df <- read.csv("C:/Users/ASUS/Downloads/Data Final.csv", stringsAsFactors = FALSE)
cat(" INFO DATASET \n")
## INFO DATASET
cat("Jumlah responden :", nrow(df), "\n")
## Jumlah responden : 355
cat("Jumlah kolom :", ncol(df), "\n")
## Jumlah kolom : 36
cat("Missing values :", sum(is.na(df)), "\n\n")
## Missing values : 0
item_cols <- c("MAD1","MAD2","MAD3",
"MAU1","MAU2","MAU3",
"MAE1","MAE2","MAE3",
"BIF1","BIF2",
"BIS1","BIS2","BIS3",
"BIE1","BIE2","BIE3",
"LPF1","LPF2","LPF3",
"LPP1","LPP2","LPP3",
"LPA1","LPA2","LPA3",
"CLB1","CLB2","CLB3",
"CLA1","CLA2","CLA3")
df_item <- df[, item_cols]
cat("STATISTIK DESKRIPTIF\n")
## STATISTIK DESKRIPTIF
desc <- describe(df_item)
print(round(desc[, c("n","mean","sd","median","skew","kurtosis")], 3))
## n mean sd median skew kurtosis
## MAD1 355 4.39 0.60 4 -0.41 -0.69
## MAD2 355 4.37 0.60 4 -0.43 -0.23
## MAD3 355 4.38 0.60 4 -0.38 -0.70
## MAU1 355 4.44 0.62 5 -0.70 -0.13
## MAU2 355 4.41 0.65 4 -0.76 0.07
## MAU3 355 4.39 0.62 4 -0.48 -0.66
## MAE1 355 4.40 0.60 4 -0.52 -0.22
## MAE2 355 4.39 0.59 4 -0.37 -0.71
## MAE3 355 4.38 0.62 4 -0.52 -0.27
## BIF1 355 4.43 0.60 4 -0.59 -0.16
## BIF2 355 4.41 0.62 4 -0.71 0.20
## BIS1 355 4.39 0.62 4 -0.51 -0.65
## BIS2 355 4.43 0.62 4 -0.59 -0.60
## BIS3 355 4.37 0.63 4 -0.55 -0.32
## BIE1 355 4.36 0.63 4 -0.47 -0.68
## BIE2 355 4.35 0.63 4 -0.43 -0.69
## BIE3 355 4.40 0.62 4 -0.54 -0.64
## LPF1 355 4.39 0.62 4 -0.50 -0.66
## LPF2 355 4.49 0.62 5 -0.82 -0.35
## LPF3 355 4.47 0.63 5 -0.78 -0.42
## LPP1 355 4.40 0.60 4 -0.47 -0.67
## LPP2 355 4.48 0.61 5 -0.75 -0.43
## LPP3 355 4.46 0.61 5 -0.65 -0.54
## LPA1 355 4.45 0.62 5 -0.66 -0.54
## LPA2 355 4.41 0.62 4 -0.54 -0.63
## LPA3 355 4.41 0.62 4 -0.52 -0.64
## CLB1 355 4.35 0.62 4 -0.40 -0.68
## CLB2 355 4.36 0.61 4 -0.38 -0.68
## CLB3 355 4.32 0.63 4 -0.45 -0.35
## CLA1 355 4.39 0.60 4 -0.50 -0.23
## CLA2 355 4.38 0.61 4 -0.52 -0.24
## CLA3 355 4.34 0.63 4 -0.48 -0.32
cat("\n CRONBACH ALPHA PER KONSTRUK\n")
##
## CRONBACH ALPHA PER KONSTRUK
konstruk_items <- list(
MAD = c("MAD1","MAD2","MAD3"), # Mobile App Design
MAU = c("MAU1","MAU2","MAU3"), # Mobile App Usability
MAE = c("MAE1","MAE2","MAE3"), # Mobile App Engagement
BIF = c("BIF1","BIF2"), # Brand Image - Feature
BIS = c("BIS1","BIS2","BIS3"), # Brand Image - Social
BIE = c("BIE1","BIE2","BIE3"), # Brand Image - Emotional
LPF = c("LPF1","LPF2","LPF3"), # Loyalty Program - Financial
LPP = c("LPP1","LPP2","LPP3"), # Loyalty Program - Partnership
LPA = c("LPA1","LPA2","LPA3"), # Loyalty Program - Aspirational
CLB = c("CLB1","CLB2","CLB3"), # Customer Loyalty - Behavioral
CLA = c("CLA1","CLA2","CLA3") # Customer Loyalty - Attitudinal
)
for (nm in names(konstruk_items)) {
a <- alpha(df_item[, konstruk_items[[nm]]])$total$raw_alpha
status <- ifelse(a >= 0.7, "✓ Reliabel", "✗ Perlu perhatian")
cat(sprintf(" %-4s : α = %.3f %s\n", nm, a, status))
}
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## MAD : α = 0.825 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## MAU : α = 0.738 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## MAE : α = 0.809 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## BIF : α = 0.704 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## BIS : α = 0.763 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## BIE : α = 0.771 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## LPF : α = 0.796 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## LPP : α = 0.746 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## LPA : α = 0.724 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## CLB : α = 0.764 ✓ Reliabel
## Warning in response.frequencies(x, max = max): response.frequency has been
## deprecated and replaced with responseFrequecy. Please fix your call
## CLA : α = 0.782 ✓ Reliabel
cat("\n MATRIKS KORELASI (item numerik) \n")
##
## MATRIKS KORELASI (item numerik)
cor_matrix <- cor(df_item, use = "complete.obs")
# Visualisasi korelasi
corrplot(cor_matrix,
method = "color",
type = "upper",
tl.cex = 0.7,
tl.col = "black",
addCoef.col = "black",
number.cex = 0.45,
title = "Matriks Korelasi Antar Item",
mar = c(0,0,1,0))
model_sem <- '
MobileApp =~ MAD1 + MAD2 + MAD3 +
MAU1 + MAU2 + MAU3 +
MAE1 + MAE2 + MAE3
BrandImage =~ BIF1 + BIF2 +
BIS1 + BIS2 + BIS3 +
BIE1 + BIE2 + BIE3
LoyaltyPrg =~ LPF1 + LPF2 + LPF3 +
LPP1 + LPP2 + LPP3 +
LPA1 + LPA2 + LPA3
CustLoyalty =~ CLB1 + CLB2 + CLB3 +
CLA1 + CLA2 + CLA3
# H1: Mobile App berpengaruh terhadap Brand Image
BrandImage ~ MobileApp
# H2: Loyalty Program berpengaruh terhadap Brand Image
BrandImage ~ LoyaltyPrg
# H3: Mobile App berpengaruh langsung terhadap Customer Loyalty
CustLoyalty ~ MobileApp
# H4: Loyalty Program berpengaruh terhadap Customer Loyalty
CustLoyalty ~ LoyaltyPrg
# H5: Brand Image berpengaruh terhadap Customer Loyalty (mediasi)
CustLoyalty ~ BrandImage
'
cat("\n ESTIMASI MODEL SEM \n")
##
## ESTIMASI MODEL SEM
fit <- sem(model_sem,
data = df_item,
estimator = "MLR",
se = "robust")
cat("Status estimasi:", lavInspect(fit, "converged"), "\n\n")
## Status estimasi: TRUE
cat(" GOODNESS OF FIT INDICES \n")
## GOODNESS OF FIT INDICES
fit_indices <- fitMeasures(fit, c(
"chisq", "df", "pvalue",
"cfi", "tli", "rmsea", "rmsea.ci.lower", "rmsea.ci.upper",
"srmr", "aic", "bic"
))
cat(sprintf("\n Chi-square (χ²) : %.3f (df = %d, p = %.4f)\n",
fit_indices["chisq"], fit_indices["df"], fit_indices["pvalue"]))
##
## Chi-square (χ²) : 1768.406 (df = 458, p = 0.0000)
cat(sprintf(" CFI : %.3f (≥ 0.90 = acceptable, ≥ 0.95 = good)\n",
fit_indices["cfi"]))
## CFI : 0.811 (≥ 0.90 = acceptable, ≥ 0.95 = good)
cat(sprintf(" TLI : %.3f (≥ 0.90 = acceptable)\n",
fit_indices["tli"]))
## TLI : 0.795 (≥ 0.90 = acceptable)
cat(sprintf(" RMSEA : %.3f (≤ 0.08 = acceptable, ≤ 0.06 = good)\n",
fit_indices["rmsea"]))
## RMSEA : 0.090 (≤ 0.08 = acceptable, ≤ 0.06 = good)
cat(sprintf(" RMSEA 90%% CI : [%.3f, %.3f]\n",
fit_indices["rmsea.ci.lower"], fit_indices["rmsea.ci.upper"]))
## RMSEA 90% CI : [0.085, 0.094]
cat(sprintf(" SRMR : %.3f (≤ 0.08 = acceptable)\n",
fit_indices["srmr"]))
## SRMR : 0.058 (≤ 0.08 = acceptable)
cat(sprintf(" AIC : %.1f\n", fit_indices["aic"]))
## AIC : 15689.9
cat(sprintf(" BIC : %.1f\n", fit_indices["bic"]))
## BIC : 15961.0
cat("\n Interpretasi \n")
##
## Interpretasi
cfi_ok <- fit_indices["cfi"] >= 0.90
tli_ok <- fit_indices["tli"] >= 0.90
rmsea_ok <- fit_indices["rmsea"] <= 0.08
srmr_ok <- fit_indices["srmr"] <= 0.08
cat(sprintf(" CFI : %s\n", ifelse(cfi_ok, "✓ Acceptable", "✗ Kurang baik")))
## CFI : ✗ Kurang baik
cat(sprintf(" TLI : %s\n", ifelse(tli_ok, "✓ Acceptable", "✗ Kurang baik")))
## TLI : ✗ Kurang baik
cat(sprintf(" RMSEA : %s\n", ifelse(rmsea_ok, "✓ Acceptable", "✗ Kurang baik")))
## RMSEA : ✗ Kurang baik
cat(sprintf(" SRMR : %s\n", ifelse(srmr_ok, "✓ Acceptable", "✗ Kurang baik")))
## SRMR : ✓ Acceptable
cat("\n RINGKASAN MODEL LENGKAP \n")
##
## RINGKASAN MODEL LENGKAP
summary(fit,
fit.measures = TRUE,
standardized = TRUE,
rsquare = TRUE)
## lavaan 0.6-21 ended normally after 85 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 70
##
## Number of observations 355
##
## Model Test User Model:
## Standard Scaled
## Test Statistic 1768.406 1665.819
## Degrees of freedom 458 458
## P-value (Chi-square) 0.000 0.000
## Scaling correction factor 1.062
## Yuan-Bentler correction (Mplus variant)
##
## Model Test Baseline Model:
##
## Test statistic 7431.474 7095.772
## Degrees of freedom 496 496
## P-value 0.000 0.000
## Scaling correction factor 1.047
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.811 0.817
## Tucker-Lewis Index (TLI) 0.795 0.802
##
## Robust Comparative Fit Index (CFI) 0.814
## Robust Tucker-Lewis Index (TLI) 0.799
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -7774.954 -7774.954
## Scaling correction factor 0.832
## for the MLR correction
## Loglikelihood unrestricted model (H1) -6890.751 -6890.751
## Scaling correction factor 1.031
## for the MLR correction
##
## Akaike (AIC) 15689.909 15689.909
## Bayesian (BIC) 15960.957 15960.957
## Sample-size adjusted Bayesian (SABIC) 15738.886 15738.886
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.090 0.086
## 90 Percent confidence interval - lower 0.085 0.082
## 90 Percent confidence interval - upper 0.094 0.091
## P-value H_0: RMSEA <= 0.050 0.000 0.000
## P-value H_0: RMSEA >= 0.080 1.000 0.991
##
## Robust RMSEA 0.089
## 90 Percent confidence interval - lower 0.084
## 90 Percent confidence interval - upper 0.093
## P-value H_0: Robust RMSEA <= 0.050 0.000
## P-value H_0: Robust RMSEA >= 0.080 0.999
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.058 0.058
##
## Parameter Estimates:
##
## Standard errors Robust.sem
## Information Expected
## Information saturated (h1) model Structured
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## MobileApp =~
## MAD1 1.000 0.374 0.622
## MAD2 1.092 0.074 14.666 0.000 0.408 0.684
## MAD3 1.039 0.099 10.467 0.000 0.388 0.652
## MAU1 1.051 0.101 10.413 0.000 0.392 0.635
## MAU2 1.123 0.107 10.456 0.000 0.419 0.646
## MAU3 1.177 0.106 11.076 0.000 0.440 0.710
## MAE1 1.129 0.100 11.289 0.000 0.422 0.700
## MAE2 1.102 0.108 10.187 0.000 0.412 0.696
## MAE3 1.134 0.101 11.242 0.000 0.424 0.686
## BrandImage =~
## BIF1 1.000 0.381 0.638
## BIF2 1.079 0.072 14.922 0.000 0.412 0.660
## BIS1 1.179 0.102 11.563 0.000 0.450 0.724
## BIS2 1.044 0.085 12.238 0.000 0.398 0.646
## BIS3 1.198 0.088 13.552 0.000 0.457 0.721
## BIE1 1.248 0.120 10.391 0.000 0.476 0.753
## BIE2 1.128 0.105 10.707 0.000 0.430 0.688
## BIE3 1.170 0.096 12.181 0.000 0.446 0.717
## LoyaltyPrg =~
## LPF1 1.000 0.424 0.680
## LPF2 1.100 0.075 14.661 0.000 0.467 0.752
## LPF3 1.257 0.086 14.572 0.000 0.533 0.848
## LPP1 1.004 0.074 13.515 0.000 0.426 0.705
## LPP2 1.142 0.082 13.853 0.000 0.484 0.792
## LPP3 0.981 0.089 11.085 0.000 0.416 0.682
## LPA1 1.064 0.082 12.928 0.000 0.451 0.729
## LPA2 0.964 0.070 13.817 0.000 0.409 0.665
## LPA3 0.909 0.079 11.452 0.000 0.385 0.628
## CustLoyalty =~
## CLB1 1.000 0.466 0.756
## CLB2 0.912 0.066 13.897 0.000 0.425 0.702
## CLB3 0.987 0.070 14.152 0.000 0.460 0.728
## CLA1 0.919 0.067 13.774 0.000 0.428 0.712
## CLA2 0.958 0.061 15.741 0.000 0.446 0.733
## CLA3 1.067 0.069 15.409 0.000 0.497 0.792
##
## Regressions:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## BrandImage ~
## MobileApp 0.303 0.066 4.574 0.000 0.297 0.297
## LoyaltyPrg 0.625 0.070 8.880 0.000 0.695 0.695
## CustLoyalty ~
## MobileApp 0.539 0.130 4.140 0.000 0.432 0.432
## LoyaltyPrg 0.534 0.197 2.715 0.007 0.485 0.485
## BrandImage -0.142 0.258 -0.550 0.582 -0.116 -0.116
##
## Covariances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## MobileApp ~~
## LoyaltyPrg 0.117 0.019 6.280 0.000 0.739 0.739
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .MAD1 0.221 0.011 20.447 0.000 0.221 0.613
## .MAD2 0.190 0.010 18.514 0.000 0.190 0.533
## .MAD3 0.204 0.011 19.404 0.000 0.204 0.575
## .MAU1 0.227 0.015 14.727 0.000 0.227 0.596
## .MAU2 0.246 0.015 16.015 0.000 0.246 0.583
## .MAU3 0.190 0.011 16.805 0.000 0.190 0.496
## .MAE1 0.185 0.011 16.630 0.000 0.185 0.510
## .MAE2 0.180 0.011 16.944 0.000 0.180 0.516
## .MAE3 0.201 0.013 16.025 0.000 0.201 0.529
## .BIF1 0.212 0.011 19.823 0.000 0.212 0.593
## .BIF2 0.220 0.012 18.219 0.000 0.220 0.565
## .BIS1 0.183 0.013 13.641 0.000 0.183 0.475
## .BIS2 0.222 0.011 19.497 0.000 0.222 0.583
## .BIS3 0.192 0.012 16.698 0.000 0.192 0.480
## .BIE1 0.173 0.016 11.121 0.000 0.173 0.433
## .BIE2 0.206 0.012 17.320 0.000 0.206 0.527
## .BIE3 0.188 0.011 16.412 0.000 0.188 0.486
## .LPF1 0.209 0.012 16.884 0.000 0.209 0.538
## .LPF2 0.167 0.013 12.872 0.000 0.167 0.435
## .LPF3 0.111 0.011 10.259 0.000 0.111 0.282
## .LPP1 0.183 0.012 15.535 0.000 0.183 0.503
## .LPP2 0.139 0.011 12.496 0.000 0.139 0.373
## .LPP3 0.199 0.015 13.024 0.000 0.199 0.535
## .LPA1 0.179 0.014 12.425 0.000 0.179 0.468
## .LPA2 0.211 0.012 17.100 0.000 0.211 0.557
## .LPA3 0.228 0.014 15.777 0.000 0.228 0.606
## .CLB1 0.163 0.013 12.257 0.000 0.163 0.429
## .CLB2 0.186 0.013 14.404 0.000 0.186 0.507
## .CLB3 0.187 0.017 11.229 0.000 0.187 0.470
## .CLA1 0.178 0.013 14.167 0.000 0.178 0.492
## .CLA2 0.172 0.013 13.296 0.000 0.172 0.463
## .CLA3 0.147 0.014 10.310 0.000 0.147 0.373
## MobileApp 0.140 0.024 5.843 0.000 1.000 1.000
## .BrandImage 0.018 0.004 4.772 0.000 0.125 0.125
## LoyaltyPrg 0.180 0.027 6.679 0.000 1.000 1.000
## .CustLoyalty 0.095 0.014 6.934 0.000 0.439 0.439
##
## R-Square:
## Estimate
## MAD1 0.387
## MAD2 0.467
## MAD3 0.425
## MAU1 0.404
## MAU2 0.417
## MAU3 0.504
## MAE1 0.490
## MAE2 0.484
## MAE3 0.471
## BIF1 0.407
## BIF2 0.435
## BIS1 0.525
## BIS2 0.417
## BIS3 0.520
## BIE1 0.567
## BIE2 0.473
## BIE3 0.514
## LPF1 0.462
## LPF2 0.565
## LPF3 0.718
## LPP1 0.497
## LPP2 0.627
## LPP3 0.465
## LPA1 0.532
## LPA2 0.443
## LPA3 0.394
## CLB1 0.571
## CLB2 0.493
## CLB3 0.530
## CLA1 0.508
## CLA2 0.537
## CLA3 0.627
## BrandImage 0.875
## CustLoyalty 0.561
cat("\n VALIDITAS KONVERGEN (AVE & CR) \n")
##
## VALIDITAS KONVERGEN (AVE & CR)
# Ambil standardized loadings
params <- parameterEstimates(fit, standardized = TRUE)
loadings <- params[params$op == "=~", c("lhs","rhs","std.all")]
konstruk_laten <- c("MobileApp","BrandImage","LoyaltyPrg","CustLoyalty")
for (k in konstruk_laten) {
lam <- loadings[loadings$lhs == k, "std.all"]
n_item <- length(lam)
# AVE (Average Variance Extracted)
ave <- mean(lam^2)
# CR (Composite Reliability)
cr <- (sum(lam))^2 / ((sum(lam))^2 + sum(1 - lam^2))
ave_ok <- ifelse(ave >= 0.50, "✓", "✗")
cr_ok <- ifelse(cr >= 0.70, "✓", "✗")
cat(sprintf(" %-12s | AVE = %.3f %s (≥0.50) | CR = %.3f %s (≥0.70) | n_item = %d\n",
k, ave, ave_ok, cr, cr_ok, n_item))
}
## MobileApp | AVE = 0.450 ✗ (≥0.50) | CR = 0.880 ✓ (≥0.70) | n_item = 9
## BrandImage | AVE = 0.482 ✗ (≥0.50) | CR = 0.881 ✓ (≥0.70) | n_item = 8
## LoyaltyPrg | AVE = 0.523 ✓ (≥0.50) | CR = 0.907 ✓ (≥0.70) | n_item = 9
## CustLoyalty | AVE = 0.544 ✓ (≥0.50) | CR = 0.877 ✓ (≥0.70) | n_item = 6
cat("\n UJI HIPOTESIS (PATH COEFFICIENTS) \n")
##
## UJI HIPOTESIS (PATH COEFFICIENTS)
cat(sprintf("%-35s %8s %8s %8s %8s %s\n",
"Jalur", "β std", "SE", "z", "p-value", "Ket"))
## Jalur β std SE z p-value Ket
cat(strrep("-", 80), "\n")
## --------------------------------------------------------------------------------
paths <- params[params$op == "~",]
for (i in 1:nrow(paths)) {
p <- paths[i,]
sig <- ifelse(p$pvalue < 0.001, "***",
ifelse(p$pvalue < 0.01, "**",
ifelse(p$pvalue < 0.05, "*", "ns")))
ket <- ifelse(p$pvalue < 0.05, "✓ Signifikan", "✗ Tidak Signifikan")
cat(sprintf("%-15s → %-15s %8.3f %8.3f %8.3f %8.4f %s %s\n",
p$rhs, p$lhs,
p$std.all, p$se, p$z, p$pvalue,
sig, ket))
}
## MobileApp → BrandImage 0.297 0.066 4.574 0.0000 *** ✓ Signifikan
## LoyaltyPrg → BrandImage 0.695 0.070 8.880 0.0000 *** ✓ Signifikan
## MobileApp → CustLoyalty 0.432 0.130 4.140 0.0000 *** ✓ Signifikan
## LoyaltyPrg → CustLoyalty 0.485 0.197 2.715 0.0066 ** ✓ Signifikan
## BrandImage → CustLoyalty -0.116 0.258 -0.550 0.5822 ns ✗ Tidak Signifikan
cat("\n UJI HIPOTESIS (PATH COEFFICIENTS) \n")
##
## UJI HIPOTESIS (PATH COEFFICIENTS)
cat(sprintf("%-35s %8s %8s %8s %8s %s\n",
"Jalur", "β std", "SE", "z", "p-value", "Ket"))
## Jalur β std SE z p-value Ket
cat(strrep("-", 80), "\n")
## --------------------------------------------------------------------------------
paths <- params[params$op == "~",]
for (i in 1:nrow(paths)) {
p <- paths[i,]
sig <- ifelse(p$pvalue < 0.001, "***",
ifelse(p$pvalue < 0.01, "**",
ifelse(p$pvalue < 0.05, "*", "ns")))
ket <- ifelse(p$pvalue < 0.05, "✓ Signifikan", "✗ Tidak Signifikan")
cat(sprintf("%-15s → %-15s %8.3f %8.3f %8.3f %8.4f %s %s\n",
p$rhs, p$lhs,
p$std.all, p$se, p$z, p$pvalue,
sig, ket))
}
## MobileApp → BrandImage 0.297 0.066 4.574 0.0000 *** ✓ Signifikan
## LoyaltyPrg → BrandImage 0.695 0.070 8.880 0.0000 *** ✓ Signifikan
## MobileApp → CustLoyalty 0.432 0.130 4.140 0.0000 *** ✓ Signifikan
## LoyaltyPrg → CustLoyalty 0.485 0.197 2.715 0.0066 ** ✓ Signifikan
## BrandImage → CustLoyalty -0.116 0.258 -0.550 0.5822 ns ✗ Tidak Signifikan
cat("\n R-SQUARED (Variance Explained) \n")
##
## R-SQUARED (Variance Explained)
r2 <- lavInspect(fit, "r2")
for (nm in names(r2)) {
cat(sprintf(" %-15s : R² = %.3f (%.1f%% varians dijelaskan)\n",
nm, r2[nm], r2[nm]*100))
}
## MAD1 : R² = 0.387 (38.7% varians dijelaskan)
## MAD2 : R² = 0.467 (46.7% varians dijelaskan)
## MAD3 : R² = 0.425 (42.5% varians dijelaskan)
## MAU1 : R² = 0.404 (40.4% varians dijelaskan)
## MAU2 : R² = 0.417 (41.7% varians dijelaskan)
## MAU3 : R² = 0.504 (50.4% varians dijelaskan)
## MAE1 : R² = 0.490 (49.0% varians dijelaskan)
## MAE2 : R² = 0.484 (48.4% varians dijelaskan)
## MAE3 : R² = 0.471 (47.1% varians dijelaskan)
## BIF1 : R² = 0.407 (40.7% varians dijelaskan)
## BIF2 : R² = 0.435 (43.5% varians dijelaskan)
## BIS1 : R² = 0.525 (52.5% varians dijelaskan)
## BIS2 : R² = 0.417 (41.7% varians dijelaskan)
## BIS3 : R² = 0.520 (52.0% varians dijelaskan)
## BIE1 : R² = 0.567 (56.7% varians dijelaskan)
## BIE2 : R² = 0.473 (47.3% varians dijelaskan)
## BIE3 : R² = 0.514 (51.4% varians dijelaskan)
## LPF1 : R² = 0.462 (46.2% varians dijelaskan)
## LPF2 : R² = 0.565 (56.5% varians dijelaskan)
## LPF3 : R² = 0.718 (71.8% varians dijelaskan)
## LPP1 : R² = 0.497 (49.7% varians dijelaskan)
## LPP2 : R² = 0.627 (62.7% varians dijelaskan)
## LPP3 : R² = 0.465 (46.5% varians dijelaskan)
## LPA1 : R² = 0.532 (53.2% varians dijelaskan)
## LPA2 : R² = 0.443 (44.3% varians dijelaskan)
## LPA3 : R² = 0.394 (39.4% varians dijelaskan)
## CLB1 : R² = 0.571 (57.1% varians dijelaskan)
## CLB2 : R² = 0.493 (49.3% varians dijelaskan)
## CLB3 : R² = 0.530 (53.0% varians dijelaskan)
## CLA1 : R² = 0.508 (50.8% varians dijelaskan)
## CLA2 : R² = 0.537 (53.7% varians dijelaskan)
## CLA3 : R² = 0.627 (62.7% varians dijelaskan)
## BrandImage : R² = 0.875 (87.5% varians dijelaskan)
## CustLoyalty : R² = 0.561 (56.1% varians dijelaskan)
cat("\n TOP 10 MODIFICATION INDICES \n")
##
## TOP 10 MODIFICATION INDICES
mi <- modindices(fit, sort. = TRUE, maximum.number = 10)
print(mi[, c("lhs","op","rhs","mi","epc")])
## lhs op rhs mi epc
## 171 MAD1 ~~ MAD2 135.373 0.140
## 316 MAU3 ~~ MAE1 119.590 0.125
## 138 LoyaltyPrg =~ BIE3 93.044 2.095
## 514 BIE1 ~~ BIE2 84.484 0.103
## 646 LPA3 ~~ CLB1 60.208 0.088
## 639 LPA2 ~~ LPA3 41.066 0.079
## 232 MAD3 ~~ MAU1 38.994 0.078
## 496 BIS3 ~~ BIE1 38.987 0.069
## 477 BIS2 ~~ BIS3 38.968 -0.075
## 622 LPP3 ~~ LPA1 36.660 0.066
cat("\n MEMBUAT PATH DIAGRAM \n")
##
## MEMBUAT PATH DIAGRAM
# Path diagram standar
semPaths(fit,
what = "std", # tampilkan koefisien terstandarisasi
whatLabels = "std",
style = "ram",
layout = "tree2",
rotation = 2,
edge.label.cex = 0.7,
node.label.cex = 0.8,
color = list(lat = "#4A90D9", man = "#7ED321"),
title = TRUE,
thresholds = FALSE,
residuals = FALSE,
nCharNodes = 0,
mar = c(3,3,3,3))
## Warning in qgraph::qgraph(Edgelist, labels = nLab, bidirectional = Bidir, : The
## following arguments are not documented and likely not arguments of qgraph and
## thus ignored: node.label.cex
title("Path Diagram SEM: Mobile App & Loyalty Program → Customer Loyalty",
cex.main = 0.9)
cat("\n UJI MEDIASI (INDIRECT EFFECTS via BrandImage) \n")
##
## UJI MEDIASI (INDIRECT EFFECTS via BrandImage)
model_mediasi <- '
MobileApp =~ MAD1 + MAD2 + MAD3 + MAU1 + MAU2 + MAU3 + MAE1 + MAE2 + MAE3
BrandImage =~ BIF1 + BIF2 + BIS1 + BIS2 + BIS3 + BIE1 + BIE2 + BIE3
LoyaltyPrg =~ LPF1 + LPF2 + LPF3 + LPP1 + LPP2 + LPP3 + LPA1 + LPA2 + LPA3
CustLoyalty =~ CLB1 + CLB2 + CLB3 + CLA1 + CLA2 + CLA3
BrandImage ~ a1 * MobileApp + a2 * LoyaltyPrg
CustLoyalty ~ b * BrandImage
CustLoyalty ~ c1 * MobileApp + c2 * LoyaltyPrg
ind_MA := a1 * b # Indirect: MobileApp → BrandImage → CustLoyalty
ind_LP := a2 * b # Indirect: LoyaltyPrg → BrandImage → CustLoyalty
tot_MA := c1 + (a1 * b)
tot_LP := c2 + (a2 * b)
'
fit_med <- sem(model_mediasi,
data = df_item,
estimator = "ML",
se = "bootstrap",
bootstrap = 1000)
cat("\nHasil Indirect Effects (Bootstrap 1000x):\n")
##
## Hasil Indirect Effects (Bootstrap 1000x):
indirect <- parameterEstimates(fit_med, boot.ci.type = "perc", level = 0.95)
med_rows <- indirect[indirect$label %in% c("ind_MA","ind_LP","tot_MA","tot_LP"), ]
print(med_rows[, c("label","est","se","z","pvalue","ci.lower","ci.upper")])
## label est se z pvalue ci.lower ci.upper
## 75 ind_MA -0.043 0.090 -0.478 0.633 -0.229 0.132
## 76 ind_LP -0.089 0.186 -0.475 0.635 -0.472 0.252
## 77 tot_MA 0.496 0.112 4.413 0.000 0.280 0.717
## 78 tot_LP 0.445 0.094 4.734 0.000 0.276 0.642
cat("\n ind_MA = Indirect effect MobileApp → BrandImage → CustLoyalty\n")
##
## ind_MA = Indirect effect MobileApp → BrandImage → CustLoyalty
cat(" ind_LP = Indirect effect LoyaltyPrg → BrandImage → CustLoyalty\n")
## ind_LP = Indirect effect LoyaltyPrg → BrandImage → CustLoyalty
cat(" tot_MA = Total effect MobileApp → CustLoyalty\n")
## tot_MA = Total effect MobileApp → CustLoyalty
cat(" tot_LP = Total effect LoyaltyPrg → CustLoyalty\n")
## tot_LP = Total effect LoyaltyPrg → CustLoyalty
cat("\n Mediasi terjadi jika: CI tidak melewati 0 (p < 0.05)\n")
##
## Mediasi terjadi jika: CI tidak melewati 0 (p < 0.05)