LavaanModellingOct15
Alex Brooks
10/15/2018
Let’s just look at trying to make some single factor lavaan models work
Start with 2016 only data
#read in the data for 2016 only
Data <-read.csv('../cleaned_data/2016_data.csv')
str(Data)## 'data.frame': 7474 obs. of 54 variables:
## $ postcode : int 800 800 800 810 810 810 812 812 812 820 ...
## $ state : Factor w/ 12 levels "ACT","NSW","NSW/ACT",..: 6 6 6 6 6 6 6 6 6 6 ...
## $ year : int 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 ...
## $ age : int 1 2 5 5 1 2 1 5 2 2 ...
## $ pc_immun : Factor w/ 9 levels "<70.0","70.0-74.9",..: 6 3 6 6 7 6 6 7 5 4 ...
## $ caution : int 1 1 1 0 0 0 0 0 0 0 ...
## $ pc_immun_class : int 6 3 6 6 7 6 6 7 5 4 ...
## $ PHN_code : Factor w/ 31 levels "PHN101","PHN102",..: 30 30 30 30 30 30 30 30 30 30 ...
## $ PHN_number : int 701 701 701 701 701 701 701 701 701 701 ...
## $ IEO_MAXS : int 1163 1163 1163 1181 1181 1181 1061 1061 1061 1214 ...
## $ IEO_MINS : int 1023 1023 1023 858 858 858 916 916 916 622 ...
## $ IEO_RWAD : int 9 9 9 8 8 8 6 6 6 9 ...
## $ IEO_RWAP : int 87 87 87 77 77 77 58 58 58 87 ...
## $ IEO_RWAR : int 2287 2287 2287 2001 2001 2001 1520 1520 1520 2263 ...
## $ IEO_RWSD : int 10 10 10 9 9 9 6 6 6 9 ...
## $ IEO_RWSP : int 92 92 92 84 84 84 56 56 56 89 ...
## $ IEO_RWSR : int 33 33 33 30 30 30 20 20 20 32 ...
## $ IEO_SCORE : int 1089 1089 1089 1045 1045 1045 997 997 997 1085 ...
## $ IEO_URP : int 6464 6464 6464 33302 33302 33302 18873 18873 18873 19447 ...
## $ IER_MAXS : int 1019 1019 1019 1162 1162 1162 1122 1122 1122 1159 ...
## $ IER_MINS : int 751 751 751 779 779 779 735 735 735 407 ...
## $ IER_RWAD : int 2 2 2 6 6 6 6 6 6 6 ...
## $ IER_RWAP : int 16 16 16 58 58 58 57 57 57 56 ...
## $ IER_RWAR : int 412 412 412 1510 1510 1510 1485 1485 1485 1453 ...
## $ IER_RWSD : int 6 6 6 8 8 8 7 7 7 7 ...
## $ IER_RWSP : int 51 51 51 76 76 76 70 70 70 67 ...
## $ IER_RWSR : int 18 18 18 27 27 27 25 25 25 24 ...
## $ IER_SCORE : int 946 946 946 1014 1014 1014 1013 1013 1013 1011 ...
## $ IER_URP : int 6464 6464 6464 33302 33302 33302 18873 18873 18873 19447 ...
## $ IRSAD_MAXS : int 1167 1167 1167 1168 1168 1168 1130 1130 1130 1211 ...
## $ IRSAD_MINS : int 914 914 914 868 868 868 782 782 782 523 ...
## $ IRSAD_RWAD : int 10 10 10 9 9 9 7 7 7 10 ...
## $ IRSAD_RWAP : int 92 92 92 81 81 81 69 69 69 91 ...
## $ IRSAD_RWAR : int 2398 2398 2398 2121 2121 2121 1800 1800 1800 2389 ...
## $ IRSAD_RWSD : int 10 10 10 8 8 8 7 7 7 9 ...
## $ IRSAD_RWSP : int 92 92 92 73 73 73 62 62 62 89 ...
## $ IRSAD_RWSR : int 33 33 33 26 26 26 22 22 22 32 ...
## $ IRSAD_SCORE : int 1096 1096 1096 1052 1052 1052 1020 1020 1020 1094 ...
## $ IRSAD_URP : int 6464 6464 6464 33302 33302 33302 18873 18873 18873 19447 ...
## $ IRSD_MAXS : int 1125 1125 1125 1146 1146 1146 1111 1111 1111 1167 ...
## $ IRSD_MINS : int 842 842 842 832 832 832 720 720 720 403 ...
## $ IRSD_RWAD : int 9 9 9 8 8 8 6 6 6 9 ...
## $ IRSD_RWAP : int 86 86 86 72 72 72 59 59 59 89 ...
## $ IRSD_RWAR : int 2243 2243 2243 1883 1883 1883 1535 1535 1535 2323 ...
## $ IRSD_RWSD : int 8 8 8 7 7 7 6 6 6 9 ...
## $ IRSD_RWSP : int 78 78 78 67 67 67 59 59 59 87 ...
## $ IRSD_RWSR : int 28 28 28 24 24 24 21 21 21 31 ...
## $ IRSD_SCORE : int 1066 1066 1066 1037 1037 1037 1013 1013 1013 1073 ...
## $ IRSD_URP : int 6464 6464 6464 33302 33302 33302 18873 18873 18873 19447 ...
## $ electorate : Factor w/ 150 levels "Adelaide","Aston",..: 136 136 136 136 136 136 136 136 136 136 ...
## $ total_tax : num 389375576 389375576 389375576 1367379789 1367379789 ...
## $ num_individuals: int 5464 5464 5464 21128 21128 21128 11509 11509 11509 13252 ...
## $ mean_tax_000s : num 71.3 71.3 71.3 64.7 64.7 ...
## $ political_score: num 4 4 4 4 4 4 4 4 4 4 ...#Filter and clean data to get only the variables we really want
geo_data <- Data %>%
select(postcode, state, year, age, pc_immun, caution, pc_immun_class, PHN_number, IEO_SCORE, IER_SCORE, IRSAD_SCORE, IRSD_SCORE, electorate, mean_tax_000s, political_score)
str(geo_data)## 'data.frame': 7474 obs. of 15 variables:
## $ postcode : int 800 800 800 810 810 810 812 812 812 820 ...
## $ state : Factor w/ 12 levels "ACT","NSW","NSW/ACT",..: 6 6 6 6 6 6 6 6 6 6 ...
## $ year : int 2016 2016 2016 2016 2016 2016 2016 2016 2016 2016 ...
## $ age : int 1 2 5 5 1 2 1 5 2 2 ...
## $ pc_immun : Factor w/ 9 levels "<70.0","70.0-74.9",..: 6 3 6 6 7 6 6 7 5 4 ...
## $ caution : int 1 1 1 0 0 0 0 0 0 0 ...
## $ pc_immun_class : int 6 3 6 6 7 6 6 7 5 4 ...
## $ PHN_number : int 701 701 701 701 701 701 701 701 701 701 ...
## $ IEO_SCORE : int 1089 1089 1089 1045 1045 1045 997 997 997 1085 ...
## $ IER_SCORE : int 946 946 946 1014 1014 1014 1013 1013 1013 1011 ...
## $ IRSAD_SCORE : int 1096 1096 1096 1052 1052 1052 1020 1020 1020 1094 ...
## $ IRSD_SCORE : int 1066 1066 1066 1037 1037 1037 1013 1013 1013 1073 ...
## $ electorate : Factor w/ 150 levels "Adelaide","Aston",..: 136 136 136 136 136 136 136 136 136 136 ...
## $ mean_tax_000s : num 71.3 71.3 71.3 64.7 64.7 ...
## $ political_score: num 4 4 4 4 4 4 4 4 4 4 ...Remove zeros from pc_immun_class
This seems an impossible task - ask Kirsty or Jarod tonight
#make electorate and state take a numerical class value in some way
#ASK TONIGHTFirst 2016 model - not very good
pc.model <- 'pc_immun_class =~ postcode + PHN_number + IRSAD_SCORE'
pc.fit <- cfa(model=pc.model,
data = geo_data)## Warning in lav_data_full(data = data, group = group, cluster = cluster, : lavaan WARNING: some observed variances are larger than 1000000
## lavaan NOTE: use varTable(fit) to investigatesummary(pc.fit, standardized = TRUE, fit.measures = TRUE)## lavaan 0.6-3 ended normally after 8 iterations
##
## Optimization method NLMINB
## Number of free parameters 6
##
## Used Total
## Number of observations 7471 7474
##
## Estimator ML
## Model Fit Test Statistic 6917.557
## Degrees of freedom 0
## Minimum Function Value 0.4629606128181
##
## Model test baseline model:
##
## Minimum Function Test Statistic 6917.558
## Degrees of freedom 3
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 0.000
## Tucker-Lewis Index (TLI) 1.000
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -156861.729
## Loglikelihood unrestricted model (H1) -153402.950
##
## Number of free parameters 6
## Akaike (AIC) 313735.458
## Bayesian (BIC) 313776.970
## Sample-size adjusted Bayesian (BIC) 313757.903
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.000
## 90 Percent Confidence Interval 0.000 0.000
## P-value RMSEA <= 0.05 NA
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.324
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv
## pc_immun_class =~
## postcode 1.000 0.224
## PHN_number 0.050 543802.767 0.000 1.000 0.011
## IRSAD_SCORE -0.006 58350.984 -0.000 1.000 -0.001
## Std.all
##
## 0.000
## 0.000
## -0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv
## .postcode 2247616.205 553050.810 4.064 0.000 2247616.205
## .PHN_number 26460.640 1421.910 18.609 0.000 26460.640
## .IRSAD_SCORE 5823.976 97.088 59.986 0.000 5823.976
## pc_immun_class 0.050 551826.810 0.000 1.000 1.000
## Std.all
## 1.000
## 1.000
## 1.000
## 1.000lavInspect(pc.fit, what="estimates")## $lambda
## pc_mm_
## postcode 1.000
## PHN_number 0.050
## IRSAD_SCORE -0.006
##
## $theta
## postcd PHN_nm IRSAD_
## postcode 2247616.205
## PHN_number 0.000 26460.640
## IRSAD_SCORE 0.000 0.000 5823.976
##
## $psi
## pc_mm_
## pc_immun_class 0.05semPaths(pc.fit, what = "stand", rotation = 4)
You can also embed plots, for example:
semCors(pc.fit)
Try the same single factor model as first model usng 2015 data
#read in the data for 2015 only
#try this to stop data confusion
char_data <- read.csv('../cleaned_data/2015_data.csv', stringsAsFactors=F)
num_data <- data.frame(data.matrix(char_data))## Warning in data.matrix(char_data): NAs introduced by coercion
## Warning in data.matrix(char_data): NAs introduced by coercion
## Warning in data.matrix(char_data): NAs introduced by coercion
## Warning in data.matrix(char_data): NAs introduced by coercionnumeric_columns <- sapply(num_data,function(x){mean(as.numeric(is.na(x)))<0.5})
final_data <- data.frame(num_data[,numeric_columns], char_data[,!numeric_columns])
str(final_data)## 'data.frame': 7260 obs. of 54 variables:
## $ postcode : num 800 800 800 810 810 810 812 812 812 820 ...
## $ year : num 2015 2015 2015 2015 2015 ...
## $ age : num 1 5 2 2 1 5 2 5 1 2 ...
## $ caution : num 1 1 1 0 0 0 0 0 0 0 ...
## $ pc_immun_class : num 5 5 3 6 6 6 6 6 7 6 ...
## $ PHN_number : num 701 701 701 701 701 701 701 701 701 701 ...
## $ IEO_MAXS : num 1167 1167 1167 1252 1252 ...
## $ IEO_MINS : num 1066 1066 1066 872 872 ...
## $ IEO_RWAD : num 9 9 9 8 8 8 7 7 7 9 ...
## $ IEO_RWAP : num 85 85 85 79 79 79 65 65 65 85 ...
## $ IEO_RWAR : num 2089 2089 2089 1958 1958 ...
## $ IEO_RWSD : num 10 10 10 10 10 10 8 8 8 10 ...
## $ IEO_RWSP : num NA NA NA NA NA NA NA NA NA NA ...
## $ IEO_RWSR : num 28 28 28 27 27 27 21 21 21 29 ...
## $ IEO_SCORE : num 1077 1077 1077 1051 1051 ...
## $ IEO_URP : num 0 0 0 0 0 0 0 0 0 3 ...
## $ IER_MAXS : num 986 986 986 1123 1123 ...
## $ IER_MINS : num 927 927 927 743 743 743 734 734 734 404 ...
## $ IER_RWAD : num 3 3 3 6 6 6 7 7 7 6 ...
## $ IER_RWAP : num 21 21 21 56 56 56 61 61 61 57 ...
## $ IER_RWAR : num 501 501 501 1388 1388 ...
## $ IER_RWSD : num 6 6 6 8 8 8 8 8 8 8 ...
## $ IER_RWSP : num NA NA NA NA NA NA NA NA NA NA ...
## $ IER_RWSR : num 15 15 15 21 21 21 23 23 23 22 ...
## $ IER_SCORE : num 952 952 952 1008 1008 ...
## $ IER_URP : num 0 0 0 0 0 0 0 0 0 3 ...
## $ IRSAD_MAXS : num 1144 1144 1144 1131 1131 ...
## $ IRSAD_MINS : num 1051 1051 1051 803 803 ...
## $ IRSAD_RWAD : num 9 9 9 8 8 8 7 7 7 9 ...
## $ IRSAD_RWAP : num 87 87 87 75 75 75 67 67 67 88 ...
## $ IRSAD_RWAR : num 2136 2136 2136 1840 1840 ...
## $ IRSAD_RWSD : num 9 9 9 8 8 8 8 8 8 10 ...
## $ IRSAD_RWSP : num NA NA NA NA NA NA NA NA NA NA ...
## $ IRSAD_RWSR : num 26 26 26 22 22 22 21 21 21 28 ...
## $ IRSAD_SCORE : num 1072 1072 1072 1037 1037 ...
## $ IRSAD_URP : num 4564 4564 4564 29725 29725 ...
## $ IRSD_MAXS : num 1127 1127 1127 1123 1123 ...
## $ IRSD_MINS : num 1032 1032 1032 768 768 ...
## $ IRSD_RWAD : num 9 9 9 7 7 7 6 6 6 9 ...
## $ IRSD_RWAP : num 83 83 83 67 67 67 60 60 60 86 ...
## $ IRSD_RWAR : num 2053 2053 2053 1656 1656 ...
## $ IRSD_RWSD : num 9 9 9 8 8 8 8 8 8 9 ...
## $ IRSD_RWSP : num NA NA NA NA NA NA NA NA NA NA ...
## $ IRSD_RWSR : num 25 25 25 22 22 22 21 21 21 26 ...
## $ IRSD_SCORE : num 1060 1060 1060 1027 1027 ...
## $ IRSD_URP : num 0 0 0 0 0 0 0 0 0 3 ...
## $ total_tax : num 389375576 389375576 389375576 1367379789 1367379789 ...
## $ num_individuals: num 5464 5464 5464 21128 21128 ...
## $ mean_tax_000s : num 71.3 71.3 71.3 64.7 64.7 ...
## $ political_score: num 6.5 6.5 6.5 6.5 6.5 6.5 6.5 6.5 6.5 6.5 ...
## $ state : chr "NT" "NT" "NT" "NT" ...
## $ pc_immun : chr "85.0-89.9" "85.0-89.9" "75.0-79.9" "90.0-92.4" ...
## $ PHN_code : chr "PHN701" "PHN701" "PHN701" "PHN701" ...
## $ electorate : chr "Solomon" "Solomon" "Solomon" "Solomon" ...#Our geo factor is measured by PHN, IRSAD score and pc immun class
geo_2015.model <- 'pc_immun_class =~ postcode + PHN_number + IRSAD_SCORE'
geo_2015.fit <- cfa(model = geo_2015.model,
data = final_data)## Warning in lav_data_full(data = data, group = group, cluster = cluster, : lavaan WARNING: some observed variances are larger than 1000000
## lavaan NOTE: use varTable(fit) to investigatesummary(geo_2015.fit, standardized = TRUE, fit.measures = TRUE)## lavaan 0.6-3 ended normally after 8 iterations
##
## Optimization method NLMINB
## Number of free parameters 6
##
## Used Total
## Number of observations 7257 7260
##
## Estimator ML
## Model Fit Test Statistic 6725.242
## Degrees of freedom 0
## Minimum Function Value 0.4633624320893
##
## Model test baseline model:
##
## Minimum Function Test Statistic 6725.242
## Degrees of freedom 3
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 0.000
## Tucker-Lewis Index (TLI) 1.000
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -152404.198
## Loglikelihood unrestricted model (H1) -149041.576
##
## Number of free parameters 6
## Akaike (AIC) 304820.395
## Bayesian (BIC) 304861.734
## Sample-size adjusted Bayesian (BIC) 304842.667
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.000
## 90 Percent Confidence Interval 0.000 0.000
## P-value RMSEA <= 0.05 NA
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.320
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv
## pc_immun_class =~
## postcode 1.000 0.224
## PHN_number 0.043 938907.382 0.000 1.000 0.010
## IRSAD_SCORE -0.003 68164.660 -0.000 1.000 -0.001
## Std.all
##
## 0.000
## 0.000
## -0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv
## .postcode 2236493.080 1090367.414 2.051 0.040 2236493.080
## .PHN_number 26388.243 2076.885 12.706 0.000 26388.243
## .IRSAD_SCORE 5926.914 99.164 59.769 0.000 5926.914
## pc_immun_class 0.050 1089735.103 0.000 1.000 1.000
## Std.all
## 1.000
## 1.000
## 1.000
## 1.000lavInspect(geo_2015.fit, what="estimates")## $lambda
## pc_mm_
## postcode 1.000
## PHN_number 0.043
## IRSAD_SCORE -0.003
##
## $theta
## postcd PHN_nm IRSAD_
## postcode 2236493.080
## PHN_number 0.000 26388.243
## IRSAD_SCORE 0.000 0.000 5926.914
##
## $psi
## pc_mm_
## pc_immun_class 0.05semPaths(geo_2015.fit, what = "stand", rotation = 4)
You can also embed plots, for example:
semCors(geo_2015.fit)
#modification indices (rmsea and srmr need to be below 0.10)
#look at loading and variances to check whether they are improbable or just don't relate to the latent variable, if loadings are over point 3 (variance will be large in comparison to the data)
#var(final_data)Not a great geo model
QUESTIONS: How can I correctly set the weight of these to become properly measured manifest variables?
Third model
#Try a single factor model on the sentiment of trust
trust.model <- 'score =~ IRSAD_SCORE + political_score + mean_tax_000s + postcode'
trust.fit <- cfa(model=trust.model,
data = geo_data)## Warning in lav_data_full(data = data, group = group, cluster = cluster, :
## lavaan WARNING: some observed variances are (at least) a factor 1000 times
## larger than others; use varTable(fit) to investigate## Warning in lav_data_full(data = data, group = group, cluster = cluster, : lavaan WARNING: some observed variances are larger than 1000000
## lavaan NOTE: use varTable(fit) to investigate## Warning in lavaan::lavaan(model = trust.model, data = geo_data, model.type
## = "cfa", : lavaan WARNING: the optimizer warns that a solution has NOT been
## found!summary(trust.fit, standardized = TRUE, fit.measures = TRUE)## lavaan 0.6-3 did NOT end normally after 10000 iterations
## ** WARNING ** Estimates below are most likely unreliable
##
## Optimization method NLMINB
## Number of free parameters 8
##
## Used Total
## Number of observations 7286 7474
##
## Estimator ML
## Model Fit Test Statistic NA
## Degrees of freedom NA
## P-value NA## Warning in .local(object, ...): lavaan WARNING: fit measures not available if model did not converge##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv
## score =~
## IRSAD_SCORE 1.000 1.906
## political_scor -0.003 NA -0.006
## mean_tax_000s 237.702 NA 453.073
## postcode 1.152 NA 2.195
## Std.all
##
## 0.025
## -0.004
## 26.874
## 0.001
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv
## .IRSAD_SCORE 5747.355 NA 5747.355
## .political_scor 2.715 NA 2.715
## .mean_tax_000s -204990.801 NA -204990.801
## .postcode 2227970.825 NA 2227970.825
## score 3.633 NA 1.000
## Std.all
## 0.999
## 1.000
## -721.214
## 1.000
## 1.000semCors(trust.fit)
semPaths(trust.fit, what = "stand", rotation = 4)
semCors(trust.fit, include="difference")
##This isn’t really working Let’s see if we can replicate the geo modelling at SA3 level ##Fourth model
#try this to stop data confusion
char_data <- read.csv('../raw_data/sa3_immune.csv', stringsAsFactors=F)
num_data <- data.frame(data.matrix(char_data))## Warning in data.matrix(char_data): NAs introduced by coercion
## Warning in data.matrix(char_data): NAs introduced by coercion
## Warning in data.matrix(char_data): NAs introduced by coercion
## Warning in data.matrix(char_data): NAs introduced by coercion
## Warning in data.matrix(char_data): NAs introduced by coercion
## Warning in data.matrix(char_data): NAs introduced by coercionnumeric_columns <- sapply(num_data,function(x){mean(as.numeric(is.na(x)))<0.5})
final_data <- data.frame(num_data[,numeric_columns], char_data[,!numeric_columns])
str(final_data)## 'data.frame': 5997 obs. of 10 variables:
## $ SA3_code : num 10101 10102 10103 10104 10201 ...
## $ year_start : num 2016 2016 2016 2016 2016 ...
## $ year_close : num 2017 2017 2017 2017 2017 ...
## $ age : num 1 1 1 1 1 1 1 1 1 1 ...
## $ num_registered : num 844 797 204 624 2014 ...
## $ num_fully_immunised : num 808 767 193 581 1904 ...
## $ num_not_fully_immunised: num 36 30 11 43 110 68 24 28 29 34 ...
## $ percent_fully_immunised: num 95.7 96.2 94.6 93.1 94.5 96.6 95.9 96.2 94.9 96.2 ...
## $ state : chr "NSW" "NSW" "NSW" "NSW" ...
## $ SA3_name : chr "Goulburn - Yass" "Queanbeyan" "Snowy Mountains" "South Coast" ...#Try a single factor model
#Our geo factor is measured by PHN, IRSAD score and pc immun class
sa3.model <- 'sa3 =~ SA3_code + year_start + age + num_not_fully_immunised + num_fully_immunised + percent_fully_immunised'
sa3.fit <- cfa(model=sa3.model,
data = final_data)## Warning in lav_data_full(data = data, group = group, cluster = cluster, :
## lavaan WARNING: some observed variances are (at least) a factor 1000 times
## larger than others; use varTable(fit) to investigate## Warning in lav_data_full(data = data, group = group, cluster = cluster, : lavaan WARNING: some observed variances are larger than 1000000
## lavaan NOTE: use varTable(fit) to investigate## Warning in lav_model_vcov(lavmodel = lavmodel, lavsamplestats = lavsamplestats, : lavaan WARNING:
## Could not compute standard errors! The information matrix could
## not be inverted. This may be a symptom that the model is not
## identified.summary(sa3.fit, standardized = TRUE, fit.measures = TRUE)## lavaan 0.6-3 ended normally after 8 iterations
##
## Optimization method NLMINB
## Number of free parameters 12
##
## Used Total
## Number of observations 5823 5997
##
## Estimator ML
## Model Fit Test Statistic 15133.237
## Degrees of freedom 9
## P-value (Chi-square) 0.000
##
## Model test baseline model:
##
## Minimum Function Test Statistic 15133.237
## Degrees of freedom 15
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 0.000
## Tucker-Lewis Index (TLI) -0.667
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -179463.826
## Loglikelihood unrestricted model (H1) -171897.207
##
## Number of free parameters 12
## Akaike (AIC) 358951.652
## Bayesian (BIC) 359031.687
## Sample-size adjusted Bayesian (BIC) 358993.554
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.537
## 90 Percent Confidence Interval 0.530 0.544
## P-value RMSEA <= 0.05 0.000
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.220
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv
## sa3 =~
## SA3_code 1.000 0.224
## year_start -0.000 NA -0.000
## age -0.000 NA -0.000
## nm_nt_flly_mmn -0.009 NA -0.002
## num_flly_mmnsd -0.115 NA -0.026
## prcnt_flly_mmn -0.000 NA -0.000
## Std.all
##
## 0.000
## -0.000
## -0.000
## -0.000
## -0.000
## -0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv
## .SA3_code 301155142.396 NA 301155142.396
## .year_start 2.914 NA 2.914
## .age 2.888 NA 2.888
## .nm_nt_flly_mmn 3580.560 NA 3580.560
## .num_flly_mmnsd 350482.667 NA 350482.667
## .prcnt_flly_mmn 7.456 NA 7.456
## sa3 0.050 NA 1.000
## Std.all
## 1.000
## 1.000
## 1.000
## 1.000
## 1.000
## 1.000
## 1.000#plot it
lavInspect(sa3.fit, what="estimates")## $lambda
## sa3
## SA3_code 1.000
## year_start 0.000
## age 0.000
## num_not_fully_immunised -0.009
## num_fully_immunised -0.115
## percent_fully_immunised 0.000
##
## $theta
## SA3_cd yr_str age
## SA3_code 301155142.396
## year_start 0.000 2.914
## age 0.000 0.000 2.888
## num_not_fully_immunised 0.000 0.000 0.000
## num_fully_immunised 0.000 0.000 0.000
## percent_fully_immunised 0.000 0.000 0.000
## nm_n__ nm_fl_ prcn__
## SA3_code
## year_start
## age
## num_not_fully_immunised 3580.560
## num_fully_immunised 0.000 350482.667
## percent_fully_immunised 0.000 0.000 7.456
##
## $psi
## sa3
## sa3 0.05semPaths(sa3.fit, what = "stand", rotation = 4)
semCors(sa3.fit, include="difference")