MLQ from PN no Negative Question Compared with Youth Dataset with Negative Questions
Levi Brackman
March 18, 2016
- MLQ_1: I understand my life’s meaning.
- MLQ_4: My life has a clear sense of purpose.
- MLQ_5: I have a good sense of what makes my life meaningful.
- MLQ_6: I have discovered a satisfying life purpose.
- MLQ_11: I have definite purpose in my life.
data<-read.csv("/Users/levibrackman/Dropbox/Git/stats/Schools Study/Students self assessment data.csv")
require(dplyr)## Loading required package: dplyr##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionsectionone<-select(data, MLQ_1_1, MLQ_4_1, MLQ_5_1, MLQ_6_1, MLQ_11_1)
sectiononecomp<- sectionone[complete.cases(sectionone[,]),]Analysis of Purpsoe Navigator data with meaning purpose as two factors
## Loading required package: lavaan## This is lavaan 0.5-20## lavaan is BETA software! Please report any bugs.## Warning in lav_data_full(data = data, group = group, group.label =
## group.label, : lavaan WARNING: unordered factor(s) with more than 2 levels
## detected in data: MLQ_1_1 MLQ_5_1 MLQ_4_1 MLQ_6_1 MLQ_11_1## Loading required package: semPlot
## lavaan (0.5-20) converged normally after 54 iterations
##
## Number of observations 748
##
## Estimator ML
## Minimum Function Test Statistic 15.561
## Degrees of freedom 4
## P-value (Chi-square) 0.004
##
## Parameter Estimates:
##
## Information Expected
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err Z-value P(>|z|) Std.lv Std.all
## Purpose =~
## MLQ_1_1 1.000 3.653 0.945
## MLQ_5_1 1.151 0.022 52.256 0.000 4.204 0.936
## Meaning =~
## MLQ_4_1 1.000 4.084 0.963
## MLQ_6_1 0.998 0.017 60.080 0.000 4.078 0.946
## MLQ_11_1 0.805 0.017 46.078 0.000 3.288 0.891
##
## Covariances:
## Estimate Std.Err Z-value P(>|z|) Std.lv Std.all
## Purpose ~~
## Meaning 14.876 0.809 18.382 0.000 0.997 0.997
##
## Variances:
## Estimate Std.Err Z-value P(>|z|) Std.lv Std.all
## MLQ_1_1 1.612 0.118 13.684 0.000 1.612 0.108
## MLQ_5_1 2.493 0.170 14.698 0.000 2.493 0.124
## MLQ_4_1 1.319 0.105 12.528 0.000 1.319 0.073
## MLQ_6_1 1.950 0.131 14.858 0.000 1.950 0.105
## MLQ_11_1 2.807 0.161 17.384 0.000 2.807 0.206
## Purpose 13.346 0.773 17.255 0.000 1.000 1.000
## Meaning 16.679 0.932 17.904 0.000 1.000 1.000
##
## R-Square:
## Estimate
## MLQ_1_1 0.892
## MLQ_5_1 0.876
## MLQ_4_1 0.927
## MLQ_6_1 0.895
## MLQ_11_1 0.794## $type
## [1] "cor.bollen"
##
## $cor
## MLQ_1_ MLQ_5_ MLQ_4_ MLQ_6_ MLQ_11
## MLQ_1_1 0.000
## MLQ_5_1 0.000 0.000
## MLQ_4_1 -0.003 0.007 0.000
## MLQ_6_1 0.003 -0.009 0.000 0.000
## MLQ_11_1 0.003 -0.003 -0.006 0.009 0.000
##
## $mean
## MLQ_1_1 MLQ_5_1 MLQ_4_1 MLQ_6_1 MLQ_11_1
## 0 0 0 0 0## $type
## [1] "standardized"
##
## $cov
## MLQ_1_ MLQ_5_ MLQ_4_ MLQ_6_ MLQ_11
## MLQ_1_1 0.000
## MLQ_5_1 0.000 0.000
## MLQ_4_1 NA 1.570 0.000
## MLQ_6_1 0.816 NA 0.026 NA
## MLQ_11_1 0.551 -0.761 NA 1.461 NA
##
## $mean
## MLQ_1_1 MLQ_5_1 MLQ_4_1 MLQ_6_1 MLQ_11_1
## 0 0 0 0 0## Found more than one class "Model" in cache; using the first, from namespace 'lavaan'## lhs op rhs mi epc sepc.lv sepc.all sepc.nox
## 23 MLQ_5_1 ~~ MLQ_4_1 13.353 0.432 0.432 0.023 0.023
## 24 MLQ_5_1 ~~ MLQ_6_1 10.561 -0.397 -0.397 -0.021 -0.021
## 14 Purpose =~ MLQ_4_1 4.616 3.366 12.297 2.899 2.899
## 28 MLQ_6_1 ~~ MLQ_11_1 4.616 0.236 0.236 0.015 0.015
## 15 Purpose =~ MLQ_6_1 4.203 -3.021 -11.036 -2.560 -2.560
## 27 MLQ_4_1 ~~ MLQ_11_1 4.203 -0.213 -0.213 -0.014 -0.014Analysis of Purpsoe Navigator data with purpose as one factors
## Warning in lav_data_full(data = data, group = group, group.label =
## group.label, : lavaan WARNING: unordered factor(s) with more than 2 levels
## detected in data: MLQ_1_1 MLQ_5_1 MLQ_4_1 MLQ_6_1 MLQ_11_1
## lavaan (0.5-20) converged normally after 34 iterations
##
## Number of observations 748
##
## Estimator ML
## Minimum Function Test Statistic 16.300
## Degrees of freedom 5
## P-value (Chi-square) 0.006
##
## Parameter Estimates:
##
## Information Expected
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err Z-value P(>|z|) Std.lv Std.all
## Purpose =~
## MLQ_1_1 1.000 3.647 0.943
## MLQ_5_1 1.151 0.022 51.719 0.000 4.197 0.935
## MLQ_4_1 1.119 0.019 58.995 0.000 4.082 0.962
## MLQ_6_1 1.117 0.021 54.343 0.000 4.075 0.946
## MLQ_11_1 0.901 0.021 43.348 0.000 3.287 0.891
##
## Variances:
## Estimate Std.Err Z-value P(>|z|) Std.lv Std.all
## MLQ_1_1 1.654 0.108 15.282 0.000 1.654 0.111
## MLQ_5_1 2.545 0.160 15.863 0.000 2.545 0.126
## MLQ_4_1 1.335 0.102 13.022 0.000 1.335 0.074
## MLQ_6_1 1.968 0.131 15.074 0.000 1.968 0.106
## MLQ_11_1 2.813 0.161 17.438 0.000 2.813 0.207
## Purpose 13.304 0.772 17.242 0.000 1.000 1.000
##
## R-Square:
## Estimate
## MLQ_1_1 0.889
## MLQ_5_1 0.874
## MLQ_4_1 0.926
## MLQ_6_1 0.894
## MLQ_11_1 0.793## $type
## [1] "cor.bollen"
##
## $cor
## MLQ_1_ MLQ_5_ MLQ_4_ MLQ_6_ MLQ_11
## MLQ_1_1 0.000
## MLQ_5_1 0.003 0.000
## MLQ_4_1 -0.004 0.006 0.000
## MLQ_6_1 0.002 -0.010 0.001 0.000
## MLQ_11_1 0.002 -0.004 -0.006 0.010 0.000
##
## $mean
## MLQ_1_1 MLQ_5_1 MLQ_4_1 MLQ_6_1 MLQ_11_1
## 0 0 0 0 0## $type
## [1] "standardized"
##
## $cov
## MLQ_1_ MLQ_5_ MLQ_4_ MLQ_6_ MLQ_11
## MLQ_1_1 0.000
## MLQ_5_1 0.674 NA
## MLQ_4_1 NA 1.423 NA
## MLQ_6_1 0.639 NA 0.401 NA
## MLQ_11_1 0.384 -1.046 NA 1.524 NA
##
## $mean
## MLQ_1_1 MLQ_5_1 MLQ_4_1 MLQ_6_1 MLQ_11_1
## 0 0 0 0 0## lhs op rhs mi epc sepc.lv sepc.all sepc.nox
## 17 MLQ_5_1 ~~ MLQ_6_1 10.939 -0.381 -0.381 -0.020 -0.020
## 16 MLQ_5_1 ~~ MLQ_4_1 8.585 0.320 0.320 0.017 0.017
## 21 MLQ_6_1 ~~ MLQ_11_1 5.084 0.244 0.244 0.015 0.015Compare with other MLQ from large youth dataset with negatively worded item
data <- read.csv("~/Dropbox/Git/stats/allsurveysYT1_Jan2016.csv", header=T)
data<-tbl_df(data)
MLQ<-select(data, MLQ_1, MLQ_2, MLQ_3, MLQ_4, MLQ_5, MLQ_6,MLQ_7, MLQ_8, MLQ_9, MLQ_10)
MLQ$MLQ_9 <- 8- MLQ$MLQ_9
purposemodelMLQ= ' Purpose =~ MLQ_1 + MLQ_5 + MLQ_4 + MLQ_6 + MLQ_9'
purposefit1=cfa(purposemodelMLQ, data=MLQ)
require(semPlot)
semPaths(purposefit1, whatLabels = "std", layout = "tree")
summary(purposefit1, standardized = TRUE, rsquare=TRUE)## lavaan (0.5-20) converged normally after 25 iterations
##
## Used Total
## Number of observations 970 1288
##
## Estimator ML
## Minimum Function Test Statistic 33.938
## Degrees of freedom 5
## P-value (Chi-square) 0.000
##
## Parameter Estimates:
##
## Information Expected
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err Z-value P(>|z|) Std.lv Std.all
## Purpose =~
## MLQ_1 1.000 1.398 0.813
## MLQ_5 0.825 0.032 25.569 0.000 1.154 0.781
## MLQ_4 0.899 0.035 25.729 0.000 1.257 0.785
## MLQ_6 0.947 0.037 25.872 0.000 1.324 0.789
## MLQ_9 0.563 0.047 12.037 0.000 0.788 0.401
##
## Variances:
## Estimate Std.Err Z-value P(>|z|) Std.lv Std.all
## MLQ_1 1.000 0.064 15.566 0.000 1.000 0.338
## MLQ_5 0.851 0.050 16.882 0.000 0.851 0.390
## MLQ_4 0.983 0.059 16.729 0.000 0.983 0.383
## MLQ_6 1.062 0.064 16.586 0.000 1.062 0.377
## MLQ_9 3.246 0.152 21.416 0.000 3.246 0.840
## Purpose 1.955 0.134 14.567 0.000 1.000 1.000
##
## R-Square:
## Estimate
## MLQ_1 0.662
## MLQ_5 0.610
## MLQ_4 0.617
## MLQ_6 0.623
## MLQ_9 0.160correlMP = residuals(purposefit1, type="cor")
correlMP## $type
## [1] "cor.bollen"
##
## $cor
## MLQ_1 MLQ_5 MLQ_4 MLQ_6 MLQ_9
## MLQ_1 0.000
## MLQ_5 0.019 0.000
## MLQ_4 -0.021 -0.011 0.000
## MLQ_6 -0.004 -0.018 0.037 0.000
## MLQ_9 0.030 0.033 -0.007 -0.059 0.000
##
## $mean
## MLQ_1 MLQ_5 MLQ_4 MLQ_6 MLQ_9
## 0 0 0 0 0zcorrelsMP = residuals(purposefit1, type = "standardized")
zcorrelsMP## $type
## [1] "standardized"
##
## $cov
## MLQ_1 MLQ_5 MLQ_4 MLQ_6 MLQ_9
## MLQ_1 0.000
## MLQ_5 2.272 0.000
## MLQ_4 -5.229 -1.657 NA
## MLQ_6 -0.727 -3.039 3.585 0.000
## MLQ_9 2.014 2.005 -0.439 -4.177 NA
##
## $mean
## MLQ_1 MLQ_5 MLQ_4 MLQ_6 MLQ_9
## 0 0 0 0 0modindices(purposefit1, sort. = TRUE, minimum.value = 3.84)## lhs op rhs mi epc sepc.lv sepc.all sepc.nox
## 19 MLQ_4 ~~ MLQ_6 23.755 0.258 0.258 0.096 0.096
## 21 MLQ_6 ~~ MLQ_9 14.849 -0.271 -0.271 -0.082 -0.082
## 13 MLQ_1 ~~ MLQ_4 10.238 -0.178 -0.178 -0.065 -0.065
## 12 MLQ_1 ~~ MLQ_5 8.034 0.145 0.145 0.057 0.057
## 17 MLQ_5 ~~ MLQ_6 5.668 -0.116 -0.116 -0.047 -0.047
## 15 MLQ_1 ~~ MLQ_9 4.485 0.150 0.150 0.044 0.044
## 18 MLQ_5 ~~ MLQ_9 4.384 0.131 0.131 0.045 0.045options(scipen=999)
MLQpurpose_fit<-fitmeasures(purposefit1)[c('chisq', 'df', 'pvalue', 'cfi', 'rmsea', 'srmr', 'tli')]#Models purpose and meaningTable of Fit Measures
Colunm names explained:
- MLQ_With_9 is the large Youth data frame from the study which includes MLQ_9 a negatively worded item
- Meaning_Purpose_PN is the data from the purpose navigator with two factors
- Purpose_PN is the data from the purpose navigator with just one factor
It seems that when we have a positive question instead of the negatively worded one we get a much better fit for the data.
## Source: local data frame [7 x 4]
##
## Fit_Measure MLQ_With_9 Meaning_Purpose_PN Purpose_PN
## (fctr) (dbl) (dbl) (dbl)
## 1 Chisq 33.937974012844 15.560836718 16.300424242
## 2 DF 5.000000000000 4.000000000 5.000000000
## 3 P-Value 0.000002449668 0.003668787 0.006036693
## 4 CFI 0.985671972577 0.997790999 0.997840758
## 5 RMSEA 0.077243721945 0.062160406 0.054968207
## 6 SRMR 0.023296983085 0.004403999 0.004560752
## 7 TLI 0.971343945155 0.994477498 0.995681515