alpha greater than 0.7, and roh_A near 1.00 are indicative of high consistency (Hair et al., 2017). Table 3 demonstrations that the eleven constructs met the criteria for internal consistency. To show convergent validity, the value of the average variance extracted (AVE) should be greater than 0.5. This means that constructs are responsible for more than 50% of the items in the suggested model (Hair et al., 2017). All of the latent variables attained convergent validity because their AVEs exceeded the 0.5 threshold (Table 3). Table 3: Construct Reliability and AVE Construct CR Cronbach’s Alpha Roh_A AVE CUI 0.797 0.879 0.815 0.523 PE 0.743 0.821 0.893 0.596 EE 0.801 0.911 0.967 0.667 SI 0.845 0.823 0.852 0.684 FC 0.783 0.846 0.812 0.521 HM 0.861 0.934 0.943 0.679 PV 0.754 0.784 0.921 0.583 HB 0.732 0.847 0.891 0.563 IQ 0.869 0.880 0.901 0.642 TS 0.812 0.921 0.957 0.691 CO 0.799 0.847 0.881 0.562 Note: CUI= Continuance Usage Intention; PE = Performance Expectancy; EE = Effort Expectancy; SI= Social Influence; FC = Facilitating Conditions; HM = Hedonic Motivation; PV = Price Value; HB = Habit; TS = Time Savings; CO = Convenience; IQ = Information Quality; AVE = Average Variance Extracted; CR = Composite Reliability. ” Table 4 presents an illustration of the Fornell and Lacker criteria for demonstrating discriminant validity (Fornell & Larcker, 1981). In order to accomplish this, we compared the square root of the average variance extracted (AVE) of each construct and the correlation of coefficients with other constructs. According to Hair et al. (2017), in the correlation matrix, the diagonal values (square root of AVE) must be greater than the off-diagonal values (the variables’ correlations). The evidence presented in Table 4 demonstrated that all of the diagonal values were superior to the values that were found outside of the diagonal. As a consequence of this, the discriminant validity of the study constructs was demonstrated to