alt calc
anjie
12/16/2020
same anlaysis smaller set
alt_raw_id <- alt_raw %>%
filter(alternative_calc == "between") %>%
select(unique_id) %>%
pull
all_paper_trans <- read_csv(here("data/processed/syntactic_bootstrapping_tidy_data.csv")) %>%
filter(unique_id %in% alt_raw_id) %>%
filter(sentence_structure == "transitive")## Parsed with column specification:
## cols(
## .default = col_character(),
## se = col_double(),
## mean_age = col_double(),
## productive_vocab_mean = col_double(),
## productive_vocab_median = col_double(),
## n_train_test_pair = col_double(),
## n_test_trial_per_pair = col_double(),
## n_repetitions_sentence = col_double(),
## n_repetitions_video = col_double(),
## inclusion_certainty = col_double(),
## publication_year = col_double(),
## n_1 = col_double(),
## x_1 = col_double(),
## x_2 = col_double(),
## x_2_raw = col_double(),
## sd_1 = col_double(),
## sd_2 = col_double(),
## sd_2_raw = col_double(),
## t = col_double(),
## d = col_double(),
## d_calc = col_double()
## # ... with 3 more columns
## )## See spec(...) for full column specifications.generate_funnel_plot(all_paper_trans)
alpha = 0.05
model <- rma.mv(d_calc,
V = d_var_calc,
random = ~ 1 | short_cite/same_infant/row_id,
method = "REML",
data = all_paper_trans)
this_moderator_estimate <- model$b[1]
this_moderator_SE <- model$se[1]
this_moderator_estimate.cil <- model$ci.lb[1]
this_moderator_estimate.cih <- model$ci.ub[1]
this_moderator_z <- model$zval[1]
this_moderator_p <- model$pval[1]
abbreviate_label <- function(original_label){
label_vec <- unlist(strsplit(original_label, ","))
# get the first label
first_author_name <- label_vec[[1]]
year <- gsub("(-).*","",tail(label_vec, n=1))
number_label <- gsub(".*-","",tail(label_vec, n=1))
abbreviated_label <- paste0(first_author_name, " et al.,", year, "-", number_label)
#print(number_label)
#print(abbreviated_label)
}
individual_data <- all_paper_trans %>%
select(short_cite, unique_id,d_calc,d_var_calc, n_1, plot_label) %>%
rowwise() %>%
mutate(cil = d_calc - qnorm(alpha / 2, lower.tail = FALSE) * sqrt(d_var_calc),
cil = case_when(
(cil < -8) ~ -8, # truncate error bar for visibility reason
TRUE ~ cil
),
ciu = d_calc +
qnorm(alpha / 2, lower.tail = FALSE) * sqrt(d_var_calc),
ciu = case_when(
(ciu > 3 ) ~ 3, # truncate error bar for visibility reason
TRUE ~ ciu
),
sentence_structure = "individual",
meta = "no",
label_color = "black",
print_full = paste(round(d_calc,2), " [",round(cil,2),", ",round(ciu,2), "]", sep = ""),
plot_label = case_when(
str_count(plot_label,",") >= 4 ~ abbreviate_label(plot_label),
TRUE ~ plot_label
)
)
cumulative_data <- tibble_row(
short_cite = "Meta-Analytic Effect Size",
plot_label = "Meta-Analytic Effect Size",
d_calc = this_moderator_estimate,
d_var_calc = NA,
n_1 = 99,
expt_num = "",
expt_condition = "",
cil = this_moderator_estimate.cil,
ciu = this_moderator_estimate.cih,
sentence_structure = "cumulative",
print_full = paste(round(d_calc,2), " [",round(cil,2),", ",round(ciu,2), "]", sep = ""),
meta = "yes",
label_color = "red"
)
forest_data_all_trans <- bind_rows(individual_data, cumulative_data) forest_data_all_trans$sentence_structure <- as.factor(forest_data_all_trans$sentence_structure)
forest_data_all_trans$meta <- as.factor(forest_data_all_trans$meta)
forest_data_all_trans <- forest_data_all_trans %>%
rowid_to_column() %>%
mutate(
rowid = if_else(rowid == 0, 99, as.double(rowid)) #to always put the MA ES at bottom
) %>%
group_by(sentence_structure) %>% arrange(-rowid, .by_group = TRUE)
forest_data_all_trans$plot_label <- factor(forest_data_all_trans$plot_label, levels = forest_data_all_trans$plot_label)
mm_to_point = 18/6.5
label_size = 5
x_axis_title <- expression(paste("Cohen\'s ", italic('d')))
# set the neighbourhood levels in the order the occur in the data frame
label_colors <- forest_data_all_trans$label_color[order(forest_data_all_trans$plot_label)]
forest_data_all_trans %>% # First sort by val. This sort the dataframe but NOT the factor levels
ggplot(aes(x = plot_label, y = d_calc)) +
geom_hline(aes(yintercept = 0), color = "gray44",linetype = 2, size =.3) +
geom_hline(aes(yintercept = filter(forest_data_all_trans, sentence_structure == "cumulative")$d_calc),
color = "red", linetype = 2, size = .3) +
geom_point(data = forest_data_all_trans,
aes(size=(n_1/100), shape = sentence_structure, color = sentence_structure)) +
scale_color_manual(breaks = c("cumulative", "individual"),
values = c("red", "black"))+
scale_size(guide = 'none', range = c(0.3,3)) +
scale_shape_manual(breaks = c("cumulative", "individual"),
values=c(18,16, 17)) +
guides(colour = guide_legend(override.aes = list(size=3))) +
geom_linerange(aes(ymin = cil, ymax = ciu, color = sentence_structure), show.legend = FALSE) +
geom_segment(aes(x = plot_label, y = d_calc, xend = plot_label, yend = ciu),
linejoin = "round",
lineend = "round",
size = 0.2,
arrow = arrow(length = unit(0.02, "inches")),
data = filter(forest_data_all_trans,ciu == 3))+
geom_text(aes(label = print_full, x = plot_label, y = 4.2),
size = label_size / mm_to_point, colour = label_colors) +
scale_y_continuous(breaks = seq(-10, 7, 1))+
coord_cartesian(clip = 'on') +
coord_flip() +
ylim(-1.8, 5)+ #doesn't seem to help a lot
ylab(x_axis_title) +
labs(color = "Effect Size Type",shape = "Effect Size Type") + # merge two legends
theme(text = element_text(size=label_size),
legend.position="bottom",
legend.text=element_text(size=label_size*1.5),
plot.margin = unit(c(1,1,1,1), "lines"),
legend.title = element_blank(),
panel.background = element_blank(),
#panel.background = element_rect(fill = "white", colour = "grey50"),
axis.title.y = element_blank(),
axis.title.x = element_text(size=label_size*1.5),
axis.text.y = element_text(colour = label_colors,
size = label_size),
axis.text.x = element_text(size = label_size * 2)) ## Coordinate system already present. Adding new coordinate system, which will replace the existing one.## Scale for 'y' is already present. Adding another scale for 'y', which will
## replace the existing scale.## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.
alt_calculated <- alt_raw %>%
filter(alternative_calc == "between") %>%
mutate(
pooled_SD = sqrt(((n_1 - 1) * sd_1 ^ 2 + (n_2 - 1) * sd_2 ^ 2) / (n_1 + n_2 - 2)),
d_calc = (x_1 - x_2) / pooled_SD,
d_var_calc = ((n_1 + n_2) / (n_1 * n_2)) + (d_calc ^ 2 / (2 * (n_1 + n_2)))
)model <- rma.mv(d_calc,
V = d_var_calc,
random = ~ 1 | short_cite/unique_infant/rowid,
method = "REML",
data = alt_calculated)
this_moderator_estimate <- model$b[1]
this_moderator_SE <- model$se[1]
this_moderator_estimate.cil <- model$ci.lb[1]
this_moderator_estimate.cih <- model$ci.ub[1]
this_moderator_z <- model$zval[1]
this_moderator_p <- model$pval[1]alpha = 0.05
abbreviate_label <- function(original_label){
label_vec <- unlist(strsplit(original_label, ","))
# get the first label
first_author_name <- label_vec[[1]]
year <- gsub("(-).*","",tail(label_vec, n=1))
number_label <- gsub(".*-","",tail(label_vec, n=1))
abbreviated_label <- paste0(first_author_name, " et al.,", year, "-", number_label)
#print(number_label)
#print(abbreviated_label)
}
individual_data <- alt_calculated %>%
select(short_cite, unique_id,d_calc,d_var_calc, n_1, plot_label) %>%
rowwise() %>%
mutate(cil = d_calc - qnorm(alpha / 2, lower.tail = FALSE) * sqrt(d_var_calc),
cil = case_when(
(cil < -8) ~ -8, # truncate error bar for visibility reason
TRUE ~ cil
),
ciu = d_calc +
qnorm(alpha / 2, lower.tail = FALSE) * sqrt(d_var_calc),
ciu = case_when(
(ciu > 3 ) ~ 3, # truncate error bar for visibility reason
TRUE ~ ciu
),
sentence_structure = "individual",
meta = "no",
label_color = "black",
print_full = paste(round(d_calc,2), " [",round(cil,2),", ",round(ciu,2), "]", sep = ""),
plot_label = case_when(
str_count(plot_label,",") >= 4 ~ abbreviate_label(plot_label),
TRUE ~ plot_label
)
)
cumulative_data <- tibble_row(
short_cite = "Meta-Analytic Effect Size",
plot_label = "Meta-Analytic Effect Size",
d_calc = this_moderator_estimate,
d_var_calc = NA,
n_1 = 99,
expt_num = "",
expt_condition = "",
cil = this_moderator_estimate.cil,
ciu = this_moderator_estimate.cih,
sentence_structure = "cumulative",
print_full = paste(round(d_calc,2), " [",round(cil,2),", ",round(ciu,2), "]", sep = ""),
meta = "yes",
label_color = "red"
)
forest_data <- bind_rows(individual_data, cumulative_data)Forest plot alt calc
the weird ones Gertner & Fisher 2012 are weird. original reports transitive vs agent -1st and agent-1st vs patient-1st. the values are calculated from transitive vs agent -1st.
image
forest_data$sentence_structure <- as.factor(forest_data$sentence_structure)
forest_data$meta <- as.factor(forest_data$meta)
forest_data <- forest_data %>%
rowid_to_column() %>%
mutate(
rowid = if_else(rowid == 0, 99, as.double(rowid)) #to always put the MA ES at bottom
) %>%
group_by(sentence_structure) %>% arrange(-rowid, .by_group = TRUE)
forest_data$plot_label <- factor(forest_data$plot_label, levels = forest_data$plot_label)
mm_to_point = 18/6.5
label_size = 5
x_axis_title <- expression(paste("Cohen\'s ", italic('d')))
# set the neighbourhood levels in the order the occur in the data frame
label_colors <- forest_data$label_color[order(forest_data$plot_label)]
forest_data %>% # First sort by val. This sort the dataframe but NOT the factor levels
ggplot(aes(x = plot_label, y = d_calc)) +
geom_hline(aes(yintercept = 0), color = "gray44",linetype = 2, size =.3) +
geom_hline(aes(yintercept = filter(forest_data, sentence_structure == "cumulative")$d_calc),
color = "red", linetype = 2, size = .3) +
geom_point(data = forest_data,
aes(size=(n_1/100), shape = sentence_structure, color = sentence_structure)) +
scale_color_manual(breaks = c("cumulative", "individual"),
values = c("red", "black"))+
scale_size(guide = 'none', range = c(0.3,3)) +
scale_shape_manual(breaks = c("cumulative", "individual"),
values=c(18,16)) +
guides(colour = guide_legend(override.aes = list(size=3))) +
geom_linerange(aes(ymin = cil, ymax = ciu, color = sentence_structure), show.legend = FALSE) +
geom_segment(aes(x = plot_label, y = d_calc, xend = plot_label, yend = ciu),
linejoin = "round",
lineend = "round",
size = 0.2,
arrow = arrow(length = unit(0.02, "inches")),
data = filter(forest_data,ciu == 3))+
geom_text(aes(label = print_full, x = plot_label, y = 4.2),
size = label_size / mm_to_point, colour = label_colors) +
#scale_y_continuous(breaks = seq(-10, 7, 1))+
coord_cartesian(clip = 'on') +
coord_flip() +
ylim(-3, 5)+ #doesn't seem to help a lot
ylab(x_axis_title) +
labs(color = "Effect Size Type",shape = "Effect Size Type") + # merge two legends
theme(text = element_text(size=label_size),
legend.position="bottom",
legend.text=element_text(size=label_size*1.5),
plot.margin = unit(c(1,1,1,1), "lines"),
legend.title = element_blank(),
panel.background = element_blank(),
#panel.background = element_rect(fill = "white", colour = "grey50"),
axis.title.y = element_blank(),
axis.title.x = element_text(size=label_size*1.5),
axis.text.y = element_text(colour = label_colors,
size = label_size),
axis.text.x = element_text(size = label_size * 2)) ## Coordinate system already present. Adding new coordinate system, which will replace the existing one.## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.
generate_funnel_plot <- function(data){
CRIT_95 = 1.96
ma_data <- data
model <- rma.mv(d_calc ~ 1, d_var_calc,
random = ~ 1 | short_cite/unique_infant/rowid,
method = "REML",
data=ma_data)
predicted_val <- predict.rma(model)
intercept_in_model <- model$beta[1]
ma_funnel <- ma_data %>%
mutate(
se = sqrt(d_var_calc),
es = d_calc,
center = intercept_in_model,
lower_lim = max(se) + .05 * max(se),
type = "non_moderated"
) %>%
select(es, se, center, lower_lim)
funnel_shape_wide <- ma_funnel %>%
mutate(x1 = (center-lower_lim * CRIT_95)[1],
x2 = center[1],
x3 = center[1] + lower_lim[1] * CRIT_95,
y1 = -lower_lim[1],
y2 = 0,
y3 = -lower_lim[1])
funnel95_data_x = funnel_shape_wide %>%
select(dplyr::contains("x")) %>%
gather("coordx", "x", 1:3) %>%
arrange(coordx) %>%
select(-coordx)
funnel95_data_Y = funnel_shape_wide %>%
select(dplyr::contains("y")) %>%
gather("coordy", "y", 1:3) %>%
arrange(coordy) %>%
select(-coordy)
funnel95.data = bind_cols(funnel95_data_x, funnel95_data_Y)
ggplot(ma_funnel, aes(x = es, y = -se)) +
xlab(expression(paste("Effect Size (Cohen's ", italic(d), ")"))) +
ylab("Standard Error\n") +
geom_polygon(aes(x = x, y = y),
data = funnel95.data,
fill = "grey90") +
geom_vline(aes(xintercept=intercept_in_model),
linetype = "dashed", color = "red", size = .8, data = funnel_shape_wide) +
geom_vline(xintercept = 0, linetype = "dashed", color = "grey44", size = .8) +
scale_y_continuous(labels = function(x){abs(x)}) +
geom_point(data = ma_funnel, size = 1.5, alpha = .7) +
theme(text = element_text(size = 11),
panel.grid.major = element_line(colour = "white", size = 0.2),
panel.grid.minor = element_line(colour = "white", size = 0.5),
strip.text.x = element_text(size = 9),
strip.background = element_rect(fill="white"))
}alt only funnel plot
generate_funnel_plot(alt_calculated)
alt only reg test
reg_result <- rma.mv(d_calc ~ sqrt(d_var_calc), d_var_calc,
random = ~ 1 | short_cite/unique_infant/rowid, data=alt_calculated) # access zval and pval from here
reg_z <- round(reg_result$zval[2], digits = 2)
p_val <- reg_result$pval[2]
reg_result##
## Multivariate Meta-Analysis Model (k = 18; method: REML)
##
## Variance Components:
##
## estim sqrt nlvls fixed factor
## sigma^2.1 0.3023 0.5498 7 no short_cite
## sigma^2.2 0.0000 0.0000 18 no short_cite/unique_infant
## sigma^2.3 0.0000 0.0000 18 no short_cite/unique_infant/rowid
##
## Test for Residual Heterogeneity:
## QE(df = 16) = 27.9815, p-val = 0.0318
##
## Test of Moderators (coefficient 2):
## QM(df = 1) = 2.4686, p-val = 0.1161
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## intrcpt -0.4207 0.8946 -0.4703 0.6382 -2.1741 1.3326
## sqrt(d_var_calc) 2.9689 1.8896 1.5712 0.1161 -0.7347 6.6725
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1non alt studies vanilla way (vs. chance)
alt_calc_paper <- alt_raw %>% filter(alternative_calc == "between") %>% pull(unique_id)
all_paper <- read_csv(here("data/processed/syntactic_bootstrapping_tidy_data.csv")) %>%
filter(!(unique_id %in% alt_calc_paper))## Parsed with column specification:
## cols(
## .default = col_character(),
## se = col_double(),
## mean_age = col_double(),
## productive_vocab_mean = col_double(),
## productive_vocab_median = col_double(),
## n_train_test_pair = col_double(),
## n_test_trial_per_pair = col_double(),
## n_repetitions_sentence = col_double(),
## n_repetitions_video = col_double(),
## inclusion_certainty = col_double(),
## publication_year = col_double(),
## n_1 = col_double(),
## x_1 = col_double(),
## x_2 = col_double(),
## x_2_raw = col_double(),
## sd_1 = col_double(),
## sd_2 = col_double(),
## sd_2_raw = col_double(),
## t = col_double(),
## d = col_double(),
## d_calc = col_double()
## # ... with 3 more columns
## )## See spec(...) for full column specifications.all_model <- rma.mv(d_calc,
V = d_var_calc,
random = ~ 1 | short_cite/same_infant/row_id,
method = "REML",
data = all_paper)
this_moderator_estimate <- all_model$b[1]
this_moderator_SE <- all_model$se[1]
this_moderator_estimate.cil <- all_model$ci.lb[1]
this_moderator_estimate.cih <- all_model$ci.ub[1]
this_moderator_z <- all_model$zval[1]
this_moderator_p <- all_model$pval[1]individual_data_all <- all_paper %>%
select(short_cite, unique_id,d_calc,d_var_calc, n_1, plot_label) %>%
rowwise() %>%
mutate(cil = d_calc - qnorm(alpha / 2, lower.tail = FALSE) * sqrt(d_var_calc),
cil = case_when(
(cil < -8) ~ -8, # truncate error bar for visibility reason
TRUE ~ cil
),
ciu = d_calc +
qnorm(alpha / 2, lower.tail = FALSE) * sqrt(d_var_calc),
ciu = case_when(
(ciu > 3 ) ~ 3, # truncate error bar for visibility reason
TRUE ~ ciu
),
sentence_structure = "individual",
meta = "no",
label_color = "black",
print_full = paste(round(d_calc,2), " [",round(cil,2),", ",round(ciu,2), "]", sep = ""),
plot_label = case_when(
str_count(plot_label,",") >= 4 ~ abbreviate_label(plot_label),
TRUE ~ plot_label
)
)
cumulative_data_all <- tibble_row(
short_cite = "Meta-Analytic Effect Size",
plot_label = "Meta-Analytic Effect Size",
d_calc = this_moderator_estimate,
d_var_calc = NA,
n_1 = 99,
expt_num = "",
expt_condition = "",
cil = this_moderator_estimate.cil,
ciu = this_moderator_estimate.cih,
sentence_structure = "cumulative",
print_full = paste(round(d_calc,2), " [",round(cil,2),", ",round(ciu,2), "]", sep = ""),
meta = "yes",
label_color = "red"
)
forest_data_all <- bind_rows(individual_data_all, cumulative_data_all)non-alt forest plot
forest_data_all$sentence_structure <- as.factor(forest_data_all$sentence_structure)
forest_data_all$meta <- as.factor(forest_data_all$meta)
forest_data_all <- forest_data_all %>%
rowid_to_column() %>%
mutate(
rowid = if_else(rowid == 0, 99, as.double(rowid)) #to always put the MA ES at bottom
) %>%
group_by(sentence_structure) %>% arrange(-rowid, .by_group = TRUE)
forest_data_all$plot_label <- factor(forest_data_all$plot_label, levels = forest_data_all$plot_label)
mm_to_point = 18/6.5
label_size = 5
x_axis_title <- expression(paste("Cohen\'s ", italic('d')))
# set the neighbourhood levels in the order the occur in the data frame
label_colors <- forest_data_all$label_color[order(forest_data_all$plot_label)]
forest_data_all %>% # First sort by val. This sort the dataframe but NOT the factor levels
ggplot(aes(x = plot_label, y = d_calc)) +
geom_hline(aes(yintercept = 0), color = "gray44",linetype = 2, size =.3) +
geom_hline(aes(yintercept = filter(forest_data_all, sentence_structure == "cumulative")$d_calc),
color = "red", linetype = 2, size = .3) +
geom_point(data = forest_data_all,
aes(size=(n_1/100), shape = sentence_structure, color = sentence_structure)) +
scale_color_manual(breaks = c("cumulative", "individual"),
values = c("red", "black"))+
scale_size(guide = 'none', range = c(0.3,3)) +
scale_shape_manual(breaks = c("cumulative", "individual"),
values=c(18,16, 17)) +
guides(colour = guide_legend(override.aes = list(size=3))) +
geom_linerange(aes(ymin = cil, ymax = ciu, color = sentence_structure), show.legend = FALSE) +
geom_segment(aes(x = plot_label, y = d_calc, xend = plot_label, yend = ciu),
linejoin = "round",
lineend = "round",
size = 0.2,
arrow = arrow(length = unit(0.02, "inches")),
data = filter(forest_data_all,ciu == 3))+
geom_text(aes(label = print_full, x = plot_label, y = 4.2),
size = label_size / mm_to_point, colour = label_colors) +
scale_y_continuous(breaks = seq(-10, 7, 1))+
coord_cartesian(clip = 'on') +
coord_flip() +
ylim(-1.8, 5)+ #doesn't seem to help a lot
ylab(x_axis_title) +
labs(color = "Effect Size Type",shape = "Effect Size Type") + # merge two legends
theme(text = element_text(size=label_size),
legend.position="bottom",
legend.text=element_text(size=label_size*1.5),
plot.margin = unit(c(1,1,1,1), "lines"),
legend.title = element_blank(),
panel.background = element_blank(),
#panel.background = element_rect(fill = "white", colour = "grey50"),
axis.title.y = element_blank(),
axis.title.x = element_text(size=label_size*1.5),
axis.text.y = element_text(colour = label_colors,
size = label_size),
axis.text.x = element_text(size = label_size * 2)) ## Coordinate system already present. Adding new coordinate system, which will replace the existing one.## Scale for 'y' is already present. Adding another scale for 'y', which will
## replace the existing scale.## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.
non alt funnel plot
generate_funnel_plot <- function(data){
CRIT_95 = 1.96
ma_data <- data
model <- rma.mv(d_calc ~ 1, d_var_calc,
random = ~ 1 | short_cite/same_infant/row_id,
method = "REML",
data=ma_data)
predicted_val <- predict.rma(model)
intercept_in_model <- model$beta[1]
ma_funnel <- ma_data %>%
mutate(
se = sqrt(d_var_calc),
es = d_calc,
center = intercept_in_model,
lower_lim = max(se) + .05 * max(se),
type = "non_moderated"
) %>%
select(es, se, center, lower_lim)
funnel_shape_wide <- ma_funnel %>%
mutate(x1 = (center-lower_lim * CRIT_95)[1],
x2 = center[1],
x3 = center[1] + lower_lim[1] * CRIT_95,
y1 = -lower_lim[1],
y2 = 0,
y3 = -lower_lim[1])
funnel95_data_x = funnel_shape_wide %>%
select(dplyr::contains("x")) %>%
gather("coordx", "x", 1:3) %>%
arrange(coordx) %>%
select(-coordx)
funnel95_data_Y = funnel_shape_wide %>%
select(dplyr::contains("y")) %>%
gather("coordy", "y", 1:3) %>%
arrange(coordy) %>%
select(-coordy)
funnel95.data = bind_cols(funnel95_data_x, funnel95_data_Y)
ggplot(ma_funnel, aes(x = es, y = -se)) +
xlab(expression(paste("Effect Size (Cohen's ", italic(d), ")"))) +
ylab("Standard Error\n") +
geom_polygon(aes(x = x, y = y),
data = funnel95.data,
fill = "grey90") +
geom_vline(aes(xintercept=intercept_in_model),
linetype = "dashed", color = "red", size = .8, data = funnel_shape_wide) +
geom_vline(xintercept = 0, linetype = "dashed", color = "grey44", size = .8) +
scale_y_continuous(labels = function(x){abs(x)}) +
geom_point(data = ma_funnel, size = 1.5, alpha = .7) +
theme(text = element_text(size = 11),
panel.grid.major = element_line(colour = "white", size = 0.2),
panel.grid.minor = element_line(colour = "white", size = 0.5),
strip.text.x = element_text(size = 9),
strip.background = element_rect(fill="white"))
}
generate_funnel_plot(all_paper)
non alt reg test
reg_result <- rma.mv(d_calc ~ sqrt(d_var_calc), d_var_calc,
random = ~ 1 | short_cite/same_infant/row_id, data=all_paper) # access zval and pval from here
reg_z <- round(reg_result$zval[2], digits = 2)
p_val <- reg_result$pval[2]
reg_result##
## Multivariate Meta-Analysis Model (k = 20; method: REML)
##
## Variance Components:
##
## estim sqrt nlvls fixed factor
## sigma^2.1 0.1740 0.4171 10 no short_cite
## sigma^2.2 0.0379 0.1948 18 no short_cite/same_infant
## sigma^2.3 0.0000 0.0000 20 no short_cite/same_infant/row_id
##
## Test for Residual Heterogeneity:
## QE(df = 18) = 79.7451, p-val < .0001
##
## Test of Moderators (coefficient 2):
## QM(df = 1) = 0.1827, p-val = 0.6691
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## intrcpt 0.4687 0.9107 0.5147 0.6068 -1.3162 2.2537
## sqrt(d_var_calc) -1.5719 3.6779 -0.4274 0.6691 -8.7805 5.6367
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1published or not is not moderator
PUB_PATH <- here("exploratory_analyses/06_alt_ES/published_moderator.csv")
pub_data <- read_csv(PUB_PATH)## Parsed with column specification:
## cols(
## .default = col_character(),
## se = col_double(),
## mean_age = col_double(),
## productive_vocab_mean = col_double(),
## productive_vocab_median = col_double(),
## n_train_test_pair = col_double(),
## n_test_trial_per_pair = col_double(),
## n_repetitions_sentence = col_double(),
## n_repetitions_video = col_double(),
## inclusion_certainty = col_double(),
## publication_year = col_double(),
## n_1 = col_double(),
## x_1 = col_double(),
## x_2 = col_double(),
## x_2_raw = col_double(),
## sd_1 = col_double(),
## sd_2 = col_double(),
## sd_2_raw = col_double(),
## t = col_double(),
## d = col_double(),
## d_calc = col_double()
## # ... with 3 more columns
## )## See spec(...) for full column specifications.pub_moderator <- rma.mv(d_calc ~ published, d_var_calc,
random = ~ 1 | short_cite/same_infant/row_id, data=pub_data) #
pub_moderator##
## Multivariate Meta-Analysis Model (k = 60; method: REML)
##
## Variance Components:
##
## estim sqrt nlvls fixed factor
## sigma^2.1 0.1096 0.3310 17 no short_cite
## sigma^2.2 0.1090 0.3302 58 no short_cite/same_infant
## sigma^2.3 0.0000 0.0000 60 no short_cite/same_infant/row_id
##
## Test for Residual Heterogeneity:
## QE(df = 58) = 189.5166, p-val < .0001
##
## Test of Moderators (coefficient 2):
## QM(df = 1) = 0.3659, p-val = 0.5452
##
## Model Results:
##
## estimate se zval pval ci.lb ci.ub
## intrcpt 0.3499 0.2162 1.6183 0.1056 -0.0739 0.7737
## publishedyes -0.1499 0.2477 -0.6049 0.5452 -0.6354 0.3357
##
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1