This dataset contains substance use behavior data from the National Longitudinal Survey of Youth 1997 (NLSY97) for three years: 1998, 2003, and 2008. The dataset focuses on youth born in 1984 and tracks three types of substance use behaviors: tobacco/cigarette smoking, alcohol drinking, and marijuana use.
Format
A data frame with 1004 rows and 38 columns:
SEXRespondent's sex
RACERespondent's race
ESMK_98,ESMK_03,ESMK_08(Ever smoked) Ever smoked in 1998, 2003, and 2008 (0: No, 1: Yes)
FSMK_98,FSMK_03,FSMK_08(Frequent smoke) Monthly smoking in 1998, 2003, and 2008 (0: No, 1: Yes)
DSMK_98,DSMK_03,DSMK_08(Daily smoke) Daily smoking in 1998, 2003, and 2008 (0: No, 1: Yes)
HSMK_98,HSMK_03,HSMK_08(Heavy smoke) 10+ cigarettes per day in 1998, 2003, and 2008 (0: No, 1: Yes)
EDRK_98,EDRK_03,EDRK_08(Ever drunk) Ever drunk in 1998, 2003, and 2008? (0: No, 1: Yes)
CDRK_98,CDRK_03,CDRK_08(Current drinker) Monthly drinking in 1998, 2003, and 2008 (0: No, 1: Yes)
WDRK_98,WDRK_03,WDRK_08(Weakly drinker) 5+ days drinking in a month in 1998, 2003, and 2008 (0: No, 1: Yes)
BDRK_98,BDRK_03,BDRK_08(Binge drinker) 5+ drinks on the same day at least one time in the last 30 day (0: No, 1: Yes)
EMRJ_98,EMRJ_03,EMRJ_08(Ever marijuana used) Have you ever used marijuana in 1998, 2003, and 2008? (0: No, 1: Yes)
CMRJ_98,CMRJ_03,CMRJ_08(Current marijuana user) Monthly marijuana use in 1998, 2003, and 2008 (0: No, 1: Yes)
OMRJ_98,OMRJ_03,OMRJ_08(Occasional marijuana user) 10+ days marijuana use in a month in 1998, 2003, and 2008 (0: No, 1: Yes)
SMRJ_98,SMRJ_03,SMRJ_08(School/work marijuana user) Marijuana use before/during school or work in 1998, 2003, and 2008 (0: No, 1: Yes)
References
Bureau of Labor Statistics, U.S. Department of Labor. National Longitudinal Survey of Youth 1997 cohort, 1997-2017 (rounds 1-18). Produced and distributed by the Center for Human Resource Research (CHRR), The Ohio State University. Columbus, OH: 2019.
Examples
library(magrittr)
nlsy_smoke <- slca(SMK_98(3) ~ ESMK_98 + FSMK_98 + DSMK_98 + HSMK_98) %>%
estimate(data = nlsy97, control = list(verbose = FALSE))
summary(nlsy_smoke)
#> Structural Latent Class Model
#>
#> Summary of model structure
#>
#> Number of manifest variables 4
#> Number of latent class variables 1
#>
#> Latent variables (Root*):
#> Label: SMK_98*
#> nclass: 3
#>
#> Measurement model:
#> SMK_98 -> { ESMK_98, FSMK_98, DSMK_98, HSMK_98 } a
#>
#>
#> Summary of manifest variables
#>
#> Categories for each variable:
#> response
#> 1 2
#> ESMK_98 Yes No
#> FSMK_98 Yes No
#> DSMK_98 Yes No
#> HSMK_98 Yes No
#>
#> Frequencies for each categories:
#> response
#> 1 2 <NA>
#> ESMK_98 558 446 0
#> FSMK_98 413 591 0
#> DSMK_98 179 825 0
#> HSMK_98 115 889 0
#>
#>
#> Summary of model fit
#>
#> Number of observations 1004
#> Number of free parameters 14
#> Log-likelihood -1473.365
#> Information criteria
#> Akaike (AIC) 2974.731
#> Bayesian (BIC) 3043.495
#> Chi-squared Tests
#> Residual degree of freedom (df) 1
#> Pearson Chi-squared (X-squared) 0.000
#> P(>Chi) 1.000
#> Likelihood Ratio (G-squared) 0.000
#> P(>Chi) 1.000
# A reproducible subset and two-class models keep these syntax examples small.
# Choose class counts and fit the full data separately for substantive inference.
set.seed(20261017)
example_data <- nlsy97[sample.int(nrow(nlsy97), 500), ]
# JLCA
model_jlca <- slca(
SMK_98(2) ~ ESMK_98 + FSMK_98 + DSMK_98 + HSMK_98,
DRK_98(2) ~ EDRK_98 + CDRK_98 + WDRK_98 + BDRK_98,
MRJ_98(2) ~ EMRJ_98 + CMRJ_98 + OMRJ_98 + SMRJ_98,
SUB_98(2) ~ SMK_98 + DRK_98 + MRJ_98
) %>% estimate(data = example_data, control = slcaControl(em.tol = 1e-6))
summary(model_jlca)
#> Structural Latent Class Model
#>
#> Summary of model structure
#>
#> Number of manifest variables 12
#> Number of latent class variables 4
#>
#> Latent variables (Root*):
#> Label: SMK_98 DRK_98 MRJ_98 SUB_98*
#> nclass: 2 2 2 2
#>
#> Measurement model:
#> SMK_98 -> { ESMK_98, FSMK_98, DSMK_98, HSMK_98 } a
#> DRK_98 -> { EDRK_98, CDRK_98, WDRK_98, BDRK_98 } b
#> MRJ_98 -> { EMRJ_98, CMRJ_98, OMRJ_98, SMRJ_98 } c
#>
#> Structural model:
#> SUB_98 -> { SMK_98, DRK_98, MRJ_98 }
#>
#> Dependency constraints:
#> A B C
#> SUB_98 -> SMK_98 SUB_98 -> DRK_98 SUB_98 -> MRJ_98
#>
#> Tree of structural model:
#> SUB_98 -> SMK_98
#> -> DRK_98
#> -> MRJ_98
#>
#>
#> Summary of manifest variables
#>
#> Categories for each variable:
#> response
#> 1 2
#> ESMK_98 Yes No
#> FSMK_98 Yes No
#> DSMK_98 Yes No
#> HSMK_98 Yes No
#> EDRK_98 Yes No
#> CDRK_98 Yes No
#> WDRK_98 Yes No
#> BDRK_98 Yes No
#> EMRJ_98 Yes No
#> CMRJ_98 Yes No
#> OMRJ_98 Yes No
#> SMRJ_98 Yes No
#>
#> Frequencies for each categories:
#> response
#> 1 2 <NA>
#> ESMK_98 280 220 0
#> FSMK_98 204 296 0
#> DSMK_98 93 407 0
#> HSMK_98 52 448 0
#> EDRK_98 374 126 0
#> CDRK_98 265 235 0
#> WDRK_98 110 390 0
#> BDRK_98 137 363 0
#> EMRJ_98 191 309 0
#> CMRJ_98 110 390 0
#> OMRJ_98 42 458 0
#> SMRJ_98 44 456 0
#>
#>
#> Summary of model fit
#>
#> Number of observations 500
#> Number of free parameters 31
#> Log-likelihood -2156.738
#> Information criteria
#> Akaike (AIC) 4375.477
#> Bayesian (BIC) 4506.130
#> Chi-squared Tests
#> Residual degree of freedom (df) 4064
#> Pearson Chi-squared (X-squared) 492.146
#> P(>Chi) 1.000
#> Likelihood Ratio (G-squared) 419.510
#> P(>Chi) 1.000
param(model_jlca)
#> PI :
#> (SUB_98)
#> class
#> 1 2
#> 0.5853 0.4147
#>
#> TAU :
#> (A)
#> parent
#> child 1 2
#> 1 0.1386 0.7883
#> 2 0.8614 0.2117
#>
#> parent SUB_98
#> child SMK_98
#> (B)
#> parent
#> child 1 2
#> 1 0.2782 0.8855
#> 2 0.7218 0.1145
#>
#> parent SUB_98
#> child DRK_98
#> (C)
#> parent
#> child 1 2
#> 1 0.0248 0.4956
#> 2 0.9752 0.5044
#>
#> parent SUB_98
#> child MRJ_98
#>
#> RHO :
#> (a)
#> class
#> response 1 2
#> 1(V1) 1.0000 0.2568
#> 2 0.0000 0.7432
#> 1(V2) 1.0000 0.0000
#> 2 0.0000 1.0000
#> 1(V3) 0.4559 0.0000
#> 2 0.5441 1.0000
#> 1(V4) 0.2549 0.0000
#> 2 0.7451 1.0000
#>
#> V1 V2 V3 V4
#> SMK_98 ESMK_98 FSMK_98 DSMK_98 HSMK_98
#> (b)
#> class
#> response 1 2
#> 1(V1) 1.0000 0.4638
#> 2 0.0000 0.5362
#> 1(V2) 1.0000 0.0000
#> 2 0.0000 1.0000
#> 1(V3) 0.4151 0.0000
#> 2 0.5849 1.0000
#> 1(V4) 0.5170 0.0000
#> 2 0.4830 1.0000
#>
#> V1 V2 V3 V4
#> DRK_98 EDRK_98 CDRK_98 WDRK_98 BDRK_98
#> (c)
#> class
#> response 1 2
#> 1(V1) 1.0000 0.2077
#> 2 0.0000 0.7923
#> 1(V2) 1.0000 0.0000
#> 2 0.0000 1.0000
#> 1(V3) 0.3818 0.0000
#> 2 0.6182 1.0000
#> 1(V4) 0.4000 0.0000
#> 2 0.6000 1.0000
#>
#> V1 V2 V3 V4
#> MRJ_98 EMRJ_98 CMRJ_98 OMRJ_98 SMRJ_98
# JLCPA
nlsy_jlcpa <- slca(
SMK_98(2) ~ ESMK_98 + FSMK_98 + DSMK_98 + HSMK_98,
DRK_98(2) ~ EDRK_98 + CDRK_98 + WDRK_98 + BDRK_98,
MRJ_98(2) ~ EMRJ_98 + CMRJ_98 + OMRJ_98 + SMRJ_98,
SUB_98(2) ~ SMK_98 + DRK_98 + MRJ_98,
SMK_03(2) ~ ESMK_03 + FSMK_03 + DSMK_03 + HSMK_03,
DRK_03(2) ~ EDRK_03 + CDRK_03 + WDRK_03 + BDRK_03,
MRJ_03(2) ~ EMRJ_03 + CMRJ_03 + OMRJ_03 + SMRJ_03,
SUB_03(2) ~ SMK_03 + DRK_03 + MRJ_03,
SMK_08(2) ~ ESMK_08 + FSMK_08 + DSMK_08 + HSMK_08,
DRK_08(2) ~ EDRK_08 + CDRK_08 + WDRK_08 + BDRK_08,
MRJ_08(2) ~ EMRJ_08 + CMRJ_08 + OMRJ_08 + SMRJ_08,
SUB_08(2) ~ SMK_08 + DRK_08 + MRJ_08,
PROF(2) ~ SUB_98 + SUB_03 + SUB_08,
constraints = list(
c("SMK_98", "SMK_03", "SMK_08"),
c("DRK_98", "DRK_03", "DRK_08"),
c("MRJ_98", "MRJ_03", "MRJ_08"),
c("SUB_98 ~ SMK_98", "SUB_03 ~ SMK_03", "SUB_08 ~ SMK_08"),
c("SUB_98 ~ DRK_98", "SUB_03 ~ DRK_03", "SUB_08 ~ DRK_08"),
c("SUB_98 ~ MRJ_98", "SUB_03 ~ MRJ_03", "SUB_08 ~ MRJ_08")
)
) %>% estimate(example_data, control = slcaControl(em.tol = 1e-6))
nlsy_jlcpa$convergence
#> EM nlm
#> TRUE NA