HLSaC Scale: CFA
Ivan Ropovik
Mon Nov 13 13:50:41 2017
Príprava dát
Potrebné R libraries
library(lavaan, quietly = TRUE, warn.conflicts = FALSE)
library(semPlot, quietly = TRUE, warn.conflicts = FALSE)
library(dplyr, quietly = TRUE, warn.conflicts = FALSE)
library(psych, quietly = TRUE, warn.conflicts = FALSE)
library(ICC, quietly = TRUE, warn.conflicts = FALSE)
library(Amelia, quietly = TRUE, warn.conflicts = FALSE)## Warning in as.POSIXlt.POSIXct(x, tz): unknown timezone 'zone/tz/2017c.1.0/
## zoneinfo/Europe/Bratislava'library(BaylorEdPsych, quietly = TRUE, warn.conflicts = FALSE)Načítanie dátového súboru
#rm(list = ls())
setwd(dir = "/Users/ivanropovik/OneDrive/MANUSCRIPTS/2017 HQL Validizacna studia")
full.data <- read.csv(file = "data_raw.csv", header = TRUE, sep = ";")
#View(full.data)Selekcia premenných pre CFA nástroja
HLQ.collumns <- full.data %>% select(HL1:HL25, M1, M2, M3, M4)
delete.na <- function(HLQ.collumns, n=NULL) {
HLQ.collumns[rowSums(is.na(HLQ.collumns)) <= n,]
}Vizualizácia chýbajucich dát
data.na.rm <- delete.na(HLQ.collumns, n = 10)
missmap(data.na.rm, rank.order = TRUE)
Demografia
N
nrow(data.na.rm)## [1] 291Pohlavie
table(data.na.rm$M1) # 1 = chlapec, 2 = dievča##
## 1 2
## 140 150table(data.na.rm$M1[data.na.rm$M2 == 1])/length(data.na.rm$M2[data.na.rm$M2 == 1])*100 # Proporcia chlapcov v 7.ročníku##
## 1 2
## 42.77457 56.64740table(data.na.rm$M1[data.na.rm$M2 == 2])/length(data.na.rm$M2[data.na.rm$M2 == 2])*100 # Proporcia chlapcov v 9.ročníku##
## 1 2
## 55.9322 44.0678Vek v rokoch
data.na.rm$M4[data.na.rm$M4 == 1] <- 1997
data.na.rm$M4[data.na.rm$M4 == 2] <- 1998
data.na.rm$M4[data.na.rm$M4 == 3] <- 1999
data.na.rm$M4[data.na.rm$M4 == 4] <- 2000
data.na.rm$M4[data.na.rm$M4 == 5] <- 2001
data.na.rm$M4[data.na.rm$M4 == 6] <- 2002
data.na.rm$M4[data.na.rm$M4 == 7] <- 2003
data.na.rm$M4[data.na.rm$M4 == 8] <- 2004
data.na.rm$M4[data.na.rm$M4 == 9] <- 2005
zber <- 2016.4 #Máj 2016Priemer
mean((zber - data.na.rm$M4)*12 - (12 - data.na.rm$M3), na.rm = TRUE)/12 #roky## [1] 14.24856mean((zber - data.na.rm$M4[data.na.rm$M2 == 1])*12 - (12 - data.na.rm$M3), na.rm = TRUE)/12 # Priemerný vek v 7. ročníku## Warning in (zber - data.na.rm$M4[data.na.rm$M2 == 1]) * 12 - (12 -
## data.na.rm$M3): longer object length is not a multiple of shorter object
## length## [1] 13.41638mean((zber - data.na.rm$M4[data.na.rm$M2 == 2])*12 - (12 - data.na.rm$M3), na.rm = TRUE)/12 # Priemerný vek v 9. ročníku## Warning in (zber - data.na.rm$M4[data.na.rm$M2 == 2]) * 12 - (12 -
## data.na.rm$M3): longer object length is not a multiple of shorter object
## length## [1] 15.50109SD
sd((zber - data.na.rm$M4)*12 - (12 - data.na.rm$M3), na.rm = TRUE)/12 #roky## [1] 1.233906sd((zber - data.na.rm$M4[data.na.rm$M2 == 1])*12 - (12 - data.na.rm$M3), na.rm = TRUE)/12 # SD veku v 7. ročníku## Warning in (zber - data.na.rm$M4[data.na.rm$M2 == 1]) * 12 - (12 -
## data.na.rm$M3): longer object length is not a multiple of shorter object
## length## [1] 0.5583029sd((zber - data.na.rm$M4[data.na.rm$M2 == 2])*12 - (12 - data.na.rm$M3), na.rm = TRUE)/12 # SD veku v 9. ročníku## Warning in (zber - data.na.rm$M4[data.na.rm$M2 == 2]) * 12 - (12 -
## data.na.rm$M3): longer object length is not a multiple of shorter object
## length## [1] 0.5885805Ročník
1 = 7. ročník; 2 = 9.ročník
table(data.na.rm$M2)##
## 1 2
## 173 118Priemerný počet žiakov v skupine
cluster.size <- nrow(data.na.rm)/21
cluster.size## [1] 13.85714Percento chýbajúcich dát
paste(round(sum(is.na(data.na.rm))/prod(dim(data.na.rm))*100, 3), "%", sep = "")## [1] "0.391%"Imputácia chýbajúcich dát
Bootstraped expected maximization
set.seed(123)
data_imput <- amelia(data.na.rm, ords = c("HL1", "HL2", "HL3", "HL4", "HL5",
"HL6", "HL7", "HL8", "HL9", "HL10",
"HL11", "HL12", "HL13", "HL14", "HL15",
"HL16", "HL17", "HL18", "HL19", "HL20",
"HL21", "HL22", "HL23", "HL24", "HL25"), m = 1)## -- Imputation 1 --
##
## 1 2 3 4data <- as.data.frame(data_imput$imputations)
names(data) <- c("HL1", "HL2", "HL3", "HL4", "HL5", "HL6", "HL7", "HL8", "HL9",
"HL10", "HL11", "HL12", "HL13", "HL14", "HL15", "HL16", "HL17",
"HL18", "HL19", "HL20", "HL21", "HL22", "HL23", "HL24", "HL25",
"Pohlavie", "Rocnik", "Narodenie_mesiac", "Narodenie_rok")Finálne dáta
#View(data)Deskriptívna analýza
Frekvenčné tabuľky
lapply(data[,c("HL1", "HL2", "HL3", "HL4", "HL5", "HL6", "HL7", "HL8", "HL9",
"HL10", "HL11", "HL12", "HL13", "HL14", "HL15", "HL16", "HL17",
"HL18", "HL19", "HL20", "HL21", "HL22", "HL23", "HL24", "HL25")],
function(x){table(x, useNA = "ifany")})## $HL1
## x
## 1 2 3 4
## 1 28 176 86
##
## $HL2
## x
## 1 2 3 4
## 4 31 127 129
##
## $HL3
## x
## 1 2 3 4
## 21 83 121 66
##
## $HL4
## x
## 1 2 3 4
## 18 62 112 99
##
## $HL5
## x
## 1 2 3 4
## 5 51 126 109
##
## $HL6
## x
## 1 2 3 4
## 5 33 95 158
##
## $HL7
## x
## 1 2 3 4
## 18 76 137 60
##
## $HL8
## x
## 1 2 3 4
## 7 57 116 111
##
## $HL9
## x
## 1 2 3 4
## 8 46 109 128
##
## $HL10
## x
## 1 2 3 4
## 8 39 115 129
##
## $HL11
## x
## 1 2 3 4
## 2 23 105 161
##
## $HL12
## x
## 1 2 3 4
## 22 77 124 68
##
## $HL13
## x
## 1 2 3 4
## 16 84 142 49
##
## $HL14
## x
## 1 2 3 4
## 9 47 114 121
##
## $HL15
## x
## 1 2 3 4
## 8 40 116 127
##
## $HL16
## x
## 1 2 3 4
## 15 63 148 65
##
## $HL17
## x
## 1 2 3 4
## 5 19 78 189
##
## $HL18
## x
## 1 2 3 4
## 1 15 98 177
##
## $HL19
## x
## 1 2 3 4
## 10 35 134 112
##
## $HL20
## x
## 1 2 3 4
## 13 64 138 76
##
## $HL21
## x
## 1 2 3 4
## 5 53 139 94
##
## $HL22
## x
## 1 2 3 4
## 6 32 98 155
##
## $HL23
## x
## 1 2 3 4
## 7 50 152 82
##
## $HL24
## x
## 1 2 3 4
## 13 84 148 46
##
## $HL25
## x
## 1 2 3 4
## 19 65 117 90Deskriptívne štatistiky
describe(data[,1:25], na.rm = TRUE, skew = TRUE, ranges = TRUE, type = 2, trim = 0)## vars n mean sd median trimmed mad min max range skew kurtosis
## HL1 1 291 3.19 0.61 3 3.19 0.00 1 4 3 -0.22 -0.03
## HL2 2 291 3.31 0.71 3 3.31 1.48 1 4 3 -0.76 0.14
## HL3 3 291 2.80 0.87 3 2.80 1.48 1 4 3 -0.25 -0.66
## HL4 4 291 3.00 0.90 3 3.00 1.48 1 4 3 -0.53 -0.57
## HL5 5 291 3.16 0.77 3 3.16 1.48 1 4 3 -0.52 -0.46
## HL6 6 291 3.40 0.76 4 3.40 0.00 1 4 3 -1.04 0.35
## HL7 7 291 2.82 0.83 3 2.82 1.48 1 4 3 -0.32 -0.42
## HL8 8 291 3.14 0.81 3 3.14 1.48 1 4 3 -0.53 -0.56
## HL9 9 291 3.23 0.81 3 3.23 1.48 1 4 3 -0.75 -0.22
## HL10 10 291 3.25 0.79 3 3.25 1.48 1 4 3 -0.82 0.07
## HL11 11 291 3.46 0.67 4 3.46 0.00 1 4 3 -1.00 0.39
## HL12 12 291 2.82 0.88 3 2.82 1.48 1 4 3 -0.31 -0.61
## HL13 13 291 2.77 0.79 3 2.77 1.48 1 4 3 -0.24 -0.35
## HL14 14 291 3.19 0.82 3 3.19 1.48 1 4 3 -0.71 -0.22
## HL15 15 291 3.24 0.79 3 3.24 1.48 1 4 3 -0.80 0.02
## HL16 16 291 2.90 0.80 3 2.90 0.00 1 4 3 -0.44 -0.16
## HL17 17 291 3.55 0.69 4 3.55 0.00 1 4 3 -1.55 2.04
## HL18 18 291 3.55 0.61 4 3.55 0.00 1 4 3 -1.11 0.65
## HL19 19 291 3.20 0.78 3 3.20 1.48 1 4 3 -0.80 0.33
## HL20 20 291 2.95 0.81 3 2.95 1.48 1 4 3 -0.42 -0.34
## HL21 21 291 3.11 0.75 3 3.11 1.48 1 4 3 -0.42 -0.40
## HL22 22 291 3.38 0.76 4 3.38 0.00 1 4 3 -1.05 0.45
## HL23 23 291 3.06 0.74 3 3.06 0.00 1 4 3 -0.46 -0.06
## HL24 24 291 2.78 0.76 3 2.78 0.00 1 4 3 -0.22 -0.26
## HL25 25 291 2.96 0.89 3 2.96 1.48 1 4 3 -0.47 -0.58
## se
## HL1 0.04
## HL2 0.04
## HL3 0.05
## HL4 0.05
## HL5 0.05
## HL6 0.04
## HL7 0.05
## HL8 0.05
## HL9 0.05
## HL10 0.05
## HL11 0.04
## HL12 0.05
## HL13 0.05
## HL14 0.05
## HL15 0.05
## HL16 0.05
## HL17 0.04
## HL18 0.04
## HL19 0.05
## HL20 0.05
## HL21 0.04
## HL22 0.04
## HL23 0.04
## HL24 0.04
## HL25 0.05Redukcia počtu kategórií v prípade nízkej frekvencie odpoveďovej kategórie
Pred výpočtom matice polychorických kovariancií však bolo nutné redukovať počet kategórií v prípade nízkej frekvencie odpoveďovej kategórie. Odhad polychorickej kovariančnej matice totiž predpokladá absenciu buniek xa a yb s nulovou frekvenciou (kde x, y sú ktorýmkoľvek párom premenných a a, b ktorýmkoľvek párom odpoveďových kategórií). Väčšina premenných vykazovala silne negatívne zošikmenie, pričom subjekty zriedka volili najmä odpoveďovú kategóriu [vôbec nie je pravda]. Vo veľkej väčšine prípadov postačovalo zlúčiť odpoveďovú kategóriu [vôbec nie je pravda] s kategóriou [nie tak celkom pravda], ak tá mala nižšiu celkovú frekvenciu ako 20.
data$HL1 <- ifelse(data$HL1 == 1, yes = 2, no = data$HL1)
data$HL2 <- ifelse(data$HL2 == 1, yes = 2, no = data$HL2)
data$HL4 <- ifelse(data$HL4 == 1, yes = 2, no = data$HL4)
data$HL5 <- ifelse(data$HL5 == 1, yes = 2, no = data$HL5)
data$HL6 <- ifelse(data$HL6 == 1, yes = 2, no = data$HL6)
data$HL8 <- ifelse(data$HL8 == 1, yes = 2, no = data$HL8)
data$HL9 <- ifelse(data$HL9 == 1, yes = 2, no = data$HL9)
data$HL10 <- ifelse(data$HL10 == 1, yes = 2, no = data$HL10)
data$HL11 <- ifelse(data$HL11 == 1, yes = 2, no = data$HL11)
data$HL13 <- ifelse(data$HL13 == 1, yes = 2, no = data$HL13)
data$HL14 <- ifelse(data$HL14 == 1, yes = 2, no = data$HL14)
data$HL15 <- ifelse(data$HL15 == 1, yes = 2, no = data$HL15)
data$HL17 <- ifelse(data$HL17 == 1, yes = 2, no = data$HL17)
data$HL18 <- ifelse(data$HL18 == 1, yes = 2, no = data$HL18)
data$HL18 <- ifelse(data$HL18 == 2, yes = 3, no = data$HL18)
data$HL19 <- ifelse(data$HL19 == 1, yes = 2, no = data$HL19)
data$HL20 <- ifelse(data$HL20 == 1, yes = 2, no = data$HL20)
data$HL21 <- ifelse(data$HL21 == 1, yes = 2, no = data$HL21)
data$HL22 <- ifelse(data$HL22 == 1, yes = 2, no = data$HL22)
data$HL23 <- ifelse(data$HL23 == 1, yes = 2, no = data$HL23)
data$HL24 <- ifelse(data$HL24 == 1, yes = 2, no = data$HL24)
data$HL25 <- ifelse(data$HL25 == 1, yes = 2, no = data$HL25)Matica polychorických korelácií
polychoric.cor <- polychoric(data[1:25], correct = FALSE, smooth = TRUE,
global = FALSE, na.rm = TRUE)
round(polychoric.cor$rho, 2)## HL1 HL2 HL3 HL4 HL5 HL6 HL7 HL8 HL9 HL10 HL11 HL12 HL13 HL14
## HL1 1.00 0.39 0.28 0.16 0.23 0.30 0.30 0.39 0.34 0.29 0.27 0.31 0.37 0.24
## HL2 0.39 1.00 0.40 0.30 0.29 0.37 0.21 0.19 0.29 0.33 0.30 0.27 0.22 0.25
## HL3 0.28 0.40 1.00 0.46 0.40 0.21 0.27 0.43 0.29 0.42 0.25 0.33 0.49 0.18
## HL4 0.16 0.30 0.46 1.00 0.52 0.20 0.35 0.39 0.31 0.61 0.27 0.55 0.27 0.21
## HL5 0.23 0.29 0.40 0.52 1.00 0.24 0.31 0.38 0.39 0.47 0.30 0.43 0.31 0.32
## HL6 0.30 0.37 0.21 0.20 0.24 1.00 0.32 0.19 0.26 0.29 0.22 0.35 0.28 0.29
## HL7 0.30 0.21 0.27 0.35 0.31 0.32 1.00 0.37 0.36 0.40 0.18 0.43 0.29 0.10
## HL8 0.39 0.19 0.43 0.39 0.38 0.19 0.37 1.00 0.43 0.43 0.35 0.35 0.45 0.16
## HL9 0.34 0.29 0.29 0.31 0.39 0.26 0.36 0.43 1.00 0.50 0.31 0.33 0.30 0.24
## HL10 0.29 0.33 0.42 0.61 0.47 0.29 0.40 0.43 0.50 1.00 0.33 0.55 0.31 0.22
## HL11 0.27 0.30 0.25 0.27 0.30 0.22 0.18 0.35 0.31 0.33 1.00 0.31 0.37 0.34
## HL12 0.31 0.27 0.33 0.55 0.43 0.35 0.43 0.35 0.33 0.55 0.31 1.00 0.38 0.33
## HL13 0.37 0.22 0.49 0.27 0.31 0.28 0.29 0.45 0.30 0.31 0.37 0.38 1.00 0.30
## HL14 0.24 0.25 0.18 0.21 0.32 0.29 0.10 0.16 0.24 0.22 0.34 0.33 0.30 1.00
## HL15 0.37 0.30 0.41 0.33 0.33 0.32 0.24 0.43 0.29 0.33 0.42 0.30 0.44 0.24
## HL16 0.27 0.35 0.41 0.39 0.25 0.30 0.50 0.41 0.33 0.47 0.40 0.48 0.38 0.26
## HL17 0.27 0.36 0.35 0.33 0.34 0.44 0.32 0.28 0.29 0.34 0.33 0.33 0.27 0.21
## HL18 0.31 0.24 0.38 0.45 0.40 0.21 0.18 0.50 0.47 0.47 0.43 0.41 0.32 0.18
## HL19 0.22 0.36 0.24 0.32 0.25 0.23 0.31 0.28 0.31 0.36 0.48 0.33 0.33 0.20
## HL20 0.44 0.38 0.42 0.51 0.45 0.34 0.45 0.54 0.35 0.54 0.42 0.50 0.50 0.31
## HL21 0.33 0.31 0.43 0.45 0.56 0.30 0.30 0.44 0.36 0.45 0.33 0.42 0.33 0.21
## HL22 0.29 0.19 0.22 0.14 0.16 0.27 0.10 0.13 0.25 0.15 0.26 0.21 0.19 0.17
## HL23 0.27 0.30 0.40 0.41 0.47 0.26 0.37 0.40 0.53 0.48 0.44 0.34 0.37 0.13
## HL24 0.31 0.22 0.44 0.47 0.46 0.17 0.25 0.44 0.39 0.44 0.47 0.47 0.46 0.36
## HL25 0.21 0.26 0.32 0.28 0.29 0.23 0.16 0.21 0.22 0.29 0.22 0.43 0.37 0.61
## HL15 HL16 HL17 HL18 HL19 HL20 HL21 HL22 HL23 HL24 HL25
## HL1 0.37 0.27 0.27 0.31 0.22 0.44 0.33 0.29 0.27 0.31 0.21
## HL2 0.30 0.35 0.36 0.24 0.36 0.38 0.31 0.19 0.30 0.22 0.26
## HL3 0.41 0.41 0.35 0.38 0.24 0.42 0.43 0.22 0.40 0.44 0.32
## HL4 0.33 0.39 0.33 0.45 0.32 0.51 0.45 0.14 0.41 0.47 0.28
## HL5 0.33 0.25 0.34 0.40 0.25 0.45 0.56 0.16 0.47 0.46 0.29
## HL6 0.32 0.30 0.44 0.21 0.23 0.34 0.30 0.27 0.26 0.17 0.23
## HL7 0.24 0.50 0.32 0.18 0.31 0.45 0.30 0.10 0.37 0.25 0.16
## HL8 0.43 0.41 0.28 0.50 0.28 0.54 0.44 0.13 0.40 0.44 0.21
## HL9 0.29 0.33 0.29 0.47 0.31 0.35 0.36 0.25 0.53 0.39 0.22
## HL10 0.33 0.47 0.34 0.47 0.36 0.54 0.45 0.15 0.48 0.44 0.29
## HL11 0.42 0.40 0.33 0.43 0.48 0.42 0.33 0.26 0.44 0.47 0.22
## HL12 0.30 0.48 0.33 0.41 0.33 0.50 0.42 0.21 0.34 0.47 0.43
## HL13 0.44 0.38 0.27 0.32 0.33 0.50 0.33 0.19 0.37 0.46 0.37
## HL14 0.24 0.26 0.21 0.18 0.20 0.31 0.21 0.17 0.13 0.36 0.61
## HL15 1.00 0.38 0.33 0.53 0.35 0.52 0.51 0.17 0.42 0.38 0.16
## HL16 0.38 1.00 0.45 0.40 0.34 0.48 0.38 0.21 0.41 0.39 0.28
## HL17 0.33 0.45 1.00 0.42 0.31 0.30 0.35 0.32 0.32 0.24 0.29
## HL18 0.53 0.40 0.42 1.00 0.40 0.51 0.48 0.16 0.43 0.40 0.40
## HL19 0.35 0.34 0.31 0.40 1.00 0.46 0.43 0.28 0.42 0.34 0.25
## HL20 0.52 0.48 0.30 0.51 0.46 1.00 0.50 0.27 0.53 0.55 0.34
## HL21 0.51 0.38 0.35 0.48 0.43 0.50 1.00 0.27 0.40 0.48 0.31
## HL22 0.17 0.21 0.32 0.16 0.28 0.27 0.27 1.00 0.38 0.32 0.27
## HL23 0.42 0.41 0.32 0.43 0.42 0.53 0.40 0.38 1.00 0.57 0.26
## HL24 0.38 0.39 0.24 0.40 0.34 0.55 0.48 0.32 0.57 1.00 0.42
## HL25 0.16 0.28 0.29 0.40 0.25 0.34 0.31 0.27 0.26 0.42 1.00Priemerná korelácia
polychoric.cor.low <- polychoric.cor$rho[lower.tri(polychoric.cor$rho)]
mean(abs(polychoric.cor.low))## [1] 0.340758Výpočet polychorickej kovariančnej matice
SDs <- describe(data[1:25], na.rm = TRUE)$sd
polychoric.cov <- cor2cov(R = polychoric.cor$rho, sds = SDs)5-faktorový model
Špecifikácia 5-faktorového modelu
model <- '
Teor_ved =~ HL1 + HL8 + HL14 + HL18 + HL25
Prakt_ved =~ HL2 + HL4 + HL6 + HL10 + HL17
Krit_mysl =~ HL7 + HL12 + HL16 + HL21 + HL24
Sebauved =~ HL5 + HL11 + HL15 + HL19 + HL22
Občianstvo =~ HL3 + HL9 + HL13 + HL20 + HL23
'Estimácia 5-faktorového modelu
Podľa Muthén, 1984
Daný typ estimátora je (1) robustný voči porušeniam predpokladu normálneho rozloženia premenných (nerealistické u Likertových škál), (2) produkuje značne menšie skreslenie pri odhade parametrov a teste zhody modelu a dát pri chybne špecifikovaných modeloch, a (3) proporcia chýb I. radu pri posudzovaní korektne špecifikovaných modelov je u daného typu dát výrazne bližšia vopred stanovenej nominálnej hodnote α, ako je to napríklad v prípade metódy maximálnej vierohodnosti (Beauducel, Herzberg, 2009).
fitted.model <- cfa(model = model, data = data, meanstructure = TRUE, std.lv = TRUE, mimic = "Mplus",
estimator = "WLSMVS", test = "Satterthwaite", orthogonal = FALSE, bootstrap = 5000,
ordered = c("HL1", "HL2", "HL3", "HL4", "HL5", "HL6", "HL7", "HL8", "HL9", "HL10",
"HL11", "HL12", "HL13", "HL14", "HL15", "HL16", "HL17", "HL18", "HL19",
"HL20", "HL21", "HL22", "HL23", "HL24", "HL25"))## Warning in lav_samplestats_from_data(lavdata = lavdata, missing = lavoptions$missing, : lavaan WARNING: 5 bivariate tables have empty cells; to see them, use:
## lavInspect(fit, "zero.cell.tables")## Warning in lav_object_post_check(object): lavaan WARNING: covariance matrix of latent variables
## is not positive definite;
## use inspect(fit,"cov.lv") to investigate.Počet buniek HLa x HLb s nulovou frekvenciou je už zanedbateľný.
Analýza štatistickej sily
Pre test blízkej zhody (na základe RMSEA distribúcie). Štatistická sila pre detekciu chybného modelu (RMSEA > .08)
df <- fitted.model@test[[1]]$df
alfa <- .05
n <- nrow(data)
rmsea0 <- .05 # RMSEA za predpokladu H0
rmseaa <- .08 # RMSEA za predpokladu H1
ncp0 <- (n-1)*df*rmsea0**2 ;
ncpa <-(n-1)*df*rmseaa**2 ;
if(rmsea0 < rmseaa) {
cval <- qchisq(1-alfa,df=df,ncp=ncp0)
sila.rmsea <- 1 - pchisq(cval,df=df,ncp=ncpa)
} else {
cval <- qchisq(alfa,df=df,ncp=ncp0)
sila.rmsea <- pchisq(cval,df=df,ncp=ncpa)
}
rm(ncp0, ncpa, cval)
print(round(sila.rmsea,10))## [1] 0.9999999Kovariančná matica je non-positive definite, pravdepodobne z dôvodu, že viaceré z definovaných latentných premenných su kolineárne (de facto identické). Jedna z korelácií v rámci štrukturálneho modelu (korelácia medzi citizenship a slf_wr) je väčšia ako 1.
eigen(inspect(fitted.model, "cov.lv") )$values## [1] 4.657176051 0.224744196 0.094689073 0.033357754 -0.009967075Negatívna piata eigenvalue má ale relatívne nízku hodnotu, takže výsledky testu modelu sú interpretovateľné.
Test modelu, odhady voľných parametrov
Stačí si všímať “Robust” test, Latent variable, Covariances a R-square. Intercepts, Thresholds, Intercepts (…) môžte kľudne ignorovať.
summary(fitted.model, standardized = TRUE, rsquare = TRUE)## lavaan (0.5-23.1097) converged normally after 35 iterations
##
## Number of observations 291
##
## Estimator DWLS Robust
## Minimum Function Test Statistic 370.444 219.259
## Degrees of freedom 265 100
## P-value (Chi-square) 0.000 0.000
## Scaling correction factor 1.690
## for the mean and variance adjusted correction (WLSMV)
##
## Parameter Estimates:
##
## Information Expected
## Standard Errors Robust.sem
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## Teor_ved =~
## HL1 0.540 0.057 9.417 0.000 0.540 0.540
## HL8 0.684 0.047 14.690 0.000 0.684 0.684
## HL14 0.491 0.054 9.034 0.000 0.491 0.491
## HL18 0.721 0.053 13.688 0.000 0.721 0.721
## HL25 0.567 0.052 10.918 0.000 0.567 0.567
## Prakt_ved =~
## HL2 0.521 0.059 8.880 0.000 0.521 0.521
## HL4 0.705 0.043 16.316 0.000 0.705 0.705
## HL6 0.484 0.057 8.550 0.000 0.484 0.484
## HL10 0.752 0.038 19.646 0.000 0.752 0.752
## HL17 0.575 0.057 10.009 0.000 0.575 0.575
## Krit_mysl =~
## HL7 0.533 0.047 11.385 0.000 0.533 0.533
## HL12 0.678 0.037 18.476 0.000 0.678 0.678
## HL16 0.652 0.038 16.946 0.000 0.652 0.652
## HL21 0.684 0.039 17.751 0.000 0.684 0.684
## HL24 0.697 0.039 17.779 0.000 0.697 0.697
## Sebauved =~
## HL5 0.634 0.045 14.158 0.000 0.634 0.634
## HL11 0.571 0.052 11.069 0.000 0.571 0.571
## HL15 0.617 0.046 13.427 0.000 0.617 0.617
## HL19 0.555 0.049 11.284 0.000 0.555 0.555
## HL22 0.378 0.062 6.139 0.000 0.378 0.378
## Občianstvo =~
## HL3 0.613 0.045 13.674 0.000 0.613 0.613
## HL9 0.587 0.048 12.272 0.000 0.587 0.587
## HL13 0.603 0.045 13.419 0.000 0.603 0.603
## HL20 0.774 0.034 22.881 0.000 0.774 0.774
## HL23 0.681 0.040 16.912 0.000 0.681 0.681
##
## Covariances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## Teor_ved ~~
## Prakt_ved 0.787 0.058 13.565 0.000 0.787 0.787
## Krit_mysl 0.871 0.044 19.778 0.000 0.871 0.871
## Sebauved 0.893 0.051 17.502 0.000 0.893 0.893
## Občianstvo 0.908 0.037 24.435 0.000 0.908 0.908
## Prakt_ved ~~
## Krit_mysl 0.943 0.038 24.686 0.000 0.943 0.943
## Sebauved 0.903 0.064 14.082 0.000 0.903 0.903
## Občianstvo 0.897 0.045 20.079 0.000 0.897 0.897
## Krit_mysl ~~
## Sebauved 0.963 0.047 20.481 0.000 0.963 0.963
## Občianstvo 0.959 0.030 31.704 0.000 0.959 0.959
## Sebauved ~~
## Občianstvo 1.009 0.044 22.901 0.000 1.009 1.009
##
## Intercepts:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .HL1 0.000 0.000 0.000
## .HL8 0.000 0.000 0.000
## .HL14 0.000 0.000 0.000
## .HL18 0.000 0.000 0.000
## .HL25 0.000 0.000 0.000
## .HL2 0.000 0.000 0.000
## .HL4 0.000 0.000 0.000
## .HL6 0.000 0.000 0.000
## .HL10 0.000 0.000 0.000
## .HL17 0.000 0.000 0.000
## .HL7 0.000 0.000 0.000
## .HL12 0.000 0.000 0.000
## .HL16 0.000 0.000 0.000
## .HL21 0.000 0.000 0.000
## .HL24 0.000 0.000 0.000
## .HL5 0.000 0.000 0.000
## .HL11 0.000 0.000 0.000
## .HL15 0.000 0.000 0.000
## .HL19 0.000 0.000 0.000
## .HL22 0.000 0.000 0.000
## .HL3 0.000 0.000 0.000
## .HL9 0.000 0.000 0.000
## .HL13 0.000 0.000 0.000
## .HL20 0.000 0.000 0.000
## .HL23 0.000 0.000 0.000
## Teor_ved 0.000 0.000 0.000
## Prakt_ved 0.000 0.000 0.000
## Krit_mysl 0.000 0.000 0.000
## Sebauved 0.000 0.000 0.000
## Občianstvo 0.000 0.000 0.000
##
## Thresholds:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## HL1|t1 -1.284 0.100 -12.774 0.000 -1.284 -1.284
## HL1|t2 0.537 0.078 6.925 0.000 0.537 0.537
## HL8|t1 -0.772 0.082 -9.401 0.000 -0.772 -0.772
## HL8|t2 0.302 0.075 4.032 0.000 0.302 0.302
## HL14|t1 -0.869 0.085 -10.266 0.000 -0.869 -0.869
## HL14|t2 0.213 0.074 2.865 0.004 0.213 0.213
## HL18|t1 -0.275 0.075 -3.682 0.000 -0.275 -0.275
## HL25|t1 -0.557 0.078 -7.154 0.000 -0.557 -0.557
## HL25|t2 0.498 0.077 6.465 0.000 0.498 0.498
## HL2|t1 -1.174 0.095 -12.311 0.000 -1.174 -1.174
## HL2|t2 0.143 0.074 1.931 0.054 0.143 0.143
## HL4|t1 -0.598 0.079 -7.610 0.000 -0.598 -0.598
## HL4|t2 0.412 0.076 5.426 0.000 0.412 0.412
## HL6|t1 -1.124 0.093 -12.051 0.000 -1.124 -1.124
## HL6|t2 -0.108 0.074 -1.463 0.144 -0.108 -0.108
## HL10|t1 -0.988 0.088 -11.196 0.000 -0.988 -0.988
## HL10|t2 0.143 0.074 1.931 0.054 0.143 0.143
## HL17|t1 -1.389 0.106 -13.077 0.000 -1.389 -1.389
## HL17|t2 -0.384 0.076 -5.078 0.000 -0.384 -0.384
## HL7|t1 -1.539 0.116 -13.276 0.000 -1.539 -1.539
## HL7|t2 -0.459 0.076 -6.004 0.000 -0.459 -0.459
## HL7|t3 0.820 0.083 9.837 0.000 0.820 0.820
## HL12|t1 -1.435 0.109 -13.168 0.000 -1.435 -1.435
## HL12|t2 -0.412 0.076 -5.426 0.000 -0.412 -0.412
## HL12|t3 0.727 0.081 8.960 0.000 0.727 0.727
## HL16|t1 -1.630 0.123 -13.266 0.000 -1.630 -1.630
## HL16|t2 -0.619 0.079 -7.837 0.000 -0.619 -0.619
## HL16|t3 0.761 0.082 9.292 0.000 0.761 0.761
## HL21|t1 -0.844 0.084 -10.053 0.000 -0.844 -0.844
## HL21|t2 0.459 0.076 6.004 0.000 0.459 0.459
## HL24|t1 -0.431 0.076 -5.658 0.000 -0.431 -0.431
## HL24|t2 1.002 0.089 11.295 0.000 1.002 1.002
## HL5|t1 -0.869 0.085 -10.266 0.000 -0.869 -0.869
## HL5|t2 0.320 0.075 4.265 0.000 0.320 0.320
## HL11|t1 -1.366 0.105 -13.024 0.000 -1.366 -1.366
## HL11|t2 -0.134 0.074 -1.814 0.070 -0.134 -0.134
## HL15|t1 -0.974 0.088 -11.095 0.000 -0.974 -0.974
## HL15|t2 0.160 0.074 2.164 0.030 0.160 0.160
## HL19|t1 -1.017 0.089 -11.394 0.000 -1.017 -1.017
## HL19|t2 0.293 0.075 3.915 0.000 0.293 0.293
## HL22|t1 -1.124 0.093 -12.051 0.000 -1.124 -1.124
## HL22|t2 -0.082 0.074 -1.112 0.266 -0.082 -0.082
## HL3|t1 -1.460 0.111 -13.205 0.000 -1.460 -1.460
## HL3|t2 -0.365 0.075 -4.846 0.000 -0.365 -0.365
## HL3|t3 0.749 0.082 9.181 0.000 0.749 0.749
## HL9|t1 -0.894 0.085 -10.477 0.000 -0.894 -0.894
## HL9|t2 0.151 0.074 2.047 0.041 0.151 0.151
## HL13|t1 -0.403 0.076 -5.310 0.000 -0.403 -0.403
## HL13|t2 0.961 0.087 10.994 0.000 0.961 0.961
## HL20|t1 -0.629 0.079 -7.950 0.000 -0.629 -0.629
## HL20|t2 0.640 0.079 8.063 0.000 0.640 0.640
## HL23|t1 -0.856 0.084 -10.160 0.000 -0.856 -0.856
## HL23|t2 0.578 0.078 7.382 0.000 0.578 0.578
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .HL1 0.708 0.708 0.708
## .HL8 0.532 0.532 0.532
## .HL14 0.759 0.759 0.759
## .HL18 0.480 0.480 0.480
## .HL25 0.678 0.678 0.678
## .HL2 0.728 0.728 0.728
## .HL4 0.503 0.503 0.503
## .HL6 0.766 0.766 0.766
## .HL10 0.435 0.435 0.435
## .HL17 0.669 0.669 0.669
## .HL7 0.716 0.716 0.716
## .HL12 0.541 0.541 0.541
## .HL16 0.575 0.575 0.575
## .HL21 0.532 0.532 0.532
## .HL24 0.514 0.514 0.514
## .HL5 0.598 0.598 0.598
## .HL11 0.673 0.673 0.673
## .HL15 0.620 0.620 0.620
## .HL19 0.692 0.692 0.692
## .HL22 0.857 0.857 0.857
## .HL3 0.624 0.624 0.624
## .HL9 0.655 0.655 0.655
## .HL13 0.636 0.636 0.636
## .HL20 0.401 0.401 0.401
## .HL23 0.536 0.536 0.536
## Teor_ved 1.000 1.000 1.000
## Prakt_ved 1.000 1.000 1.000
## Krit_mysl 1.000 1.000 1.000
## Sebauved 1.000 1.000 1.000
## Občianstvo 1.000 1.000 1.000
##
## Scales y*:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## HL1 1.000 1.000 1.000
## HL8 1.000 1.000 1.000
## HL14 1.000 1.000 1.000
## HL18 1.000 1.000 1.000
## HL25 1.000 1.000 1.000
## HL2 1.000 1.000 1.000
## HL4 1.000 1.000 1.000
## HL6 1.000 1.000 1.000
## HL10 1.000 1.000 1.000
## HL17 1.000 1.000 1.000
## HL7 1.000 1.000 1.000
## HL12 1.000 1.000 1.000
## HL16 1.000 1.000 1.000
## HL21 1.000 1.000 1.000
## HL24 1.000 1.000 1.000
## HL5 1.000 1.000 1.000
## HL11 1.000 1.000 1.000
## HL15 1.000 1.000 1.000
## HL19 1.000 1.000 1.000
## HL22 1.000 1.000 1.000
## HL3 1.000 1.000 1.000
## HL9 1.000 1.000 1.000
## HL13 1.000 1.000 1.000
## HL20 1.000 1.000 1.000
## HL23 1.000 1.000 1.000
##
## R-Square:
## Estimate
## HL1 0.292
## HL8 0.468
## HL14 0.241
## HL18 0.520
## HL25 0.322
## HL2 0.272
## HL4 0.497
## HL6 0.234
## HL10 0.565
## HL17 0.331
## HL7 0.284
## HL12 0.459
## HL16 0.425
## HL21 0.468
## HL24 0.486
## HL5 0.402
## HL11 0.327
## HL15 0.380
## HL19 0.308
## HL22 0.143
## HL3 0.376
## HL9 0.345
## HL13 0.364
## HL20 0.599
## HL23 0.464Priemerný faktorový náboj
mean(inspect(fitted.model,what="std")$lambda[inspect(fitted.model,what="std")$lambda > .0])## [1] 0.6118905table(inspect(fitted.model,what="std")$lambda[inspect(fitted.model,what="std")$lambda > .0] < .7)/25##
## FALSE TRUE
## 0.16 0.84Indexy blízkej zhody modelu a dát
Treba si všímať .scaled indexy
fitMeasures(fitted.model)## npar fmin
## 88.000 0.637
## chisq df
## 370.444 265.000
## pvalue chisq.scaled
## 0.000 219.259
## df.scaled pvalue.scaled
## 100.000 0.000
## chisq.scaling.factor baseline.chisq
## 1.690 10075.119
## baseline.df baseline.pvalue
## 300.000 0.000
## baseline.chisq.scaled baseline.df.scaled
## 1347.981 40.000
## baseline.pvalue.scaled baseline.chisq.scaling.factor
## 0.000 7.474
## cfi tli
## 0.989 0.988
## nnfi rfi
## 0.988 0.958
## nfi pnfi
## 0.963 0.851
## ifi rni
## 0.989 0.989
## cfi.scaled tli.scaled
## 0.909 0.964
## cfi.robust tli.robust
## NA NA
## nnfi.scaled nnfi.robust
## 0.964 NA
## rfi.scaled nfi.scaled
## 0.935 0.837
## ifi.scaled rni.scaled
## 0.837 0.988
## rni.robust rmsea
## NA 0.037
## rmsea.ci.lower rmsea.ci.upper
## 0.028 0.046
## rmsea.pvalue rmsea.scaled
## 0.994 0.064
## rmsea.ci.lower.scaled rmsea.ci.upper.scaled
## 0.055 0.073
## rmsea.pvalue.scaled rmsea.robust
## 0.005 NA
## rmsea.ci.lower.robust rmsea.ci.upper.robust
## NA NA
## rmsea.pvalue.robust rmr
## NA 0.065
## rmr_nomean srmr
## 0.067 0.065
## srmr_bentler srmr_bentler_nomean
## 0.065 0.067
## srmr_bollen srmr_bollen_nomean
## 0.065 0.067
## srmr_mplus srmr_mplus_nomean
## 0.065 0.067
## cn_05 cn_01
## 238.961 252.667
## gfi agfi
## 0.974 0.965
## pgfi mfi
## 0.731 0.834Diagram
semPaths(fitted.model, style = "mx", edge.label.cex = 0.7,
sizeLat = 7, nCharNodes = 0, nDigits = 2, "Standardized",
intercepts = FALSE, residuals = FALSE,
what = "path", edge.label.position = .5,
curvature = 4, layout = "circle",
node.width = 1.1, color = "white", thresholds = FALSE)## Warning in qgraph(Edgelist, labels = nLab, bidirectional = Bidir, directed
## = Directed, : The following arguments are not documented and likely not
## arguments of qgraph and thus ignored: loopRotation; residuals; residScale;
## residEdge; CircleEdgeEnd
Test modelu indikuje prítomnosť chybnej špecifikácie modelu. Popri indexoch blízkej zhody je preto potrebné analyzovať lokálne zdroje chybnej špecifikácie na základe matice reziduálnych korelácií
Matica reziduálnych korelácií
residuals <- residuals(fitted.model, type = "cor")$cor
residuals## HL1 HL8 HL14 HL18 HL25 HL2 HL4 HL6 HL10 HL17
## HL1 0.000
## HL8 0.017 0.000
## HL14 -0.025 -0.177 0.000
## HL18 -0.076 0.007 -0.176 0.000
## HL25 -0.097 -0.175 0.330 -0.012 0.000
## HL2 0.169 -0.090 0.045 -0.059 0.026 0.000
## HL4 -0.137 0.011 -0.060 0.054 -0.031 -0.072 0.000
## HL6 0.090 -0.071 0.104 -0.063 0.009 0.119 -0.141 0.000
## HL10 -0.030 0.030 -0.070 0.044 -0.050 -0.060 0.082 -0.071 0.000
## HL17 0.023 -0.029 -0.014 0.093 0.030 0.061 -0.073 0.163 -0.089 0.000
## HL7 0.046 0.056 -0.127 -0.153 -0.108 -0.054 0.000 0.075 0.018 0.032
## HL12 -0.009 -0.050 0.041 -0.013 0.097 -0.061 0.103 0.041 0.074 -0.042
## HL16 -0.032 0.018 -0.020 -0.009 -0.040 0.032 -0.042 0.001 0.006 0.091
## HL21 0.007 0.036 -0.087 0.050 -0.028 -0.022 -0.009 -0.016 -0.040 -0.025
## HL24 -0.022 0.022 0.059 -0.035 0.081 -0.123 0.009 -0.151 -0.050 -0.137
## HL5 -0.079 -0.011 0.045 -0.009 -0.036 -0.007 0.120 -0.034 0.043 0.012
## HL11 -0.004 -0.001 0.089 0.059 -0.071 0.026 -0.093 -0.027 -0.054 0.034
## HL15 0.074 0.053 -0.026 0.133 -0.155 0.006 -0.059 0.053 -0.085 0.011
## HL19 -0.049 -0.059 -0.048 0.040 -0.027 0.100 -0.036 -0.015 -0.017 0.021
## HL22 0.112 -0.101 0.000 -0.079 0.078 0.015 -0.098 0.103 -0.105 0.124
## HL3 -0.019 0.052 -0.089 -0.025 -0.001 0.113 0.077 -0.060 0.009 0.031
## HL9 0.048 0.064 -0.026 0.083 -0.081 0.019 -0.064 0.004 0.103 -0.010
## HL13 0.070 0.077 0.028 -0.078 0.063 -0.061 -0.111 0.017 -0.100 -0.045
## HL20 0.057 0.056 -0.037 0.006 -0.055 0.018 0.025 0.003 0.013 -0.095
## HL23 -0.064 -0.025 -0.170 -0.016 -0.088 -0.023 -0.023 -0.031 0.024 -0.029
## HL7 HL12 HL16 HL21 HL24 HL5 HL11 HL15 HL19 HL22
## HL1
## HL8
## HL14
## HL18
## HL25
## HL2
## HL4
## HL6
## HL10
## HL17
## HL7 0.000
## HL12 0.067 0.000
## HL16 0.154 0.041 0.000
## HL21 -0.065 -0.044 -0.068 0.000
## HL24 -0.121 -0.003 -0.061 0.004 0.000
## HL5 -0.011 0.012 -0.146 0.141 0.037 0.000
## HL11 -0.117 -0.062 0.037 -0.044 0.084 -0.064 0.000
## HL15 -0.078 -0.106 -0.007 0.102 -0.031 -0.057 0.063 0.000
## HL19 0.028 -0.031 -0.009 0.067 -0.035 -0.102 0.158 0.007 0.000
## HL22 -0.092 -0.039 -0.022 0.021 0.068 -0.081 0.046 -0.062 0.070 0.000
## HL3 -0.047 -0.064 0.023 0.026 0.034 0.006 -0.107 0.029 -0.105 -0.017
## HL9 0.056 -0.054 -0.034 -0.025 -0.001 0.017 -0.024 -0.071 -0.020 0.021
## HL13 -0.017 -0.010 0.007 -0.069 0.059 -0.075 0.023 0.067 -0.007 -0.035
## HL20 0.055 -0.005 0.000 -0.003 0.030 -0.048 -0.023 0.036 0.022 -0.029
## HL23 0.021 -0.104 -0.016 -0.046 0.117 0.038 0.043 -0.003 0.035 0.117
## HL3 HL9 HL13 HL20 HL23
## HL1
## HL8
## HL14
## HL18
## HL25
## HL2
## HL4
## HL6
## HL10
## HL17
## HL7
## HL12
## HL16
## HL21
## HL24
## HL5
## HL11
## HL15
## HL19
## HL22
## HL3 0.000
## HL9 -0.070 0.000
## HL13 0.119 -0.051 0.000
## HL20 -0.056 -0.102 0.030 0.000
## HL23 -0.017 0.132 -0.040 -0.002 0.000Priemer reziduálnych korelácií
mean(abs(residuals))## [1] 0.05282712Pre prehladnejšiu vizualizáciu, matica reziduí s vyznačenými reziduálnymi hodnotami > .1 (štandardizované z-reziduá je možné odhadnúť iba v prípade použitia estimátora z rodiny maximum likelihood. Arbitrárna hodnota .1 preto, lebo neumožní produkt dvoch nábojov > .3)
Počet premenných
p = 25Ak máme v matici (p(p+1)/2 - p) = 300 elementov (bez diagonály), tak
(p*(p+1)/2 - p)*.05## [1] 15z nich môže byť signifikantných na hladine alfa = .05
HRUBÁ APROXIMÁCIA - približne tolko elementov môže byť > .1 Diag = diagonála, >.1 = reziduálna hodnota vyššia ako .1
ifelse(residuals == 0, "Diag", ifelse(residuals > .1, ">.1", "."))## HL1 HL8 HL14 HL18 HL25 HL2 HL4 HL6 HL10 HL17
## HL1 "Diag" "." "." "." "." ">.1" "." "." "." "."
## HL8 "." "Diag" "." "." "." "." "." "." "." "."
## HL14 "." "." "Diag" "." ">.1" "." "." ">.1" "." "."
## HL18 "." "." "." "Diag" "." "." "." "." "." "."
## HL25 "." "." ">.1" "." "Diag" "." "." "." "." "."
## HL2 ">.1" "." "." "." "." "Diag" "." ">.1" "." "."
## HL4 "." "." "." "." "." "." "Diag" "." "." "."
## HL6 "." "." ">.1" "." "." ">.1" "." "Diag" "." ">.1"
## HL10 "." "." "." "." "." "." "." "." "Diag" "."
## HL17 "." "." "." "." "." "." "." ">.1" "." "Diag"
## HL7 "." "." "." "." "." "." "." "." "." "."
## HL12 "." "." "." "." "." "." ">.1" "." "." "."
## HL16 "." "." "." "." "." "." "." "." "." "."
## HL21 "." "." "." "." "." "." "." "." "." "."
## HL24 "." "." "." "." "." "." "." "." "." "."
## HL5 "." "." "." "." "." "." ">.1" "." "." "."
## HL11 "." "." "." "." "." "." "." "." "." "."
## HL15 "." "." "." ">.1" "." "." "." "." "." "."
## HL19 "." "." "." "." "." ">.1" "." "." "." "."
## HL22 ">.1" "." "." "." "." "." "." ">.1" "." ">.1"
## HL3 "." "." "." "." "." ">.1" "." "." "." "."
## HL9 "." "." "." "." "." "." "." "." ">.1" "."
## HL13 "." "." "." "." "." "." "." "." "." "."
## HL20 "." "." "." "." "." "." "." "." "." "."
## HL23 "." "." "." "." "." "." "." "." "." "."
## HL7 HL12 HL16 HL21 HL24 HL5 HL11 HL15 HL19 HL22
## HL1 "." "." "." "." "." "." "." "." "." ">.1"
## HL8 "." "." "." "." "." "." "." "." "." "."
## HL14 "." "." "." "." "." "." "." "." "." "."
## HL18 "." "." "." "." "." "." "." ">.1" "." "."
## HL25 "." "." "." "." "." "." "." "." "." "."
## HL2 "." "." "." "." "." "." "." "." ">.1" "."
## HL4 "." ">.1" "." "." "." ">.1" "." "." "." "."
## HL6 "." "." "." "." "." "." "." "." "." ">.1"
## HL10 "." "." "." "." "." "." "." "." "." "."
## HL17 "." "." "." "." "." "." "." "." "." ">.1"
## HL7 "Diag" "." ">.1" "." "." "." "." "." "." "."
## HL12 "." "Diag" "." "." "." "." "." "." "." "."
## HL16 ">.1" "." "Diag" "." "." "." "." "." "." "."
## HL21 "." "." "." "Diag" "." ">.1" "." ">.1" "." "."
## HL24 "." "." "." "." "Diag" "." "." "." "." "."
## HL5 "." "." "." ">.1" "." "Diag" "." "." "." "."
## HL11 "." "." "." "." "." "." "Diag" "." ">.1" "."
## HL15 "." "." "." ">.1" "." "." "." "Diag" "." "."
## HL19 "." "." "." "." "." "." ">.1" "." "Diag" "."
## HL22 "." "." "." "." "." "." "." "." "." "Diag"
## HL3 "." "." "." "." "." "." "." "." "." "."
## HL9 "." "." "." "." "." "." "." "." "." "."
## HL13 "." "." "." "." "." "." "." "." "." "."
## HL20 "." "." "." "." "." "." "." "." "." "."
## HL23 "." "." "." "." ">.1" "." "." "." "." ">.1"
## HL3 HL9 HL13 HL20 HL23
## HL1 "." "." "." "." "."
## HL8 "." "." "." "." "."
## HL14 "." "." "." "." "."
## HL18 "." "." "." "." "."
## HL25 "." "." "." "." "."
## HL2 ">.1" "." "." "." "."
## HL4 "." "." "." "." "."
## HL6 "." "." "." "." "."
## HL10 "." ">.1" "." "." "."
## HL17 "." "." "." "." "."
## HL7 "." "." "." "." "."
## HL12 "." "." "." "." "."
## HL16 "." "." "." "." "."
## HL21 "." "." "." "." "."
## HL24 "." "." "." "." ">.1"
## HL5 "." "." "." "." "."
## HL11 "." "." "." "." "."
## HL15 "." "." "." "." "."
## HL19 "." "." "." "." "."
## HL22 "." "." "." "." ">.1"
## HL3 "Diag" "." ">.1" "." "."
## HL9 "." "Diag" "." "." ">.1"
## HL13 ">.1" "." "Diag" "." "."
## HL20 "." "." "." "Diag" "."
## HL23 "." ">.1" "." "." "Diag"Odhad reliability
rel.HLQ.collumns <- full.data %>% select(HL1:HL25,RHL1:RHL25, ID3)
delete.na <- function(rel.HLQ.collumns, n=NULL) {
rel.HLQ.collumns[rowSums(is.na(rel.HLQ.collumns)) <= n,]
}
rel.HLQ.na.rm <- delete.na(rel.HLQ.collumns, n = 10)Odhad internej konzistencie - McDonaldova Omega
round(omega(rel.HLQ.na.rm[,c("HL1", "HL8", "HL14", "HL18", "HL25")], poly = TRUE, plot = FALSE)$omega.tot, 2) # pre theor_know## 10 cells were adjusted for 0 values using the correction for continuity. Examine your data carefully.## Loading required namespace: GPArotation## [1] 0.79round(omega(rel.HLQ.na.rm[,c("HL2", "HL4", "HL6", "HL10", "HL17")], poly = TRUE, plot = FALSE)$omega.tot, 2) # pre prac_know## 10 cells were adjusted for 0 values using the correction for continuity. Examine your data carefully.## [1] 0.79round(omega(rel.HLQ.na.rm[,c("HL7", "HL12", "HL16", "HL21", "HL24")], poly = TRUE, plot = FALSE)$omega.tot, 2) # pre crit_think## 10 cells were adjusted for 0 values using the correction for continuity. Examine your data carefully.## [1] 0.81round(omega(rel.HLQ.na.rm[,c("HL5", "HL11", "HL15", "HL19", "HL22")], poly = TRUE, plot = FALSE)$omega.tot, 2) # pre self_aware## 10 cells were adjusted for 0 values using the correction for continuity. Examine your data carefully.## [1] 0.77round(omega(rel.HLQ.na.rm[,c("HL3", "HL9", "HL13", "HL20", "HL23")], poly = TRUE, plot = FALSE)$omega.tot, 2) # pre citizenship## 10 cells were adjusted for 0 values using the correction for continuity. Examine your data carefully.## [1] 0.85Odhad stability v čase
Výpočet sumárneho skóre každej z 5 dimenzií HL
rel.HLQ.na.rm$theor_know <- rowMeans(rel.HLQ.na.rm[,c("HL1", "HL8", "HL14", "HL18", "HL25")], na.rm = TRUE)
rel.HLQ.na.rm$prac_know <- rowMeans(rel.HLQ.na.rm[,c("HL2", "HL4", "HL6", "HL10", "HL17")], na.rm = TRUE)
rel.HLQ.na.rm$crit_think <- rowMeans(rel.HLQ.na.rm[,c("HL7", "HL12", "HL16", "HL21", "HL24")], na.rm = TRUE)
rel.HLQ.na.rm$self_aware <- rowMeans(rel.HLQ.na.rm[,c("HL5", "HL11", "HL15", "HL19", "HL22")], na.rm = TRUE)
rel.HLQ.na.rm$citizenship <- rowMeans(rel.HLQ.na.rm[,c("HL3", "HL9", "HL13", "HL20", "HL23")], na.rm = TRUE)Výpočet sumárneho skóre každej z 5 dimenzií HL - RETEST
rel.HLQ.na.rm$Rtheor_know <- rowMeans(rel.HLQ.na.rm[,c("RHL1", "RHL8", "RHL14", "RHL18", "RHL25")], na.rm = TRUE)
rel.HLQ.na.rm$Rprac_know <- rowMeans(rel.HLQ.na.rm[,c("RHL2", "RHL4", "RHL6", "RHL10", "RHL17")], na.rm = TRUE)
rel.HLQ.na.rm$Rcrit_think <- rowMeans(rel.HLQ.na.rm[,c("RHL7", "RHL12", "RHL16", "RHL21", "RHL24")], na.rm = TRUE)
rel.HLQ.na.rm$Rself_aware <- rowMeans(rel.HLQ.na.rm[,c("RHL5", "RHL11", "RHL15", "RHL19", "RHL22")], na.rm = TRUE)
rel.HLQ.na.rm$Rcitizenship <- rowMeans(rel.HLQ.na.rm[,c("RHL3", "RHL9", "RHL13", "RHL20", "RHL23")], na.rm = TRUE)Test-retest korelácia
round(with(rel.HLQ.na.rm, cor.test(theor_know, Rtheor_know))$estimate, 2) # pre theor_know## cor
## 0.66round(with(rel.HLQ.na.rm, cor.test(prac_know, Rprac_know))$estimate, 2) # pre prac_know## cor
## 0.66round(with(rel.HLQ.na.rm, cor.test(crit_think, Rcrit_think))$estimate, 2) # pre crit_think## cor
## 0.64round(with(rel.HLQ.na.rm, cor.test(self_aware, Rself_aware))$estimate, 2) # pre self_aware## cor
## 0.65round(with(rel.HLQ.na.rm, cor.test(citizenship, Rcitizenship))$estimate, 2) # pre citizenship## cor
## 0.67Intra-class korelácie pre 5 dimenzií HL
Overenie prítomnosti hierarchickej štruktúry v dátach, ktorá mohla vznikúť použitým spôsobom vzorkovania populácie (cluster sampling). Cluster = školská trieda (premenná ID3).
ICCs <- (lapply(rel.HLQ.na.rm[,c("theor_know", "prac_know", "crit_think", "self_aware", "citizenship")],
function(x){ICCest(ID3, x, rel.HLQ.na.rm)}))## Warning in ICCest(ID3, x, rel.HLQ.na.rm): 'x' has been coerced to a factor
## Warning in ICCest(ID3, x, rel.HLQ.na.rm): 'x' has been coerced to a factor
## Warning in ICCest(ID3, x, rel.HLQ.na.rm): 'x' has been coerced to a factor
## Warning in ICCest(ID3, x, rel.HLQ.na.rm): 'x' has been coerced to a factor
## Warning in ICCest(ID3, x, rel.HLQ.na.rm): 'x' has been coerced to a factorICCs.theor_know <- abs(ICCs$theor_know$ICC) # ICC pre theor_know
ICCs.theor_know## [1] 0.1028632ICCs.prac_know <- abs(ICCs$prac_know$ICC) # ICC pre prac_know
ICCs.prac_know## [1] 0.06648811ICCs.crit_think <- abs(ICCs$crit_think$ICC) # ICC pre crit_think
ICCs.crit_think## [1] 0.09452927ICCs.self_aware <- abs(ICCs$self_aware$ICC) # ICC pre self_aware
ICCs.self_aware## [1] 0.0004898439ICCs.citizenship <- abs(ICCs$citizenship$ICC) # ICC pre citizenship
ICCs.citizenship## [1] 0.07762138Priemerná hodnota design effect
1 + (cluster.size - 1)*((ICCs.theor_know+ICCs.prac_know+ICCs.crit_think+ICCs.self_aware+ICCs.citizenship)/5)## [1] 1.879408Intra-class korelácie a z nich vychádzajúca priemerná hodnota efektu výskumného dizajnu (Muthén & Sattora, 1995) marginálne indikuje prípustnosť považovať dáta za jednoúrovňové.
1-faktorový model
Špecifikácia 1-faktorového modelu merania
model2 <- '
HLQ =~ a*HL1 + b*HL2 + c*HL3 + d*HL4 + e*HL5 + f*HL6 + g*HL7 + h*HL8 + i*HL9 +
j*HL10 + k*HL11 + l*HL12 + m*HL13 + n*HL14 + o*HL15 + p*HL16 + q*HL17 + r*HL18 +
s*HL19 + t*HL20 + u*HL21 + v*HL22 + x*HL23 + y*HL24 + z*HL25
'Estimácia 1-faktorového modelu
Podľa Muthén, 1984
fitted.model2 <- cfa(model = model2, data = data, meanstructure = TRUE, std.lv = TRUE, mimic = "Mplus",
estimator = "WLSMVS", test = "Satterthwaite", bootstrap = 50000,
ordered = c("HL1", "HL2", "HL3", "HL4", "HL5", "HL6", "HL7", "HL8", "HL9", "HL10",
"HL11", "HL12", "HL13", "HL14", "HL15", "HL16", "HL17", "HL18", "HL19",
"HL20", "HL21", "HL22", "HL23", "HL24", "HL25"))Je v poriadku ignorovať toto chybové hlásenie. Počet buniek HLa x HLb s nulovou frekvenciou je už zanedbateľný.
Test modelu, odhady voľných parametrov
Stačí si všímať “Robust” test, Latent variable, Covariances a R-square. Intercepts, Thresholds, Intercepts (…) môžte kľudne ignorovať.
summary(fitted.model2, standardized = TRUE, rsquare = TRUE)## lavaan (0.5-23.1097) converged normally after 20 iterations
##
## Number of observations 291
##
## Estimator DWLS Robust
## Minimum Function Test Statistic 391.193 219.010
## Degrees of freedom 275 99
## P-value (Chi-square) 0.000 0.000
## Scaling correction factor 1.786
## for the mean and variance adjusted correction (WLSMV)
##
## Parameter Estimates:
##
## Information Expected
## Standard Errors Robust.sem
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## HLQ =~
## HL1 (a) 0.497 0.054 9.278 0.000 0.497 0.497
## HL2 (b) 0.491 0.056 8.783 0.000 0.491 0.491
## HL3 (c) 0.606 0.043 13.949 0.000 0.606 0.606
## HL4 (d) 0.661 0.040 16.500 0.000 0.661 0.661
## HL5 (e) 0.629 0.041 15.464 0.000 0.629 0.629
## HL6 (f) 0.455 0.055 8.344 0.000 0.455 0.455
## HL7 (g) 0.524 0.046 11.328 0.000 0.524 0.524
## HL8 (h) 0.627 0.041 15.275 0.000 0.627 0.627
## HL9 (i) 0.581 0.047 12.378 0.000 0.581 0.581
## HL10 (j) 0.705 0.036 19.596 0.000 0.705 0.705
## HL11 (k) 0.566 0.050 11.318 0.000 0.566 0.566
## HL12 (l) 0.667 0.036 18.440 0.000 0.667 0.667
## HL13 (m) 0.596 0.045 13.358 0.000 0.596 0.596
## HL14 (n) 0.448 0.053 8.391 0.000 0.448 0.448
## HL15 (o) 0.611 0.044 13.922 0.000 0.611 0.611
## HL16 (p) 0.642 0.038 17.014 0.000 0.642 0.642
## HL17 (q) 0.539 0.054 10.030 0.000 0.539 0.539
## HL18 (r) 0.662 0.050 13.283 0.000 0.662 0.662
## HL19 (s) 0.549 0.047 11.683 0.000 0.549 0.549
## HL20 (t) 0.765 0.033 23.458 0.000 0.765 0.765
## HL21 (u) 0.673 0.038 17.921 0.000 0.673 0.673
## HL22 (v) 0.374 0.060 6.232 0.000 0.374 0.374
## HL23 (x) 0.674 0.039 17.472 0.000 0.674 0.674
## HL24 (y) 0.687 0.039 17.832 0.000 0.687 0.687
## HL25 (z) 0.517 0.051 10.237 0.000 0.517 0.517
##
## Intercepts:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .HL1 0.000 0.000 0.000
## .HL2 0.000 0.000 0.000
## .HL3 0.000 0.000 0.000
## .HL4 0.000 0.000 0.000
## .HL5 0.000 0.000 0.000
## .HL6 0.000 0.000 0.000
## .HL7 0.000 0.000 0.000
## .HL8 0.000 0.000 0.000
## .HL9 0.000 0.000 0.000
## .HL10 0.000 0.000 0.000
## .HL11 0.000 0.000 0.000
## .HL12 0.000 0.000 0.000
## .HL13 0.000 0.000 0.000
## .HL14 0.000 0.000 0.000
## .HL15 0.000 0.000 0.000
## .HL16 0.000 0.000 0.000
## .HL17 0.000 0.000 0.000
## .HL18 0.000 0.000 0.000
## .HL19 0.000 0.000 0.000
## .HL20 0.000 0.000 0.000
## .HL21 0.000 0.000 0.000
## .HL22 0.000 0.000 0.000
## .HL23 0.000 0.000 0.000
## .HL24 0.000 0.000 0.000
## .HL25 0.000 0.000 0.000
## HLQ 0.000 0.000 0.000
##
## Thresholds:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## HL1|t1 -1.284 0.100 -12.774 0.000 -1.284 -1.284
## HL1|t2 0.537 0.078 6.925 0.000 0.537 0.537
## HL2|t1 -1.174 0.095 -12.311 0.000 -1.174 -1.174
## HL2|t2 0.143 0.074 1.931 0.054 0.143 0.143
## HL3|t1 -1.460 0.111 -13.205 0.000 -1.460 -1.460
## HL3|t2 -0.365 0.075 -4.846 0.000 -0.365 -0.365
## HL3|t3 0.749 0.082 9.181 0.000 0.749 0.749
## HL4|t1 -0.598 0.079 -7.610 0.000 -0.598 -0.598
## HL4|t2 0.412 0.076 5.426 0.000 0.412 0.412
## HL5|t1 -0.869 0.085 -10.266 0.000 -0.869 -0.869
## HL5|t2 0.320 0.075 4.265 0.000 0.320 0.320
## HL6|t1 -1.124 0.093 -12.051 0.000 -1.124 -1.124
## HL6|t2 -0.108 0.074 -1.463 0.144 -0.108 -0.108
## HL7|t1 -1.539 0.116 -13.276 0.000 -1.539 -1.539
## HL7|t2 -0.459 0.076 -6.004 0.000 -0.459 -0.459
## HL7|t3 0.820 0.083 9.837 0.000 0.820 0.820
## HL8|t1 -0.772 0.082 -9.401 0.000 -0.772 -0.772
## HL8|t2 0.302 0.075 4.032 0.000 0.302 0.302
## HL9|t1 -0.894 0.085 -10.477 0.000 -0.894 -0.894
## HL9|t2 0.151 0.074 2.047 0.041 0.151 0.151
## HL10|t1 -0.988 0.088 -11.196 0.000 -0.988 -0.988
## HL10|t2 0.143 0.074 1.931 0.054 0.143 0.143
## HL11|t1 -1.366 0.105 -13.024 0.000 -1.366 -1.366
## HL11|t2 -0.134 0.074 -1.814 0.070 -0.134 -0.134
## HL12|t1 -1.435 0.109 -13.168 0.000 -1.435 -1.435
## HL12|t2 -0.412 0.076 -5.426 0.000 -0.412 -0.412
## HL12|t3 0.727 0.081 8.960 0.000 0.727 0.727
## HL13|t1 -0.403 0.076 -5.310 0.000 -0.403 -0.403
## HL13|t2 0.961 0.087 10.994 0.000 0.961 0.961
## HL14|t1 -0.869 0.085 -10.266 0.000 -0.869 -0.869
## HL14|t2 0.213 0.074 2.865 0.004 0.213 0.213
## HL15|t1 -0.974 0.088 -11.095 0.000 -0.974 -0.974
## HL15|t2 0.160 0.074 2.164 0.030 0.160 0.160
## HL16|t1 -1.630 0.123 -13.266 0.000 -1.630 -1.630
## HL16|t2 -0.619 0.079 -7.837 0.000 -0.619 -0.619
## HL16|t3 0.761 0.082 9.292 0.000 0.761 0.761
## HL17|t1 -1.389 0.106 -13.077 0.000 -1.389 -1.389
## HL17|t2 -0.384 0.076 -5.078 0.000 -0.384 -0.384
## HL18|t1 -0.275 0.075 -3.682 0.000 -0.275 -0.275
## HL19|t1 -1.017 0.089 -11.394 0.000 -1.017 -1.017
## HL19|t2 0.293 0.075 3.915 0.000 0.293 0.293
## HL20|t1 -0.629 0.079 -7.950 0.000 -0.629 -0.629
## HL20|t2 0.640 0.079 8.063 0.000 0.640 0.640
## HL21|t1 -0.844 0.084 -10.053 0.000 -0.844 -0.844
## HL21|t2 0.459 0.076 6.004 0.000 0.459 0.459
## HL22|t1 -1.124 0.093 -12.051 0.000 -1.124 -1.124
## HL22|t2 -0.082 0.074 -1.112 0.266 -0.082 -0.082
## HL23|t1 -0.856 0.084 -10.160 0.000 -0.856 -0.856
## HL23|t2 0.578 0.078 7.382 0.000 0.578 0.578
## HL24|t1 -0.431 0.076 -5.658 0.000 -0.431 -0.431
## HL24|t2 1.002 0.089 11.295 0.000 1.002 1.002
## HL25|t1 -0.557 0.078 -7.154 0.000 -0.557 -0.557
## HL25|t2 0.498 0.077 6.465 0.000 0.498 0.498
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .HL1 0.753 0.753 0.753
## .HL2 0.759 0.759 0.759
## .HL3 0.633 0.633 0.633
## .HL4 0.563 0.563 0.563
## .HL5 0.605 0.605 0.605
## .HL6 0.793 0.793 0.793
## .HL7 0.726 0.726 0.726
## .HL8 0.607 0.607 0.607
## .HL9 0.662 0.662 0.662
## .HL10 0.503 0.503 0.503
## .HL11 0.680 0.680 0.680
## .HL12 0.555 0.555 0.555
## .HL13 0.645 0.645 0.645
## .HL14 0.799 0.799 0.799
## .HL15 0.627 0.627 0.627
## .HL16 0.588 0.588 0.588
## .HL17 0.709 0.709 0.709
## .HL18 0.562 0.562 0.562
## .HL19 0.698 0.698 0.698
## .HL20 0.415 0.415 0.415
## .HL21 0.547 0.547 0.547
## .HL22 0.860 0.860 0.860
## .HL23 0.546 0.546 0.546
## .HL24 0.527 0.527 0.527
## .HL25 0.733 0.733 0.733
## HLQ 1.000 1.000 1.000
##
## Scales y*:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## HL1 1.000 1.000 1.000
## HL2 1.000 1.000 1.000
## HL3 1.000 1.000 1.000
## HL4 1.000 1.000 1.000
## HL5 1.000 1.000 1.000
## HL6 1.000 1.000 1.000
## HL7 1.000 1.000 1.000
## HL8 1.000 1.000 1.000
## HL9 1.000 1.000 1.000
## HL10 1.000 1.000 1.000
## HL11 1.000 1.000 1.000
## HL12 1.000 1.000 1.000
## HL13 1.000 1.000 1.000
## HL14 1.000 1.000 1.000
## HL15 1.000 1.000 1.000
## HL16 1.000 1.000 1.000
## HL17 1.000 1.000 1.000
## HL18 1.000 1.000 1.000
## HL19 1.000 1.000 1.000
## HL20 1.000 1.000 1.000
## HL21 1.000 1.000 1.000
## HL22 1.000 1.000 1.000
## HL23 1.000 1.000 1.000
## HL24 1.000 1.000 1.000
## HL25 1.000 1.000 1.000
##
## R-Square:
## Estimate
## HL1 0.247
## HL2 0.241
## HL3 0.367
## HL4 0.437
## HL5 0.395
## HL6 0.207
## HL7 0.274
## HL8 0.393
## HL9 0.338
## HL10 0.497
## HL11 0.320
## HL12 0.445
## HL13 0.355
## HL14 0.201
## HL15 0.373
## HL16 0.412
## HL17 0.291
## HL18 0.438
## HL19 0.302
## HL20 0.585
## HL21 0.453
## HL22 0.140
## HL23 0.454
## HL24 0.473
## HL25 0.267Priemerný faktorový náboj
mean(inspect(fitted.model2,what="std")$lambda)## [1] 0.5898883table(inspect(fitted.model2,what="std")$lambda < .7)/25##
## FALSE TRUE
## 0.08 0.92Indexy blízkej zhody modelu a dát
Treba si všímať .scaled indexy
fitMeasures(fitted.model2)## npar fmin
## 78.000 0.672
## chisq df
## 391.193 275.000
## pvalue chisq.scaled
## 0.000 219.010
## df.scaled pvalue.scaled
## 99.000 0.000
## chisq.scaling.factor baseline.chisq
## 1.786 10075.119
## baseline.df baseline.pvalue
## 300.000 0.000
## baseline.chisq.scaled baseline.df.scaled
## 1347.981 40.000
## baseline.pvalue.scaled baseline.chisq.scaling.factor
## 0.000 7.474
## cfi tli
## 0.988 0.987
## nnfi rfi
## 0.987 0.958
## nfi pnfi
## 0.961 0.881
## ifi rni
## 0.988 0.988
## cfi.scaled tli.scaled
## 0.908 0.963
## cfi.robust tli.robust
## NA NA
## nnfi.scaled nnfi.robust
## 0.963 NA
## rfi.scaled nfi.scaled
## 0.934 0.838
## ifi.scaled rni.scaled
## 0.838 0.988
## rni.robust rmsea
## NA 0.038
## rmsea.ci.lower rmsea.ci.upper
## 0.029 0.047
## rmsea.pvalue rmsea.scaled
## 0.991 0.065
## rmsea.ci.lower.scaled rmsea.ci.upper.scaled
## 0.056 0.073
## rmsea.pvalue.scaled rmsea.robust
## 0.003 NA
## rmsea.ci.lower.robust rmsea.ci.upper.robust
## NA NA
## rmsea.pvalue.robust rmr
## NA 0.066
## rmr_nomean srmr
## 0.068 0.066
## srmr_bentler srmr_bentler_nomean
## 0.066 0.068
## srmr_bollen srmr_bollen_nomean
## 0.066 0.068
## srmr_mplus srmr_mplus_nomean
## 0.066 0.068
## cn_05 cn_01
## 234.278 247.475
## gfi agfi
## 0.972 0.964
## pgfi mfi
## 0.757 0.818\(χ^{2}\) test rozdielov medzi modelmi - p-hodnota
pchisq((fitted.model2@test[[2]]$stat - fitted.model@test[[2]]$stat),
(fitted.model2@test[[1]]$df - fitted.model@test[[1]]$df),
lower.tail = FALSE)## [1] 1Diagram
semPaths(fitted.model2, style = "mx", layout = "circle",
edge.label.cex = 0.5, sizeLat = 5, nCharNodes = 0,
nDigits = 2, "Standardized",
intercepts = FALSE, residuals = TRUE, exoVar = FALSE,
fade = TRUE, groups = "latents", pastel = TRUE)## Warning in qgraph(Edgelist, labels = nLab, bidirectional = Bidir, directed
## = Directed, : The following arguments are not documented and likely not
## arguments of qgraph and thus ignored: loopRotation; residuals; residScale;
## residEdge; CircleEdgeEnd
Test modelu indikuje prítomnosť chybnej špecifikácie modelu. Popri indexoch blízkej zhody je preto potrebné analyzovať lokálne zdroje chybnej špecifikácie na základe matice reziduálnych korelácií
Matica reziduálnych korelácií
residuals.m2 <- residuals(fitted.model2, type = "cor" )$cor
residuals.m2## HL1 HL2 HL3 HL4 HL5 HL6 HL7 HL8 HL9 HL10
## HL1 0.000
## HL2 0.147 0.000
## HL3 -0.020 0.102 0.000
## HL4 -0.166 -0.029 0.064 0.000
## HL5 -0.085 -0.017 0.018 0.108 0.000
## HL6 0.069 0.148 -0.070 -0.100 -0.043 0.000
## HL7 0.036 -0.049 -0.051 0.008 -0.015 0.080 0.000
## HL8 0.074 -0.117 0.053 -0.024 -0.017 -0.095 0.045 0.000
## HL9 0.046 0.009 -0.063 -0.077 0.028 -0.006 0.052 0.064 0.000
## HL10 -0.061 -0.014 -0.005 0.145 0.030 -0.028 0.027 -0.007 0.089 0.000
## HL11 -0.010 0.017 -0.097 -0.104 -0.057 -0.035 -0.121 -0.006 -0.015 -0.065
## HL12 -0.022 -0.055 -0.070 0.113 0.006 0.047 0.078 -0.064 -0.060 0.084
## HL13 0.069 -0.072 0.127 -0.124 -0.064 0.008 -0.021 0.078 -0.043 -0.113
## HL14 0.017 0.026 -0.087 -0.084 0.041 0.088 -0.134 -0.122 -0.025 -0.095
## HL15 0.068 -0.003 0.040 -0.071 -0.050 0.045 -0.081 0.046 -0.060 -0.097
## HL16 -0.044 0.037 0.017 -0.033 -0.151 0.007 0.165 0.004 -0.040 0.016
## HL17 -0.001 0.096 0.020 -0.024 0.003 0.196 0.039 -0.058 -0.021 -0.037
## HL18 -0.016 -0.088 -0.025 0.016 -0.017 -0.089 -0.166 0.086 0.083 0.004
## HL19 -0.054 0.092 -0.094 -0.046 -0.096 -0.023 0.025 -0.065 -0.010 -0.028
## HL20 0.056 0.004 -0.046 0.008 -0.034 -0.009 0.050 0.057 -0.092 -0.004
## HL21 -0.006 -0.016 0.020 0.000 0.136 -0.010 -0.053 0.022 -0.031 -0.030
## HL22 0.108 0.010 -0.010 -0.105 -0.077 0.098 -0.094 -0.105 0.028 -0.113
## HL23 -0.065 -0.035 -0.007 -0.037 0.050 -0.042 0.016 -0.024 0.140 0.008
## HL24 -0.036 -0.118 0.027 0.018 0.030 -0.146 -0.110 0.006 -0.009 -0.041
## HL25 -0.048 0.005 0.002 -0.059 -0.040 -0.010 -0.116 -0.111 -0.080 -0.079
## HL11 HL12 HL13 HL14 HL15 HL16 HL17 HL18 HL19 HL20
## HL1
## HL2
## HL3
## HL4
## HL5
## HL6
## HL7
## HL8
## HL9
## HL10
## HL11 0.000
## HL12 -0.066 0.000
## HL13 0.034 -0.016 0.000
## HL14 0.086 0.031 0.030 0.000
## HL15 0.070 -0.111 0.079 -0.029 0.000
## HL16 0.033 0.055 0.001 -0.029 -0.011 0.000
## HL17 0.025 -0.034 -0.056 -0.034 0.002 0.099 0.000
## HL18 0.053 -0.029 -0.078 -0.119 0.126 -0.024 0.063 0.000
## HL19 0.164 -0.035 0.003 -0.051 0.013 -0.013 0.013 0.033 0.000
## HL20 -0.010 -0.012 0.040 -0.035 0.050 -0.007 -0.108 0.006 0.035 0.000
## HL21 -0.049 -0.029 -0.074 -0.096 0.097 -0.053 -0.017 0.034 0.062 -0.010
## HL22 0.050 -0.042 -0.028 -0.002 -0.057 -0.025 0.119 -0.083 0.075 -0.020
## HL23 0.055 -0.111 -0.030 -0.168 0.010 -0.022 -0.041 -0.016 0.046 0.010
## HL24 0.079 0.011 0.052 0.049 -0.037 -0.047 -0.129 -0.052 -0.040 0.021
## HL25 -0.074 0.087 0.065 0.376 -0.158 -0.050 0.008 0.055 -0.030 -0.052
## HL21 HL22 HL23 HL24 HL25
## HL1
## HL2
## HL3
## HL4
## HL5
## HL6
## HL7
## HL8
## HL9
## HL10
## HL11
## HL12
## HL13
## HL14
## HL15
## HL16
## HL17
## HL18
## HL19
## HL20
## HL21 0.000
## HL22 0.019 0.000
## HL23 -0.052 0.125 0.000
## HL24 0.018 0.064 0.109 0.000
## HL25 -0.039 0.076 -0.086 0.069 0.000Priemer reziduálnych korelácií
mean(abs(residuals.m2))## [1] 0.0528761Pre prehladnejšiu vizualizáciu, matica reziduí s vyznačenými reziduálnymi hodnotami > .1 (štandardizované z-reziduá je možné odhadnúť iba v prípade použitia estimátora z rodiny maximum likelihood. Arbitrárna hodnota .1 preto, lebo neumožní produkt dvoch nábojov > .3)
Počet premenných
p = 25Ak máme v matici (p(p+1)/2 - p) = 300 elementov (bez diagonály), tak
(p*(p+1)/2 - p)*.05## [1] 15z nich môže byť signifikantných na hladine alfa = .05
HRUBÁ APROXIMÁCIA - približne toľko elementov môže byť > .1 Diag = diagonála, >.1 = reziduálna hodnota vyššia ako .1
ifelse(residuals.m2 == 0, "Diag", ifelse(residuals.m2 > .1, ">.1", "."))## HL1 HL2 HL3 HL4 HL5 HL6 HL7 HL8 HL9 HL10
## HL1 "Diag" ">.1" "." "." "." "." "." "." "." "."
## HL2 ">.1" "Diag" ">.1" "." "." ">.1" "." "." "." "."
## HL3 "." ">.1" "Diag" "." "." "." "." "." "." "."
## HL4 "." "." "." "Diag" ">.1" "." "." "." "." ">.1"
## HL5 "." "." "." ">.1" "Diag" "." "." "." "." "."
## HL6 "." ">.1" "." "." "." "Diag" "." "." "." "."
## HL7 "." "." "." "." "." "." "Diag" "." "." "."
## HL8 "." "." "." "." "." "." "." "Diag" "." "."
## HL9 "." "." "." "." "." "." "." "." "Diag" "."
## HL10 "." "." "." ">.1" "." "." "." "." "." "Diag"
## HL11 "." "." "." "." "." "." "." "." "." "."
## HL12 "." "." "." ">.1" "." "." "." "." "." "."
## HL13 "." "." ">.1" "." "." "." "." "." "." "."
## HL14 "." "." "." "." "." "." "." "." "." "."
## HL15 "." "." "." "." "." "." "." "." "." "."
## HL16 "." "." "." "." "." "." ">.1" "." "." "."
## HL17 "." "." "." "." "." ">.1" "." "." "." "."
## HL18 "." "." "." "." "." "." "." "." "." "."
## HL19 "." "." "." "." "." "." "." "." "." "."
## HL20 "." "." "." "." "." "." "." "." "." "."
## HL21 "." "." "." "." ">.1" "." "." "." "." "."
## HL22 ">.1" "." "." "." "." "." "." "." "." "."
## HL23 "." "." "." "." "." "." "." "." ">.1" "."
## HL24 "." "." "." "." "." "." "." "." "." "."
## HL25 "." "." "." "." "." "." "." "." "." "."
## HL11 HL12 HL13 HL14 HL15 HL16 HL17 HL18 HL19 HL20
## HL1 "." "." "." "." "." "." "." "." "." "."
## HL2 "." "." "." "." "." "." "." "." "." "."
## HL3 "." "." ">.1" "." "." "." "." "." "." "."
## HL4 "." ">.1" "." "." "." "." "." "." "." "."
## HL5 "." "." "." "." "." "." "." "." "." "."
## HL6 "." "." "." "." "." "." ">.1" "." "." "."
## HL7 "." "." "." "." "." ">.1" "." "." "." "."
## HL8 "." "." "." "." "." "." "." "." "." "."
## HL9 "." "." "." "." "." "." "." "." "." "."
## HL10 "." "." "." "." "." "." "." "." "." "."
## HL11 "Diag" "." "." "." "." "." "." "." ">.1" "."
## HL12 "." "Diag" "." "." "." "." "." "." "." "."
## HL13 "." "." "Diag" "." "." "." "." "." "." "."
## HL14 "." "." "." "Diag" "." "." "." "." "." "."
## HL15 "." "." "." "." "Diag" "." "." ">.1" "." "."
## HL16 "." "." "." "." "." "Diag" "." "." "." "."
## HL17 "." "." "." "." "." "." "Diag" "." "." "."
## HL18 "." "." "." "." ">.1" "." "." "Diag" "." "."
## HL19 ">.1" "." "." "." "." "." "." "." "Diag" "."
## HL20 "." "." "." "." "." "." "." "." "." "Diag"
## HL21 "." "." "." "." "." "." "." "." "." "."
## HL22 "." "." "." "." "." "." ">.1" "." "." "."
## HL23 "." "." "." "." "." "." "." "." "." "."
## HL24 "." "." "." "." "." "." "." "." "." "."
## HL25 "." "." "." ">.1" "." "." "." "." "." "."
## HL21 HL22 HL23 HL24 HL25
## HL1 "." ">.1" "." "." "."
## HL2 "." "." "." "." "."
## HL3 "." "." "." "." "."
## HL4 "." "." "." "." "."
## HL5 ">.1" "." "." "." "."
## HL6 "." "." "." "." "."
## HL7 "." "." "." "." "."
## HL8 "." "." "." "." "."
## HL9 "." "." ">.1" "." "."
## HL10 "." "." "." "." "."
## HL11 "." "." "." "." "."
## HL12 "." "." "." "." "."
## HL13 "." "." "." "." "."
## HL14 "." "." "." "." ">.1"
## HL15 "." "." "." "." "."
## HL16 "." "." "." "." "."
## HL17 "." ">.1" "." "." "."
## HL18 "." "." "." "." "."
## HL19 "." "." "." "." "."
## HL20 "." "." "." "." "."
## HL21 "Diag" "." "." "." "."
## HL22 "." "Diag" ">.1" "." "."
## HL23 "." ">.1" "Diag" ">.1" "."
## HL24 "." "." ">.1" "Diag" "."
## HL25 "." "." "." "." "Diag"Odhad reliability
Odhad internej konzistencie - McDonaldova Omega
round(omega(rel.HLQ.na.rm[,1:25], poly = TRUE)$omega.tot, 2) # pre jednofaktorovu (25 polozkovu) skalu## 295 cells were adjusted for 0 values using the correction for continuity. Examine your data carefully.
## [1] 0.94Odhad stability v čase
Výpočet sumárneho skóre HLQ
rel.HLQ.na.rm$HLQ_sum <- rowMeans(rel.HLQ.na.rm[,c("HL1", "HL2", "HL3", "HL4", "HL5", "HL6", "HL7", "HL8", "HL9",
"HL10", "HL11", "HL12", "HL13", "HL14", "HL15", "HL16", "HL17",
"HL18", "HL19", "HL20", "HL21", "HL22", "HL23", "HL24", "HL25")], na.rm = TRUE)Výpočet sumárného skóre retestu RHLQ
rel.HLQ.na.rm$RHLQ_sum <- rowMeans(rel.HLQ.na.rm[,c("RHL1", "RHL2", "RHL3", "RHL4", "RHL5", "RHL6", "RHL7", "RHL8", "RHL9",
"RHL10", "RHL11", "RHL12", "RHL13", "RHL14", "RHL15", "RHL16", "RHL17",
"RHL18", "RHL19", "RHL20", "RHL21", "RHL22", "RHL23", "RHL24", "RHL25")], na.rm = TRUE)Test-retest korelácia
round(with(rel.HLQ.na.rm, cor.test(HLQ_sum, RHLQ_sum))$estimate, 2) # pre jednofaktorovu (25 polozkovu) skalu## cor
## 0.76sessionInfo()## R version 3.3.2 (2016-10-31)
## Platform: x86_64-apple-darwin13.4.0 (64-bit)
## Running under: macOS 10.13.2
##
## locale:
## [1] sk_SK.UTF-8/sk_SK.UTF-8/sk_SK.UTF-8/C/sk_SK.UTF-8/sk_SK.UTF-8
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] BaylorEdPsych_0.5 Amelia_1.7.4 Rcpp_0.12.12
## [4] ICC_2.3.0 psych_1.7.3.21 dplyr_0.7.2
## [7] semPlot_1.1 lavaan_0.5-23.1097
##
## loaded via a namespace (and not attached):
## [1] nlme_3.1-131 pbkrtest_0.4-7 RColorBrewer_1.1-2
## [4] rprojroot_1.2 mi_1.0 tools_3.3.2
## [7] backports_1.0.5 R6_2.2.2 rpart_4.1-10
## [10] d3Network_0.5.2.1 Hmisc_4.0-3 lazyeval_0.2.0
## [13] mgcv_1.8-17 colorspace_1.3-2 nnet_7.3-12
## [16] gridExtra_2.2.1 mnormt_1.5-5 qgraph_1.4.3
## [19] fdrtool_1.2.15 quantreg_5.29 htmlTable_1.9
## [22] SparseM_1.76 network_1.13.0 scales_0.4.1
## [25] checkmate_1.8.2 quadprog_1.5-5 sem_3.1-9
## [28] stringr_1.2.0 digest_0.6.12 pbivnorm_0.6.0
## [31] foreign_0.8-67 minqa_1.2.4 rmarkdown_1.3
## [34] base64enc_0.1-3 jpeg_0.1-8 pkgconfig_2.0.1
## [37] htmltools_0.3.6 lme4_1.1-13 lisrelToR_0.1.4
## [40] htmlwidgets_0.8 rlang_0.1.1 huge_1.2.7
## [43] bindr_0.1 statnet.common_3.3.0 gtools_3.5.0
## [46] acepack_1.4.1 car_2.1-5 magrittr_1.5
## [49] OpenMx_2.7.16 Formula_1.2-2 Matrix_1.2-8
## [52] munsell_0.4.3 abind_1.4-5 rockchalk_1.8.101
## [55] stringi_1.1.2 whisker_0.3-2 yaml_2.1.14
## [58] MASS_7.3-45 plyr_1.8.4 matrixcalc_1.0-3
## [61] grid_3.3.2 parallel_3.3.2 lattice_0.20-34
## [64] splines_3.3.2 sna_2.4 knitr_1.17
## [67] igraph_1.1.2 boot_1.3-18 rjson_0.2.15
## [70] corpcor_1.6.9 reshape2_1.4.2 stats4_3.3.2
## [73] GPArotation_2014.11-1 glue_1.1.1 XML_3.98-1.9
## [76] evaluate_0.10.1 latticeExtra_0.6-28 data.table_1.10.4
## [79] png_0.1-7 nloptr_1.0.4 MatrixModels_0.4-1
## [82] gtable_0.2.0 assertthat_0.1 ggplot2_2.2.1
## [85] semTools_0.4-14 coda_0.19-1 survival_2.40-1
## [88] glasso_1.8 tibble_1.3.3 arm_1.9-3
## [91] ggm_2.3 ellipse_0.3-8 bindrcpp_0.2
## [94] cluster_2.0.5