DESCRIPTIVE PLOTS
# DESCRIPTIVES ====
#random sample for plots
sample = sample(data$ID, size = 100)
sample <- data %>% filter(ID %in% c(sample))
sample <- reshape(sample, direction = "long",
varying = list(c("age_1", "age_2", "age_3"),
c("AC_1", "AC_2", "AC_3"),
c("AG_1", "AG_2", "AG_3"),
c("CON_1", "CON_2", "CON_3"),
c("HA_1", "HA_2", "HA_3"),
c("SP_1", "SP_2", "SP_3"),
c("TR_1", "TR_2", "TR_3")),
timevar = "time",
times = c(0,1,2),
v.names = c("age","AC","AG","CON","HA","SP","TR"),
idvar = c("ID"))
row.names(sample) <- 1:nrow(sample)
# > Plot of change over timepoints ----
pAC <- ggplot(data = sample,
aes(x = time, y = AC, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$AC,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
pAG <- ggplot(data = sample,
aes(x = time, y = AG, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$AG,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
pCON <- ggplot(data = sample,
aes(x = time, y = CON, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$CON,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
pHA <- ggplot(data = sample,
aes(x = time, y = HA, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$HA,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
pSP <- ggplot(data = sample,
aes(x = time, y = SP, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$SP,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
pTR <- ggplot(data = sample,
aes(x = time, y = TR, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$TR,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
cowplot::plot_grid(pAC, pAG, pCON, pHA, pSP, pTR,
nrow = 3, ncol = 2)
# > Plot of change with age ----
paAC <- ggplot(data = sample,
aes(x = age, y = AC, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$AC,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
paAG <- ggplot(data = sample,
aes(x = age, y = AG, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$AG,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
paCON <- ggplot(data = sample,
aes(x = age, y = CON, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$CON,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
paHA <- ggplot(data = sample,
aes(x = age, y = HA, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$HA,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
paSP <- ggplot(data = sample,
aes(x = age, y = SP, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$SP,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
paTR <- ggplot(data = sample,
aes(x = age, y = TR, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$TR,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
cowplot::plot_grid(paAC, paAG, paCON, paHA, paSP, paTR,
nrow = 3, ncol = 2)
rm(sample, paAC, paAG, paCON, paHA, paSP, paTR, pAC, pAG, pCON, pHA, pSP, pTR)
MISSINGNESS
# MISSINGNESS ====
## patterns for response variables
data %>% dplyr::select(AC_1, AC_2, AC_3) %>% md.pattern()
## AC_2 AC_3 AC_1
## 132 1 1 1 0
## 352 1 1 0 1
## 66 1 0 1 1
## 217 1 0 0 2
## 47 0 1 1 1
## 159 0 1 0 2
## 49 0 0 1 2
## 52 0 0 0 3
## 307 384 780 1471
data %>% dplyr::select(AG_1, AG_2, AG_3) %>% md.pattern()
## AG_2 AG_3 AG_1
## 135 1 1 1 0
## 351 1 1 0 1
## 61 1 0 1 1
## 225 1 0 0 2
## 47 0 1 1 1
## 157 0 1 0 2
## 49 0 0 1 2
## 49 0 0 0 3
## 302 384 782 1468
data %>% dplyr::select(CON_1, CON_2, CON_3) %>% md.pattern()
## CON_2 CON_3 CON_1
## 133 1 1 1 0
## 353 1 1 0 1
## 65 1 0 1 1
## 223 1 0 0 2
## 47 0 1 1 1
## 157 0 1 0 2
## 48 0 0 1 2
## 48 0 0 0 3
## 300 384 781 1465
data %>% dplyr::select(HA_1, HA_2, HA_3) %>% md.pattern()
## HA_2 HA_3 HA_1
## 135 1 1 1 0
## 348 1 1 0 1
## 66 1 0 1 1
## 222 1 0 0 2
## 46 0 1 1 1
## 160 0 1 0 2
## 48 0 0 1 2
## 49 0 0 0 3
## 303 385 779 1467
data %>% dplyr::select(SP_1, SP_2, SP_3) %>% md.pattern()
## SP_2 SP_3 SP_1
## 127 1 1 1 0
## 356 1 1 0 1
## 66 1 0 1 1
## 219 1 0 0 2
## 47 0 1 1 1
## 158 0 1 0 2
## 50 0 0 1 2
## 51 0 0 0 3
## 306 386 784 1476
data %>% dplyr::select(TR_1, TR_2, TR_3) %>% md.pattern()
## TR_2 TR_3 TR_1
## 120 1 1 1 0
## 345 1 1 0 1
## 53 1 0 1 1
## 216 1 0 0 2
## 45 0 1 1 1
## 169 0 1 0 2
## 53 0 0 1 2
## 73 0 0 0 3
## 340 395 803 1538
## compare missing and non-missing
explanatory = c("IDAB", "IDSEX", "age")
long %>%
missing_compare("AC", explanatory)
long %>%
missing_compare("AG", explanatory)
long %>%
missing_compare("CON", explanatory)
long %>%
missing_compare("HA", explanatory)
long %>%
missing_compare("SP", explanatory)
long %>%
missing_compare("TR", explanatory)
ANALYSES
1. Linear change patterns
# ANALYSIS ====
# > Distinguishable dyads ----
# >> Mean-level change ----
## Achievement
mACd <- lme(fixed = AC ~ -1 + young + young:age + old + old:age,
random = ~ -1 + young + old| IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mACd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2436.448 2491.128 -1208.224
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3082724 young
## old 0.3424229 -0.076
## Residual 0.3795695
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1387964
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.00000 1.06258
## Fixed effects: AC ~ -1 + young + young:age + old + old:age
## Value Std.Error DF t-value p-value
## young 2.3276309 0.05869231 1214 39.65819 0
## old 2.2817953 0.04541684 1214 50.24117 0
## young:age 0.0202866 0.00229157 1214 8.85269 0
## age:old 0.0201256 0.00177022 1214 11.36898 0
## Correlation:
## young old yong:g
## old 0.034
## young:age -0.936 -0.045
## age:old -0.020 -0.901 0.036
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.878617291 -0.548692802 -0.002274224 0.548885126 2.461665966
##
## Number of Observations: 1751
## Number of Groups: 534
## Aggression
mAGd <- lme(fixed = AG ~ -1 + young + young:age + old + old:age,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mAGd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2617.876 2672.572 -1298.938
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.4066366 young
## old 0.4141205 0.19
## Residual 0.3961539
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1570479
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.0000000 0.8878833
## Fixed effects: AG ~ -1 + young + young:age + old + old:age
## Value Std.Error DF t-value p-value
## young 2.4453938 0.05415273 1218 45.15735 0
## old 2.4577534 0.04829683 1218 50.88850 0
## young:age -0.0238512 0.00205628 1218 -11.59922 0
## age:old -0.0212711 0.00185033 1218 -11.49583 0
## Correlation:
## young old yong:g
## old 0.067
## young:age -0.908 -0.051
## age:old -0.022 -0.884 0.041
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.3280654 -0.5357146 -0.0746261 0.4416845 3.2080815
##
## Number of Observations: 1754
## Number of Groups: 533
## Control
mCONd <- lme(fixed = CON ~ -1 + young + young:age + old + old:age,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mCONd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2017.453 2072.167 -998.7264
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.2935685 young
## old 0.3127515 0.004
## Residual 0.3309577
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.08152349
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.028143
## Fixed effects: CON ~ -1 + young + young:age + old + old:age
## Value Std.Error DF t-value p-value
## young 2.3325172 0.05059566 1221 46.10113 0
## old 2.2669570 0.04007293 1221 56.57078 0
## young:age 0.0198194 0.00197139 1221 10.05356 0
## age:old 0.0211922 0.00156048 1221 13.58056 0
## Correlation:
## young old yong:g
## old 0.023
## young:age -0.931 -0.026
## age:old -0.012 -0.898 0.021
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -3.71577264 -0.50588692 0.01334313 0.55963014 2.63398308
##
## Number of Observations: 1757
## Number of Groups: 533
## Harm Avoidance
mHAd <- lme(fixed = HA ~ -1 + young + young:age + old + old:age,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mHAd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2932.339 2987.041 -1456.17
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3845670 young
## old 0.4909243 0.236
## Residual 0.3957237
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.0515229
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.090358
## Fixed effects: HA ~ -1 + young + young:age + old + old:age
## Value Std.Error DF t-value p-value
## young 2.2111913 0.06482722 1218 34.10899 0
## old 2.2769154 0.05042065 1218 45.15839 0
## young:age 0.0262789 0.00252441 1218 10.40990 0
## age:old 0.0216881 0.00189909 1218 11.42026 0
## Correlation:
## young old yong:g
## old 0.040
## young:age -0.930 -0.017
## age:old -0.008 -0.866 0.014
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.51064675 -0.51647705 0.02990568 0.55753397 2.42997097
##
## Number of Observations: 1755
## Number of Groups: 534
## Social Potency
mSPd <- lme(fixed = SP ~ -1 + young + young:age + old + old:age,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mSPd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2136.781 2191.432 -1058.391
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3155051 young
## old 0.4005251 0.026
## Residual 0.3200959
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.08469
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.024266
## Fixed effects: SP ~ -1 + young + young:age + old + old:age
## Value Std.Error DF t-value p-value
## young 2.8902776 0.04987075 1210 57.95537 0
## old 2.8578113 0.04089201 1210 69.88678 0
## young:age -0.0135269 0.00192820 1210 -7.01527 0
## age:old -0.0095613 0.00153728 1210 -6.21959 0
## Correlation:
## young old yong:g
## old 0.026
## young:age -0.925 -0.027
## age:old -0.013 -0.864 0.023
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.75931141 -0.47887263 0.02941493 0.49507079 2.90521243
##
## Number of Observations: 1746
## Number of Groups: 533
## Traditionalism
mTRd <- lme(fixed = TR ~ -1 + young + young:age + old + old:age,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mTRd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 1809.403 1863.693 -894.7017
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.2739266 young
## old 0.3024467 0.454
## Residual 0.3447630
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1523673
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.0000000 0.9270239
## Fixed effects: TR ~ -1 + young + young:age + old + old:age
## Value Std.Error DF t-value p-value
## young 2.8409146 0.04742318 1146 59.90562 0
## old 2.8077394 0.04182368 1146 67.13277 0
## young:age -0.0103666 0.00183418 1146 -5.65188 0
## age:old -0.0083830 0.00162155 1146 -5.16978 0
## Correlation:
## young old yong:g
## old 0.079
## young:age -0.929 -0.050
## age:old -0.020 -0.904 0.039
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -3.2997751 -0.5069002 0.0131458 0.4885317 3.2893505
##
## Number of Observations: 1684
## Number of Groups: 535
2. Quadratic change patterns
# >> Quad change ----
## Achievement
mpACd <- lme(fixed = AC ~ -1 + young + young:poly(age,2) + old + old:poly(age,2),
random = ~ -1 + young + old| IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mpACd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2429.76 2495.375 -1202.88
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3121907 young
## old 0.3429936 -0.069
## Residual 0.3776353
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1306086
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.054796
## Fixed effects: AC ~ -1 + young + young:poly(age, 2) + old + old:poly(age, 2)
## Value Std.Error DF t-value p-value
## young 2.799580 0.0209869 1212 133.39657 0.0000
## old 2.761010 0.0198052 1212 139.40814 0.0000
## young:poly(age, 2)1 5.073745 0.7289206 1212 6.96063 0.0000
## young:poly(age, 2)2 -2.030551 0.7920887 1212 -2.56354 0.0105
## poly(age, 2)1:old 6.158764 0.5329056 1212 11.55695 0.0000
## poly(age, 2)2:old -1.147722 0.5438553 1212 -2.11034 0.0350
## Correlation:
## young old y:(,2)1 y:(,2)2 p(,2)1
## old 0.031
## young:poly(age, 2)1 0.033 -0.026
## young:poly(age, 2)2 0.155 -0.002 0.420
## poly(age, 2)1:old 0.037 0.069 0.018 -0.012
## poly(age, 2)2:old -0.014 -0.085 0.036 0.027 -0.268
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.89559388 -0.55790965 -0.01570711 0.54187493 2.53836000
##
## Number of Observations: 1751
## Number of Groups: 534
anova(mACd, mpACd)
## Aggression
mpAGd <- lme(fixed = AG ~ -1 + young + young:poly(age,2) + old + old:poly(age,2),
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mpAGd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2588.275 2653.911 -1282.137
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.4112961 young
## old 0.4147771 0.205
## Residual 0.3901300
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1451594
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.0000000 0.8751877
## Fixed effects: AG ~ -1 + young + young:poly(age, 2) + old + old:poly(age, 2)
## Value Std.Error DF t-value p-value
## young 1.893169 0.0228966 1216 82.68357 0
## old 1.948192 0.0225861 1216 86.25633 0
## young:poly(age, 2)1 -5.748114 0.6436813 1216 -8.93006 0
## young:poly(age, 2)2 2.948985 0.7245185 1216 4.07027 0
## poly(age, 2)1:old -6.832697 0.5527507 1216 -12.36126 0
## poly(age, 2)2:old 2.425107 0.5611317 1216 4.32181 0
## Correlation:
## young old y:(,2)1 y:(,2)2 p(,2)1
## old 0.181
## young:poly(age, 2)1 0.035 -0.026
## young:poly(age, 2)2 0.140 0.000 0.415
## poly(age, 2)1:old 0.033 0.064 0.020 -0.014
## poly(age, 2)2:old -0.012 -0.074 0.044 0.034 -0.269
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.39460769 -0.51895415 -0.07387035 0.45216285 3.10491171
##
## Number of Observations: 1754
## Number of Groups: 533
anova(mAGd, mpAGd)
## Control
mpCONd <- lme(fixed = CON ~ -1 + young + young:poly(age,2) + old + old:poly(age,2),
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mpCONd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2008.868 2074.525 -992.4341
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.2965023 young
## old 0.3125449 0.015
## Residual 0.3309932
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.08310539
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.010886
## Fixed effects: CON ~ -1 + young + young:poly(age, 2) + old + old:poly(age, 2)
## Value Std.Error DF t-value p-value
## young 2.791069 0.0186189 1219 149.90478 0.0000
## old 2.768615 0.0177022 1219 156.39975 0.0000
## young:poly(age, 2)1 4.837273 0.6246079 1219 7.74450 0.0000
## young:poly(age, 2)2 -2.383480 0.6838614 1219 -3.48533 0.0005
## poly(age, 2)1:old 6.286252 0.4734497 1219 13.27755 0.0000
## poly(age, 2)2:old -0.396581 0.4818648 1219 -0.82301 0.4107
## Correlation:
## young old y:(,2)1 y:(,2)2 p(,2)1
## old 0.046
## young:poly(age, 2)1 0.036 -0.016
## young:poly(age, 2)2 0.149 -0.001 0.423
## poly(age, 2)1:old 0.023 0.069 0.011 -0.008
## poly(age, 2)2:old -0.009 -0.078 0.024 0.019 -0.279
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -3.721878321 -0.492422091 0.002128635 0.558578312 2.790272356
##
## Number of Observations: 1757
## Number of Groups: 533
anova(mCONd, mpCONd)
## Harm Avoidance
mpHAd <- lme(fixed = HA ~ -1 + young + young:poly(age,2) + old + old:poly(age,2),
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mpHAd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2933.898 2999.541 -1454.949
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3872230 young
## old 0.4912781 0.235
## Residual 0.3952571
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.04932726
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.085256
## Fixed effects: HA ~ -1 + young + young:poly(age, 2) + old + old:poly(age, 2)
## Value Std.Error DF t-value p-value
## young 2.836869 0.0241031 1216 117.69722 0.0000
## old 2.790452 0.0253431 1216 110.10698 0.0000
## young:poly(age, 2)1 8.099389 0.8073394 1216 10.03220 0.0000
## young:poly(age, 2)2 1.191161 0.8820982 1216 1.35037 0.1771
## poly(age, 2)1:old 6.435612 0.5767664 1216 11.15809 0.0000
## poly(age, 2)2:old -0.454092 0.5908759 1216 -0.76851 0.4423
## Correlation:
## young old y:(,2)1 y:(,2)2 p(,2)1
## old 0.158
## young:poly(age, 2)1 0.037 -0.009
## young:poly(age, 2)2 0.150 0.001 0.423
## poly(age, 2)1:old 0.016 0.065 0.006 -0.005
## poly(age, 2)2:old -0.005 -0.067 0.019 0.017 -0.288
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.48969460 -0.51135178 0.03285518 0.55725126 2.44945347
##
## Number of Observations: 1755
## Number of Groups: 534
anova(mHAd, mpHAd)
## Social Potency
mpSPd <- lme(fixed = SP ~ -1 + young + young:poly(age,2) + old + old:poly(age,2),
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mpSPd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2134.065 2199.646 -1055.032
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3167144 young
## old 0.3998405 0.029
## Residual 0.3185928
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.08167836
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.026258
## Fixed effects: SP ~ -1 + young + young:poly(age, 2) + old + old:poly(age, 2)
## Value Std.Error DF t-value p-value
## young 2.569510 0.0192103 1208 133.75660 0.0000
## old 2.635566 0.0206057 1208 127.90496 0.0000
## young:poly(age, 2)1 -4.109017 0.6193898 1208 -6.63398 0.0000
## young:poly(age, 2)2 -0.388261 0.6844700 1208 -0.56724 0.5707
## poly(age, 2)1:old -2.442685 0.4639074 1208 -5.26546 0.0000
## poly(age, 2)2:old -1.210384 0.4764997 1208 -2.54016 0.0112
## Correlation:
## young old y:(,2)1 y:(,2)2 p(,2)1
## old 0.048
## young:poly(age, 2)1 0.038 -0.013
## young:poly(age, 2)2 0.149 -0.001 0.429
## poly(age, 2)1:old 0.021 0.066 0.011 -0.008
## poly(age, 2)2:old -0.008 -0.071 0.024 0.019 -0.283
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.76506564 -0.48675424 0.02936957 0.50077625 2.86544205
##
## Number of Observations: 1746
## Number of Groups: 533
anova(mSPd, mpSPd)
## Traditionalism
mpTRd <- lme(fixed = TR ~ -1 + young + young:poly(age,2) + old + old:poly(age,2),
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mpTRd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 1812.759 1877.906 -894.3795
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.2731863 young
## old 0.3020960 0.448
## Residual 0.3450778
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1552855
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.0000000 0.9266525
## Fixed effects: TR ~ -1 + young + young:poly(age, 2) + old + old:poly(age, 2)
## Value Std.Error DF t-value p-value
## young 2.5968905 0.0177304 1144 146.46518 0.0000
## old 2.6090080 0.0179969 1144 144.96977 0.0000
## young:poly(age, 2)1 -2.7847412 0.5785067 1144 -4.81367 0.0000
## young:poly(age, 2)2 0.4866105 0.6298540 1144 0.77258 0.4399
## poly(age, 2)1:old -2.3764550 0.4826046 1144 -4.92423 0.0000
## poly(age, 2)2:old -0.0913254 0.4888883 1144 -0.18680 0.8518
## Correlation:
## young old y:(,2)1 y:(,2)2 p(,2)1
## old 0.297
## young:poly(age, 2)1 0.037 -0.032
## young:poly(age, 2)2 0.152 0.000 0.418
## poly(age, 2)1:old 0.046 0.062 0.022 -0.012
## poly(age, 2)2:old -0.013 -0.089 0.050 0.049 -0.269
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -3.283382781 -0.513444834 0.006709692 0.488324080 3.267275350
##
## Number of Observations: 1684
## Number of Groups: 535
anova(mTRd, mpTRd)
3. Gender as moderator
# >> Gender as moderators ----
#male = 0, female = 1
long <- long %>% mutate(
IDSEX = ifelse(IDSEX == 1, 0, 1))
## Achievement
mgACd <- lme(fixed = AC ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mgACd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2433.174 2498.789 -1204.587
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3094524 young
## old 0.3396406 -0.082
## Residual 0.3796024
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1361955
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.057253
## Fixed effects: AC ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX
## Value Std.Error DF t-value p-value
## young 2.3203930 0.05867007 1212 39.54986 0.0000
## old 2.2902508 0.04560664 1212 50.21748 0.0000
## young:age 0.0224357 0.00258605 1212 8.67567 0.0000
## age:old 0.0179794 0.00206450 1212 8.70883 0.0000
## young:age:IDSEX -0.0029714 0.00168110 1212 -1.76753 0.0774
## age:old:IDSEX 0.0031680 0.00156042 1212 2.03023 0.0426
## Correlation:
## young old yong:g age:ld y::IDS
## old 0.032
## young:age -0.856 -0.039
## age:old -0.015 -0.818 0.027
## young:age:IDSEX 0.067 0.001 -0.468 -0.003
## age:old:IDSEX -0.002 0.094 -0.001 -0.514 0.007
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.882196992 -0.562386413 -0.008149212 0.550027059 2.458071124
##
## Number of Observations: 1751
## Number of Groups: 534
anova(mACd, mgACd)
## Aggression
mgAGd <- lme(fixed = AG ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mgAGd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2545.44 2611.076 -1260.72
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3677290 young
## old 0.3705066 0.18
## Residual 0.4040065
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1680723
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.0000000 0.9038783
## Fixed effects: AG ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX
## Value Std.Error DF t-value p-value
## young 2.4369854 0.05482332 1216 44.45162 0
## old 2.4315246 0.04839687 1216 50.24136 0
## young:age -0.0173274 0.00242022 1216 -7.15940 0
## age:old -0.0134539 0.00218266 1216 -6.16396 0
## young:age:IDSEX -0.0098510 0.00171292 1216 -5.75100 0
## age:old:IDSEX -0.0120394 0.00165776 1216 -7.26245 0
## Correlation:
## young old yong:g age:ld y::IDS
## old 0.066
## young:age -0.824 -0.048
## age:old -0.019 -0.813 0.038
## young:age:IDSEX 0.056 0.001 -0.495 -0.014
## age:old:IDSEX -0.003 0.093 -0.013 -0.516 0.033
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.24414925 -0.52867121 -0.08375458 0.46239279 3.24808954
##
## Number of Observations: 1754
## Number of Groups: 533
anova(mAGd, mgAGd)
## Control
mgCONd <- lme(fixed = CON ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mgCONd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 1993.489 2059.145 -984.7445
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.2840547 young
## old 0.3026628 0
## Residual 0.3316112
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.08177836
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.032462
## Fixed effects: CON ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX
## Value Std.Error DF t-value p-value
## young 2.3414594 0.05074763 1219 46.13928 0.0000
## old 2.2822596 0.04013875 1219 56.85927 0.0000
## young:age 0.0164638 0.00224194 1219 7.34355 0.0000
## age:old 0.0172461 0.00181921 1219 9.48004 0.0000
## young:age:IDSEX 0.0047699 0.00147827 1219 3.22670 0.0013
## age:old:IDSEX 0.0058750 0.00137537 1219 4.27154 0.0000
## Correlation:
## young old yong:g age:ld y::IDS
## old 0.023
## young:age -0.851 -0.024
## age:old -0.009 -0.818 0.017
## young:age:IDSEX 0.063 0.001 -0.471 -0.004
## age:old:IDSEX -0.001 0.093 -0.002 -0.513 0.008
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -3.72295264 -0.50998111 0.01620036 0.54920625 2.56960557
##
## Number of Observations: 1757
## Number of Groups: 533
anova(mCONd, mgCONd)
## Harm Avoidance
mgHAd <- lme(fixed = HA ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mgHAd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2848.246 2913.889 -1412.123
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3666552 young
## old 0.4611298 0.271
## Residual 0.3928596
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.04188171
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.089754
## Fixed effects: HA ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX
## Value Std.Error DF t-value p-value
## young 2.2326849 0.06413936 1216 34.80990 0
## old 2.3120205 0.04976422 1216 46.45949 0
## young:age 0.0183513 0.00282369 1216 6.49906 0
## age:old 0.0124002 0.00225294 1216 5.50400 0
## young:age:IDSEX 0.0112708 0.00185331 1216 6.08144 0
## age:old:IDSEX 0.0139600 0.00184318 1216 7.57383 0
## Correlation:
## young old yong:g age:ld y::IDS
## old 0.039
## young:age -0.851 -0.013
## age:old -0.005 -0.780 0.015
## young:age:IDSEX 0.059 0.000 -0.465 -0.015
## age:old:IDSEX 0.000 0.097 -0.013 -0.550 0.032
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.63968997 -0.52424576 0.04285149 0.54519800 2.42287784
##
## Number of Observations: 1755
## Number of Groups: 534
anova(mHAd, mgHAd)
## Social Potency
mgSPd <- lme(fixed = SP ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mgSPd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2139.061 2204.642 -1057.53
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3151553 young
## old 0.4010750 0.029
## Residual 0.3196475
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.08860991
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.00000 1.02675
## Fixed effects: SP ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX
## Value Std.Error DF t-value p-value
## young 2.8897802 0.04997101 1208 57.82913 0.0000
## old 2.8526775 0.04105981 1208 69.47614 0.0000
## young:age -0.0133706 0.00220375 1208 -6.06721 0.0000
## age:old -0.0081902 0.00185899 1208 -4.40572 0.0000
## young:age:IDSEX -0.0002194 0.00152445 1208 -0.14394 0.8856
## age:old:IDSEX -0.0020622 0.00157535 1208 -1.30902 0.1908
## Correlation:
## young old yong:g age:ld y::IDS
## old 0.027
## young:age -0.834 -0.025
## age:old -0.010 -0.764 0.019
## young:age:IDSEX 0.053 0.001 -0.483 -0.005
## age:old:IDSEX -0.002 0.096 -0.003 -0.564 0.011
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.76631806 -0.47146696 0.02929323 0.49217553 2.85531402
##
## Number of Observations: 1746
## Number of Groups: 533
anova(mSPd, mgSPd)
## Traditionalism
mgTRd <- lme(fixed = TR ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(mgTRd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 1809.354 1874.501 -892.6769
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.2702731 young
## old 0.3009721 0.459
## Residual 0.3451784
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1522969
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.0000000 0.9296463
## Fixed effects: TR ~ -1 + young + young:age + old + old:age + young:age:IDSEX + old:age:IDSEX
## Value Std.Error DF t-value p-value
## young 2.8460719 0.04762873 1144 59.75537 0.0000
## old 2.8121352 0.04208072 1144 66.82717 0.0000
## young:age -0.0120355 0.00208378 1144 -5.77578 0.0000
## age:old -0.0094375 0.00187888 1144 -5.02291 0.0000
## young:age:IDSEX 0.0023393 0.00135433 1144 1.72726 0.0844
## age:old:IDSEX 0.0015487 0.00135496 1144 1.14296 0.2533
## Correlation:
## young old yong:g age:ld y::IDS
## old 0.077
## young:age -0.851 -0.043
## age:old -0.016 -0.829 0.037
## young:age:IDSEX 0.068 0.000 -0.470 -0.023
## age:old:IDSEX -0.002 0.102 -0.020 -0.503 0.057
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -3.332172791 -0.516147629 0.009496759 0.494548221 3.259495394
##
## Number of Observations: 1684
## Number of Groups: 535
anova(mTRd, mgTRd)
4. Adoption status as moderator
# >> Adoption status as moderators ----
#bio = 0, adop = 1
long <- long %>% mutate(
IDAB = ifelse(IDAB == 1, 1, 0))
## Achievement
maACd <- lme(fixed = AC ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(maACd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2439.379 2504.994 -1207.689
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3069992 young
## old 0.3426590 -0.07
## Residual 0.3792424
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1358322
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.064094
## Fixed effects: AC ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB
## Value Std.Error DF t-value p-value
## young 2.3271299 0.05876186 1212 39.60273 0.0000
## old 2.2822807 0.04543903 1212 50.22732 0.0000
## young:age 0.0210291 0.00243841 1212 8.62412 0.0000
## age:old 0.0196357 0.00204059 1212 9.62257 0.0000
## young:age:IDAB -0.0014606 0.00161553 1212 -0.90409 0.3661
## age:old:IDAB 0.0007625 0.00158341 1212 0.48158 0.6302
## Correlation:
## young old yong:g age:ld y::IDA
## old 0.033
## young:age -0.884 -0.041
## age:old -0.015 -0.792 0.032
## young:age:IDAB 0.011 0.001 -0.338 -0.012
## age:old:IDAB -0.003 0.022 -0.006 -0.497 0.026
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.875425684 -0.555607053 -0.004827308 0.546583548 2.462914858
##
## Number of Observations: 1751
## Number of Groups: 534
anova(mACd, maACd)
## Aggression
maAGd <- lme(fixed = AG ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(maAGd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2605.949 2671.585 -1290.974
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.4076677 young
## old 0.4111097 0.216
## Residual 0.3941475
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.1570875
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.0000000 0.8896306
## Fixed effects: AG ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB
## Value Std.Error DF t-value p-value
## young 2.4435980 0.05405320 1216 45.20728 0.0000
## old 2.4617407 0.04806207 1216 51.22003 0.0000
## young:age -0.0248564 0.00222326 1216 -11.18014 0.0000
## age:old -0.0256783 0.00215387 1216 -11.92191 0.0000
## young:age:IDAB 0.0021288 0.00171108 1216 1.24415 0.2137
## age:old:IDAB 0.0068843 0.00175186 1216 3.92970 0.0001
## Correlation:
## young old yong:g age:ld y::IDA
## old 0.070
## young:age -0.839 -0.047
## age:old -0.014 -0.766 0.050
## young:age:IDAB 0.004 0.000 -0.386 -0.053
## age:old:IDAB -0.010 0.020 -0.035 -0.519 0.109
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.3113230 -0.5335534 -0.0730538 0.4522537 3.2018594
##
## Number of Observations: 1754
## Number of Groups: 533
anova(mAGd, maAGd)
## Control
maCONd <- lme(fixed = CON ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(maCONd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2020.507 2086.163 -998.2533
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.2939346 young
## old 0.3132572 0.008
## Residual 0.3305191
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.07993586
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.028306
## Fixed effects: CON ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB
## Value Std.Error DF t-value p-value
## young 2.3324060 0.05059898 1219 46.09591 0.0000
## old 2.2663642 0.04007329 1219 56.55548 0.0000
## young:age 0.0202174 0.00210266 1219 9.61514 0.0000
## age:old 0.0219229 0.00180234 1219 12.16355 0.0000
## young:age:IDAB -0.0007945 0.00143698 1219 -0.55292 0.5804
## age:old:IDAB -0.0011508 0.00141292 1219 -0.81449 0.4155
## Correlation:
## young old yong:g age:ld y::IDA
## old 0.023
## young:age -0.877 -0.024
## age:old -0.009 -0.787 0.020
## young:age:IDAB 0.011 0.000 -0.348 -0.012
## age:old:IDAB -0.002 0.021 -0.007 -0.501 0.026
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -3.72044349 -0.49981642 0.01178782 0.56415046 2.63060411
##
## Number of Observations: 1757
## Number of Groups: 533
anova(mCONd, maCONd)
## Harm Avoidance
maHAd <- lme(fixed = HA ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(maHAd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2928.541 2994.183 -1452.27
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3787724 young
## old 0.4916011 0.236
## Residual 0.3954184
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.05255881
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.091461
## Fixed effects: HA ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB
## Value Std.Error DF t-value p-value
## young 2.2133566 0.06472351 1216 34.19710 0.0000
## old 2.2764548 0.05043363 1216 45.13764 0.0000
## young:age 0.0237705 0.00268666 1216 8.84758 0.0000
## age:old 0.0223201 0.00225819 1216 9.88406 0.0000
## young:age:IDAB 0.0049314 0.00182795 1216 2.69778 0.0071
## age:old:IDAB -0.0009851 0.00194708 1216 -0.50592 0.6130
## Correlation:
## young old yong:g age:ld y::IDA
## old 0.040
## young:age -0.878 -0.016
## age:old -0.003 -0.733 0.025
## young:age:IDAB 0.010 -0.001 -0.344 -0.046
## age:old:IDAB -0.007 0.009 -0.025 -0.541 0.089
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.51150701 -0.50682618 0.03016124 0.55083881 2.42784070
##
## Number of Observations: 1755
## Number of Groups: 534
anova(mHAd, maHAd)
## Social Potency
maSPd <- lme(fixed = SP ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(maSPd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 2137.567 2203.148 -1056.784
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.3154804 young
## old 0.3998781 0.036
## Residual 0.3195298
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.08242774
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.000000 1.025885
## Fixed effects: SP ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB
## Value Std.Error DF t-value p-value
## young 2.8901595 0.04990754 1208 57.91028 0.0000
## old 2.8586357 0.04085335 1208 69.97311 0.0000
## young:age -0.0135404 0.00206339 1208 -6.56221 0.0000
## age:old -0.0113605 0.00183446 1208 -6.19287 0.0000
## young:age:IDAB 0.0000323 0.00146563 1208 0.02206 0.9824
## age:old:IDAB 0.0028552 0.00159077 1208 1.79484 0.0729
## Correlation:
## young old yong:g age:ld y::IDA
## old 0.027
## young:age -0.867 -0.024
## age:old -0.009 -0.730 0.022
## young:age:IDAB 0.004 0.000 -0.354 -0.017
## age:old:IDAB -0.003 0.012 -0.010 -0.547 0.033
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -2.7585751 -0.4817614 0.0363319 0.4981193 2.8692835
##
## Number of Observations: 1746
## Number of Groups: 533
anova(mSPd, maSPd)
## Traditionalism
maTRd <- lme(fixed = TR ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB,
random = ~ -1 + young + old | IDYRFAM,
correlation = corAR1(),
weights=varIdent(form = ~1 | yo),
na.action = "na.omit",
data = long,
control = list(maxIter = 1000),
method = "ML")
summary(maTRd)
## Linear mixed-effects model fit by maximum likelihood
## Data: long
## AIC BIC logLik
## 1801.908 1867.055 -888.9538
##
## Random effects:
## Formula: ~-1 + young + old | IDYRFAM
## Structure: General positive-definite, Log-Cholesky parametrization
## StdDev Corr
## young 0.2744407 young
## old 0.2934818 0.447
## Residual 0.3456808
##
## Correlation Structure: AR(1)
## Formula: ~1 | IDYRFAM
## Parameter estimate(s):
## Phi
## 0.151767
## Variance function:
## Structure: Different standard deviations per stratum
## Formula: ~1 | yo
## Parameter estimates:
## o y
## 1.0000000 0.9234463
## Fixed effects: TR ~ -1 + young + young:age + old + old:age + young:age:IDAB + old:age:IDAB
## Value Std.Error DF t-value p-value
## young 2.8432494 0.04744580 1144 59.92626 0.0000
## old 2.8040473 0.04178600 1144 67.10494 0.0000
## young:age -0.0105412 0.00195439 1144 -5.39361 0.0000
## age:old -0.0053664 0.00185259 1144 -2.89670 0.0038
## young:age:IDAB 0.0002262 0.00132638 1144 0.17054 0.8646
## age:old:IDAB -0.0045942 0.00137490 1144 -3.34151 0.0009
## Correlation:
## young old yong:g age:ld y::IDA
## old 0.077
## young:age -0.877 -0.045
## age:old -0.012 -0.806 0.053
## young:age:IDAB 0.013 -0.002 -0.345 -0.074
## age:old:IDAB -0.011 0.023 -0.050 -0.481 0.170
##
## Standardized Within-Group Residuals:
## Min Q1 Med Q3 Max
## -3.28738460 -0.52365551 0.01386345 0.50601908 3.30210265
##
## Number of Observations: 1684
## Number of Groups: 535
anova(mTRd, maTRd)