=========================================================
研究題目：數位世代青少年現實心理需求滿足對網路成癮的影響：
核心負向情緒與網路代償的串聯中介效應分析
資料來源：TIGPS 2024 Wave
2 國八學生問卷
========================================================
##############################
### 0. data & package
##############################
if (!require("tidyverse")) install.packages("tidyverse")
if (!require("haven")) install.packages("haven")
if (!require("lavaan")) install.packages("lavaan")
if (!require("semTools")) install.packages("semTools")
if (!require("psych")) install.packages("psych")
library(tidyverse)
library(haven)
library(lavaan)
library(semTools)
library(psych)
raw_data <- read_sav("TIGPSw2_s.sav")##############################
### 1. 資料整理
##############################
# SDT
Relatedness_items <- c("bs5e", "bs5f", "bs12b", "bs9a")
Autonomy_items <- c("bs55q", "bs55r", "bs55s")
Competence_items <- c("bs55g", "bs55h", "bs55i")
SDT_items <- c(Relatedness_items, Autonomy_items, Competence_items)
# 負向情緒
# bs56g、bs56i 本身是負向情緒，不反向
# bs53a 是正向自我價值，需反向
NegEmo_items <- c("bs56c", "bs56d", "bs56e", "bs56f",
"bs56g", "bs56i", "bs56j",
"bs56k", "bs56l", "bs56m", "bs56n","bs53a")
# 網路代償調節
CybComp_items <- c("bs25m", "bs44b", "bs25i", "bs25l",
"bs25c", "bs22e", "bs22d", "bs25n")
# 網路成癮
IntAdd_items <- c( "bs28a", "bs28b", "bs28c", "bs28d", "bs28e",
"bs28f", "bs28g", "bs28h", "bs28i", "bs28j")
all_raw_items <- c( SDT_items, NegEmo_items, CybComp_items, IntAdd_items )
selected_df <- raw_data %>%
select( nstudent_id, bs1, bs3, bs4a, bs4b, bs4c, all_of(all_raw_items))
clean_df <- selected_df %>%
mutate(across(everything(), ~ ifelse(. < 0, NA, .))) %>%
mutate(bs53a = 5 - bs53a) # 反向題處理
apply(apply(clean_df, 1, is.na), 1, sum)## nstudent_id bs1 bs3 bs4a bs4b bs4c
## 0 0 17 28 28 28
## bs5e bs5f bs12b bs9a bs55q bs55r
## 49 49 57 9 86 86
## bs55s bs55g bs55h bs55i bs56c bs56d
## 86 86 86 86 89 90
## bs56e bs56f bs56g bs56i bs56j bs56k
## 89 89 89 89 89 89
## bs56l bs56m bs56n bs53a bs25m bs44b
## 89 89 89 84 50 88
## bs25i bs25l bs25c bs22e bs22d bs25n
## 50 50 50 51 51 50
## bs28a bs28b bs28c bs28d bs28e bs28f
## 55 54 54 54 54 54
## bs28g bs28h bs28i bs28j
## 54 54 54 54##############################
### 2. CFA
##############################
cfa_data <- clean_df
cat("CFA有效樣本數 =", nrow(cfa_data), "\n") # n= 8537## CFA有效樣本數 = 8537# model
cfa_model <- '
Relatedness =~
bs5e + bs5f + bs12b + bs9a
Autonomy =~
bs55q + bs55r + bs55s
Competence =~
bs55g + bs55h + bs55i
X_SDT =~
Relatedness + Autonomy + Competence
M1_NegEmo =~
bs56c + bs56d + bs56e + bs56f +
bs56g + bs56i + bs56j +
bs56k + bs56l + bs56m +
bs56n + bs53a
M2_CybComp =~
bs25m + bs44b + bs25i + bs25l +
bs25c + bs22e + bs22d + bs25n
Y_IntAdd =~
bs28a + bs28b + bs28c + bs28d + bs28e +
bs28f + bs28g + bs28h + bs28i + bs28j
'
cfa_fit <- cfa( model = cfa_model, data = cfa_data, estimator = "ML", std.lv = TRUE )
summary( cfa_fit, fit.measures = TRUE, standardized = TRUE)## lavaan 0.6-21 ended normally after 59 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 89
##
## Number of observations 8537
##
## Model Test User Model:
##
## Test statistic 30409.605
## Degrees of freedom 731
## P-value (Chi-square) 0.000
##
## Model Test Baseline Model:
##
## Test statistic 211847.141
## Degrees of freedom 780
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.859
## Tucker-Lewis Index (TLI) 0.850
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -351565.414
## Loglikelihood unrestricted model (H1) -336360.611
##
## Akaike (AIC) 703308.827
## Bayesian (BIC) 703936.470
## Sample-size adjusted Bayesian (SABIC) 703653.644
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.069
## 90 Percent confidence interval - lower 0.068
## 90 Percent confidence interval - upper 0.070
## P-value H_0: RMSEA <= 0.050 0.000
## P-value H_0: RMSEA >= 0.080 0.000
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.067
##
## Parameter Estimates:
##
## Standard errors Standard
## Information Expected
## Information saturated (h1) model Structured
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## Relatedness =~
## bs5e 0.734 0.010 76.715 0.000 0.788 0.888
## bs5f 0.655 0.009 75.634 0.000 0.703 0.860
## bs12b 0.170 0.007 24.682 0.000 0.182 0.281
## bs9a 0.221 0.009 24.115 0.000 0.237 0.275
## Autonomy =~
## bs55q 0.382 0.008 45.639 0.000 0.561 0.801
## bs55r 0.350 0.008 45.694 0.000 0.515 0.804
## bs55s 0.399 0.009 46.838 0.000 0.586 0.873
## Competence =~
## bs55g 0.305 0.021 14.692 0.000 0.707 0.875
## bs55h 0.304 0.021 14.691 0.000 0.705 0.874
## bs55i 0.293 0.020 14.693 0.000 0.680 0.879
## X_SDT =~
## Relatedness 0.390 0.015 26.257 0.000 0.364 0.364
## Autonomy 1.076 0.040 26.782 0.000 0.733 0.733
## Competence 2.089 0.175 11.970 0.000 0.902 0.902
## M1_NegEmo =~
## bs56c 0.872 0.010 88.933 0.000 0.872 0.805
## bs56d 0.662 0.009 76.137 0.000 0.662 0.722
## bs56e 0.618 0.008 80.823 0.000 0.618 0.754
## bs56f 0.646 0.008 79.643 0.000 0.646 0.746
## bs56g 0.720 0.010 70.203 0.000 0.720 0.680
## bs56i 0.783 0.009 84.904 0.000 0.783 0.780
## bs56j 0.815 0.010 83.313 0.000 0.815 0.770
## bs56k 0.766 0.008 93.868 0.000 0.766 0.833
## bs56l 0.793 0.009 84.931 0.000 0.793 0.780
## bs56m 0.914 0.009 96.504 0.000 0.914 0.848
## bs56n 0.912 0.010 94.868 0.000 0.912 0.839
## bs53a 0.368 0.009 42.336 0.000 0.368 0.445
## M2_CybComp =~
## bs25m 0.803 0.009 87.791 0.000 0.803 0.818
## bs44b 0.345 0.009 37.444 0.000 0.345 0.411
## bs25i 0.613 0.009 65.277 0.000 0.613 0.659
## bs25l 0.678 0.009 75.234 0.000 0.678 0.734
## bs25c 0.552 0.009 58.982 0.000 0.552 0.609
## bs22e 0.317 0.012 26.223 0.000 0.317 0.295
## bs22d 0.227 0.011 19.798 0.000 0.227 0.225
## bs25n 0.812 0.009 91.361 0.000 0.812 0.840
## Y_IntAdd =~
## bs28a 0.682 0.008 86.352 0.000 0.682 0.792
## bs28b 0.683 0.008 89.153 0.000 0.683 0.809
## bs28c 0.690 0.008 87.674 0.000 0.690 0.800
## bs28d 0.726 0.008 85.392 0.000 0.726 0.786
## bs28e 0.717 0.009 84.267 0.000 0.717 0.779
## bs28f 0.712 0.009 82.544 0.000 0.712 0.768
## bs28g 0.609 0.008 74.911 0.000 0.609 0.717
## bs28h 0.559 0.008 72.966 0.000 0.559 0.703
## bs28i 0.567 0.008 69.751 0.000 0.567 0.680
## bs28j 0.568 0.008 68.058 0.000 0.568 0.667
##
## Covariances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## X_SDT ~~
## M1_NegEmo -0.209 0.012 -17.219 0.000 -0.209 -0.209
## M2_CybComp -0.095 0.013 -7.405 0.000 -0.095 -0.095
## Y_IntAdd -0.190 0.012 -15.468 0.000 -0.190 -0.190
## M1_NegEmo ~~
## M2_CybComp 0.306 0.011 28.240 0.000 0.306 0.306
## Y_IntAdd 0.306 0.011 29.137 0.000 0.306 0.306
## M2_CybComp ~~
## Y_IntAdd 0.500 0.009 53.843 0.000 0.500 0.500
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .bs5e 0.167 0.010 15.954 0.000 0.167 0.212
## .bs5f 0.174 0.009 20.471 0.000 0.174 0.261
## .bs12b 0.389 0.006 64.507 0.000 0.389 0.921
## .bs9a 0.689 0.011 64.547 0.000 0.689 0.925
## .bs55q 0.176 0.004 48.252 0.000 0.176 0.358
## .bs55r 0.145 0.003 47.921 0.000 0.145 0.354
## .bs55s 0.107 0.003 35.413 0.000 0.107 0.238
## .bs55g 0.152 0.003 43.651 0.000 0.152 0.234
## .bs55h 0.153 0.003 43.950 0.000 0.153 0.236
## .bs55i 0.135 0.003 42.775 0.000 0.135 0.227
## .bs56c 0.414 0.007 58.827 0.000 0.414 0.353
## .bs56d 0.402 0.007 61.482 0.000 0.402 0.479
## .bs56e 0.290 0.005 60.682 0.000 0.290 0.432
## .bs56f 0.333 0.005 60.898 0.000 0.333 0.444
## .bs56g 0.604 0.010 62.299 0.000 0.604 0.538
## .bs56i 0.395 0.007 59.837 0.000 0.395 0.392
## .bs56j 0.457 0.008 60.185 0.000 0.457 0.407
## .bs56k 0.258 0.005 57.264 0.000 0.258 0.305
## .bs56l 0.404 0.007 59.831 0.000 0.404 0.392
## .bs56m 0.326 0.006 56.232 0.000 0.326 0.281
## .bs56n 0.350 0.006 56.891 0.000 0.350 0.296
## .bs53a 0.548 0.008 64.461 0.000 0.548 0.802
## .bs25m 0.320 0.007 47.333 0.000 0.320 0.332
## .bs44b 0.586 0.009 63.585 0.000 0.586 0.831
## .bs25i 0.489 0.008 58.633 0.000 0.489 0.565
## .bs25l 0.395 0.007 55.094 0.000 0.395 0.462
## .bs25c 0.517 0.009 60.231 0.000 0.517 0.629
## .bs22e 1.057 0.016 64.518 0.000 1.057 0.913
## .bs22d 0.971 0.015 64.878 0.000 0.971 0.949
## .bs25n 0.275 0.006 44.065 0.000 0.275 0.295
## .bs28a 0.276 0.005 57.425 0.000 0.276 0.372
## .bs28b 0.246 0.004 56.423 0.000 0.246 0.345
## .bs28c 0.267 0.005 56.970 0.000 0.267 0.359
## .bs28d 0.325 0.006 57.737 0.000 0.325 0.382
## .bs28e 0.333 0.006 58.085 0.000 0.333 0.393
## .bs28f 0.352 0.006 58.583 0.000 0.352 0.410
## .bs28g 0.351 0.006 60.382 0.000 0.351 0.486
## .bs28h 0.319 0.005 60.756 0.000 0.319 0.505
## .bs28i 0.374 0.006 61.314 0.000 0.374 0.538
## .bs28j 0.402 0.007 61.581 0.000 0.402 0.555
## .Relatedness 1.000 0.868 0.868
## .Autonomy 1.000 0.463 0.463
## .Competence 1.000 0.186 0.186
## X_SDT 1.000 1.000 1.000
## M1_NegEmo 1.000 1.000 1.000
## M2_CybComp 1.000 1.000 1.000
## Y_IntAdd 1.000 1.000 1.000## chisq df cfi tli rmsea srmr
## 30409.605 731.000 0.859 0.850 0.069 0.067##############################
### 3. reliability
##############################
rel <- semTools::reliability(cfa_fit)
table_reliability <- data.frame(
Construct = c(
"Relatedness","Autonomy","Competence",
"Negative Emotion","Cyber Compensation","Internet Addiction"),
Alpha = c(
psych::alpha(cfa_data[, Relatedness_items], check.keys = TRUE)$total$raw_alpha,
psych::alpha(cfa_data[, Autonomy_items], check.keys = TRUE)$total$raw_alpha,
psych::alpha(cfa_data[, Competence_items], check.keys = TRUE)$total$raw_alpha,
psych::alpha(cfa_data[, NegEmo_items], check.keys = TRUE)$total$raw_alpha,
psych::alpha(cfa_data[, CybComp_items], check.keys = TRUE)$total$raw_alpha,
psych::alpha(cfa_data[, IntAdd_items], check.keys = TRUE)$total$raw_alpha ),
CR = c(
rel["omega", "Relatedness"],
rel["omega", "Autonomy"],
rel["omega", "Competence"],
rel["omega", "M1_NegEmo"],
rel["omega", "M2_CybComp"],
rel["omega", "Y_IntAdd"]),
AVE = c(
rel["avevar", "Relatedness"],
rel["avevar", "Autonomy"],
rel["avevar", "Competence"],
rel["avevar", "M1_NegEmo"],
rel["avevar", "M2_CybComp"],
rel["avevar", "Y_IntAdd"])) %>%
mutate(across(where(is.numeric), ~ round(.x, 3)))
table_reliability## Construct Alpha CR AVE
## 1 Relatedness 0.664 0.720 0.459
## 2 Autonomy 0.865 0.866 0.683
## 3 Competence 0.908 0.908 0.768
## 4 Negative Emotion 0.939 0.943 0.588
## 5 Cyber Compensation 0.801 0.804 0.371
## 6 Internet Addiction 0.928 0.929 0.569alpha_SDT<- psych::alpha(cfa_data[, SDT_items], check.keys = TRUE)$total$raw_alpha
# 是否要放二階因子的因素負荷量??
# 另外手算
standardizedSolution(cfa_fit) %>% filter(lhs=="X_SDT", op=="=~")## lhs op rhs est.std se z pvalue ci.lower ci.upper
## 1 X_SDT =~ Relatedness 0.364 0.012 30.256 0 0.340 0.387
## 2 X_SDT =~ Autonomy 0.733 0.013 57.806 0 0.708 0.757
## 3 X_SDT =~ Competence 0.902 0.014 64.210 0 0.874 0.930lambda <- standardizedSolution(cfa_fit) %>% filter(lhs == "X_SDT", op == "=~") %>%
pull(est.std)
CR_SDT <- (sum(lambda)^2) / ((sum(lambda)^2) + sum(1 - lambda^2))
AVE_SDT <- mean(lambda^2)
data.frame(Alpha = round(alpha_SDT, 3),
CR = round(CR_SDT, 3),
AVE = round(AVE_SDT, 3))## Alpha CR AVE
## 1 0.839 0.725 0.494##############################
### 4. 描述統計
##############################
# 算總分
analysis_df <- clean_df %>%
mutate(
GRIT = bs4a + bs4b + bs4c,
Relatedness = bs5e + bs5f + bs12b + bs9a,
Autonomy = bs55q + bs55r + bs55s,
Competence = bs55g + bs55h + bs55i,
X_SDT = Relatedness + Autonomy + Competence,
M1_NegEmo =
bs56c + bs56d + bs56e + bs56f +
bs56g + bs56i + bs56j +
bs56k + bs56l + bs56m + bs56n + bs53a,
M2_CybComp =
bs25m + bs44b + bs25i + bs25l +
bs25c + bs22e + bs22d + bs25n,
Y_IntAdd =
bs28a + bs28b + bs28c + bs28d + bs28e +
bs28f + bs28g + bs28h + bs28i + bs28j )
names(analysis_df)## [1] "nstudent_id" "bs1" "bs3" "bs4a" "bs4b"
## [6] "bs4c" "bs5e" "bs5f" "bs12b" "bs9a"
## [11] "bs55q" "bs55r" "bs55s" "bs55g" "bs55h"
## [16] "bs55i" "bs56c" "bs56d" "bs56e" "bs56f"
## [21] "bs56g" "bs56i" "bs56j" "bs56k" "bs56l"
## [26] "bs56m" "bs56n" "bs53a" "bs25m" "bs44b"
## [31] "bs25i" "bs25l" "bs25c" "bs22e" "bs22d"
## [36] "bs25n" "bs28a" "bs28b" "bs28c" "bs28d"
## [41] "bs28e" "bs28f" "bs28g" "bs28h" "bs28i"
## [46] "bs28j" "GRIT" "Relatedness" "Autonomy" "Competence"
## [51] "X_SDT" "M1_NegEmo" "M2_CybComp" "Y_IntAdd"# 背景變項
table_gender <- analysis_df %>%
count(bs1) %>% #編碼簿中 1女 2男
mutate( Gender = c("女生", "男生"), Percent = round(100 * n / sum(n), 2)) %>%
select(Gender, N = n, Percent)
table_gender## # A tibble: 2 × 3
## Gender N Percent
## <chr> <int> <dbl>
## 1 女生 4211 49.3
## 2 男生 4326 50.7## # A tibble: 10 × 3
## bs3 n Percent
## <dbl> <int> <dbl>
## 1 1 154 1.8
## 2 2 170 1.99
## 3 3 314 3.68
## 4 4 673 7.88
## 5 5 2268 26.6
## 6 6 2093 24.5
## 7 7 1587 18.6
## 8 8 695 8.14
## 9 9 189 2.21
## 10 10 394 4.62# 連續變項
table_descriptive <- psych::describe(analysis_df %>%
select(X_SDT, Relatedness, Autonomy, Competence,
M1_NegEmo, M2_CybComp, Y_IntAdd )) %>%
as.data.frame() %>%
rownames_to_column("Variable") %>%
select( Variable, n, M = mean, SD = sd, Skewness = skew, Kurtosis = kurtosis ) %>%
mutate(across(where(is.numeric), ~ round(.x, 3)))
table_descriptive## Variable n M SD Skewness Kurtosis
## 1 X_SDT 8537 30.847 4.897 -0.457 0.824
## 2 Relatedness 8537 12.138 2.286 -0.519 0.285
## 3 Autonomy 8537 9.754 1.788 -0.727 1.517
## 4 Competence 8537 8.956 2.194 -0.634 0.469
## 5 M1_NegEmo 8537 7.523 9.138 2.157 4.704
## 6 M2_CybComp 8537 19.561 4.947 0.028 -0.151
## 7 Y_IntAdd 8537 20.211 6.767 0.297 -0.081##############################
### 5. correlation
##############################
corr_data <- analysis_df %>% select(X_SDT, M1_NegEmo, M2_CybComp, Y_IntAdd)
corr_result <- psych::corr.test(corr_data, method = "pearson", adjust = "none")
r_matrix <- corr_result$r
p_matrix <- corr_result$p
vars <- c("X_SDT", "M1_NegEmo", "M2_CybComp", "Y_IntAdd")
formatted_matrix <- matrix( "", nrow = length(vars), ncol = length(vars) )
rownames(formatted_matrix) <- c(
"1. SDT",
"2. Negative Emotion",
"3. Cyber Compensation",
"4. Internet Addiction" )
colnames(formatted_matrix) <- c("1", "2", "3", "4")
for (i in 1:length(vars)) {
for (j in 1:length(vars)) {
if (i > j) {
r_val <- r_matrix[i, j]
p_val <- p_matrix[i, j]
stars <- case_when( p_val < .001 ~ "***", p_val < .01 ~ "**", p_val < .05 ~ "*", TRUE ~ "")
formatted_matrix[i, j] <- paste0(sprintf("%.3f", r_val), stars)}}}
table_correlation <- formatted_matrix %>% as.data.frame() %>% rownames_to_column("Variable")
table_correlation## Variable 1 2 3 4
## 1 1. SDT
## 2 2. Negative Emotion -0.267***
## 3 3. Cyber Compensation -0.029** 0.244***
## 4 4. Internet Addiction -0.183*** 0.295*** 0.451***##############################
### 6. 串聯中介模型
##############################
cat("正式中介分析有效樣本數 =", nrow(analysis_df), "\n")## 正式中介分析有效樣本數 = 8537mediation_model <- '
# M1（Negative Emotion）
M1_NegEmo ~ a1*X_SDT + bs1 + bs3
# M2（Cyber Compensation）
M2_CybComp ~ a2*X_SDT + b1*M1_NegEmo + bs1 + bs3
# Y（Internet Addiction）
Y_IntAdd ~ cp*X_SDT + c1*M1_NegEmo + b2*M2_CybComp + bs1 + bs3
# 定義效果量 (Defined Parameters) #
# 1. (X -> M1 -> M2 -> Y)
serial_indirect := a1 * b1 * b2
# 2. (X -> M1 -> Y)
m1_indirect := a1 * c1
# 3. (X -> M2 -> Y)
m2_indirect := a2 * b2
# 4. TOTAL
total_indirect := (a1 * b1 * b2) + (a1 * c1) + (a2 * b2)
# 5. c prime
direct_effect := cp
# 6. (Total Effect = 直接效果 + 總間接效果)
total_effect := cp + total_indirect
'
set.seed(2026)
fit_model <- sem( model = mediation_model, data = analysis_df, se = "bootstrap", bootstrap = 1000)
summary( fit_model, fit.measures = TRUE, standardized = TRUE, rsquare = TRUE)## lavaan 0.6-21 ended normally after 1 iteration
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 15
##
## Number of observations 8537
##
## Model Test User Model:
##
## Test statistic 0.000
## Degrees of freedom 0
##
## Model Test Baseline Model:
##
## Test statistic 3911.877
## Degrees of freedom 12
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 1.000
## Tucker-Lewis Index (TLI) 1.000
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -83242.487
## Loglikelihood unrestricted model (H1) -83242.487
##
## Akaike (AIC) 166514.973
## Bayesian (BIC) 166620.756
## Sample-size adjusted Bayesian (SABIC) 166573.089
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.000
## 90 Percent confidence interval - lower 0.000
## 90 Percent confidence interval - upper 0.000
## P-value H_0: RMSEA <= 0.050 NA
## P-value H_0: RMSEA >= 0.080 NA
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.000
##
## Parameter Estimates:
##
## Standard errors Bootstrap
## Number of requested bootstrap draws 1000
## Number of successful bootstrap draws 1000
##
## Regressions:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## M1_NegEmo ~
## X_SDT (a1) -0.475 0.024 -19.659 0.000 -0.475 -0.255
## bs1 -2.387 0.195 -12.258 0.000 -2.387 -0.131
## bs3 -0.356 0.066 -5.356 0.000 -0.356 -0.069
## M2_CybComp ~
## X_SDT (a2) 0.043 0.014 2.983 0.003 0.043 0.042
## M1_NegEmo (b1) 0.141 0.007 20.962 0.000 0.141 0.261
## bs1 0.492 0.105 4.684 0.000 0.492 0.050
## bs3 -0.035 0.034 -1.048 0.295 -0.035 -0.013
## Y_IntAdd ~
## X_SDT (cp) -0.178 0.015 -11.648 0.000 -0.178 -0.129
## M1_NegEmo (c1) 0.120 0.009 13.478 0.000 0.120 0.162
## M2_CybCmp (b2) 0.557 0.016 35.920 0.000 0.557 0.408
## bs1 0.102 0.128 0.793 0.428 0.102 0.008
## bs3 0.014 0.042 0.330 0.741 0.014 0.004
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .M1_NegEmo 75.696 2.197 34.462 0.000 75.696 0.907
## .M2_CybComp 22.907 0.343 66.864 0.000 22.907 0.936
## .Y_IntAdd 34.116 0.589 57.912 0.000 34.116 0.745
##
## R-Square:
## Estimate
## M1_NegEmo 0.093
## M2_CybComp 0.064
## Y_IntAdd 0.255
##
## Defined Parameters:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## serial_indirct -0.037 0.003 -14.044 0.000 -0.037 -0.027
## m1_indirect -0.057 0.005 -12.076 0.000 -0.057 -0.041
## m2_indirect 0.024 0.008 2.948 0.003 0.024 0.017
## total_indirect -0.071 0.009 -7.556 0.000 -0.071 -0.051
## direct_effect -0.178 0.015 -11.648 0.000 -0.178 -0.129
## total_effect -0.249 0.018 -13.699 0.000 -0.249 -0.180# Confidence Intervals
parameterEstimates( fit_model, boot.ci.type = "perc", standardized = TRUE) %>%
filter(op == ":=")## lhs op rhs label est se
## 1 serial_indirect := a1*b1*b2 serial_indirect -0.037 0.003
## 2 m1_indirect := a1*c1 m1_indirect -0.057 0.005
## 3 m2_indirect := a2*b2 m2_indirect 0.024 0.008
## 4 total_indirect := (a1*b1*b2)+(a1*c1)+(a2*b2) total_indirect -0.071 0.009
## 5 direct_effect := cp direct_effect -0.178 0.015
## 6 total_effect := cp+total_indirect total_effect -0.249 0.018
## z pvalue ci.lower ci.upper std.lv std.all std.nox
## 1 -14.044 0.000 -0.043 -0.032 -0.037 -0.027 -0.006
## 2 -12.076 0.000 -0.067 -0.049 -0.057 -0.041 -0.008
## 3 2.948 0.003 0.008 0.040 0.024 0.017 0.004
## 4 -7.556 0.000 -0.089 -0.050 -0.071 -0.051 -0.010
## 5 -11.648 0.000 -0.207 -0.147 -0.178 -0.129 -0.026
## 6 -13.699 0.000 -0.284 -0.212 -0.249 -0.180 -0.037##############################
### 7. Post survey
##############################
post_corr_data <- analysis_df %>% select(Relatedness, Autonomy, Competence,
M1_NegEmo, M2_CybComp, Y_IntAdd)
post_corr_result <- psych::corr.test(post_corr_data, method = "pearson", adjust = "none")
r_matrix <- post_corr_result$r
p_matrix <- post_corr_result$p
vars <- c("Relatedness", "Autonomy", "Competence", "M1_NegEmo", "M2_CybComp", "Y_IntAdd")
postformatted_matrix <- matrix( "", nrow = length(vars), ncol = length(vars) )
rownames(postformatted_matrix) <- c(
"1. Relatedness",
"2. Autonomy",
"3. Competence",
"4. Negative Emotion",
"5. Cyber Compensation",
"6. Internet Addiction" )
colnames(postformatted_matrix) <- c("1", "2", "3", "4", "5", "6")
for (i in 1:length(vars)) {
for (j in 1:length(vars)) {
if (i > j) {
r_val <- r_matrix[i, j]
p_val <- p_matrix[i, j]
stars <- case_when( p_val < .001 ~ "***", p_val < .01 ~ "**", p_val < .05 ~ "*", TRUE ~ "")
postformatted_matrix[i, j] <- paste0(sprintf("%.3f", r_val), stars)}}}
posttable_correlation <- postformatted_matrix %>% as.data.frame() %>% rownames_to_column("Variable")
posttable_correlation## Variable 1 2 3 4 5 6
## 1 1. Relatedness
## 2 2. Autonomy 0.319***
## 3 3. Competence 0.352*** 0.586***
## 4 4. Negative Emotion -0.313*** -0.079*** -0.205***
## 5 5. Cyber Compensation -0.080*** 0.039*** -0.014 0.244***
## 6 6. Internet Addiction -0.153*** -0.121*** -0.150*** 0.295*** 0.451***