Robustness checks
Ruben Arslan
10/4/2021
Load data
vcs <- rio::import("data_complete_2021_zscored.rds")
vcs <- vcs %>%
mutate(
age = if_else(dataset == 9, 20, age)/10,
age_se = if_else(dataset == 9, 3, 0.5)/10
)
vcs <- vcs %>% filter(!is.na(f0), !is.na(pf), !is.na(age))
vcs_novowels <- vcs %>% filter(! dataset %in% c(1, 10))
xtabs(~ is.na(dominance) + sex, vcs_novowels)## sex
## is.na(dominance) female male
## FALSE 491 494
## TRUE 481 337xtabs(~ is.na(extra) + sex, vcs_novowels)## sex
## is.na(extra) female male
## FALSE 816 529
## TRUE 156 302vcs_within <- rio::import("data_complete_2021_within_zscored.rds")
vcs_nooutlier <- rio::import("data_complete_2021_zscored_no_outliers.rds")
var_label(vcs$dominance) <- "Dominance"
var_label(vcs$neuro) <- "Neuroticism"
var_label(vcs$agree) <- "Agreeableness"
var_label(vcs$extra) <- "Extraversion"
var_label(vcs$openn) <- "Openness"
var_label(vcs$consc) <- "Conscientiousness"
var_label(vcs$soir_full) <- "Unrestricted sociosexuality"
var_label(vcs$f0) <- "Voice pitch"
var_label(vcs$pf) <- "Formants"
vcs$sex_c <- vcs$sex
contrasts(vcs$sex) <- contr.helmert(2)
var_label(vcs$sex) <- "Sex"
set.seed(1)
var_label(vcs$age) <- "Age"
vcs_nooutlier <- vcs_nooutlier %>%
mutate(
age = if_else(dataset == 9, 20, age)/10,
age_se = if_else(dataset == 9, 3, 0.5)/10
)
vcs_within <- vcs_within %>%
mutate(
age = if_else(dataset == 9, 20, age)/10,
age_se = if_else(dataset == 9, 3, 0.5)/10
)
warmup <- 2000
iter <- warmup + 2000
chains <- 4
control <- list(adapt_delta = 0.99)
priors <- c(
prior(normal(0, 3), class = b)
)
library(brms)
options(mc.cores = parallel::detectCores(), brms.backend = "cmdstanr")
rstan::rstan_options(auto_write = TRUE)
variable_labels <- c("Intercept"= "Intercept", "sex1" = "Sex [male]", "bsp_meageage_se" = "Age±SE", "f0" = "Voice pitch (f0)", "pf" = " Formant position (Pf)", "age" = "Age", "f0:sex1" = "Voice pitch (f0) × Sex")
effect_labels <- c("b_Intercept"= "Intercept", "b_sex1" = "Sex [male]", "bsp_meageage_se" = "Age±SE", "b_f0" = "Voice pitch (f0)", "b_pf" = " Formant position (Pf)", "b_age" = "Age", "b_f0.sex1" = "Voice pitch (f0) × Sex")Check Hypothesis 1: Dominance
Participants with lower voice pitch will have a more dominant personality.
Interpretation: Little difference across dataset specifications (whether standardized globally, within dataset, whether outliers or vowel-based data were omitted). No clear effect of individual formants (f1-f4).
Models
h1_simple <- brm(dominance ~ f0 + pf + sex + age + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, save_mevars = TRUE,
file = "models/dominance/h1_simple")
h1_simple_within <- update(h1_simple, newdata = vcs_within, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 3 finished in 3.5 seconds.
## Chain 2 finished in 3.5 seconds.
## Chain 4 finished in 3.7 seconds.
## Chain 1 finished in 4.1 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 3.7 seconds.
## Total execution time: 4.7 seconds.h1_simple_nooutlier <- update(h1_simple, newdata = vcs_nooutlier, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 4 finished in 3.8 seconds.
## Chain 1 finished in 4.3 seconds.
## Chain 2 finished in 4.3 seconds.
## Chain 3 finished in 4.2 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 4.1 seconds.
## Total execution time: 4.6 seconds.h1_simple_novowels <- update(h1_simple, newdata = vcs_novowels, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 1 finished in 3.8 seconds.
## Chain 2 finished in 4.0 seconds.
## Chain 4 finished in 4.3 seconds.
## Chain 3 finished in 4.5 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 4.2 seconds.
## Total execution time: 4.9 seconds.h1_simple_formants <- update(h1_simple, newdata = vcs, formula = . ~ . - pf + f1 + f2 + f3 + f4, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 4 finished in 5.1 seconds.
## Chain 2 finished in 5.2 seconds.
## Chain 3 finished in 5.3 seconds.
## Chain 1 finished in 5.5 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 5.3 seconds.
## Total execution time: 5.9 seconds.list(`Preregistered` = h1_simple,
`Standardized within` = h1_simple_within,
`No outliers` = h1_simple_nooutlier,
`No vowels` = h1_simple_novowels) %>%
map(~ broom.mixed::tidy(., conf.int = T)) %>%
bind_rows(.id = "Specification") %>%
filter(!term %in% c("(Intercept)"),
!term %in% c("sd__(Intercept)"),
!term %in% c("sd__Observation")
) %>%
ggplot(aes(x = Specification, y = estimate, ymin = conf.low, ymax = conf.high)) +
geom_pointrange() +
ggrepel::geom_text_repel(aes(label = sprintf("%.2f [%.2f;%.2f]", estimate, conf.low, conf.high )), nudge_x = 0.15, size = 3) +
facet_wrap(~ term) +
coord_flip()
list(`Preregistered` = h1_simple,
`Single formants` = h1_simple_formants) %>%
map(~ broom.mixed::tidy(., conf.int = T)) %>%
bind_rows(.id = "Specification") %>%
filter(term %in% c("pf", "f1", "f2", "f3", "f4")) %>%
ggplot(aes(x = term, y = estimate, ymin = conf.low, ymax = conf.high, color = Specification)) +
geom_pointrange() +
ggrepel::geom_text_repel(aes(label = sprintf("%.2f [%.2f;%.2f]", estimate, conf.low, conf.high )), nudge_x = 0.15, size = 3) +
coord_flip()
Model details
summary(h1_simple)## Warning: There were 3 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: dominance ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs (Number of observations: 985)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 4)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.53 0.41 0.16 1.68 1.00 1993 2686
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.07 0.37 -0.80 0.68 1.00 2616 2533
## f0 -0.27 0.09 -0.45 -0.08 1.00 5357 5343
## pf -0.00 0.07 -0.13 0.12 1.00 6495 5685
## sex1 -0.30 0.11 -0.52 -0.08 1.00 4465 4603
## age 0.04 0.06 -0.08 0.17 1.00 7482 5631
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.93 1.02 1.00 8271 5405
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_within)## Warning: There were 335 divergent transitions after warmup. Increasing
## adapt_delta above may help. See http://mc-stan.org/misc/warnings.html#divergent-
## transitions-after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: dominance ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_within (Number of observations: 985)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 4)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.54 0.47 0.15 2.26 1.05 99 18
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.12 0.38 -1.19 0.48 1.04 107 22
## f0 -0.26 0.09 -0.45 -0.08 1.01 779 666
## pf 0.00 0.07 -0.13 0.13 1.01 3711 4065
## sex1 -0.28 0.11 -0.51 -0.06 1.01 1331 648
## age 0.04 0.06 -0.08 0.16 1.00 2072 3893
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.93 1.01 1.01 341 3835
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_nooutlier)## Warning: There were 57 divergent transitions after warmup. Increasing
## adapt_delta above may help. See http://mc-stan.org/misc/warnings.html#divergent-
## transitions-after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: dominance ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_nooutlier (Number of observations: 983)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 4)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.47 0.25 0.17 1.11 1.00 2129 2847
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.09 0.29 -0.70 0.49 1.00 2614 2761
## f0 -0.29 0.10 -0.47 -0.10 1.00 5047 5169
## pf -0.00 0.07 -0.13 0.13 1.00 5144 4646
## sex1 -0.31 0.11 -0.53 -0.09 1.00 4383 5240
## age 0.05 0.06 -0.08 0.17 1.00 6374 5010
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.93 1.02 1.00 6992 5123
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_novowels)## Warning: There were 143 divergent transitions after warmup. Increasing
## adapt_delta above may help. See http://mc-stan.org/misc/warnings.html#divergent-
## transitions-after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: dominance ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_novowels (Number of observations: 985)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 4)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.51 0.32 0.17 1.46 1.01 462 122
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.07 0.33 -0.75 0.67 1.00 486 112
## f0 -0.27 0.09 -0.46 -0.09 1.00 2545 5273
## pf -0.00 0.06 -0.13 0.13 1.00 4953 5217
## sex1 -0.30 0.11 -0.52 -0.09 1.00 3495 3520
## age 0.04 0.06 -0.08 0.16 1.00 6277 5489
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.93 1.02 1.00 6219 4984
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_formants)## Warning: There were 67 divergent transitions after warmup. Increasing
## adapt_delta above may help. See http://mc-stan.org/misc/warnings.html#divergent-
## transitions-after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: dominance ~ f0 + sex + age + (1 | dataset) + f1 + f2 + f3 + f4
## Data: vcs (Number of observations: 985)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 4)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.52 0.31 0.18 1.38 1.00 1855 1065
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.06 0.31 -0.69 0.57 1.00 2840 1367
## f0 -0.27 0.09 -0.45 -0.08 1.00 6653 6132
## sex1 -0.29 0.12 -0.53 -0.07 1.00 5842 5118
## age 0.04 0.06 -0.08 0.17 1.00 6893 4790
## f1 -0.03 0.04 -0.11 0.05 1.00 8061 4392
## f2 0.05 0.06 -0.07 0.18 1.00 7920 4879
## f3 0.04 0.08 -0.12 0.21 1.00 7753 4939
## f4 -0.06 0.09 -0.23 0.11 1.00 7110 4902
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.93 1.02 1.00 8857 5317
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).Check Hypothesis 2: Extraversion
Participants with lower voice pitch will score higher on extraversion
Interpretation: Little difference across dataset specifications (whether standardized globally, within dataset, whether outliers or vowel-based data were omitted). No clear effect of individual formants (f1-f4).
Models
h1_simple <- brm(extra ~ f0 + pf + sex + age + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, save_mevars = TRUE,
file = "models/extra/h1_simple")
h1_simple_within <- update(h1_simple, newdata = vcs_within, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 1 finished in 32.1 seconds.
## Chain 4 finished in 33.0 seconds.
## Chain 2 finished in 33.1 seconds.
## Chain 3 finished in 33.9 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 33.0 seconds.
## Total execution time: 34.2 seconds.h1_simple_nooutlier <- update(h1_simple, newdata = vcs_nooutlier, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 3 finished in 32.0 seconds.
## Chain 1 finished in 32.2 seconds.
## Chain 4 finished in 32.2 seconds.
## Chain 2 finished in 32.7 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 32.3 seconds.
## Total execution time: 33.2 seconds.h1_simple_novowels <- update(h1_simple, newdata = vcs_novowels, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 2 finished in 30.8 seconds.
## Chain 4 finished in 31.7 seconds.
## Chain 3 finished in 33.6 seconds.
## Chain 1 finished in 40.0 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 34.0 seconds.
## Total execution time: 40.4 seconds.h1_simple_formants <- update(h1_simple, newdata = vcs, formula = . ~ . - pf + f1 + f2 + f3 + f4, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 1 finished in 32.6 seconds.
## Chain 4 finished in 33.5 seconds.
## Chain 3 finished in 33.6 seconds.
## Chain 2 finished in 34.7 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 33.6 seconds.
## Total execution time: 35.2 seconds.list(`Preregistered` = h1_simple,
`Standardized within` = h1_simple_within,
`No outliers` = h1_simple_nooutlier,
`No vowels` = h1_simple_novowels) %>%
map(~ broom.mixed::tidy(., conf.int = T)) %>%
bind_rows(.id = "Specification") %>%
filter(!term %in% c("(Intercept)"),
!term %in% c("sd__(Intercept)"),
!term %in% c("sd__Observation")
) %>%
ggplot(aes(x = Specification, y = estimate, ymin = conf.low, ymax = conf.high)) +
geom_pointrange() +
ggrepel::geom_text_repel(aes(label = sprintf("%.2f [%.2f;%.2f]", estimate, conf.low, conf.high )), nudge_x = 0.15, size = 3) +
facet_wrap(~ term) +
coord_flip()## Warning in tidy.brmsfit(., conf.int = T): some parameter names contain
## underscores: term naming may be unreliable!## Warning in tidy.brmsfit(., conf.int = T): some parameter names contain
## underscores: term naming may be unreliable!
## Warning in tidy.brmsfit(., conf.int = T): some parameter names contain
## underscores: term naming may be unreliable!
## Warning in tidy.brmsfit(., conf.int = T): some parameter names contain
## underscores: term naming may be unreliable!
list(`Preregistered` = h1_simple,
`Single formants` = h1_simple_formants) %>%
map(~ broom.mixed::tidy(., conf.int = T)) %>%
bind_rows(.id = "Specification") %>%
filter(term %in% c("pf", "f1", "f2", "f3", "f4")) %>%
ggplot(aes(x = term, y = estimate, ymin = conf.low, ymax = conf.high, color = Specification)) +
geom_pointrange() +
ggrepel::geom_text_repel(aes(label = sprintf("%.2f [%.2f;%.2f]", estimate, conf.low, conf.high )), nudge_x = 0.15, size = 3) +
coord_flip()## Warning in tidy.brmsfit(., conf.int = T): some parameter names contain
## underscores: term naming may be unreliable!
## Warning in tidy.brmsfit(., conf.int = T): some parameter names contain
## underscores: term naming may be unreliable!
Model details
summary(h1_simple)## Warning: Parts of the model have not converged (some Rhats are > 1.05). Be
## careful when analysing the results! We recommend running more iterations and/or
## setting stronger priors.## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: extra ~ f0 + pf + sex + me(age, age_se) + (1 | dataset)
## Data: vcs (Number of observations: 1433)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 7)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.35 0.15 0.16 0.73 1.00 2404 4372
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.19 0.20 -0.21 0.58 1.00 3114 4163
## f0 -0.23 0.08 -0.38 -0.08 1.00 7221 6134
## pf 0.08 0.05 -0.03 0.18 1.00 8287 6673
## sex1 -0.28 0.09 -0.46 -0.10 1.00 6190 6003
## meageage_se -0.11 0.06 -0.22 0.01 1.00 7134 5697
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.94 1.01 1.00 12826 5752
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_within)## Warning: Parts of the model have not converged (some Rhats are > 1.05). Be
## careful when analysing the results! We recommend running more iterations and/or
## setting stronger priors.## Warning: There were 1 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: extra ~ f0 + pf + sex + me(age, age_se) + (1 | dataset)
## Data: vcs_within (Number of observations: 1433)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 7)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.41 0.18 0.20 0.86 1.00 3340 4542
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.19 0.22 -0.25 0.62 1.00 4422 4731
## f0 -0.19 0.07 -0.34 -0.06 1.00 13897 5874
## pf 0.09 0.06 -0.02 0.20 1.00 14738 5699
## sex1 -0.24 0.09 -0.42 -0.06 1.00 11345 5436
## meageage_se -0.10 0.06 -0.21 0.01 1.00 14493 5490
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.94 1.01 1.00 19129 5708
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_nooutlier)## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: extra ~ f0 + pf + sex + me(age, age_se) + (1 | dataset)
## Data: vcs_nooutlier (Number of observations: 1430)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 7)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.34 0.15 0.16 0.72 1.00 3474 4314
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.18 0.20 -0.23 0.57 1.00 5146 5480
## f0 -0.24 0.08 -0.39 -0.08 1.00 14255 6501
## pf 0.07 0.06 -0.04 0.17 1.00 14622 6190
## sex1 -0.29 0.09 -0.48 -0.11 1.00 10952 6690
## meageage_se -0.10 0.06 -0.22 0.01 1.00 11853 5833
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.94 1.01 1.00 21269 5824
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_novowels)## Warning: There were 3 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: extra ~ f0 + pf + sex + me(age, age_se) + (1 | dataset)
## Data: vcs_novowels (Number of observations: 1345)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 6)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.40 0.20 0.18 0.92 1.00 2901 4891
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.15 0.23 -0.32 0.61 1.00 4521 4450
## f0 -0.21 0.08 -0.37 -0.06 1.00 13855 6283
## pf 0.09 0.06 -0.02 0.20 1.00 13996 6469
## sex1 -0.25 0.10 -0.44 -0.07 1.00 10973 6129
## meageage_se -0.09 0.06 -0.21 0.01 1.00 13554 6175
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.95 0.02 0.92 0.99 1.00 19088 6132
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_formants)## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: extra ~ f0 + sex + me(age, age_se) + (1 | dataset) + f1 + f2 + f3 + f4
## Data: vcs (Number of observations: 1433)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 7)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.38 0.16 0.18 0.81 1.00 3710 4644
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.18 0.21 -0.24 0.60 1.00 5958 5200
## f0 -0.23 0.08 -0.38 -0.07 1.00 14095 6585
## sex1 -0.32 0.10 -0.51 -0.13 1.00 12943 6283
## f1 0.03 0.04 -0.04 0.10 1.00 18403 6138
## f2 0.10 0.05 -0.01 0.20 1.00 15838 5494
## f3 0.02 0.07 -0.12 0.15 1.00 17298 6094
## f4 -0.07 0.07 -0.21 0.06 1.00 18077 5747
## meageage_se -0.11 0.06 -0.22 0.00 1.00 16396 5899
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.02 0.94 1.01 1.00 21407 5625
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).Check Hypothesis 3: Agreeableness
Participants with lower voice pitch will score higher on agreeableness.
Interpretation: Little difference across dataset specifications (whether standardized globally, within dataset, whether outliers or vowel-based data were omitted). No clear effect of individual formants f1, f2, f4, the 95% CI for f3 nominally excludes zero.
Models
h1_simple <- brm(agree ~ f0 + pf + sex + age + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, save_mevars = TRUE,
file = "models/agree/h1_simple")
h1_simple_within <- update(h1_simple, newdata = vcs_within, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 4 finished in 4.1 seconds.
## Chain 3 finished in 4.3 seconds.
## Chain 2 finished in 4.7 seconds.
## Chain 1 finished in 5.0 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 4.5 seconds.
## Total execution time: 5.5 seconds.h1_simple_nooutlier <- update(h1_simple, newdata = vcs_nooutlier, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 3 finished in 4.5 seconds.
## Chain 2 finished in 4.6 seconds.
## Chain 1 finished in 4.8 seconds.
## Chain 4 finished in 5.5 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 4.8 seconds.
## Total execution time: 5.7 seconds.h1_simple_novowels <- update(h1_simple, newdata = vcs_novowels, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 4 finished in 4.5 seconds.
## Chain 1 finished in 4.8 seconds.
## Chain 3 finished in 5.1 seconds.
## Chain 2 finished in 5.4 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 5.0 seconds.
## Total execution time: 5.8 seconds.h1_simple_formants <- update(h1_simple, newdata = vcs, formula = . ~ . - pf + f1 + f2 + f3 + f4, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 1 finished in 5.6 seconds.
## Chain 3 finished in 5.9 seconds.
## Chain 2 finished in 6.3 seconds.
## Chain 4 finished in 6.7 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 6.1 seconds.
## Total execution time: 7.0 seconds.list(`Preregistered` = h1_simple,
`Standardized within` = h1_simple_within,
`No outliers` = h1_simple_nooutlier,
`No vowels` = h1_simple_novowels) %>%
map(~ broom.mixed::tidy(., conf.int = T)) %>%
bind_rows(.id = "Specification") %>%
filter(!term %in% c("(Intercept)"),
!term %in% c("sd__(Intercept)"),
!term %in% c("sd__Observation")
) %>%
ggplot(aes(x = Specification, y = estimate, ymin = conf.low, ymax = conf.high)) +
geom_pointrange() +
ggrepel::geom_text_repel(aes(label = sprintf("%.2f [%.2f;%.2f]", estimate, conf.low, conf.high )), nudge_x = 0.15, size = 3) +
facet_wrap(~ term) +
coord_flip()
list(`Preregistered` = h1_simple,
`Single formants` = h1_simple_formants) %>%
map(~ broom.mixed::tidy(., conf.int = T)) %>%
bind_rows(.id = "Specification") %>%
filter(term %in% c("pf", "f1", "f2", "f3", "f4")) %>%
ggplot(aes(x = term, y = estimate, ymin = conf.low, ymax = conf.high, color = Specification)) +
geom_pointrange() +
ggrepel::geom_text_repel(aes(label = sprintf("%.2f [%.2f;%.2f]", estimate, conf.low, conf.high )), nudge_x = 0.15, size = 3) +
coord_flip()
Model details
summary(h1_simple)## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: agree ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs (Number of observations: 1434)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 7)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.37 0.15 0.19 0.76 1.00 2125 3434
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.09 0.21 -0.31 0.49 1.00 3092 3611
## f0 0.02 0.08 -0.13 0.17 1.00 5513 5486
## pf 0.03 0.05 -0.07 0.13 1.00 5878 5807
## sex1 -0.00 0.09 -0.18 0.18 1.00 4455 4461
## age -0.06 0.06 -0.17 0.06 1.00 6591 5131
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.96 0.02 0.92 0.99 1.00 7894 5152
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_within)## Warning: There were 13 divergent transitions after warmup. Increasing
## adapt_delta above may help. See http://mc-stan.org/misc/warnings.html#divergent-
## transitions-after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: agree ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_within (Number of observations: 1434)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 7)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.37 0.15 0.18 0.76 1.00 1966 2144
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.09 0.20 -0.29 0.50 1.00 3210 3959
## f0 0.01 0.07 -0.13 0.15 1.00 6157 5206
## pf 0.02 0.06 -0.09 0.13 1.00 6769 5218
## sex1 -0.03 0.09 -0.20 0.14 1.00 5212 4940
## age -0.06 0.06 -0.17 0.05 1.00 6610 5628
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.96 0.02 0.92 0.99 1.00 9250 4800
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_nooutlier)## Warning: There were 3 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: agree ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_nooutlier (Number of observations: 1431)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 7)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.37 0.15 0.18 0.76 1.00 2080 3185
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.10 0.21 -0.30 0.52 1.00 3104 4141
## f0 0.02 0.08 -0.12 0.17 1.00 5899 5579
## pf 0.03 0.05 -0.08 0.14 1.00 6755 5351
## sex1 -0.00 0.09 -0.19 0.18 1.00 5028 5234
## age -0.06 0.06 -0.17 0.05 1.00 7832 5484
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.96 0.02 0.92 0.99 1.00 8379 5865
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_novowels)## Warning: There were 14 divergent transitions after warmup. Increasing
## adapt_delta above may help. See http://mc-stan.org/misc/warnings.html#divergent-
## transitions-after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: agree ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_novowels (Number of observations: 1346)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 6)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.42 0.19 0.20 0.92 1.00 1915 1405
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.10 0.24 -0.36 0.57 1.00 1926 1463
## f0 -0.03 0.08 -0.18 0.12 1.00 5918 5873
## pf 0.07 0.06 -0.04 0.18 1.00 6312 5368
## sex1 -0.02 0.09 -0.21 0.16 1.00 4926 4747
## age -0.06 0.06 -0.17 0.05 1.00 6435 1826
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.95 0.02 0.92 0.99 1.00 8062 5468
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_formants)## Warning: There were 6 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: agree ~ f0 + sex + age + (1 | dataset) + f1 + f2 + f3 + f4
## Data: vcs (Number of observations: 1434)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 7)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.39 0.17 0.19 0.82 1.00 1622 1400
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.11 0.22 -0.31 0.54 1.00 2208 2000
## f0 0.02 0.08 -0.13 0.16 1.00 6528 5849
## sex1 -0.01 0.10 -0.20 0.18 1.00 5906 5583
## age -0.06 0.06 -0.17 0.05 1.00 8600 5093
## f1 0.03 0.04 -0.04 0.10 1.00 8275 5862
## f2 -0.05 0.05 -0.15 0.05 1.00 6412 3006
## f3 0.14 0.07 0.01 0.27 1.00 7127 5922
## f4 -0.08 0.07 -0.22 0.06 1.00 6485 4651
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.96 0.02 0.92 0.99 1.00 8752 4731
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).Check Hypothesis 4: SOI-R behavior
Participants with lower voice pitch will report having a more unrestricted sociosexual behavior.
Interpretation: Little difference across dataset specifications (whether standardized globally, within dataset, whether outliers or vowel-based data were omitted). No clear effect of individual formants (f1-f4).
Models
h1_simple <- brm(behavior ~ f0 + pf + sex + age + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, save_mevars = TRUE,
file = "models/behavior/h1_simple")
h1_simple_within <- update(h1_simple, newdata = vcs_within, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 2 finished in 6.8 seconds.
## Chain 3 finished in 7.2 seconds.
## Chain 4 finished in 7.8 seconds.
## Chain 1 finished in 8.2 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 7.5 seconds.
## Total execution time: 8.6 seconds.h1_simple_nooutlier <- update(h1_simple, newdata = vcs_nooutlier, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 4 finished in 6.8 seconds.
## Chain 1 finished in 7.2 seconds.
## Chain 2 finished in 7.3 seconds.
## Chain 3 finished in 7.8 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 7.3 seconds.
## Total execution time: 8.2 seconds.h1_simple_novowels <- update(h1_simple, newdata = vcs_novowels, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 3 finished in 6.4 seconds.
## Chain 4 finished in 6.8 seconds.
## Chain 1 finished in 7.1 seconds.
## Chain 2 finished in 7.1 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 6.9 seconds.
## Total execution time: 7.4 seconds.h1_simple_formants <- update(h1_simple, newdata = vcs, formula = . ~ . - pf + f1 + f2 + f3 + f4, recompile = TRUE, silent = 2, refresh = 0)## Warning: Rows containing NAs were excluded from the model.## Running MCMC with 4 chains, at most 8 in parallel...
##
## Chain 2 finished in 8.3 seconds.
## Chain 1 finished in 8.6 seconds.
## Chain 4 finished in 8.6 seconds.
## Chain 3 finished in 8.9 seconds.
##
## All 4 chains finished successfully.
## Mean chain execution time: 8.6 seconds.
## Total execution time: 9.2 seconds.list(`Preregistered` = h1_simple,
`Standardized within` = h1_simple_within,
`No outliers` = h1_simple_nooutlier,
`No vowels` = h1_simple_novowels) %>%
map(~ broom.mixed::tidy(., conf.int = T)) %>%
bind_rows(.id = "Specification") %>%
filter(!term %in% c("(Intercept)"),
!term %in% c("sd__(Intercept)"),
!term %in% c("sd__Observation")
) %>%
ggplot(aes(x = Specification, y = estimate, ymin = conf.low, ymax = conf.high)) +
geom_pointrange() +
ggrepel::geom_text_repel(aes(label = sprintf("%.2f [%.2f;%.2f]", estimate, conf.low, conf.high )), nudge_x = 0.15, size = 3) +
facet_wrap(~ term) +
coord_flip()
list(`Preregistered` = h1_simple,
`Single formants` = h1_simple_formants) %>%
map(~ broom.mixed::tidy(., conf.int = T)) %>%
bind_rows(.id = "Specification") %>%
filter(term %in% c("pf", "f1", "f2", "f3", "f4")) %>%
ggplot(aes(x = term, y = estimate, ymin = conf.low, ymax = conf.high, color = Specification)) +
geom_pointrange() +
ggrepel::geom_text_repel(aes(label = sprintf("%.2f [%.2f;%.2f]", estimate, conf.low, conf.high )), nudge_x = 0.15, size = 3) +
coord_flip()
Model details
summary(h1_simple)## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: behavior ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs (Number of observations: 1996)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 9)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.41 0.14 0.23 0.75 1.00 1925 3771
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.85 0.18 -1.20 -0.50 1.00 2520 4240
## f0 -0.21 0.06 -0.33 -0.09 1.00 5719 5077
## pf -0.02 0.04 -0.10 0.07 1.00 6518 5000
## sex1 -0.24 0.07 -0.38 -0.09 1.00 4768 5122
## age 0.32 0.05 0.23 0.41 1.00 6975 5146
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.91 0.01 0.89 0.94 1.00 7722 4831
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_within)## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: behavior ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_within (Number of observations: 1996)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 9)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.43 0.14 0.24 0.79 1.00 2025 3178
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.83 0.19 -1.19 -0.47 1.00 2562 3541
## f0 -0.19 0.06 -0.30 -0.08 1.00 6108 5186
## pf -0.01 0.05 -0.11 0.08 1.00 6547 5647
## sex1 -0.21 0.07 -0.35 -0.07 1.00 5173 4373
## age 0.32 0.05 0.23 0.41 1.00 6767 5288
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.91 0.01 0.89 0.94 1.00 7506 6005
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_nooutlier)## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: behavior ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_nooutlier (Number of observations: 1994)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 9)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.41 0.14 0.23 0.75 1.00 1952 3174
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.85 0.18 -1.20 -0.49 1.00 2668 3884
## f0 -0.21 0.06 -0.32 -0.09 1.00 5984 5314
## pf -0.01 0.04 -0.10 0.07 1.00 6619 5260
## sex1 -0.23 0.07 -0.37 -0.08 1.00 5046 5147
## age 0.32 0.05 0.23 0.41 1.00 7687 5318
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.91 0.01 0.88 0.94 1.00 7780 5082
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_novowels)## Warning: There were 1 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: behavior ~ f0 + pf + sex + age + (1 | dataset)
## Data: vcs_novowels (Number of observations: 1657)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 8)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.44 0.15 0.24 0.81 1.00 2195 3572
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.81 0.21 -1.22 -0.41 1.00 2908 3654
## f0 -0.20 0.07 -0.34 -0.07 1.00 6303 5262
## pf 0.01 0.05 -0.09 0.11 1.00 6646 5218
## sex1 -0.20 0.08 -0.37 -0.04 1.00 5465 4857
## age 0.31 0.05 0.21 0.40 1.00 7028 4620
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.91 0.02 0.88 0.94 1.00 7803 5165
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).summary(h1_simple_formants)## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: behavior ~ f0 + sex + age + (1 | dataset) + f1 + f2 + f3 + f4
## Data: vcs (Number of observations: 1996)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 9)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.43 0.14 0.24 0.79 1.00 2177 3545
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.85 0.18 -1.22 -0.49 1.00 2454 4095
## f0 -0.21 0.06 -0.33 -0.09 1.00 7014 5733
## sex1 -0.27 0.08 -0.42 -0.12 1.00 6085 5349
## age 0.32 0.05 0.23 0.41 1.00 9046 6093
## f1 0.04 0.03 -0.02 0.10 1.00 9594 6080
## f2 0.00 0.03 -0.06 0.06 1.00 8704 5803
## f3 -0.02 0.05 -0.12 0.08 1.00 7749 5891
## f4 -0.05 0.05 -0.16 0.05 1.00 7522 5784
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.91 0.01 0.88 0.94 1.00 11770 5242
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).Extraversion, adjusted for dominance
extra_adj_for_dominance <- brm(extra ~ dominance + f0 + pf + sex + me(age, age_se) + (1 | dataset), data = vcs,
iter = iter + 1500, warmup = warmup + 1500, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/extra/extra_adj_for_dominance2")## Warning: Rows containing NAs were excluded from the model.## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.extra_adj_for_dominance## Warning: Parts of the model have not converged (some Rhats are > 1.05). Be
## careful when analysing the results! We recommend running more iterations and/or
## setting stronger priors.## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: extra ~ dominance + f0 + pf + sex + me(age, age_se) + (1 | dataset)
## Data: vcs (Number of observations: 970)
## Samples: 4 chains, each with iter = 5500; warmup = 3500; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 4)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.25 0.26 0.03 0.92 1.00 1649 2050
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.29 0.21 -0.09 0.71 1.00 2755 3113
## dominance 0.49 0.03 0.44 0.54 1.00 8796 5920
## f0 -0.07 0.08 -0.23 0.08 1.00 5260 5393
## pf 0.13 0.05 0.02 0.24 1.00 5626 5912
## sex1 -0.08 0.09 -0.26 0.11 1.00 4407 4801
## meageage_se -0.10 0.05 -0.20 0.00 1.00 4867 5170
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.79 0.02 0.76 0.83 1.00 9001 5842
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).Dominance, adjusted for extraversion
dominance_adj_for_extra <- brm(dominance ~ extra + f0 + pf + sex + age + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/dominance/dominance_adj_for_extra")## Warning: Rows containing NAs were excluded from the model.## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.dominance_adj_for_extra## Warning: There were 4 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: dominance ~ extra + f0 + pf + sex + age + (1 | dataset)
## Data: vcs (Number of observations: 970)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 4)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.49 0.37 0.16 1.51 1.00 1949 2701
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.22 0.32 -0.87 0.39 1.00 3299 3365
## extra 0.54 0.03 0.48 0.60 1.00 8459 6010
## f0 -0.14 0.08 -0.30 0.02 1.00 5696 5429
## pf -0.08 0.06 -0.19 0.03 1.00 6699 5181
## sex1 -0.17 0.10 -0.35 0.02 1.00 4852 4716
## age 0.09 0.05 -0.02 0.19 1.00 7039 5434
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.83 0.02 0.79 0.87 1.00 9052 5588
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).SOI-R Behavior adj. for extraversion
behavior_adj_for_extra <- brm(behavior ~ extra + f0 + pf + sex + age + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/behavior/behavior_adj_for_extra")## Warning: Rows containing NAs were excluded from the model.## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.behavior_adj_for_extra## Warning: There were 1 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: behavior ~ extra + f0 + pf + sex + age + (1 | dataset)
## Data: vcs (Number of observations: 1236)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 5)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.48 0.28 0.20 1.22 1.00 1976 2101
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.91 0.28 -1.47 -0.37 1.00 2894 3255
## extra 0.23 0.03 0.18 0.29 1.00 9045 5931
## f0 -0.15 0.08 -0.31 0.01 1.00 5483 5249
## pf -0.04 0.06 -0.15 0.07 1.00 7390 5614
## sex1 -0.18 0.10 -0.37 0.01 1.00 5136 5211
## age 0.34 0.06 0.23 0.45 1.00 8327 5965
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.93 0.02 0.89 0.96 1.00 8979 4936
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).Extraversion, separate by sex
h1_men <- brm(extra ~ f0 + pf + me(age, age_se) + (1 | dataset), data = vcs %>% filter(sex == "male"),
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/extra/h1_men")## Warning: Rows containing NAs were excluded from the model.## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_women <- brm(extra ~ f0 + pf + me(age, age_se) + (1 | dataset), data = vcs %>% filter(sex != "male"),
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/extra/h1_women")## Warning: Rows containing NAs were excluded from the model.
## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_f0bysex <- brm(extra ~ f0*sex + me(age, age_se) + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/extra/h1_f0bysex")## Warning: Rows containing NAs were excluded from the model.
## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_men## Warning: Parts of the model have not converged (some Rhats are > 1.05). Be
## careful when analysing the results! We recommend running more iterations and/or
## setting stronger priors.## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: extra ~ f0 + pf + me(age, age_se) + (1 | dataset)
## Data: vcs %>% filter(sex == "male") (Number of observations: 617)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 5)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.33 0.26 0.05 0.99 1.00 1533 1799
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.09 0.32 -0.71 0.56 1.00 3667 4132
## f0 -0.21 0.13 -0.47 0.06 1.00 8527 6150
## pf 0.06 0.09 -0.11 0.24 1.00 7574 5932
## meageage_se -0.09 0.08 -0.25 0.07 1.00 4739 5047
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 1.00 0.03 0.94 1.06 1.00 10115 5306
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).h1_women## Warning: Parts of the model have not converged (some Rhats are > 1.05). Be
## careful when analysing the results! We recommend running more iterations and/or
## setting stronger priors.## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: extra ~ f0 + pf + me(age, age_se) + (1 | dataset)
## Data: vcs %>% filter(sex != "male") (Number of observations: 816)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 5)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.50 0.31 0.19 1.33 1.00 2038 3102
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.41 0.34 -0.27 1.06 1.00 2843 3355
## f0 -0.24 0.10 -0.43 -0.05 1.00 10974 5540
## pf 0.10 0.07 -0.03 0.23 1.00 10580 5563
## meageage_se -0.11 0.08 -0.26 0.05 1.00 6457 5693
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.96 0.02 0.91 1.01 1.00 10529 5958
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).equivalence_test(h1_f0bysex, ci = 0.89)## # Test for Practical Equivalence
##
## ROPE: [-0.10 0.10]
##
## Parameter | H0 | inside ROPE | 89% HDI
## -----------------------------------------------------
## Intercept | Undecided | 25.73 % | [-0.15 0.56]
## f0 | Undecided | 2.06 % | [-0.34 -0.09]
## sex1 | Rejected | 0.00 % | [-0.48 -0.20]
## f0.sex1 | Undecided | 87.57 % | [-0.10 0.15]
## meageage_se | Undecided | 44.50 % | [-0.20 -0.01]Agreeableness, separate by sex
h1_men <- brm(agree ~ f0 + pf + me(age, age_se) + (1 | dataset), data = vcs %>% filter(sex == "male"),
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/agree/h1_men") %>%
add_criterion("loo") %>%
add_criterion("bayes_R2") %>%
add_criterion("loo_R2") ## Warning: Rows containing NAs were excluded from the model.## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_women <- brm(agree ~ f0 + pf + me(age, age_se) + (1 | dataset), data = vcs %>% filter(sex != "male"),
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/agree/h1_women") %>%
add_criterion("loo") %>%
add_criterion("bayes_R2") %>%
add_criterion("loo_R2") ## Warning: Rows containing NAs were excluded from the model.
## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_f0bysex <- brm(agree ~ f0*sex + me(age, age_se) + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/agree/h1_f0bysex")## Warning: Rows containing NAs were excluded from the model.
## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_men## Warning: Parts of the model have not converged (some Rhats are > 1.05). Be
## careful when analysing the results! We recommend running more iterations and/or
## setting stronger priors.## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: agree ~ f0 + pf + me(age, age_se) + (1 | dataset)
## Data: vcs %>% filter(sex == "male") (Number of observations: 617)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 5)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.36 0.22 0.12 0.94 1.00 2109 3721
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.10 0.31 -0.51 0.72 1.00 3097 4470
## f0 0.23 0.12 -0.02 0.48 1.00 8401 5757
## pf -0.03 0.09 -0.20 0.14 1.00 7243 6177
## meageage_se -0.01 0.08 -0.17 0.15 1.00 4621 5388
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.94 0.03 0.89 1.00 1.00 8765 5539
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).h1_women## Warning: Parts of the model have not converged (some Rhats are > 1.05). Be
## careful when analysing the results! We recommend running more iterations and/or
## setting stronger priors.## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: agree ~ f0 + pf + me(age, age_se) + (1 | dataset)
## Data: vcs %>% filter(sex != "male") (Number of observations: 817)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 5)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.63 0.38 0.26 1.55 1.00 2037 2915
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.31 0.41 -0.51 1.09 1.00 2130 2641
## f0 -0.10 0.10 -0.29 0.09 1.00 7677 6441
## pf 0.04 0.07 -0.08 0.17 1.00 8024 5535
## meageage_se -0.11 0.08 -0.27 0.06 1.00 5958 5828
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.96 0.02 0.92 1.01 1.00 9136 6149
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).equivalence_test(h1_f0bysex, ci = 0.89)## # Test for Practical Equivalence
##
## ROPE: [-0.10 0.10]
##
## Parameter | H0 | inside ROPE | 89% HDI
## ----------------------------------------------------
## Intercept | Undecided | 12.09 % | [-0.06 0.67]
## f0 | Undecided | 55.65 % | [-0.04 0.21]
## sex1 | Undecided | 66.66 % | [-0.07 0.21]
## f0.sex1 | Undecided | 1.98 % | [ 0.09 0.33]
## meageage_se | Undecided | 70.07 % | [-0.16 0.02]Dominance, separate by sex
h1_men <- brm(dominance ~ f0 + pf + age + (1 | dataset), data = vcs %>% filter(sex == "male"),
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/dominance/h1_men")## Warning: Rows containing NAs were excluded from the model.## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_women <- brm(dominance ~ f0 + pf + age + (1 | dataset), data = vcs %>% filter(sex != "male"),
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/dominance/h1_women")## Warning: Rows containing NAs were excluded from the model.
## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_f0bysex <- brm(dominance ~ f0*sex + age + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change",
file = "models/dominance/h1_f0bysex")## Warning: Rows containing NAs were excluded from the model.h1_men## Warning: There were 3 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: dominance ~ f0 + pf + age + (1 | dataset)
## Data: vcs %>% filter(sex == "male") (Number of observations: 494)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 3)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.82 0.71 0.18 2.73 1.00 2085 3522
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.35 0.62 -1.68 0.86 1.00 2549 2263
## f0 -0.26 0.16 -0.57 0.04 1.00 6563 5146
## pf 0.09 0.13 -0.17 0.35 1.00 5761 5688
## age 0.04 0.09 -0.13 0.21 1.00 6290 4832
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.98 0.03 0.92 1.04 1.00 7660 5600
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).h1_women## Warning: There were 2 divergent transitions after warmup. Increasing adapt_delta
## above may help. See http://mc-stan.org/misc/warnings.html#divergent-transitions-
## after-warmup## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: dominance ~ f0 + pf + age + (1 | dataset)
## Data: vcs %>% filter(sex != "male") (Number of observations: 491)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 3)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.70 0.65 0.13 2.60 1.00 1703 1889
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.22 0.63 -1.02 1.34 1.00 2231 2138
## f0 -0.29 0.12 -0.53 -0.05 1.00 6646 4832
## pf -0.04 0.08 -0.19 0.10 1.00 5961 5006
## age 0.04 0.09 -0.15 0.22 1.00 6006 5404
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.97 0.03 0.91 1.03 1.00 6867 5480
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).equivalence_test(h1_f0bysex, ci = 0.89)## # Test for Practical Equivalence
##
## ROPE: [-0.10 0.10]
##
## Parameter | H0 | inside ROPE | 89% HDI
## ---------------------------------------------------
## Intercept | Undecided | 27.76 % | [-0.57 0.51]
## f0 | Rejected | 0.00 % | [-0.42 -0.11]
## sex1 | Rejected | 0.00 % | [-0.45 -0.11]
## age | Undecided | 87.75 % | [-0.06 0.14]
## f0.sex1 | Undecided | 76.63 % | [-0.13 0.18]SOI-R behavior, separate by sex
h1_men <- brm(behavior ~ f0 + pf + age + (1 | dataset), data = vcs %>% filter(sex == "male"),
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/behavior/h1_men") %>%
add_criterion("loo") %>%
add_criterion("bayes_R2") %>%
add_criterion("loo_R2") ## Warning: Rows containing NAs were excluded from the model.## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_women <- brm(behavior ~ f0 + pf + age + (1 | dataset), data = vcs %>% filter(sex != "male"),
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change", save_mevars = TRUE,
file = "models/behavior/h1_women") %>%
add_criterion("loo") %>%
add_criterion("bayes_R2") %>%
add_criterion("loo_R2") ## Warning: Rows containing NAs were excluded from the model.
## Warning: Argument 'save_mevars' is deprecated. Please use argument 'latent' in
## function 'save_pars()' instead.h1_f0bysex <- brm(behavior ~ f0*sex + age + (1 | dataset), data = vcs,
iter = iter, warmup = warmup, chains = chains, cores = chains,
prior = priors,
control = control, file_refit = "on_change",
file = "models/behavior/h1_f0bysex")## Warning: Rows containing NAs were excluded from the model.h1_men## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: behavior ~ f0 + pf + age + (1 | dataset)
## Data: vcs %>% filter(sex == "male") (Number of observations: 790)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 6)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.49 0.25 0.22 1.12 1.00 1819 3212
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.95 0.32 -1.57 -0.32 1.00 3843 3412
## f0 -0.29 0.12 -0.52 -0.07 1.00 7162 5398
## pf 0.03 0.11 -0.18 0.24 1.00 6331 5105
## age 0.25 0.07 0.12 0.39 1.00 6140 5067
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.95 0.02 0.91 1.00 1.00 7084 5565
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).h1_women## Family: gaussian
## Links: mu = identity; sigma = identity
## Formula: behavior ~ f0 + pf + age + (1 | dataset)
## Data: vcs %>% filter(sex != "male") (Number of observations: 1206)
## Samples: 4 chains, each with iter = 4000; warmup = 2000; thin = 1;
## total post-warmup samples = 8000
##
## Group-Level Effects:
## ~dataset (Number of levels: 6)
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sd(Intercept) 0.47 0.22 0.22 1.05 1.00 1777 2842
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -0.95 0.27 -1.48 -0.41 1.00 3170 3625
## f0 -0.16 0.07 -0.30 -0.03 1.00 7384 5290
## pf -0.03 0.05 -0.12 0.06 1.00 7324 5509
## age 0.42 0.07 0.28 0.54 1.00 6460 5647
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## sigma 0.89 0.02 0.85 0.92 1.00 7182 5552
##
## Samples were drawn using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).equivalence_test(h1_f0bysex, ci = 0.89)## # Test for Practical Equivalence
##
## ROPE: [-0.10 0.10]
##
## Parameter | H0 | inside ROPE | 89% HDI
## ---------------------------------------------------
## Intercept | Rejected | 0.00 % | [-1.22 -0.60]
## f0 | Rejected | 0.00 % | [-0.34 -0.13]
## sex1 | Rejected | 0.00 % | [-0.37 -0.13]
## age | Rejected | 0.00 % | [ 0.25 0.40]
## f0.sex1 | Undecided | 82.66 % | [-0.15 0.05]