Bootstrapping u okviru medijacionog modela
2022-10-27
Za ilustraciju upotrebe bootstrap intervala poverenja u medijacionim
modelima, ovde ćemo prikazati kako se testiraju medijacioni modeli iz
članka:
Lazarević, Lj., Purić, D., Teovanović, P., Lukić, P., Zupan, Z. &
Knežević G. (2021). What drives us to be (ir)responsible for our health
during the COVID-19 pandemic? The role of personality, thinking styles,
and conspiracy mentality. Personality and Individual Differences, 176,
https://doi.org/10.1016/j.paid.2021.110771.
Članak se može pročitati na: https://www.sciencedirect.com/science/article/pii/S019188692100146X?via%3Dihub
Pripremne aktivnosti
Učitavanje potrebnih paketa
haven nam je potreban da lako učitamo spss fajl
lavaan je potreban za testiranje medijacionog modela
semPlot ćemo koristiti da nacrtamo model
library(haven)
library(lavaan)
library(semPlot)Učitavanje fajla sa podacima
Podaci se mogu naći na OSF stranici istraživanja: https://osf.io/9njp3/
Ovde koristimo Database_Serbian_transformed (fajl je samo preimenovan za
vežbe)
setwd(dirname(rstudioapi::getActiveDocumentContext()$path))
dat <- read_sav("mediation_delta_PSP.sav")Postavljanje random seed-a
Ovo je važno kako bismo dobili reproducibilne rezultate (svaki put kada pokrenemo analizu, bilo mi, bilo neko drugi)
set.seed(1234)Medijacioni model
Postavljanje medijacionog modela
Ovaj model je pojednostavljena verzija u odnosu na onu iz članka zato
što nisu uključeni kovarijati (rezultati se ne menjaju dramatično)
Model iz članka dostupan je isto na OSF stranici istraživanja: https://osf.io/9njp3/
Varijabla CAM_bin se odnosi na primenu pseudo-naučnih praksi
(pseudo-scientific practices PSP)
DELTA je dezintegracija
NFC je Need for cognition, odnosno racionalni stil razmišljanja
FI je Faith in intuition, odnosno intuitivni stil razmišljanja
CMQ je Conspiracy Mentality Questionnaire, odnosno zavereničko
mišljenje
DELTA_PSP <- '
CAM_bin ~ c * DELTA + b1 * NFC + b2 * FI + b3 * CMQ
NFC ~ a1 * DELTA
FI ~ a2 * DELTA
CMQ ~ a3 * DELTA
indirect1 := a1*b1
indirect2 := a2*b2
indirect3 := a3*b3
direct := c
total := c + (a1*b1) + (a2*b2) + (a3*b3)
NFC ~~ FI + CMQ
FI ~~ CMQ
'
fit_DELTA_PSP <- sem(DELTA_PSP,
data = dat,
se = "bootstrap")Rezultati medijacionog modela
summary(fit_DELTA_PSP,
rsquare=TRUE,
standardized = TRUE,
fit.measures = TRUE)## lavaan 0.6-12 ended normally after 36 iterations
##
## Estimator ML
## Optimization method NLMINB
## Number of model parameters 14
##
## Number of observations 417
##
## Model Test User Model:
##
## Test statistic 0.000
## Degrees of freedom 0
##
## Model Test Baseline Model:
##
## Test statistic 165.703
## Degrees of freedom 10
## 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) -2578.306
## Loglikelihood unrestricted model (H1) -2578.306
##
## Akaike (AIC) 5184.612
## Bayesian (BIC) 5241.075
## Sample-size adjusted Bayesian (BIC) 5196.650
##
## 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 RMSEA <= 0.05 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
## CAM_bin ~
## DELTA (c) 0.012 0.018 0.698 0.485 0.012 0.036
## NFC (b1) -0.045 0.015 -2.992 0.003 -0.045 -0.155
## FI (b2) 0.043 0.015 2.916 0.004 0.043 0.146
## CMQ (b3) 0.001 0.001 1.882 0.060 0.001 0.095
## NFC ~
## DELTA (a1) -0.337 0.056 -6.003 0.000 -0.337 -0.288
## FI ~
## DELTA (a2) 0.261 0.055 4.735 0.000 0.261 0.226
## CMQ ~
## DELTA (a3) 9.676 1.490 6.496 0.000 9.676 0.322
##
## Covariances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .NFC ~~
## .FI 0.042 0.027 1.562 0.118 0.042 0.088
## .CMQ -1.416 0.556 -2.546 0.011 -1.416 -0.117
## .FI ~~
## .CMQ 2.513 0.642 3.913 0.000 2.513 0.206
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .CAM_bin 0.041 0.002 16.953 0.000 0.041 0.924
## .NFC 0.479 0.032 15.032 0.000 0.479 0.917
## .FI 0.484 0.034 14.337 0.000 0.484 0.949
## .CMQ 307.365 21.323 14.415 0.000 307.365 0.896
##
## R-Square:
## Estimate
## CAM_bin 0.076
## NFC 0.083
## FI 0.051
## CMQ 0.104
##
## Defined Parameters:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## indirect1 0.015 0.006 2.597 0.009 0.015 0.045
## indirect2 0.011 0.004 2.537 0.011 0.011 0.033
## indirect3 0.011 0.006 1.821 0.069 0.011 0.031
## direct 0.012 0.018 0.698 0.485 0.012 0.036
## total 0.049 0.017 2.928 0.003 0.049 0.144parameterEstimates(fit_DELTA_PSP, standardized = TRUE)| lhs | op | rhs | label | est | se | z | pvalue | ci.lower | ci.upper | std.lv | std.all | std.nox |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| CAM_bin | ~ | DELTA | c | 0.0122626 | 0.0175655 | 0.6981101 | 0.4851083 | -0.0206872 | 0.0472886 | 0.0122626 | 0.0357041 | 0.0578986 |
| CAM_bin | ~ | NFC | b1 | -0.0453833 | 0.0151662 | -2.9923983 | 0.0027679 | -0.0742805 | -0.0161435 | -0.0453833 | -0.1548965 | -0.1548965 |
| CAM_bin | ~ | FI | b2 | 0.0434155 | 0.0148902 | 2.9157073 | 0.0035488 | 0.0154975 | 0.0723327 | 0.0434155 | 0.1464536 | 0.1464536 |
| CAM_bin | ~ | CMQ | b3 | 0.0010856 | 0.0005768 | 1.8821467 | 0.0598161 | -0.0000202 | 0.0022914 | 0.0010856 | 0.0949216 | 0.0949216 |
| NFC | ~ | DELTA | a1 | -0.3372013 | 0.0561717 | -6.0030437 | 0.0000000 | -0.4475090 | -0.2251410 | -0.3372013 | -0.2876586 | -0.4664740 |
| FI | ~ | DELTA | a2 | 0.2612813 | 0.0551793 | 4.7351336 | 0.0000022 | 0.1534420 | 0.3683007 | 0.2612813 | 0.2255214 | 0.3657108 |
| CMQ | ~ | DELTA | a3 | 9.6761695 | 1.4895758 | 6.4959227 | 0.0000000 | 6.7133260 | 12.5199734 | 9.6761695 | 0.3221998 | 0.5224868 |
| NFC | ~~ | FI | 0.0422842 | 0.0270773 | 1.5616116 | 0.1183795 | -0.0100225 | 0.0925257 | 0.0422842 | 0.0877477 | 0.0877477 | |
| NFC | ~~ | CMQ | -1.4162827 | 0.5563275 | -2.5457715 | 0.0109037 | -2.5583648 | -0.2505474 | -1.4162827 | -0.1166853 | -0.1166853 | |
| FI | ~~ | CMQ | 2.5126821 | 0.6421203 | 3.9131018 | 0.0000911 | 1.2798364 | 3.8893392 | 2.5126821 | 0.2059085 | 0.2059085 | |
| CAM_bin | ~~ | CAM_bin | 0.0414563 | 0.0024454 | 16.9525604 | 0.0000000 | 0.0364641 | 0.0461520 | 0.0414563 | 0.9241840 | 0.9241840 | |
| NFC | ~~ | NFC | 0.4793056 | 0.0318852 | 15.0322114 | 0.0000000 | 0.4174242 | 0.5455806 | 0.4793056 | 0.9172525 | 0.9172525 | |
| FI | ~~ | FI | 0.4844752 | 0.0337931 | 14.3365291 | 0.0000000 | 0.4203101 | 0.5516562 | 0.4844752 | 0.9491401 | 0.9491401 | |
| CMQ | ~~ | CMQ | 307.3654147 | 21.3226517 | 14.4149714 | 0.0000000 | 266.6150929 | 348.5749530 | 307.3654147 | 0.8961873 | 0.8961873 | |
| DELTA | ~~ | DELTA | 0.3802768 | 0.0000000 | NA | NA | 0.3802768 | 0.3802768 | 0.3802768 | 1.0000000 | 0.3802768 | |
| indirect1 | := | a1*b1 | indirect1 | 0.0153033 | 0.0058938 | 2.5965168 | 0.0094174 | 0.0050479 | 0.0282639 | 0.0153033 | 0.0445573 | 0.0722552 |
| indirect2 | := | a2*b2 | indirect2 | 0.0113437 | 0.0044715 | 2.5368840 | 0.0111844 | 0.0038343 | 0.0209466 | 0.0113437 | 0.0330284 | 0.0535596 |
| indirect3 | := | a3*b3 | indirect3 | 0.0105040 | 0.0057677 | 1.8211829 | 0.0685791 | -0.0002298 | 0.0224921 | 0.0105040 | 0.0305837 | 0.0495953 |
| direct | := | c | direct | 0.0122626 | 0.0175743 | 0.6977609 | 0.4853267 | -0.0206872 | 0.0472886 | 0.0122626 | 0.0357041 | 0.0578986 |
| total | := | c+(a1b1)+(a2b2)+(a3*b3) | total | 0.0494136 | 0.0168757 | 2.9280874 | 0.0034105 | 0.0176699 | 0.0834085 | 0.0494136 | 0.1438736 | 0.2333088 |
Plot modela
semPaths(fit_DELTA_PSP, what = "path", whatLabels = "std", layout = "spring", sizeMan = 8, edge.label.cex = 1)