play-pilot-summary
Rick O. Gilmore
2019-02-19 17:30:47
1 Purpose
This document summarizes the analyses conducted on the Play & Learning Across a Year (PLAY) project pilot data. The data are stored and shared at https://nyu.databrary.org/volume/444.
2 Set-up
knitr::opts_chunk$set(echo = TRUE)
# Load databraryapi package from github.com/PLAY-behaviorome/databraryapi
if (!require(databraryapi)) {
devtools::install_github("PLAY-behaviorome/databraryapi")
library(databraryapi)
}## Loading required package: databraryapi## Welcome to the databraryapi package# Load other libraries
library(tidyverse)## ── Attaching packages ─────────── tidyverse 1.2.1 ──## ✔ ggplot2 3.1.0 ✔ purrr 0.3.0
## ✔ tibble 2.0.1 ✔ dplyr 0.8.0.1
## ✔ tidyr 0.8.2 ✔ stringr 1.4.0
## ✔ readr 1.3.1 ✔ forcats 0.3.0## Warning: package 'tibble' was built under R version 3.5.2## Warning: package 'purrr' was built under R version 3.5.2## Warning: package 'dplyr' was built under R version 3.5.2## Warning: package 'stringr' was built under R version 3.5.2## ── Conflicts ────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()library(ggplot2)
## Variables for this particular volume and session.
play_vol_id <- 444
survey_data_session_name = "Parent report data"The Databrary volume is 444.
We need to log in to Databrary to have access to these data. If you have logged-in before from the machine you are running this report, then you will be prompted to enter your Databrary ID (email address). If you have not logged-in before, you will be prompted to enter your Databrary ID (email address) and your Databrary password.
databraryapi::login_db(params$db_account)## Login successful.## [1] TRUE3 Gather data
There are several data sources we need to gather and clean. Participant demographic data are stored in the Databrary spreadsheet. Parent report (survey) data are stored in the session with session_name == “Parent report data”. Let’s store that in a variable for later use: `r
3.1 Participant demographics
We’ll first gather the participant demographic data stored in the Databrary spreadsheet.
play_demog_raw <- databraryapi::download_session_csv(vol_id = play_vol_id)
if (is.null(play_demog_raw)) {
stop("No data downloaded from Databrary. Are you logged in?")
}
# Select relevant spreadsheet columns, filter those with actual session data, and rename variables to use underscores
play_demog <- play_demog_raw %>%
select(session_id, session_release, participant.gender,
participant.race, participant.ethnicity,
participant.gestational.age, participant.birth.weight,
participant.disability, participant.language, group.name,
context.language) %>%
filter(session_release %in% c('EXCERPTS', 'PRIVATE', 'SHARED')) %>%
rename(participant_gender = participant.gender,
participant_race = participant.race,
participant_ethnicity = participant.ethnicity,
participant_gestational_age = participant.gestational.age,
participant_birth_weight = participant.birth.weight,
participant_disability = participant.disability,
participant_language = participant.language,
group_name = group.name,
context_language = context.language)3.1.1 Structure of demographic data file
str(play_demog)## 'data.frame': 20 obs. of 11 variables:
## $ session_id : int 18801 18803 18805 18806 18807 18808 18810 18811 18813 18814 ...
## $ session_release : Factor w/ 4 levels "","EXCERPTS",..: 2 2 3 2 2 2 2 2 2 2 ...
## $ participant_gender : Factor w/ 3 levels "","Female","Male": 3 3 2 3 3 2 2 2 3 3 ...
## $ participant_race : Factor w/ 5 levels "","Asian","More than one",..: 4 5 3 5 5 5 5 5 5 5 ...
## $ participant_ethnicity : Factor w/ 4 levels "","Hispanic or Latino",..: 4 3 2 3 3 3 3 2 3 3 ...
## $ participant_gestational_age: num NA 40 39.5 40 40 38 40 40 41 41 ...
## $ participant_birth_weight : num NA 8.25 8 7 8.38 ...
## $ participant_disability : Factor w/ 4 levels "","Acid Reflux",..: 4 3 3 3 3 3 3 3 3 3 ...
## $ participant_language : Factor w/ 9 levels "","English","English, German",..: 9 6 9 2 2 2 7 4 3 2 ...
## $ group_name : Factor w/ 4 levels "","12 mos","18 mos",..: 3 3 3 3 4 3 3 2 3 3 ...
## $ context_language : Factor w/ 3 levels "","English","English, Spanish": 2 2 2 2 2 2 2 2 2 2 ...3.2 Parent report data
First, we need to list the data files available in the session where the parent report data are stored.
parent_rpt_session_id <- play_demog_raw %>%
filter(session_name == survey_data_session_name) %>%
select(session_id) %>%
as.numeric()Now, we list the data files in that session.
surveys <- list_assets_in_session(vol_id = play_vol_id,
session_id = parent_rpt_session_id) %>%
select(asset_id, name)
knitr::kable(surveys)| asset_id | name |
|---|---|
| 159177 | childcare |
| 117092 | locomotion |
| 116791 | child-birth |
| 116790 | family |
| 116789 | sleep |
| 116787 | language-exposure |
Let’s download each of these separately.
childcare_df <- read_csv_data_as_df(session_id = parent_rpt_session_id,
asset_id = surveys$asset_id[1])## No encoding supplied: defaulting to UTF-8.loco_df <- read_csv_data_as_df(session_id = parent_rpt_session_id, asset_id = surveys$asset_id[2])## No encoding supplied: defaulting to UTF-8.childbirth_df <- read_csv_data_as_df(session_id = parent_rpt_session_id,
asset_id = surveys$asset_id[3])## No encoding supplied: defaulting to UTF-8.## Warning in scan(file = file, what = what, sep = sep, quote = quote, dec =
## dec, : EOF within quoted stringfamily_df <- read_csv_data_as_df(session_id = parent_rpt_session_id,
asset_id = surveys$asset_id[4])## No encoding supplied: defaulting to UTF-8.sleep_df <- read_csv_data_as_df(session_id = parent_rpt_session_id,
asset_id = surveys$asset_id[5])## No encoding supplied: defaulting to UTF-8.lang_exp_df <- read_csv_data_as_df(session_id = parent_rpt_session_id,
asset_id = surveys$asset_id[6])## No encoding supplied: defaulting to UTF-8.3.2.1 Examining structure of parent report data files
3.2.1.1 Locomotor milestones
str(loco_df)## 'data.frame': 17 obs. of 7 variables:
## $ id : int 2 3 5 6 7 9 10 11 12 13 ...
## $ hkcrawl_onset_date: chr "8/13/15" "7/31/15" "5/2/15" "3/10/16" ...
## $ hkcrawl_onset_mos : num 7.53 7.03 8.48 12.59 7.2 ...
## $ walk_onset_date : chr "12/5/15" "11/30/15" "8/4/15" "7/9/16" ...
## $ walk_onset_mos : num 11.3 11.1 11.6 16.6 11.7 ...
## $ walk_onset_src : chr "phone" NA "phone" "babybook" ...
## $ Interview.Comments: chr "" "" "" "" ...3.2.1.2 Childbirth and health
str(childbirth_df)## 'data.frame': 11 obs. of 11 variables:
## $ id : int 2 3 5 6 7 9 10 11 12 13 ...
## $ bdate_child : chr "12/27/14" "12/29/14" "8/17/14" "2/21/15" ...
## $ born_duedate : chr "yes" "yes" "yes" "yes" ...
## $ child_age : num 18.3 18.3 24.3 18.2 17.8 ...
## $ agegroup : int 18 18 24 18 18 18 18 18 24 12 ...
## $ birth_weight_pounds : int 8 8 8 8 7 6 8 6 6 8 ...
## $ birth_weight_ounces : num 14 0.5 6 2 4 9 3 12 0 9 ...
## $ newborn_complications_yn : chr "no" "no" "no" "no" ...
## $ newborn_complications : chr NA NA NA NA ...
## $ illness_diagnosis_yn : chr "no" "no" "no" "no" ...
## $ illness_diagnosis_details: chr NA NA NA NA ...3.2.1.3 Family structure
str(family_df)## 'data.frame': 17 obs. of 21 variables:
## $ id : int 2 3 5 6 7 9 10 11 12 13 ...
## $ mom_bdate : chr "12/23/75" "1/9/81" "6/30/69" "1/23/82" ...
## $ mom_race : chr "5" "7" "5" "5" ...
## $ mom_ethnicity : chr "no" "yes" "no" "no" ...
## $ caregiver_country : int 4 4 1 1 1 4 1 1 1 1 ...
## $ mom_education : int 22 19 23 19 22 21 21 17 22 22 ...
## $ mom_working : chr "ft" "ft" "pt" "ft" ...
## $ mom_occupation : chr "Assistant professor" NA "Lawyer" "Editor for academic journal" ...
## $ mom_training : chr "no" "no" "no" "yes" ...
## $ caregiver_us_entry: chr "1/1/87" "1/1/89" NA NA ...
## $ partner_cohabitate: chr "yes" "yes" "yes" "yes" ...
## $ partner_race : chr "5" "7" "5" "5" ...
## $ partner_ethnicity : chr "no" "yes" "no" "no" ...
## $ partner_education : int 21 21 21 19 22 NA 21 17 23 21 ...
## $ partner_working : chr "ft" "ft" "ft" "ft" ...
## $ partner_training : chr "no" "yes" "no" "no" ...
## $ partner_occupation: chr "Finance, Asset Manager" NA "Manager" "Sales manager" ...
## $ partner_dob : chr NA "9/14/82" "3/29/80" "10/24/85" ...
## $ family_sib_dob : chr NA NA NA NA ...
## $ brothers : int NA NA NA NA NA NA NA NA 0 0 ...
## $ siblings : int NA NA NA NA NA NA NA NA 0 0 ...3.2.1.4 Sleep hygiene
str(sleep_df)## 'data.frame': 17 obs. of 5 variables:
## $ id : int 2 3 5 6 7 9 10 11 12 13 ...
## $ sleep.time : chr "20:30:00" "20:45:00" "19:30:00" "19:45:00" ...
## $ wake.time : chr "07:00:00" "07:20:00" "06:30:00" "07:15:00" ...
## $ naps.hours : num 2 2 2.5 2 2 2 2.5 1.5 2 2.5 ...
## $ sleeping.loc: chr "crib_separate" "crib_separate" "toddlerbed" "crib_separate" ...3.2.1.5 Childcare arrangements
str(childcare_df)## 'data.frame': 17 obs. of 5 variables:
## $ id : int 2 3 5 6 7 9 10 11 12 13 ...
## $ childcare_alt : chr "person" "center" "both" "center" ...
## $ childcare_hours : int 60 40 52 50 NA 50 29 21 35 20 ...
## $ childcare_other_children: int 1 11 15 9 NA 0 11 3 9 0 ...
## $ childcare_init_age : num 0 8 12 3 NA 4 1.5 14 1 6 ...3.2.1.6 Language exposure
str(lang_exp_df)## 'data.frame': 66 obs. of 3 variables:
## $ id : int 2 2 3 5 6 9 10 11 11 12 ...
## $ language : chr "Russian" "Armenian" "English" "English" ...
## $ exposure_context: chr "childcare" "childcare" "childcare" "childcare" ...4 Plot data
Let’s set some plot thematic elements.
play.palette <- scale_fill_manual(values=c("blue2", "firebrick2", "chartreuse2", "darkorchid2"))
play.theme <-
theme_classic() +
theme(legend.position = "bottom",
legend.title = element_blank(),
legend.text = element_text(size = rel(1.2)),
axis.title = element_text(size = rel(1.5), face ="bold"),
axis.text.x = element_text(size = rel(1.2)),
axis.text.y = element_text(size = rel(1.2)),
axis.line = element_blank(),
axis.title.y = element_blank())4.1 Participant demographics
play_demog %>%
ggplot(.) +
aes(x = participant_gestational_age, y = participant_birth_weight, color = participant_race) +
geom_point()## Warning: Removed 1 rows containing missing values (geom_point).
4.2 Parent surveys
These are some illustrative plots.
4.2.1 Sleep hygiene
We have to convert the sleep.time and wake.time variables so that they can be plotted. It looks like the lubridate package is a good option.
# sleep <- lubridate::ymd_hms("2001-01-01 20:30:00")
# awake <- lubridate::ymd_hms("2001-01-02 07:00:00")
# (night_hrs <- awake-sleep)
# Helper functions for conversion
sleep_time <- function(hhmmss){
lubridate::ymd_hms(paste0("2001-01-01 ", hhmmss))
}
wake_time <- function(hhmmss){
# One date later than sleep time
lubridate::ymd_hms(paste0("2001-01-02 ", hhmmss))
}
time_asleep <- function(sleep_tm, awake_tm){
wake_time(awake_tm) - sleep_time(sleep_tm)
}
sleep_df <- sleep_df %>%
mutate(sleep_tm = sleep_time(sleep.time),
wake_tm = wake_time(wake.time),
nightsleep_hrs = wake_tm - sleep_tm)Now, we can plot.
sleep_df %>%
ggplot(.) +
aes(x = nightsleep_hrs) +
geom_histogram(bins = 10) +
play.theme## Don't know how to automatically pick scale for object of type difftime. Defaulting to continuous.
4.2.2 Childcare
childcare_df %>%
ggplot(.) +
aes(x = childcare_hours, fill = childcare_alt) +
geom_histogram(bins = 30) +
play.theme## Warning: Removed 2 rows containing non-finite values (stat_bin).
4.2.3 Language exposure
lang_exp_df %>%
ggplot(.) +
aes(x = language) +
facet_grid(exposure_context ~ .) +
geom_histogram(stat='count') +
play.theme## Warning: Ignoring unknown parameters: binwidth, bins, pad
4.2.4 Locomotion onset
loco_df %>%
gather(key = milestone, value = age.mos,
hkcrawl_onset_mos, walk_onset_mos) %>%
mutate(milestone = factor(milestone,
labels = c("crawl", "walk"))) %>%
ggplot() +
aes(x = age.mos, y = id) +
geom_point(aes(shape = milestone, color = milestone)) +
geom_line(aes(group = id)) +
xlab("Age (months)") +
ylab("Participant ID") +
geom_rug(aes(x = age.mos, group = milestone, color = milestone),
sides = "b") +
play.theme## Warning: Removed 3 rows containing missing values (geom_point).## Warning: Removed 3 rows containing missing values (geom_path).