1_data_import
Laura Botzet & Ruben Arslan
Data Wrangling
Helper
Packages and functions used
source("0_helpers.R")##
## Attaching package: 'formr'## The following object is masked from 'package:rmarkdown':
##
## word_document##
## Attaching package: 'lubridate'## The following object is masked from 'package:base':
##
## date## Loading required package: carData## lattice theme set by effectsTheme()
## See ?effectsTheme for details.##
## Attaching package: 'data.table'## The following objects are masked from 'package:lubridate':
##
## hour, isoweek, mday, minute, month, quarter, second, wday, week, yday, year## The following objects are masked from 'package:formr':
##
## first, last## Loading required package: Matrix##
## Attaching package: 'lmerTest'## The following object is masked from 'package:lme4':
##
## lmer## The following object is masked from 'package:stats':
##
## step##
## Attaching package: 'psych'## The following objects are masked from 'package:ggplot2':
##
## %+%, alpha## This is lavaan 0.6-3## lavaan is BETA software! Please report any bugs.##
## Attaching package: 'lavaan'## The following object is masked from 'package:psych':
##
## cor2cov## Loading required package: lattice## Loading required package: survival## Loading required package: Formula##
## Attaching package: 'Hmisc'## The following object is masked from 'package:psych':
##
## describe## The following objects are masked from 'package:base':
##
## format.pval, units##
## Attaching package: 'tidyr'## The following object is masked from 'package:Matrix':
##
## expand##
## Attaching package: 'dplyr'## The following objects are masked from 'package:Hmisc':
##
## src, summarize## The following objects are masked from 'package:data.table':
##
## between, first, last## The following objects are masked from 'package:lubridate':
##
## intersect, setdiff, union## The following objects are masked from 'package:formr':
##
## first, last## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union##
## Attaching package: 'codebook'## The following object is masked from 'package:psych':
##
## bfi## The following objects are masked from 'package:formr':
##
## aggregate_and_document_scale, asis_knit_child, expired, paste.knit_asis##
## Attaching package: 'effsize'## The following object is masked from 'package:psych':
##
## cohen.dImport data
Analyses are based on the Indonesian Family Life Survey (IFLS). All data is retrieved from the RAND foundation.
We included the following data sets:
- IFLS Wave 1
- buk4ch1
- buk4kw2
- bukkar2
- IFLS Wave 2
- b4_ch1
- b4_kw2
- b4_cov
- IFLS Wave 3
- b4_ch1
- b4_kw3
- b4_cov
- IFLS Wave 4
- b4_ch1
- b4_kw2
- b4_cov
- bek_ek1
- bek_ek2
- IFLS Wave 5
- bk_ar1
- bk_sc1
- ptrack
- b4_ch1
- b4_kw3
- b4_cov
- ek_ek2
- ek_ek1
- b3b_cob
- b3b_co1
- b3b_psn
- b3a_si
- b3a_dl1
- b3a_dl3
- b3a_dl4
- b3a_tk2
- b3b_km
### Informations about individuals living in the household in 2014/2015
## All Individuals living in the household
bk_ar1 = read_dta("data/hh14_all_dta/bk_ar1.dta") # Book K, Section ar
# compute father pidlink
bk_ar1 = left_join(bk_ar1, bk_ar1 %>%
select(hhid14_9, pid14, pidlink) %>%
rename(ar10 = pid14, father_pidlink = pidlink), by = c("hhid14_9", "ar10"))
# compute mother pidlink
bk_ar1 = left_join(bk_ar1, bk_ar1 %>%
select(hhid14_9, pid14, pidlink) %>%
rename(ar11 = pid14, mother_pidlink = pidlink), by = c("hhid14_9", "ar11"))
bk_sc1 = read_dta("data/hh14_all_dta/bk_sc1.dta") # Location info, Book K, Section sc
# Recode Regions:
bk_sc1 <- bk_sc1 %>%
mutate(province = str_trim(recode(bk_sc1$sc01_14_14, `11` = 'N. Aceh Darussalam',
`12` = 'North Sumatera ',
`13` = 'West Sumatera',
`14` = 'Riau ',
`15` = 'Jambi ',
`16` = 'South Sumatera ',
`17` = 'Bengkulu',
`18` = 'Lampung ',
`19` = 'Bangka Belitung ',
`31` = 'Jakarta ',
`32` = 'West Java ',
`33` = 'Central Java ',
`34` = 'DI Yogyakarta',
`35` = 'East Java ',
`36` = 'Banten',
`51` = 'Bali ',
`52` = 'West Nusa Tenggara',
`53` = 'East Nusa Tenggara',
`61` = 'West Kalimantan ',
`62` = 'Central Kalimantan',
`63` = 'South Kalimantan',
`64` = 'East Kalimantan ',
`71` = 'North Sulawesi ',
`72` = 'Central Sulawesi',
`73` = 'South Sulawesi ',
`74` = 'Southeast Sulawesi',
`81` = 'Maluku',
`94` = 'Papua ',
`98` = NA_character_,
`99` = NA_character_, .default = NA_character_)))
### Informations from IFLS wave 5 to link data to earlier waves:
ptrack = read_dta("data/hh14_all_dta/ptrack.dta") # Tracking informations
### Pregnancy Informations from mother
## Wave 5 - 2014
w5_pregnancy = read_dta("data/hh14_all_dta/b4_ch1.dta") # Book 4, Section ch
## Wave 4 - 2007
w4_pregnancy = read_dta("data/hh07_all_dta/b4_ch1.dta") # Book 4, Section ch
## Wave 3 - 2000
w3_pregnancy = read_dta("data/hh00_all_dta/b4_ch1.dta") # Book 4, Section ch
## Wave 2 - 1997
w2_pregnancy = read_dta("data/hh97dta/b4_ch1.dta") # Book 4, Section ch
## Wave 1 - 1993
w1_pregnancy = read_dta("data/hh93dta/buk4ch1.dta") # Book 4, Section ch
### Marriage information from mother
## Wave 5 - 2014
w5_marriage= read_dta("data/hh14_all_dta/b4_kw3.dta") # Book 4, Section kw3
## Wave 4 - 2007
w4_marriage = read_dta("data/hh07_all_dta/b4_kw2.dta") # Book 4, Section kw2
## Wave 3 - 2000
w3_marriage = read_dta("data/hh00_all_dta/b4_kw3.dta") # Book 4, Section kw3
## Wave 2 - 1997
w2_marriage = read_dta("data/hh97dta/b4_kw2.dta") # Book 4, Section kw2
## Wave 1 - 1993
w1_marriage = read_dta("data/hh93dta/buk4kw2.dta") # Book 4, Section kw2
## Additional marriage information from mother
# Wave 5 - 2014
w5_marriage_additional = read_dta("data/hh14_all_dta/b4_cov.dta") # Book 4, Section cov
# Wave 4 - 2007
w4_marriage_additional = read_dta("data/hh07_all_dta/b4_cov.dta") # Book 4, Section cov
# Wave 3 - 2000
w3_marriage_additional = read_dta("data/hh00_all_dta/b4_cov.dta") # Book 4, Section cov
# Wave 2 - 1997
w2_marriage_additional = read_dta("data/hh97dta/b4_cov.dta") # Book 4, Section cov
# Wave 1 - 1993
w1_marriage_additional = read_dta("data/hh93dta/bukkar2.dta") # Book K, Section ar, household roaster
### IQ Information
ek_ek2 = read_dta("data/hh14_all_dta/ek_ek2.dta") # Book ek2: >15 years
ek_ek1 = read_dta("data/hh14_all_dta/ek_ek1.dta") # Book ek1: <15 years
# additional information (counting backwards, adaptive testing) for adults
b3b_cob = read_dta("data/hh14_all_dta/b3b_cob.dta") # Book 3b, Section cob
b3b_co1 = read_dta("data/hh14_all_dta/b3b_co1.dta") # Book 3b, Section co1
# additional information from earlier waves
bek_ek1 = read_dta("data/hh07_all_dta/bek_ek1.dta") # Intelligence information from wave 4 (2007): 7-14
bek_ek2 = read_dta("data/hh07_all_dta/bek_ek2.dta") # Intelligence info from wave 4 (2007): 15 - 24
### Personality Information (only for adults)
b3b_psn = read_dta("data/hh14_all_dta/b3b_psn.dta") # Book 3b, Section psn
### Risk taking
b3a_si = read_dta("data/hh14_all_dta/b3a_si.dta") # Book 3a, Section si
### Educational Attainment
b3a_dl1 = read_dta("data/hh14_all_dta/b3a_dl1.dta") # Book 3a, Section dl1
### EBTANAS/UAN/UN Score
b3a_dl3 = read_dta("data/hh14_all_dta/b3a_dl3.dta") # Book 3a, Section dl3
b3a_dl4 = read_dta("data/hh14_all_dta/b3a_dl4.dta") # Book 3a, Section dl4
### Job Information
b3a_tk2 = read_dta("data/hh14_all_dta/b3a_tk2.dta") # Book 3a, Section tk2
### Smoking behavior
b3b_km = read_dta("data/hh14_all_dta/b3b_km.dta") # Book 3b, Section kmBirth order information
Information about pregnancy
To compute birth order we need information about all pregnancies of all women who participated in IFLS wave 1 - 5
## Select data from pregnancy files
w5_pregnancy = w5_pregnancy %>%
select(pidlink, ch05, ch06, ch06a, ch08, ch09day, ch09mth, ch09yr, ch25)
w4_pregnancy = w4_pregnancy %>%
select(pidlink, ch05, ch06, ch06a, ch08, ch09day, ch09mth, ch09yr, ch25)
w3_pregnancy = w3_pregnancy %>%
select(pidlink, ch05, ch06, ch06a, ch08, ch09day, ch09mth, ch09yr, ch25)
w2_pregnancy = w2_pregnancy %>%
select(pidlink, ch05, ch06, ch06a, ch08, ch09day, ch09mth, ch09yr, ch25)
w1_pregnancy = w1_pregnancy %>%
group_by(pidlink, ch04) %>%
mutate(ch06a = if_else(!is.na(pidlink) & !is.na(ch04), if_else( n() > 1, 1, 3), 9)) %>%
ungroup() %>%
select(pidlink, ch05, ch06, ch06a, ch08, ch09day, ch09mth, ch09yr, ch25) %>%
# In the first wave the year is named wrong
mutate(ch09yr = ifelse(ch09yr <= 93, ch09yr, NA),
ch09yr = as.numeric(str_c("19", ch09yr)))
## Combine data
pregnancy = bind_rows(w1 = w1_pregnancy, w2 = w2_pregnancy, w3 = w3_pregnancy, w4 = w4_pregnancy, w5 = w5_pregnancy, .id = "wave")
pregnancy = codebook::rescue_attributes(pregnancy, w5_pregnancy)
## Rename Variables
pregnancy = pregnancy %>%
rename(chron_order_birth = ch05, lifebirths = ch06, multiple_birth = ch06a, gender = ch08,
birth_day = ch09day, birth_month = ch09mth, birth_year = ch09yr,
mother_pidlink = pidlink, alive = ch25)
# pregnancy$lifebirths values: 1 = still pregnant, 2 = livebirth, 3 = still birth, 4 = misscarriage
## Set values as NA that are missing (all values that are not logical)
pregnancy = pregnancy %>%
mutate(birth_day = ifelse(birth_day>31, NA, birth_day),
birth_month = ifelse(birth_month>12, NA, birth_month),
birth_year = ifelse(birth_year>2016, NA, birth_year),
birth_day = ifelse(is.nan(birth_day), NA, birth_day),
birth_month = ifelse(is.nan(birth_month), NA, birth_month),
birth_year = ifelse(is.nan(birth_year), NA, birth_year),
multiple_birth = ifelse(multiple_birth == 9, NA, multiple_birth),
multiple_birth = ifelse(is.nan(multiple_birth), NA, multiple_birth))
# if month of pregnancy is missing, set as first month of the year (January)
pregnancy$month = ifelse(is.na(pregnancy$birth_year), NA,
paste0(pregnancy$birth_year,"-",
ifelse(is.na(pregnancy$birth_month), "01",
pad_month(pregnancy$birth_month))))
# create birthdate variable that includes all available information about birth (year-month-day)
pregnancy = pregnancy %>%
mutate(birthdate = all_available_info_birth_date(birth_year, birth_month, birth_day),
mother_birthdate = str_c(mother_pidlink, "-", birthdate),
mother_birthorder = paste0(mother_pidlink , "-", chron_order_birth))
# some pregnancies were reported in multiple waves,
# e.g. because pregnancies changed status (still ongoing pregnancy --> alive)
# to eliminate duplicates we only use the most recent report
pregnancy = pregnancy %>%
mutate(wave = str_sub(wave, 2, 3) %>% as.numeric()) %>% # from most recent wave to oldest
arrange(desc(wave)) %>% # use most recent wave (because these will have pregnancy outcomes)
group_by(mother_birthdate) %>%
mutate(birthdate_duped_in_earlier_wave = min_rank(wave)) %>%
group_by(mother_birthorder) %>%
mutate(birthorder_duped_in_earlier_wave = min_rank(wave))
# these are pregnancy that changes status (i.e. ongoing in wave 2, miscarried/born by wave 3)
crosstabs(~ birthdate_duped_in_earlier_wave + birthorder_duped_in_earlier_wave + is.na(birthdate), pregnancy)| is.na(birthdate) | FALSE | TRUE | ||
| birthdate_duped_in_earlier_wave | birthorder_duped_in_earlier_wave | |||
| 1 | 1 | 36830 | 1339 | |
| 2 | 9330 | 0 | ||
| 3 | 2189 | 0 | ||
| 4 | 444 | 0 | ||
| 5 | 55 | 0 | ||
| 2 | 1 | 182 | 0 | |
| 2 | 2371 | 0 | ||
| 3 | 807 | 0 | ||
| 4 | 218 | 0 | ||
| 5 | 42 | 0 | ||
| 3 | 1 | 3 | 0 | |
| 2 | 80 | 0 | ||
| 3 | 191 | 0 | ||
| 4 | 58 | 0 | ||
| 5 | 16 | 0 | ||
| 4 | 1 | 1 | 0 | |
| 2 | 1 | 0 | ||
| 3 | 5 | 0 | ||
| 4 | 31 | 0 | ||
| 5 | 5 | 0 | ||
| 5 | 1 | 0 | 0 | |
| 2 | 0 | 0 | ||
| 3 | 2 | 0 | ||
| 4 | 0 | 0 | ||
| 5 | 0 | 0 | ||
| 6 | 1 | 0 | 0 | |
| 2 | 0 | 0 | ||
| 3 | 0 | 0 | ||
| 4 | 1 | 0 | ||
| 5 | 1 | 0 | ||
| 7 | 1 | 0 | 0 | |
| 2 | 0 | 0 | ||
| 3 | 0 | 0 | ||
| 4 | 2 | 0 | ||
| 5 | 0 | 0 | ||
| 1340 | 1 | 0 | 313 | |
| 2 | 0 | 151 | ||
| 3 | 0 | 0 | ||
| 4 | 0 | 0 | ||
| 5 | 0 | 0 | ||
| 1804 | 1 | 0 | 442 | |
| 2 | 0 | 160 | ||
| 3 | 0 | 57 | ||
| 4 | 0 | 0 | ||
| 5 | 0 | 0 | ||
| 2463 | 1 | 0 | 590 | |
| 2 | 0 | 56 | ||
| 3 | 0 | 23 | ||
| 4 | 0 | 9 | ||
| 5 | 0 | 0 | ||
| 3141 | 1 | 0 | 727 | |
| 2 | 0 | 171 | ||
| 3 | 0 | 65 | ||
| 4 | 0 | 42 | ||
| 5 | 0 | 10 |
# unfortunately, sometimes chron_order_birth is inconsistent with birthdates
# to eliminate duplicates from the pregnancy file (because pregnancies changed statuses)
# now, for those where we don't know the birthdate, we keep only unique birth orders
# for those, where we know the birthdate, we keep only unique birthdates (as this is higher q information)
pregnancy = pregnancy %>%
filter((is.na(birthdate) && birthorder_duped_in_earlier_wave == 1) | birthdate_duped_in_earlier_wave == 1) %>%
ungroup() # eliminate dupes across waves (same mother_birthdate), keep mult births
crosstabs(~ birthdate_duped_in_earlier_wave + birthorder_duped_in_earlier_wave + is.na(birthdate), pregnancy)| is.na(birthdate) | FALSE | TRUE | ||
| birthdate_duped_in_earlier_wave | birthorder_duped_in_earlier_wave | |||
| 1 | 1 | 36830 | 1339 | |
| 2 | 9330 | 0 | ||
| 3 | 2189 | 0 | ||
| 4 | 444 | 0 | ||
| 5 | 55 | 0 | ||
| 1340 | 1 | 0 | 170 | |
| 2 | 0 | 0 | ||
| 3 | 0 | 0 | ||
| 4 | 0 | 0 | ||
| 5 | 0 | 0 | ||
| 1804 | 1 | 0 | 167 | |
| 2 | 0 | 0 | ||
| 3 | 0 | 0 | ||
| 4 | 0 | 0 | ||
| 5 | 0 | 0 | ||
| 2463 | 1 | 0 | 252 | |
| 2 | 0 | 0 | ||
| 3 | 0 | 0 | ||
| 4 | 0 | 0 | ||
| 5 | 0 | 0 | ||
| 3141 | 1 | 0 | 727 | |
| 2 | 0 | 0 | ||
| 3 | 0 | 0 | ||
| 4 | 0 | 0 | ||
| 5 | 0 | 0 |
x = (unique(pregnancy$mother_pidlink))
# for whatever reason there are some multiple births with just one row in the data, but number are low enough
# that we consider some errors in the records the likely reason
pregnancy %>% filter(!is.na(birthdate)) %>% group_by(mother_birthdate) %>%
mutate(mult = n()) %>% crosstabs(~ mult + multiple_birth + is.na(birth_month), data = .)| is.na(birth_month) | FALSE | TRUE | ||
| mult | multiple_birth | |||
| 1 | 1 | 682 | 34 | |
| 3 | 37015 | 5670 | ||
| NA | 2613 | 1974 | ||
| 2 | 1 | 493 | 93 | |
| 3 | 28 | 100 | ||
| NA | 13 | 99 | ||
| 3 | 1 | 8 | 0 | |
| 3 | 1 | 4 | ||
| NA | 0 | 17 | ||
| 4 | 1 | 4 | 0 | |
| 3 | 0 | 0 | ||
| NA | 0 | 0 |
## Form variable for any multiple birth within a family (these families have to be excluded later)
pregnancy = pregnancy %>% group_by(mother_pidlink) %>%
mutate(any_multiple_birth = if_else(any(multiple_birth == 1, na.rm = T), 1, 0))
# Proportion of multiple births:
prop.table(crosstabs(pregnancy$multiple_birth))| 1 | 3 | NA |
|---|---|---|
| 0.02604 | 0.8549 | 0.119 |
# Proportion of families with any multiple birth:
prop.table(crosstabs(pregnancy$any_multiple_birth))| 0 | 1 |
|---|---|
| 0.9248 | 0.07524 |
Information about marriage history
For full sibling birth order we need information about all marriages of all women who participated in IFLS wave 1 - 5. We are assuming here that the husband is the father of all children who were born within this marriage.
## Select marriage data
w5_marriage = w5_marriage %>% select(pidlink, kw10mth, kw10yr, kw18mth, kw18yr, kw11, kw19)
w4_marriage = w4_marriage %>% select(pidlink, kw10mth, kw10yr, kw18mth, kw18yr, kw11, kw19)
w3_marriage = w3_marriage %>% select(pidlink, kw10mth, kw10yr, kw18mth, kw18yr, kw11, kw19)
w2_marriage = w2_marriage %>% select(pidlink, kw10mth, kw10yr, kw18mth, kw18yr, kw11, kw19)
w1_marriage = w1_marriage %>% select(pidlink, kw05a, kw05b, kw13a, kw13b, kw06, kw14age)
# In the first wave the year is named wrong
w1_marriage = w1_marriage %>%
mutate(kw05a = ifelse(kw05a <= 93, as.numeric(str_c("19", w1_marriage$kw05a)), kw05a),
kw13a = ifelse(kw13a <=93 , as.numeric(str_c("19", w1_marriage$kw13a)), kw13a))
# And the column names are different than in later waves
w1_marriage = w1_marriage %>%
rename(kw10mth = kw05b, kw10yr = kw05a, kw18mth = kw13b, kw18yr = kw13a, kw11 = kw06, kw19 = kw14age)
## Select additional marriage information (age of respondent)
w5_marriage_additional = w5_marriage_additional %>% select(pidlink, age, dob_yr)
w4_marriage_additional = w4_marriage_additional %>% select(pidlink, age, dob_yr)
w3_marriage_additional = w3_marriage_additional %>% select(pidlink, age, dob_yr)
w2_marriage_additional = w2_marriage_additional %>% select(pidlink, age, dob_yr)
w1_marriage_additional = w1_marriage_additional %>% select(pidlink, ar09yr, ar08yr)
# In the first wave the year is named wrong
w1_marriage_additional = w1_marriage_additional %>%
mutate(ar08yr = ifelse(ar08yr <= 93,
as.numeric(str_c("19", w1_marriage_additional$ar08yr)),
ar08yr))
# And the column names are different than in later waves
w1_marriage_additional = w1_marriage_additional %>% rename(age = ar09yr, dob_yr = ar08yr)
## Combine marriage information and additional marriage information:
w1_marriage = left_join(w1_marriage, w1_marriage_additional, by = "pidlink") %>%
mutate(wave = as.numeric("1993"))
w2_marriage = left_join(w2_marriage, w2_marriage_additional, by = "pidlink") %>%
mutate(wave = as.numeric("1997"))
w3_marriage = left_join(w3_marriage, w3_marriage_additional, by = "pidlink") %>%
mutate(wave = as.numeric("2000"))
w4_marriage = left_join(w4_marriage, w4_marriage_additional, by = "pidlink") %>%
mutate(wave = as.numeric("2007"))
w5_marriage = left_join(w5_marriage, w5_marriage_additional, by = "pidlink") %>%
mutate(wave = as.numeric("2014"))
## Combine marriage informations
marriage = bind_rows(w1_marriage, w2_marriage, w3_marriage, w4_marriage, w5_marriage)
# Rename columns
marriage = marriage %>%
rename(start_year = kw10yr, start_month = kw10mth, end_year = kw18yr, end_month = kw18mth, start_age = kw11, end_age = kw19, birth_year = dob_yr, birth_age = age)
# Set values as NA that are missing (all values that are not logical)
marriage$start_year[marriage$start_year < 1900] = NA
marriage$start_year[marriage$start_year > 2016] = NA
marriage$start_year[is.nan(marriage$start_year)] = NA
marriage$end_year[marriage$end_year < 1900] = NA
marriage$end_year[marriage$end_year > 2016] = NA
marriage$end_year[is.nan(marriage$end_year)] = NA
marriage$start_month[marriage$start_month > 12] = NA
marriage$start_month[is.nan(marriage$start_month)] = NA
marriage$end_month[marriage$end_month > 12] = NA
marriage$end_month[is.nan(marriage$end_month)] = NA
marriage$start_age[marriage$start_age > 97] = NA
marriage$start_age[is.nan(marriage$start_age)] = NA
marriage$end_age[marriage$end_age > 97] = NA
marriage$end_age[is.nan(marriage$end_age)] = NA
marriage$birth_year[marriage$birth_year < 1900] = NA
marriage$birth_year[marriage$birth_year > 2016] = NA
marriage$birth_year[is.nan(marriage$birth_year)] = NA
marriage$birth_age[marriage$birth_age > 97] = NA
marriage$birth_age[is.nan(marriage$birth_age)] = NA
## Reconstruct marriage start year and end year for marriages with missing year from age at marriage
marriage = marriage %>%
mutate(birth_year = ifelse(is.na(birth_year), wave - birth_age, birth_year),
start_year = ifelse(is.na(start_year), birth_year + start_age, start_year),
end_year = ifelse(is.na(end_year), birth_year + end_age, end_year))
## Arrange dataset:
marriage = marriage %>% arrange(pidlink, start_year, start_month, start_age, end_year, end_month, end_age)
## Eliminate duplicates (people who got married twice on the same day)
# propably due to some error in reporting marriages
marriage = marriage %>% filter(!duplicated(cbind(pidlink, start_year, start_month)) | is.na(start_year) | is.na(start_month)) # nobody gets married twice on the same day, right? so these are dupes.
## Calculate date for beginning of marriage:
marriage = marriage %>%
ungroup() %>%
mutate(start_string = str_c(start_year, "-", ifelse(is.na(start_month), "01",
pad_month(start_month)), "-01"),
end_string = str_c(end_year, "-", ifelse(is.na(end_month), "12", pad_month(end_month)), "-01"),
start = ymd(start_string),
end = ymd(end_string) + months(1) - days(1))
## Count number of marriages for every woman
marriage = marriage %>%
arrange(pidlink, start, end) %>%
group_by(pidlink) %>%
mutate(number_marriages = n(),
order_marriage = row_number(),
marriage_id = paste0(pidlink, "_", as.character(order_marriage), "_",
as.character(start), "/",as.character(end)))
## Marriage Timeline: reconstruct timeline of marriages for every woman
# We need this later for reconstructing who is the father
minimum_start = min(ymd(str_c(pregnancy$month, "-01")), na.rm = T)
maximum_end = max(ymd(str_c(pregnancy$month, "-01")), na.rm = T)
marriage_timeline = marriage %>%
mutate(implied_start = as.Date(ifelse(is.na(start), minimum_start , start),
origin = "1970-01-01"),
implied_end = as.Date(ifelse(is.na(end), maximum_end , end),
origin = "1970-01-01")) %>%
filter(implied_start < implied_end)
marriage_timeline = marriage_timeline %>%
rowwise() %>%
do(data.frame(
marriage_id = .$marriage_id,
mother_pidlink = .$pidlink,
order_marriage = .$order_marriage,
start = .$start,
end = .$end,
month = seq(.$implied_start,.$implied_end, by = "1 month") ))
# no duplicate mother_id - month combinations (no two marriages at the same time)
marriage_timeline = marriage_timeline %>%
arrange(mother_pidlink, start, end) %>%
distinct(mother_pidlink, month, .keep_all = TRUE)
# We only need year and month, not the exact day
marriage_timeline$month = stringr::str_sub(as.character(marriage_timeline$month),1,7)
# we assume that fathers are those to whom mothers were married in the birth month
pregnancy = pregnancy %>% left_join(marriage_timeline, by = c("mother_pidlink", "month")) %>% ungroup()Birth order calculations
Here we calculate full sibling birth order, maternal birth order and maternal pregnancy order
#### Maternal Pregnancy Order
pregnancy1 = pregnancy %>%
group_by(mother_pidlink) %>%
mutate(birthorder_uterus_preg = min_rank(birthdate),
birthorder_uterus_preg2 = ifelse(is.na(birthorder_uterus_preg), chron_order_birth,
ifelse(chron_order_birth > birthorder_uterus_preg,
chron_order_birth, birthorder_uterus_preg)),
sibling_count_uterus_preg = sum(!is.na(birthdate)),
sibling_count_uterus_preg2 = ifelse(is.na(sibling_count_uterus_preg), max(chron_order_birth, na.rm = T),
ifelse(max(chron_order_birth, na.rm = T) > sibling_count_uterus_preg,
max(chron_order_birth, na.rm = T), sibling_count_uterus_preg))
) %>%
ungroup()
cor.test(pregnancy1$birthorder_uterus_preg, pregnancy1$birthorder_uterus_preg2)| Test statistic | df | P value | Alternative hypothesis | cor |
|---|---|---|---|---|
| 1876 | 48846 | 0 * * * | two.sided | 0.9931 |
crosstabs(~ is.na(birthorder_uterus_preg) + is.na(birthorder_uterus_preg2) + is.na(birthdate), pregnancy1)| is.na(birthdate) | FALSE | TRUE | ||
| is.na(birthorder_uterus_preg) | is.na(birthorder_uterus_preg2) | |||
| FALSE | FALSE | 48848 | 0 | |
| TRUE | 0 | 0 | ||
| TRUE | FALSE | 0 | 2426 | |
| TRUE | 0 | 229 |
crosstabs(~ is.na(sibling_count_uterus_preg) + is.na(sibling_count_uterus_preg2) + is.na(birthdate), pregnancy1)| is.na(birthdate) | FALSE | TRUE | ||
| is.na(sibling_count_uterus_preg) | is.na(sibling_count_uterus_preg2) | |||
| FALSE | FALSE | 48848 | 2655 |
# our birthdate based birthorder estimates are extremely consistent with chron_order_birth
#### Maternal Birth Order
pregnancy2 = pregnancy %>%
filter(lifebirths == 2) %>%
group_by(mother_pidlink) %>%
mutate(birthorder_uterus_alive = min_rank(birthdate),
sibling_count_uterus_alive = sum(!is.na(birthdate))
) %>%
ungroup()
pregnancy2 = pregnancy2 %>% select(mother_birthdate, birthorder_uterus_alive, sibling_count_uterus_alive) %>% distinct()
#### Parental Full Sibling Birthorder (based on mother and father)
pregnancy3 = pregnancy %>%
filter(lifebirths == 2) %>%
group_by(marriage_id) %>%
mutate(birthorder_genes = min_rank(birthdate),
birthorder_genes = ifelse(is.na(marriage_id), NA, birthorder_genes),
sibling_count_genes = ifelse(is.na(marriage_id), NA, sum(!is.na(marriage_id)))) %>%
ungroup()
pregnancy3 = pregnancy3 %>% select(mother_birthdate, birthorder_genes, sibling_count_genes) %>% distinct()
# remove duplicates because of missings and twins
pregnancy1 <- pregnancy1 %>% select(-birthorder_uterus_preg2, -sibling_count_uterus_preg2) %>%
filter(!is.na(birthdate)) %>%
distinct(mother_birthdate, .keep_all = TRUE)
pregnancy2 <- pregnancy2 %>%
distinct(mother_birthdate, .keep_all = TRUE)
### Combine birthorder data
pregnancy = left_join(pregnancy1, pregnancy2, by = "mother_birthdate") %>% ungroup()
pregnancy = left_join(pregnancy, pregnancy3, by = "mother_birthdate") %>% ungroup()Birth order graphs
In order to check whether we calculated birth order correctly we created some graphs.
### Graphs are used to show differences between birth order measurements
## Biological Birthorder - Uterus_Pregnancies
ggplot(pregnancy, aes(x = sibling_count_uterus_preg, y = birthorder_uterus_preg)) +
geom_jitter(alpha = 0.1)
## Biological Birthorder - Uterus_Births
ggplot(pregnancy, aes(x = sibling_count_uterus_alive, y = birthorder_uterus_alive)) +
geom_jitter(alpha = 0.1)
## Biological Birthorder - Full Sibling Order
ggplot(pregnancy, aes(x = sibling_count_genes, y=birthorder_genes)) + geom_jitter(alpha = 0.1)
## Bio: Uterus_preg vs. Uterus_Births
ggplot(pregnancy, aes(x = birthorder_uterus_preg, y = birthorder_uterus_alive)) +
geom_jitter(alpha = 0.1)
# The birth_order_alive is always lower, which makes sense, becaus not live births (miscarriage, still births) are excluded
## Bio: Uterus_preg vs. Genes
ggplot(pregnancy, aes(x = birthorder_uterus_preg, y = birthorder_genes)) + geom_jitter(alpha = 0.1)
# The birthorder_genes is always lower, which makes sense, because different/unknown fathers are excluded
## Bio: Uterus_alive vs. Genes
ggplot(pregnancy, aes(x = birthorder_uterus_alive, y = birthorder_genes)) + geom_jitter(alpha = 0.1)
# children with the same father includes only live births
# chron order birth does not correlate perfectly, unsurprising given that we found chron orders sometimes started from 1 in new waves even though previous births were recorded
pregnancy %>% select(chron_order_birth, birthorder_uterus_alive, birthorder_uterus_preg, birthorder_genes) %>%
na.omit() %>% cor()| chron_order_birth | birthorder_uterus_alive | birthorder_uterus_preg | birthorder_genes |
|---|---|---|---|
| 1 | 0.7455 | 0.7705 | 0.7238 |
| 0.7455 | 1 | 0.9667 | 0.9679 |
| 0.7705 | 0.9667 | 1 | 0.9359 |
| 0.7238 | 0.9679 | 0.9359 | 1 |
pregnancy %>% select(chron_order_birth, birthorder_uterus_alive, birthorder_uterus_preg, birthorder_genes) %>% missingness_patterns()## index col missings
## 1 birthorder_genes 5864
## 2 birthorder_uterus_alive 5224| Pattern | Freq | Culprit |
|---|---|---|
| ___ | 42548 | _ |
| 1_2 | 5224 | |
| 1__ | 640 | birthorder_genes |
pregnancy %>% select(sibling_count_uterus_alive, sibling_count_uterus_preg, sibling_count_genes) %>% missingness_patterns()## index col missings
## 1 sibling_count_genes 5864
## 2 sibling_count_uterus_alive 5224| Pattern | Freq | Culprit |
|---|---|---|
| ___ | 42548 | _ |
| 1_2 | 5224 | |
| 1__ | 640 | sibling_count_genes |
Select individual data from IFLS 5
Here we select all variables for outcome measurements
### Individuals
individuals = bk_ar1 %>% select(hhid14_9, pidlink, father_pidlink, mother_pidlink, ar01a, ar02b, ar10, ar11, ar07, ar08day, ar08mth, ar08yr, ar09, ar18eyr, ar18emth)
individuals = left_join(individuals, bk_sc1 %>% select(hhid14_9, sc05, province, sc01_14_14), by = c("hhid14_9"))
#Rename variables:
individuals = rename(individuals, relation_to_HH_head = ar02b, fatherID = ar10, motherID = ar11, sex = ar07, age = ar09, status = ar01a, death_yr = ar18eyr, death_month = ar18emth)
# Remove duplicats (some people are mentioned in two households, e.g. because they moved in the last 12 months)
individuals = individuals %>% distinct(pidlink, .keep_all = TRUE)
individuals_unchanged = individuals
## people whose parents can not be identified have to be marked as NA:
individuals$fatherID[individuals$fatherID > 50] = NA
individuals$motherID[individuals$motherID > 50] = NA
## Create date of birth
#Set all variables missing that have not been reported:
individuals$ar08day[individuals$ar08day > 31] = NA
individuals$ar08mth[individuals$ar08mth > 12] = NA
individuals$ar08yr[individuals$ar08yr > 2016] = NA
individuals$ar08day[is.nan(individuals$ar08day)] = NA
individuals$ar08mth[is.nan(individuals$ar08mth)] = NA
individuals$ar08yr[is.nan(individuals$ar08yr)] = NA
individuals$death_month[individuals$death_month > 12] = NA
individuals$death_yr[individuals$death_yr > 2016] = NA
individuals$death_month[is.nan(individuals$death_month)] = NA
individuals$death_yr[is.nan(individuals$death_yr)] = NA
## Create variable that contains pidlink of mother and birthdate of child:
individuals = individuals %>%
mutate(birthdate = all_available_info_birth_date(ar08yr, ar08mth, ar08day),
mother_birthdate = str_c(mother_pidlink, "-", birthdate)) # mother_pidlink-YYYY-MM; is NA if birth_year is missing
## Create naive birth order(based on mother_id and living in the same household)
individuals = individuals %>% group_by(mother_pidlink) %>%
mutate(birthorder_naive_ind = if_else(!is.na(mother_pidlink), min_rank(birthdate), NA_integer_),
sibling_count_naive_ind = if_else(!is.na(mother_pidlink), n(), NA_integer_)) %>%
ungroup()
## Joing pregnancy and individuals data
alldata_pregnancy = full_join(pregnancy, individuals,
by = c("mother_pidlink", "birthdate", "mother_birthdate")) %>%
distinct(mother_pidlink, birthdate, pidlink, .keep_all = TRUE)
## check wether we have duplicated birthdates, but not reported as twins before
alldata_pregnancy = alldata_pregnancy %>%
group_by(mother_pidlink) %>%
mutate(any_multiple_birthdate = if_else(any(ifelse(!is.na(birth_month), duplicated(birthdate), NA), na.rm =T), 1, 0))
crosstabs(~alldata_pregnancy$any_multiple_birthdate + alldata_pregnancy$any_multiple_birth)| 0 | 1 | NA |
|---|---|---|
| 45034 | 2918 | 52631 |
| 34 | 560 | 64 |
#if we have duplicated birthdates within a family, we have to exclude these families, too (probably twins)
alldata_pregnancy = alldata_pregnancy %>% mutate(any_multiple_birth = ifelse(any_multiple_birthdate == 1, 1,
any_multiple_birth))
alldata_pregnancy = alldata_pregnancy %>% group_by(mother_pidlink) %>%
mutate(birthorder_naive = min_rank(if_else(!is.na(mother_pidlink), birthdate, NA_character_)),
sibling_count_naive = if_else(!is.na(mother_pidlink), n(), NA_integer_)) %>%
ungroup()Intelligence
Calculate g factor for intelligence
### IQ Informations from wave 5 (2014)
##ek2 (>14yrs, includes only individuals, that are 15 years or older)
iq2.1 = ek_ek2 %>% select(hhid14_9, pidlink, age, sex, ektype, resptype, result, reason, ek1_ans, ek2_ans, ek3_ans, ek4_ans, ek5_ans, ek6_ans, ek7_ans, ek8_ans, ek9_ans, ek10_ans, ek11_ans, ek12_ans, ek13_ans, ek14_ans, ek15_ans, ek16_ans, ek17_ans, ek18_ans, ek19_ans, ek20_ans, ek21_ans, ek22_ans)
##ek2 (<14yrs, includes all individuals, that are younger than 15 years old)
iq3.1 = ek_ek1 %>% select(hhid14_9, pidlink, age, sex, ektype, resptype, result, reason, ek1_ans, ek2_ans, ek3_ans, ek4_ans, ek5_ans, ek6_ans, ek7_ans, ek8_ans, ek9_ans, ek10_ans, ek11_ans, ek12_ans, ek13_ans, ek14_ans, ek15_ans, ek16_ans, ek17_ans, ek18_ans, ek19_ans, ek20_ans, ek21_ans, ek22_ans)
#### Raven Test (wave 2015, younger than 15 years)
answered_raven_items = iq3.1 %>% select(ek1_ans:ek12_ans)
psych::alpha(data.frame(answered_raven_items))##
## Reliability analysis
## Call: psych::alpha(x = data.frame(answered_raven_items))
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.87 0.87 0.88 0.36 6.9 0.0016 0.68 0.28 0.37
##
## lower alpha upper 95% confidence boundaries
## 0.87 0.87 0.87
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## ek1_ans 0.86 0.86 0.87 0.36 6.2 0.0017 0.020 0.35
## ek2_ans 0.86 0.86 0.87 0.36 6.1 0.0018 0.020 0.36
## ek3_ans 0.85 0.86 0.87 0.36 6.1 0.0018 0.020 0.36
## ek4_ans 0.86 0.86 0.87 0.36 6.2 0.0018 0.021 0.35
## ek5_ans 0.86 0.87 0.88 0.38 6.6 0.0016 0.021 0.39
## ek6_ans 0.87 0.88 0.88 0.39 7.1 0.0015 0.018 0.40
## ek7_ans 0.86 0.86 0.87 0.36 6.1 0.0017 0.017 0.37
## ek8_ans 0.85 0.86 0.86 0.35 6.0 0.0018 0.016 0.37
## ek9_ans 0.85 0.86 0.86 0.35 6.0 0.0018 0.017 0.35
## ek10_ans 0.86 0.86 0.87 0.36 6.2 0.0018 0.020 0.35
## ek11_ans 0.85 0.86 0.87 0.36 6.1 0.0018 0.019 0.35
## ek12_ans 0.87 0.88 0.89 0.40 7.3 0.0015 0.015 0.40
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## ek1_ans 14943 0.66 0.68 0.64 0.60 0.87 0.33
## ek2_ans 14943 0.69 0.69 0.67 0.62 0.77 0.42
## ek3_ans 14943 0.71 0.70 0.67 0.63 0.70 0.46
## ek4_ans 14943 0.69 0.68 0.65 0.61 0.69 0.46
## ek5_ans 14943 0.59 0.57 0.51 0.48 0.61 0.49
## ek6_ans 14943 0.48 0.46 0.38 0.36 0.37 0.48
## ek7_ans 14943 0.68 0.69 0.67 0.60 0.78 0.42
## ek8_ans 14943 0.71 0.72 0.71 0.64 0.81 0.39
## ek9_ans 14943 0.73 0.74 0.73 0.67 0.79 0.41
## ek10_ans 14943 0.69 0.69 0.65 0.61 0.72 0.45
## ek11_ans 14943 0.70 0.71 0.68 0.63 0.77 0.42
## ek12_ans 14943 0.41 0.41 0.31 0.29 0.24 0.43
##
## Non missing response frequency for each item
## 0 1 miss
## ek1_ans 0.13 0.87 0
## ek2_ans 0.23 0.77 0
## ek3_ans 0.30 0.70 0
## ek4_ans 0.31 0.69 0
## ek5_ans 0.39 0.61 0
## ek6_ans 0.63 0.37 0
## ek7_ans 0.22 0.78 0
## ek8_ans 0.19 0.81 0
## ek9_ans 0.21 0.79 0
## ek10_ans 0.28 0.72 0
## ek11_ans 0.23 0.77 0
## ek12_ans 0.76 0.24 0iq3.1$raven_2015_young = rowMeans( answered_raven_items, na.rm = T)
qplot(iq3.1$raven_2015_young)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
#### Math Test (wave 2015, younger than 15 years)
answered_math_items = iq3.1 %>% select(ek13_ans:ek17_ans)
psych::alpha(data.frame(answered_math_items))##
## Reliability analysis
## Call: psych::alpha(x = data.frame(answered_math_items))
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.62 0.62 0.59 0.24 1.6 0.0049 0.53 0.29 0.19
##
## lower alpha upper 95% confidence boundaries
## 0.61 0.62 0.63
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## ek13_ans 0.54 0.54 0.49 0.23 1.2 0.0061 0.0107 0.19
## ek14_ans 0.51 0.51 0.46 0.21 1.1 0.0065 0.0087 0.18
## ek15_ans 0.52 0.52 0.47 0.21 1.1 0.0064 0.0117 0.17
## ek16_ans 0.61 0.61 0.56 0.28 1.6 0.0051 0.0208 0.27
## ek17_ans 0.62 0.62 0.57 0.29 1.7 0.0050 0.0170 0.29
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## ek13_ans 14943 0.65 0.66 0.54 0.42 0.76 0.43
## ek14_ans 14943 0.70 0.70 0.61 0.47 0.66 0.47
## ek15_ans 14943 0.70 0.69 0.59 0.46 0.64 0.48
## ek16_ans 14943 0.57 0.56 0.36 0.28 0.34 0.47
## ek17_ans 14943 0.52 0.54 0.32 0.25 0.25 0.43
##
## Non missing response frequency for each item
## 0 1 miss
## ek13_ans 0.24 0.76 0
## ek14_ans 0.34 0.66 0
## ek15_ans 0.36 0.64 0
## ek16_ans 0.66 0.34 0
## ek17_ans 0.75 0.25 0iq3.1$math_2015_young = rowMeans( answered_math_items, na.rm = T)
qplot(iq3.1$math_2015_young)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
iq3.1 = iq3.1 %>% select(pidlink, age_2015_young = age, sex_2015_young = sex,
raven_2015_young, math_2015_young, reason_2015_young = reason)
##additional informations for adults: counting backwards
iq2.2 = b3b_co1 %>% select(hhid14_9, pidlink, co04a, co04b, co04c, co04d, co04e, co07count, co10count)
##additional informations for adults: adaptive number test
iq2.3 = b3b_cob %>% select(hhid14_9, pidlink, w_abil, cob18, cob19b)
## put all the informations for participants >= 15 together
iq2 = full_join(iq2.1, iq2.2, by = "pidlink")
iq2 = full_join(iq2, iq2.3, by = "pidlink")
iq = iq2
iq <- rename(iq, age_2015_old = age, reason_2015_old = reason, sex_2015_old = sex)
iq = full_join(iq, iq3.1, by = "pidlink")
### calculate iq scores
##Raven Test (wave 2015, older than 14 years)
answered_raven_items = iq %>% select(ek1_ans:ek6_ans, ek11_ans, ek12_ans)
psych::alpha(data.frame(answered_raven_items))##
## Reliability analysis
## Call: psych::alpha(x = data.frame(answered_raven_items))
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.85 0.85 0.84 0.41 5.5 0.0011 0.53 0.33 0.42
##
## lower alpha upper 95% confidence boundaries
## 0.85 0.85 0.85
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## ek1_ans 0.82 0.82 0.81 0.39 4.5 0.0013 0.019 0.34
## ek2_ans 0.81 0.81 0.80 0.38 4.3 0.0013 0.016 0.34
## ek3_ans 0.81 0.81 0.80 0.38 4.3 0.0013 0.019 0.31
## ek4_ans 0.82 0.81 0.81 0.38 4.4 0.0013 0.020 0.34
## ek5_ans 0.83 0.83 0.83 0.41 4.9 0.0012 0.025 0.34
## ek6_ans 0.85 0.85 0.84 0.44 5.6 0.0011 0.021 0.47
## ek11_ans 0.83 0.82 0.82 0.40 4.6 0.0012 0.023 0.34
## ek12_ans 0.86 0.86 0.85 0.46 5.9 0.0010 0.015 0.47
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## ek1_ans 36380 0.74 0.74 0.71 0.65 0.74 0.44
## ek2_ans 36380 0.80 0.80 0.79 0.72 0.67 0.47
## ek3_ans 36380 0.79 0.79 0.77 0.71 0.58 0.49
## ek4_ans 36380 0.78 0.77 0.74 0.69 0.59 0.49
## ek5_ans 36380 0.68 0.68 0.60 0.56 0.50 0.50
## ek6_ans 36380 0.54 0.55 0.44 0.40 0.28 0.45
## ek11_ans 36380 0.73 0.72 0.67 0.62 0.61 0.49
## ek12_ans 36380 0.47 0.49 0.36 0.34 0.22 0.42
##
## Non missing response frequency for each item
## 0 1 miss
## ek1_ans 0.26 0.74 0.18
## ek2_ans 0.33 0.67 0.18
## ek3_ans 0.42 0.58 0.18
## ek4_ans 0.41 0.59 0.18
## ek5_ans 0.50 0.50 0.18
## ek6_ans 0.72 0.28 0.18
## ek11_ans 0.39 0.61 0.18
## ek12_ans 0.78 0.22 0.18iq$raven_2015_old = rowMeans( answered_raven_items, na.rm = T)
qplot(iq$raven_2015_old)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
##Math Test (wave 2015, older than 14 years)
answered_math_items = iq %>% select(ek18_ans:ek22_ans)
psych::alpha(data.frame(answered_math_items))##
## Reliability analysis
## Call: psych::alpha(x = data.frame(answered_math_items))
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.68 0.68 0.64 0.3 2.1 0.0023 0.26 0.29 0.28
##
## lower alpha upper 95% confidence boundaries
## 0.68 0.68 0.69
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## ek18_ans 0.68 0.67 0.61 0.34 2.1 0.0025 0.0047 0.33
## ek19_ans 0.60 0.60 0.54 0.28 1.5 0.0030 0.0026 0.26
## ek20_ans 0.60 0.60 0.53 0.27 1.5 0.0031 0.0025 0.26
## ek21_ans 0.66 0.65 0.60 0.32 1.9 0.0027 0.0074 0.31
## ek22_ans 0.63 0.62 0.56 0.29 1.7 0.0029 0.0077 0.27
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## ek18_ans 36380 0.59 0.59 0.41 0.34 0.25 0.43
## ek19_ans 36380 0.71 0.71 0.61 0.50 0.28 0.45
## ek20_ans 36380 0.73 0.72 0.63 0.51 0.34 0.47
## ek21_ans 36380 0.60 0.63 0.46 0.38 0.18 0.38
## ek22_ans 36380 0.68 0.68 0.55 0.46 0.28 0.45
##
## Non missing response frequency for each item
## 0 1 miss
## ek18_ans 0.75 0.25 0.18
## ek19_ans 0.72 0.28 0.18
## ek20_ans 0.66 0.34 0.18
## ek21_ans 0.82 0.18 0.18
## ek22_ans 0.72 0.28 0.18iq$math_2015_old = rowMeans( answered_math_items, na.rm = T)
qplot(iq$math_2015_old)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
##Counting Items
# Create Right/Wrong Scores for the counting items
iq$co04aright = as.numeric(iq$co04a == 93)
iq$co04bright = as.numeric(iq$co04b == iq$co04a-7)
iq$co04cright = as.numeric(iq$co04c == iq$co04b-7)
iq$co04dright = as.numeric(iq$co04d == iq$co04c-7)
iq$co04eright = as.numeric(iq$co04e == iq$co04d-7)
answered_counting_items = iq %>% select(co04aright:co04eright)
psych::alpha(data.frame(answered_counting_items))##
## Reliability analysis
## Call: psych::alpha(x = data.frame(answered_counting_items))
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.69 0.68 0.64 0.29 2.1 0.0021 0.73 0.29 0.34
##
## lower alpha upper 95% confidence boundaries
## 0.68 0.69 0.69
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## co04aright 0.71 0.71 0.65 0.38 2.4 0.0023 0.0016 0.38
## co04bright 0.64 0.62 0.57 0.29 1.6 0.0025 0.0165 0.28
## co04cright 0.60 0.59 0.54 0.27 1.5 0.0028 0.0131 0.25
## co04dright 0.61 0.60 0.54 0.27 1.5 0.0027 0.0130 0.27
## co04eright 0.60 0.59 0.54 0.27 1.5 0.0028 0.0139 0.25
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## co04aright 30452 0.43 0.51 0.27 0.23 0.95 0.22
## co04bright 29661 0.70 0.67 0.53 0.45 0.63 0.48
## co04cright 29260 0.73 0.71 0.61 0.52 0.69 0.46
## co04dright 29078 0.73 0.71 0.61 0.51 0.69 0.46
## co04eright 28983 0.73 0.71 0.61 0.52 0.70 0.46
##
## Non missing response frequency for each item
## 0 1 miss
## co04aright 0.05 0.95 0.32
## co04bright 0.37 0.63 0.33
## co04cright 0.31 0.69 0.34
## co04dright 0.31 0.69 0.35
## co04eright 0.30 0.70 0.35iq$count_backwards = rowMeans( answered_counting_items, na.rm = T)
qplot(iq$count_backwards)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
## Word Memory
iq$words_immediate = iq$co07count
iq$words_delayed = iq$co10count
qplot(iq$words_immediate, iq$words_delayed, geom = "jitter")
answered_word_items = iq %>% select(co07count,co10count)
psych::alpha(data.frame(answered_word_items))##
## Reliability analysis
## Call: psych::alpha(x = data.frame(answered_word_items))
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.87 0.87 0.77 0.77 6.8 0.0012 4.6 1.8 0.77
##
## lower alpha upper 95% confidence boundaries
## 0.87 0.87 0.87
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## co07count 0.77 0.77 0.6 0.77 NA NA 0.77 0.77
## co10count 0.60 0.77 NA NA NA NA 0.60 0.77
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## co07count 31471 0.94 0.94 0.83 0.77 5.1 1.8
## co10count 31471 0.94 0.94 0.83 0.77 4.2 1.9iq$words_remembered_avg = rowMeans( answered_word_items, na.rm = T)
qplot(iq$words_remembered_avg)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
##Adaptive Numbering
iq$adaptive_numbering = iq$w_abil
qplot(iq$adaptive_numbering)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
# some people have reasons not to answer the test (dead, not contacted), that dont justify
# giving them zero points for that task. so i set these to NA
# additionally all participants older than 59 only answered the Raven Test, not the Math
# Test
iq = iq %>%
mutate(reason_2015_young = ifelse(reason_2015_young == 1, "refused",
ifelse(reason_2015_young == 2, "cannot read",
ifelse(reason_2015_young == 3, "unable to answer",
ifelse(reason_2015_young == 4, "not enough time",
ifelse(reason_2015_young == 5, "proxy respondent",
ifelse(reason_2015_young == 6, "other",
ifelse(reason_2015_young == 7, "could not be contacted", NA))))))),
raven_2015_young = ifelse(is.na(reason_2015_young), raven_2015_young,
ifelse(reason_2015_young == "cannot read"|
reason_2015_young == "not enough time"|
reason_2015_young == "proxy respondent"|
reason_2015_young == "other"|
reason_2015_young == "refused"|
reason_2015_young == "could not be contacted", NA,
raven_2015_young)),
math_2015_young = ifelse(is.na(reason_2015_young), math_2015_young,
ifelse(reason_2015_young == "cannot read"|
reason_2015_young == "not enough time"|
reason_2015_young == "proxy respondent"|
reason_2015_young == "other"|
reason_2015_young == "refused"|
reason_2015_young == "could not be contacted", NA,
math_2015_young)),
reason_2015_old = ifelse(reason_2015_old == 1, "refused",
ifelse(reason_2015_old == 2, "cannot read",
ifelse(reason_2015_old == 3, "unable to answer",
ifelse(reason_2015_old == 4, "not enough time",
ifelse(reason_2015_old == 5, "proxy respondent",
ifelse(reason_2015_old == 6, "other",
ifelse(reason_2015_old == 7, "could not be contacted", NA))))))),
raven_2015_old = ifelse(is.na(reason_2015_old), raven_2015_old,
ifelse(reason_2015_old == "cannot read"|
reason_2015_old == "not enough time"|
reason_2015_old == "proxy respondent"|
reason_2015_old == "other"|
reason_2015_old == "refused"|
reason_2015_old == "could not be contacted", NA,
raven_2015_old)),
math_2015_old = ifelse(is.na(reason_2015_old), math_2015_old,
ifelse(reason_2015_old == "cannot read"|
reason_2015_old == "not enough time"|
reason_2015_old == "proxy respondent"|
reason_2015_old == "other"|
reason_2015_old == "refused"|
reason_2015_old == "could not be contacted", NA,
math_2015_old)),
math_2015_old = ifelse(age_2015_old >= 60, NA, math_2015_old))
iq %>% select(raven_2015_old, math_2015_old, raven_2015_young, math_2015_young, count_backwards, words_immediate, words_delayed, adaptive_numbering) %>% tidyr::gather() %>%
ggplot(aes(value)) + geom_bar() + facet_wrap(~ key, scales = "free")
Intelligence from earlier waves
## Wave 4 - 2007
### Data
iq2007_ek2= bek_ek2 %>%
select(pidlink, ektype, reason_2007_old = reason, "age_2007_old" = age,
sex_2007_old = sex, matches("ek[0-9]x"), matches("ek[0-9][0-9]x"))
iq2007_ek1= bek_ek1 %>%
select(pidlink, ektype, reason_2007_young = reason, "age_2007_young" = age,
sex_2007_young = sex, matches("ek[0-9]x"), matches("ek[0-9][0-9]x")) %>%
filter(!(pidlink %in% iq2007_ek2$pidlink))
# some people answered both versions of the test
# (depending on whether they had seen test ek1 already in 2000)
# In order to deal with them i form an additional dataset that includes the information from
# the people that repeated the first test and their score on the first test in 2007
# that means they have already seen the test 7 years before...
# i merge this data later in the same column as the other scores in 2007 in the first test
# check with Ruben, if that is the right way to go...
iq2007_ek1_repeater = bek_ek1 %>%
select(pidlink, ektype, reason_2007_young_repeater = reason,
"age_2007_young_repeater" = age, sex_2007_young_repeater = sex,
matches("ek[0-9]x"), matches("ek[0-9][0-9]x")) %>%
filter((pidlink %in% iq2007_ek2$pidlink))
iq2007 = bind_rows(iq2007_ek1, iq2007_ek2)
### Raven
iq2007 = iq2007 %>%
mutate(ek1x = ifelse(ek1x == 1, 1,
ifelse(ek1x == 6, NA, 0)),
ek2x = ifelse(ek2x == 1, 1,
ifelse(ek2x == 6, NA, 0)),
ek3x = ifelse(ek3x == 1, 1,
ifelse(ek3x == 6, NA, 0)),
ek4x = ifelse(ek4x == 1, 1,
ifelse(ek4x == 6, NA, 0)),
ek5x = ifelse(ek5x == 1, 1,
ifelse(ek5x == 6, NA, 0)),
ek6x = ifelse(ek6x == 1, 1,
ifelse(ek6x == 6, NA, 0)),
ek7x = ifelse(ek7x == 1, 1,
ifelse(ek7x == 6, NA, 0)),
ek8x = ifelse(ek8x == 1, 1,
ifelse(ek8x == 6, NA, 0)),
ek9x = ifelse(ek9x == 1, 1,
ifelse(ek9x == 6, NA, 0)),
ek10x = ifelse(ek10x == 1, 1,
ifelse(ek10x == 6, NA, 0)),
ek11x = ifelse(ek11x == 1, 1,
ifelse(ek11x == 6, NA, 0)),
ek12x = ifelse(ek12x == 1, 1,
ifelse(ek12x == 6, NA, 0)))
answered_raven_items = iq2007 %>% select(ek1x:ek12x)
psych::alpha(data.frame(answered_raven_items))##
## Reliability analysis
## Call: psych::alpha(x = data.frame(answered_raven_items))
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.85 0.87 0.87 0.35 6.4 0.0016 0.69 0.28 0.34
##
## lower alpha upper 95% confidence boundaries
## 0.85 0.85 0.86
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## ek1x 0.84 0.85 0.86 0.34 5.7 0.0017 0.014 0.34
## ek2x 0.83 0.85 0.86 0.34 5.6 0.0018 0.014 0.33
## ek3x 0.83 0.85 0.85 0.34 5.6 0.0018 0.014 0.33
## ek4x 0.84 0.85 0.86 0.34 5.7 0.0018 0.014 0.34
## ek5x 0.85 0.86 0.87 0.36 6.1 0.0017 0.016 0.35
## ek6x 0.85 0.87 0.87 0.37 6.5 0.0016 0.013 0.37
## ek7x 0.84 0.85 0.85 0.34 5.8 0.0017 0.011 0.34
## ek8x 0.84 0.85 0.85 0.34 5.7 0.0017 0.011 0.34
## ek9x 0.84 0.85 0.86 0.35 5.8 0.0017 0.014 0.34
## ek10x 0.84 0.86 0.87 0.35 6.0 0.0017 0.016 0.34
## ek11x 0.84 0.85 0.86 0.34 5.7 0.0017 0.015 0.34
## ek12x 0.86 0.87 0.87 0.37 6.5 0.0015 0.012 0.37
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## ek1x 18508 0.63 0.69 0.65 0.58 0.89 0.31
## ek2x 18508 0.73 0.71 0.69 0.65 0.74 0.44
## ek3x 18508 0.75 0.72 0.70 0.67 0.70 0.46
## ek4x 18508 0.72 0.69 0.66 0.62 0.69 0.46
## ek5x 18508 0.63 0.59 0.52 0.50 0.64 0.48
## ek6x 18508 0.55 0.48 0.40 0.39 0.42 0.49
## ek7x 6684 0.62 0.67 0.65 0.55 0.92 0.27
## ek8x 6684 0.63 0.68 0.67 0.56 0.91 0.29
## ek9x 6684 0.63 0.65 0.61 0.54 0.85 0.36
## ek10x 6684 0.61 0.61 0.55 0.51 0.79 0.40
## ek11x 18508 0.65 0.68 0.64 0.58 0.82 0.38
## ek12x 18508 0.54 0.47 0.38 0.37 0.40 0.49
##
## Non missing response frequency for each item
## 0 1 miss
## ek1x 0.11 0.89 0.00
## ek2x 0.26 0.74 0.00
## ek3x 0.30 0.70 0.00
## ek4x 0.31 0.69 0.00
## ek5x 0.36 0.64 0.00
## ek6x 0.58 0.42 0.00
## ek7x 0.08 0.92 0.64
## ek8x 0.09 0.91 0.64
## ek9x 0.15 0.85 0.64
## ek10x 0.21 0.79 0.64
## ek11x 0.18 0.82 0.00
## ek12x 0.60 0.40 0.00iq2007$raven2007 = rowMeans(answered_raven_items, na.rm = T)
iq2007 = iq2007 %>%
mutate(raven_2007_young = ifelse(ektype == 1, raven2007, NA),
raven_2007_old = ifelse(ektype == 2, raven2007, NA))
qplot(iq2007$raven_2007_young)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
qplot(iq2007$raven_2007_old)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
##Math Test
iq2007 = iq2007 %>%
mutate(ek13x = ifelse(ek13x == 1, 1,
ifelse(ek13x == 6, NA, 0)),
ek14x = ifelse(ek14x == 1, 1,
ifelse(ek14x == 6, NA, 0)),
ek15x = ifelse(ek15x == 1, 1,
ifelse(ek15x == 6, NA, 0)),
ek16x = ifelse(ek16x == 1, 1,
ifelse(ek16x == 6, NA, 0)),
ek17x = ifelse(ek17x == 1, 1,
ifelse(ek17x == 6, NA, 0)),
ek18x = ifelse(ek18x == 1, 1,
ifelse(ek18x == 6, NA, 0)),
ek19x = ifelse(ek19x == 1, 1,
ifelse(ek19x == 6, NA, 0)),
ek20x = ifelse(ek20x == 1, 1,
ifelse(ek20x == 6, NA, 0)),
ek21x = ifelse(ek21x == 1, 1,
ifelse(ek21x == 6, NA, 0)),
ek22x = ifelse(ek22x == 1, 1,
ifelse(ek22x == 6, NA, 0)))
answered_math_items = iq2007 %>% select(ek13x:ek22x)
iq2007$math2007 = rowMeans( answered_math_items, na.rm = T)
iq2007 = iq2007 %>%
mutate(math_2007_young = ifelse(ektype == 1, math2007, NA),
math_2007_old = ifelse(ektype == 2, math2007, NA))
qplot(iq2007$math_2007_young)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
qplot(iq2007$math_2007_old)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
iq2007 = iq2007 %>% select(pidlink, age_2007_young, age_2007_old, sex_2007_young,
sex_2007_old, raven_2007_young, raven_2007_old, math_2007_young,
math_2007_old, reason_2007_old, reason_2007_young)
##Do the same for the repeater
### Raven
iq2007_ek1_repeater = iq2007_ek1_repeater %>%
mutate(ek1x = ifelse(ek1x == 1, 1,
ifelse(ek1x == 6, NA, 0)),
ek2x = ifelse(ek2x == 1, 1,
ifelse(ek2x == 6, NA, 0)),
ek3x = ifelse(ek3x == 1, 1,
ifelse(ek3x == 6, NA, 0)),
ek4x = ifelse(ek4x == 1, 1,
ifelse(ek4x == 6, NA, 0)),
ek5x = ifelse(ek5x == 1, 1,
ifelse(ek5x == 6, NA, 0)),
ek6x = ifelse(ek6x == 1, 1,
ifelse(ek6x == 6, NA, 0)),
ek7x = ifelse(ek7x == 1, 1,
ifelse(ek7x == 6, NA, 0)),
ek8x = ifelse(ek8x == 1, 1,
ifelse(ek8x == 6, NA, 0)),
ek9x = ifelse(ek9x == 1, 1,
ifelse(ek9x == 6, NA, 0)),
ek10x = ifelse(ek10x == 1, 1,
ifelse(ek10x == 6, NA, 0)),
ek11x = ifelse(ek11x == 1, 1,
ifelse(ek11x == 6, NA, 0)),
ek12x = ifelse(ek12x == 1, 1,
ifelse(ek12x == 6, NA, 0)))
answered_raven_items = iq2007_ek1_repeater %>% select(ek1x:ek12x)
psych::alpha(data.frame(answered_raven_items))##
## Reliability analysis
## Call: psych::alpha(x = data.frame(answered_raven_items))
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.87 0.89 0.9 0.41 8.2 0.0028 0.78 0.25 0.41
##
## lower alpha upper 95% confidence boundaries
## 0.87 0.87 0.88
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## ek1x 0.86 0.88 0.89 0.40 7.3 0.0031 0.020 0.40
## ek2x 0.86 0.88 0.89 0.40 7.3 0.0032 0.023 0.39
## ek3x 0.85 0.88 0.89 0.40 7.3 0.0033 0.022 0.39
## ek4x 0.86 0.88 0.89 0.40 7.5 0.0032 0.023 0.42
## ek5x 0.87 0.89 0.90 0.42 8.0 0.0030 0.023 0.43
## ek6x 0.88 0.89 0.90 0.44 8.5 0.0027 0.019 0.43
## ek7x 0.86 0.88 0.88 0.39 7.1 0.0031 0.017 0.40
## ek8x 0.86 0.88 0.88 0.39 7.1 0.0031 0.016 0.40
## ek9x 0.86 0.88 0.89 0.39 7.1 0.0032 0.019 0.40
## ek10x 0.86 0.88 0.89 0.40 7.3 0.0032 0.022 0.40
## ek11x 0.86 0.88 0.89 0.40 7.5 0.0031 0.022 0.40
## ek12x 0.88 0.90 0.90 0.44 8.6 0.0027 0.018 0.43
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## ek1x 4518 0.69 0.73 0.71 0.64 0.91 0.29
## ek2x 4518 0.72 0.72 0.69 0.65 0.81 0.40
## ek3x 4518 0.76 0.74 0.72 0.69 0.78 0.41
## ek4x 4518 0.71 0.69 0.65 0.62 0.76 0.43
## ek5x 4518 0.61 0.58 0.51 0.49 0.69 0.46
## ek6x 4518 0.53 0.48 0.40 0.39 0.48 0.50
## ek7x 4518 0.71 0.76 0.76 0.66 0.92 0.27
## ek8x 4518 0.73 0.78 0.78 0.68 0.92 0.27
## ek9x 4518 0.73 0.77 0.75 0.67 0.90 0.30
## ek10x 4518 0.70 0.72 0.69 0.64 0.87 0.34
## ek11x 4518 0.67 0.69 0.65 0.59 0.86 0.35
## ek12x 4518 0.51 0.46 0.37 0.37 0.46 0.50
##
## Non missing response frequency for each item
## 0 1 miss
## ek1x 0.09 0.91 0
## ek2x 0.19 0.81 0
## ek3x 0.22 0.78 0
## ek4x 0.24 0.76 0
## ek5x 0.31 0.69 0
## ek6x 0.52 0.48 0
## ek7x 0.08 0.92 0
## ek8x 0.08 0.92 0
## ek9x 0.10 0.90 0
## ek10x 0.13 0.87 0
## ek11x 0.14 0.86 0
## ek12x 0.54 0.46 0iq2007_ek1_repeater$raven_2007_young_repeater = rowMeans(answered_raven_items, na.rm = T)
qplot(iq2007_ek1_repeater$raven_2007_young_repeater)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
##Math Test
iq2007_ek1_repeater = iq2007_ek1_repeater %>%
mutate(ek13x = ifelse(ek13x == 1, 1,
ifelse(ek13x == 6, NA, 0)),
ek14x = ifelse(ek14x == 1, 1,
ifelse(ek14x == 6, NA, 0)),
ek15x = ifelse(ek15x == 1, 1,
ifelse(ek15x == 6, NA, 0)),
ek16x = ifelse(ek16x == 1, 1,
ifelse(ek16x == 6, NA, 0)),
ek17x = ifelse(ek17x == 1, 1,
ifelse(ek17x == 6, NA, 0)),
ek18x = ifelse(ek18x == 1, 1,
ifelse(ek18x == 6, NA, 0)),
ek19x = ifelse(ek19x == 1, 1,
ifelse(ek19x == 6, NA, 0)),
ek20x = ifelse(ek20x == 1, 1,
ifelse(ek20x == 6, NA, 0)),
ek21x = ifelse(ek21x == 1, 1,
ifelse(ek21x == 6, NA, 0)),
ek22x = ifelse(ek22x == 1, 1,
ifelse(ek22x == 6, NA, 0)))
answered_math_items = iq2007_ek1_repeater %>% select(ek13x:ek22x)
iq2007_ek1_repeater$math_2007_young_repeater = rowMeans( answered_math_items, na.rm = T)
qplot(iq2007_ek1_repeater$math_2007_young_repeater)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
iq2007_ek1_repeater = iq2007_ek1_repeater %>%
select(pidlink, raven_2007_young_repeater, math_2007_young_repeater,
reason_2007_young_repeater, age_2007_young_repeater, sex_2007_young_repeater)
iq2007_all = left_join(iq2007, iq2007_ek1_repeater, by = "pidlink")
# Now we instert raven and math scores from the repeaters as raven and math scores young
# Treat them like all the others.
crosstabs(is.na(iq2007_all$raven_2007_young) + is.na(iq2007_all$raven_2007_young_repeater))| FALSE | TRUE |
|---|---|
| 0 | 6684 |
| 4518 | 7306 |
iq2007_all = iq2007_all %>%
mutate(raven_2007_young = ifelse(!is.na(raven_2007_young_repeater),
raven_2007_young_repeater, raven_2007_young),
math_2007_young = ifelse(!is.na(math_2007_young_repeater),
math_2007_young_repeater, math_2007_young),
age_2007_young = ifelse(!is.na(age_2007_young_repeater),
age_2007_young_repeater, age_2007_young),
sex_2007_young = ifelse(!is.na(sex_2007_young_repeater),
sex_2007_young_repeater, sex_2007_young))
iq2007_all = iq2007_all %>% select(pidlink, age_2007_young, age_2007_old, raven_2007_old,
raven_2007_young, math_2007_young, math_2007_old,
reason_2007_young, reason_2007_old)
# some people have reasons not to answer the test (dead, not contacted), that dont justify
# giving them zero points for that task. so i set these to NA
# it is important to note that raven and math test were only answered by participants
# aged 7 - 24
iq2007_all = iq2007_all %>%
mutate(reason_2007_young = ifelse(reason_2007_young == 1, "refused",
ifelse(reason_2007_young == 2, "cannot read",
ifelse(reason_2007_young == 3, "unable to answer",
ifelse(reason_2007_young == 4, "not enough time",
ifelse(reason_2007_young == 5, "proxy respondent",
ifelse(reason_2007_young == 6, "other",
ifelse(reason_2007_young == 7, "could not be contacted", NA))))))),
raven_2007_young = ifelse(is.na(reason_2007_young), raven_2007_young,
ifelse(reason_2007_young == "cannot read"|
reason_2007_young == "not enough time"|
reason_2007_young == "proxy respondent"|
reason_2007_young == "other"|
reason_2007_young == "refused"|
reason_2007_young == "could not be contacted", NA,
raven_2007_young)),
math_2007_young = ifelse(is.na(reason_2007_young), math_2007_young,
ifelse(reason_2007_young == "cannot read"|
reason_2007_young == "not enough time"|
reason_2007_young == "proxy respondent"|
reason_2007_young == "other"|
reason_2007_young == "refused"|
reason_2007_young == "could not be contacted", NA,
math_2007_young)),
reason_2007_old = ifelse(reason_2007_old == 1, "refused",
ifelse(reason_2007_old == 2, "cannot read",
ifelse(reason_2007_old == 3, "unable to answer",
ifelse(reason_2007_old == 4, "not enough time",
ifelse(reason_2007_old == 5, "proxy respondent",
ifelse(reason_2007_old == 6, "other",
ifelse(reason_2007_old == 7, "could not be contacted", NA))))))),
raven_2007_old = ifelse(is.na(reason_2007_old), raven_2007_old,
ifelse(reason_2007_old == "cannot read"|
reason_2007_old == "not enough time"|
reason_2007_old == "proxy respondent"|
reason_2007_old == "other"|
reason_2007_old == "refused"|
reason_2007_old == "could not be contacted", NA,
raven_2007_old)),
math_2007_old = ifelse(is.na(reason_2007_old), math_2007_old,
ifelse(reason_2007_old == "cannot read"|
reason_2007_old == "not enough time"|
reason_2007_old == "proxy respondent"|
reason_2007_old == "other"|
reason_2007_young == "refused"|
reason_2007_old == "could not be contacted", NA,
math_2007_old)))
iq2007_all %>% select(raven_2007_old, math_2007_old, raven_2007_young, math_2007_young) %>% tidyr::gather() %>% ggplot(aes(value)) + geom_bar() + facet_wrap(~ key, scales = "free")
iq = full_join(iq, iq2007_all, by = "pidlink") %>%
select(-matches("ek[0-9]_ans"), -matches("ek[0-9][0-9]_ans"), -starts_with("hhid"),
-ektype, -resptype, -result, -starts_with("reason"), -starts_with("co"),
count_backwards, -w_abil)Correlations and Missingness Patterns for Intelligence
## IQ Tests
## Correlation of all Iq-Tests
round(cor(iq %>% select(raven_2015_old, math_2015_old, raven_2015_young, math_2015_young, count_backwards, words_immediate, words_delayed, adaptive_numbering, raven_2007_old, math_2007_young), use = "pairwise.complete.obs"), 2)| raven_2015_old | math_2015_old | raven_2015_young | math_2015_young | count_backwards | words_immediate | words_delayed | adaptive_numbering | raven_2007_old | math_2007_young |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 0.36 | 0.69 | 0.31 | 0.27 | 0.39 | 0.37 | 0.45 | 0.37 | 0.19 |
| 0.36 | 1 | 0.31 | 0.5 | 0.21 | 0.25 | 0.25 | 0.31 | 0.21 | 0.23 |
| 0.69 | 0.31 | 1 | 0.4 | 0.22 | 0.19 | 0.19 | 0.3 | 0.32 | 0.13 |
| 0.31 | 0.5 | 0.4 | 1 | 0.23 | 0.18 | 0.18 | 0.29 | 0.18 | 0.19 |
| 0.27 | 0.21 | 0.22 | 0.23 | 1 | 0.23 | 0.22 | 0.33 | 0.23 | 0.17 |
| 0.39 | 0.25 | 0.19 | 0.18 | 0.23 | 1 | 0.77 | 0.4 | 0.2 | 0.13 |
| 0.37 | 0.25 | 0.19 | 0.18 | 0.22 | 0.77 | 1 | 0.37 | 0.18 | 0.13 |
| 0.45 | 0.31 | 0.3 | 0.29 | 0.33 | 0.4 | 0.37 | 1 | 0.3 | 0.22 |
| 0.37 | 0.21 | 0.32 | 0.18 | 0.23 | 0.2 | 0.18 | 0.3 | 1 | 0.32 |
| 0.19 | 0.23 | 0.13 | 0.19 | 0.17 | 0.13 | 0.13 | 0.22 | 0.32 | 1 |
##Missingness_Patterns
formr::missingness_patterns(iq %>% select(raven_2015_old, math_2015_old, raven_2015_young, math_2015_young, count_backwards, words_immediate, words_delayed, adaptive_numbering, raven_2007_old, math_2007_young))## index col missings
## 1 math_2007_young 37147
## 2 raven_2007_old 36988
## 3 raven_2015_young 34066
## 4 math_2015_young 34066
## 5 math_2015_old 19612
## 6 count_backwards 17635
## 7 adaptive_numbering 16678
## 8 raven_2015_old 16669
## 9 words_immediate 16616
## 10 words_delayed 16616| Pattern | Freq | Culprit |
|---|---|---|
| 1_2_3_4_____________ | 15403 | |
| 1_2_____5_6_7_8_9_10 | 6889 | |
| 1_3_4___________ | 4987 | |
| 1_2_3_4_5_6_7_8_9_10 | 4129 | |
| 2_______________ | 3490 | raven_2007_old |
| ____________________ | 2629 | _ |
| 1_2_3_4_5___________ | 2593 | |
| 1___3_4_5_6_7_8_9_10 | 1616 | |
| __2_3_4_5_6_7_8_9_10 | 1547 | |
| ____3_4_5_6_7_8_9_10 | 1171 | |
| 2___5_6_7_8_9_10 | 888 | |
| 2_3_4___________ | 601 | |
| 1_2_3_4_6_______ | 482 | |
| ____3_4_____________ | 416 | |
| 1_2_3_4_5_____8_____ | 248 | |
| 1_2_3_4_5_6_________ | 188 | |
| 1_2_3_4_5_6_7___9_10 | 160 | |
| 1_2_3_4_6_7_9_10 | 140 | |
| 1_3_46________ | 121 | |
| 1_2_3_4_5_6_8___ | 66 | |
| 2_____6_________ | 54 | |
| 1_3_4_58__ | 38 | |
| __________6_________ | 38 | count_backwards |
| 1_3_4_6_7___9_10 | 33 | |
| 1_2_3_4_5_6_7_8_____ | 23 | |
| 2_3_4_58__ | 15 | |
| 2_3_46________ | 14 | |
| 2_____6_7___9_10 | 14 | |
| 1_2_3_4_5_7_8___ | 13 | |
| __________6_7___9_10 | 13 | |
| ____3_4_5_____8_____ | 12 | |
| 2_3_4_6_7___9_10 | 11 | |
| ____3_4_6_______ | 11 | |
| 1_2_3_4_6_7_____ | 8 | |
| ____3_4_5_6_7_8_____ | 3 | |
| ____3_4_6_7_9_10 | 3 | |
| 1_2_3_4_5_6_7_______ | 2 | |
| 1_2_3_4_____7_______ | 2 | |
| 2_3_4_5_6_7_8___ | 2 | |
| 2_3_4_57_8____ | 2 | |
| __________6_7_______ | 2 | |
| 1_2_____5_____8_____ | 1 | |
| 1_3_4_5_6_7_8___ | 1 | |
| 1_3_4_5_68____ | 1 | |
| 1_3_4_57_8____ | 1 | |
| 1_3_47____ | 1 | |
| 1_______5_6_7_8_9_10 | 1 | |
| 2_3_4_5_68____ | 1 | |
| 2_3_46_7______ | 1 | |
| 2_____6_7_______ | 1 | |
| ________5_6_7_8_9_10 | 1 |
iq %>% select(raven_2015_old, math_2015_old, raven_2015_young, math_2015_young, raven_2007_old, math_2007_old, raven_2007_young, math_2007_young, count_backwards, words_immediate, words_delayed, adaptive_numbering) %>% tidyr::gather() %>%
ggplot(aes(value)) + geom_bar() + facet_wrap(~ key, scales = "free")
Calculation of different g-factors
# G_factor_2015_old
nomiss = iq %>%
filter(!is.na(raven_2015_old), !is.na(math_2015_old), !is.na(count_backwards),
!is.na(words_delayed), !is.na(adaptive_numbering))
"g_factor_2015_old =~ raven_2015_old + math_2015_old + count_backwards + words_delayed+ adaptive_numbering" %>%
cfa(data = nomiss, std.lv = T, std.ov = T) -> cfa_g
summary(cfa_g, fit.measures = T, standardized = T, rsquare = TRUE)## lavaan 0.6-3 ended normally after 14 iterations
##
## Optimization method NLMINB
## Number of free parameters 10
##
## Number of observations 27526
##
## Estimator ML
## Model Fit Test Statistic 198.999
## Degrees of freedom 5
## P-value (Chi-square) 0.000
##
## Model test baseline model:
##
## Minimum Function Test Statistic 17985.333
## Degrees of freedom 10
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 0.989
## Tucker-Lewis Index (TLI) 0.978
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -186392.843
## Loglikelihood unrestricted model (H1) -186293.344
##
## Number of free parameters 10
## Akaike (AIC) 372805.686
## Bayesian (BIC) 372887.915
## Sample-size adjusted Bayesian (BIC) 372856.135
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.038
## 90 Percent Confidence Interval 0.033 0.042
## P-value RMSEA <= 0.05 1.000
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.015
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## g_factor_2015_old =~
## raven_2015_old 0.650 0.007 94.857 0.000 0.650 0.650
## math_2015_old 0.501 0.007 72.702 0.000 0.501 0.501
## count_backwrds 0.452 0.007 65.043 0.000 0.452 0.452
## words_delayed 0.478 0.007 69.072 0.000 0.478 0.478
## adaptiv_nmbrng 0.632 0.007 92.413 0.000 0.632 0.632
##
## Variances:
## Estimate Std.Err z-value P(>|z|) Std.lv Std.all
## .raven_2015_old 0.578 0.007 77.886 0.000 0.578 0.578
## .math_2015_old 0.749 0.008 99.579 0.000 0.749 0.749
## .count_backwrds 0.796 0.008 103.868 0.000 0.796 0.796
## .words_delayed 0.772 0.008 101.737 0.000 0.772 0.772
## .adaptiv_nmbrng 0.600 0.007 81.214 0.000 0.600 0.600
## g_fctr_2015_ld 1.000 1.000 1.000
##
## R-Square:
## Estimate
## raven_2015_old 0.422
## math_2015_old 0.251
## count_backwrds 0.204
## words_delayed 0.228
## adaptiv_nmbrng 0.400nomiss$g_factor_2015_old = predict(cfa_g)[,1]
qplot(nomiss$g_factor_2015_old)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
nomiss = nomiss %>% select(pidlink, g_factor_2015_old)
iq = left_join(iq, nomiss, by = "pidlink")
# G_factor_2015_young
nomiss = iq %>%
filter(!is.na(raven_2015_young), !is.na(math_2015_young))
fa.parallel(nomiss %>% select(raven_2015_young, math_2015_young) %>% data.frame())
## Parallel analysis suggests that the number of factors = 1 and the number of components = 1fa(nomiss %>% select(raven_2015_young, math_2015_young) %>% data.frame(), nfactors = 1)## Factor Analysis using method = minres
## Call: fa(r = nomiss %>% select(raven_2015_young, math_2015_young) %>%
## data.frame(), nfactors = 1)
## Standardized loadings (pattern matrix) based upon correlation matrix
## MR1 h2 u2 com
## raven_2015_young 0.63 0.4 0.6 1
## math_2015_young 0.63 0.4 0.6 1
##
## MR1
## SS loadings 0.8
## Proportion Var 0.4
##
## Mean item complexity = 1
## Test of the hypothesis that 1 factor is sufficient.
##
## The degrees of freedom for the null model are 1 and the objective function was 0.18 with Chi Square of 2456
## The degrees of freedom for the model are -1 and the objective function was 0
##
## The root mean square of the residuals (RMSR) is 0
## The df corrected root mean square of the residuals is NA
##
## The harmonic number of observations is 14021 with the empirical chi square 0 with prob < NA
## The total number of observations was 14021 with Likelihood Chi Square = 0 with prob < NA
##
## Tucker Lewis Index of factoring reliability = 1
## Fit based upon off diagonal values = 1
## Measures of factor score adequacy
## MR1
## Correlation of (regression) scores with factors 0.76
## Multiple R square of scores with factors 0.57
## Minimum correlation of possible factor scores 0.14om_results = omega(nomiss %>% select(raven_2015_young, math_2015_young) %>% data.frame(), nfactors = 1, sl = F)## Omega_h for 1 factor is not meaningful, just omega_tom_results## Omega
## Call: omega(m = nomiss %>% select(raven_2015_young, math_2015_young) %>%
## data.frame(), nfactors = 1, sl = F)
## Alpha: 0.57
## G.6: 0.4
## Omega Hierarchical: 0.57
## Omega H asymptotic: 1
## Omega Total 0.57
##
## Schmid Leiman Factor loadings greater than 0.2
## g F1* h2 u2 p2
## raven_2015_young 0.63 0.4 0.6 1
## math_2015_young 0.63 0.4 0.6 1
##
## With eigenvalues of:
## g F1*
## 0.8 0.0
##
## general/max Inf max/min = NaN
## mean percent general = 1 with sd = 0 and cv of 0
## Explained Common Variance of the general factor = 1
##
## The degrees of freedom are -1 and the fit is 0
## The number of observations was 14021 with Chi Square = 0 with prob < NA
## The root mean square of the residuals is 0
## The df corrected root mean square of the residuals is NA
##
## Compare this with the adequacy of just a general factor and no group factors
## The degrees of freedom for just the general factor are -1 and the fit is 0
## The number of observations was 14021 with Chi Square = 0 with prob < NA
## The root mean square of the residuals is 0
## The df corrected root mean square of the residuals is NA
##
## Measures of factor score adequacy
## g F1*
## Correlation of scores with factors 0.76 0
## Multiple R square of scores with factors 0.57 0
## Minimum correlation of factor score estimates 0.14 -1
##
## Total, General and Subset omega for each subset
## g F1*
## Omega total for total scores and subscales 0.57 0.57
## Omega general for total scores and subscales 0.57 0.57
## Omega group for total scores and subscales 0.00 0.00omega.diagram(om_results)
"g_factor_2015_young =~ raven_2015_young + math_2015_young" %>%
cfa(data = nomiss, std.lv = T, std.ov = T) -> cfa_g
summary(cfa_g)## lavaan 0.6-3 ended normally after 9 iterations
##
## Optimization method NLMINB
## Number of free parameters 4
##
## Number of observations 14021
##
## Estimator ML
## Model Fit Test Statistic NA
## Degrees of freedom -1
## Minimum Function Value 0.0000000000000
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## g_factor_2015_young =~
## raven_2015_yng 0.786 NA
## math_2015_yong 0.510 NA
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .raven_2015_yng 0.382 NA
## .math_2015_yong 0.740 NA
## g_fctr_2015_yn 1.000nomiss$g_factor_2015_young = predict(cfa_g)[,1]
qplot(nomiss$g_factor_2015_young)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
nomiss = nomiss %>% select(pidlink, g_factor_2015_young)
iq = left_join(iq, nomiss, by = "pidlink")
# G_factor_2007_old
nomiss = iq %>%
filter(!is.na(raven_2007_old), !is.na(math_2007_old))
fa.parallel(nomiss %>% select(raven_2007_old, math_2007_old) %>% data.frame())
## Parallel analysis suggests that the number of factors = 1 and the number of components = 1fa(nomiss %>% select(raven_2007_old, math_2007_old) %>% data.frame(), nfactors = 1)## Factor Analysis using method = minres
## Call: fa(r = nomiss %>% select(raven_2007_old, math_2007_old) %>% data.frame(),
## nfactors = 1)
## Standardized loadings (pattern matrix) based upon correlation matrix
## MR1 h2 u2 com
## raven_2007_old 0.59 0.35 0.65 1
## math_2007_old 0.59 0.35 0.65 1
##
## MR1
## SS loadings 0.70
## Proportion Var 0.35
##
## Mean item complexity = 1
## Test of the hypothesis that 1 factor is sufficient.
##
## The degrees of freedom for the null model are 1 and the objective function was 0.13 with Chi Square of 1467
## The degrees of freedom for the model are -1 and the objective function was 0
##
## The root mean square of the residuals (RMSR) is 0
## The df corrected root mean square of the residuals is NA
##
## The harmonic number of observations is 11082 with the empirical chi square 0 with prob < NA
## The total number of observations was 11082 with Likelihood Chi Square = 0 with prob < NA
##
## Tucker Lewis Index of factoring reliability = 1.001
## Fit based upon off diagonal values = 1
## Measures of factor score adequacy
## MR1
## Correlation of (regression) scores with factors 0.72
## Multiple R square of scores with factors 0.52
## Minimum correlation of possible factor scores 0.04om_results = omega(nomiss %>% select(raven_2007_old, math_2007_old) %>% data.frame(), nfactors = 1, sl = F)## Omega_h for 1 factor is not meaningful, just omega_tom_results## Omega
## Call: omega(m = nomiss %>% select(raven_2007_old, math_2007_old) %>%
## data.frame(), nfactors = 1, sl = F)
## Alpha: 0.52
## G.6: 0.35
## Omega Hierarchical: 0.52
## Omega H asymptotic: 1
## Omega Total 0.52
##
## Schmid Leiman Factor loadings greater than 0.2
## g F1* h2 u2 p2
## raven_2007_old 0.59 0.35 0.65 1
## math_2007_old 0.59 0.35 0.65 1
##
## With eigenvalues of:
## g F1*
## 0.7 0.0
##
## general/max Inf max/min = NaN
## mean percent general = 1 with sd = 0 and cv of 0
## Explained Common Variance of the general factor = 1
##
## The degrees of freedom are -1 and the fit is 0
## The number of observations was 11082 with Chi Square = 0 with prob < NA
## The root mean square of the residuals is 0
## The df corrected root mean square of the residuals is NA
##
## Compare this with the adequacy of just a general factor and no group factors
## The degrees of freedom for just the general factor are -1 and the fit is 0
## The number of observations was 11082 with Chi Square = 0 with prob < NA
## The root mean square of the residuals is 0
## The df corrected root mean square of the residuals is NA
##
## Measures of factor score adequacy
## g F1*
## Correlation of scores with factors 0.72 0
## Multiple R square of scores with factors 0.52 0
## Minimum correlation of factor score estimates 0.04 -1
##
## Total, General and Subset omega for each subset
## g F1*
## Omega total for total scores and subscales 0.52 0.52
## Omega general for total scores and subscales 0.52 0.52
## Omega group for total scores and subscales 0.00 0.00omega.diagram(om_results)
"g_factor_2007_old =~ raven_2007_old + math_2007_old" %>%
cfa(data = nomiss, std.lv = T, std.ov = T) -> cfa_g
summary(cfa_g)## lavaan 0.6-3 ended normally after 9 iterations
##
## Optimization method NLMINB
## Number of free parameters 4
##
## Number of observations 11082
##
## Estimator ML
## Model Fit Test Statistic NA
## Degrees of freedom -1
## Minimum Function Value 0.0000000000000
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## g_factor_2007_old =~
## raven_2007_old 0.784 NA
## math_2007_old 0.449 NA
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .raven_2007_old 0.386 NA
## .math_2007_old 0.798 NA
## g_fctr_2007_ld 1.000nomiss$g_factor_2007_old = predict(cfa_g)[,1]
qplot(nomiss$g_factor_2007_old)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
nomiss = nomiss %>% select(pidlink, g_factor_2007_old)
iq = left_join(iq, nomiss, by = "pidlink")
# G_factor_2007_young
nomiss = iq %>%
filter(!is.na(raven_2007_young), !is.na(math_2007_young))
fa.parallel(nomiss %>% select(raven_2007_young, math_2007_young) %>% data.frame())
## Parallel analysis suggests that the number of factors = 1 and the number of components = 1fa(nomiss %>% select(raven_2007_young, math_2007_young) %>% data.frame(), nfactors = 1)## Factor Analysis using method = minres
## Call: fa(r = nomiss %>% select(raven_2007_young, math_2007_young) %>%
## data.frame(), nfactors = 1)
## Standardized loadings (pattern matrix) based upon correlation matrix
## MR1 h2 u2 com
## raven_2007_young 0.69 0.48 0.52 1
## math_2007_young 0.69 0.48 0.52 1
##
## MR1
## SS loadings 0.96
## Proportion Var 0.48
##
## Mean item complexity = 1
## Test of the hypothesis that 1 factor is sufficient.
##
## The degrees of freedom for the null model are 1 and the objective function was 0.26 with Chi Square of 2885
## The degrees of freedom for the model are -1 and the objective function was 0
##
## The root mean square of the residuals (RMSR) is 0
## The df corrected root mean square of the residuals is NA
##
## The harmonic number of observations is 10940 with the empirical chi square 0 with prob < NA
## The total number of observations was 10940 with Likelihood Chi Square = 0 with prob < NA
##
## Tucker Lewis Index of factoring reliability = 1
## Fit based upon off diagonal values = 1
## Measures of factor score adequacy
## MR1
## Correlation of (regression) scores with factors 0.81
## Multiple R square of scores with factors 0.65
## Minimum correlation of possible factor scores 0.30om_results = omega(nomiss %>% select(raven_2007_young, math_2007_young) %>% data.frame(), nfactors = 1, sl = F)## Omega_h for 1 factor is not meaningful, just omega_tom_results## Omega
## Call: omega(m = nomiss %>% select(raven_2007_young, math_2007_young) %>%
## data.frame(), nfactors = 1, sl = F)
## Alpha: 0.65
## G.6: 0.48
## Omega Hierarchical: 0.65
## Omega H asymptotic: 1
## Omega Total 0.65
##
## Schmid Leiman Factor loadings greater than 0.2
## g F1* h2 u2 p2
## raven_2007_young 0.69 0.48 0.52 1
## math_2007_young 0.69 0.48 0.52 1
##
## With eigenvalues of:
## g F1*
## 0.96 0.00
##
## general/max Inf max/min = NaN
## mean percent general = 1 with sd = 0 and cv of 0
## Explained Common Variance of the general factor = 1
##
## The degrees of freedom are -1 and the fit is 0
## The number of observations was 10940 with Chi Square = 0 with prob < NA
## The root mean square of the residuals is 0
## The df corrected root mean square of the residuals is NA
##
## Compare this with the adequacy of just a general factor and no group factors
## The degrees of freedom for just the general factor are -1 and the fit is 0
## The number of observations was 10940 with Chi Square = 0 with prob < NA
## The root mean square of the residuals is 0
## The df corrected root mean square of the residuals is NA
##
## Measures of factor score adequacy
## g F1*
## Correlation of scores with factors 0.81 0
## Multiple R square of scores with factors 0.65 0
## Minimum correlation of factor score estimates 0.30 -1
##
## Total, General and Subset omega for each subset
## g F1*
## Omega total for total scores and subscales 0.65 0.65
## Omega general for total scores and subscales 0.65 0.65
## Omega group for total scores and subscales 0.00 0.00omega.diagram(om_results)
"g_factor_2007_young =~ raven_2007_young + math_2007_young" %>%
cfa(data = nomiss, std.lv = T, std.ov = T) -> cfa_g
summary(cfa_g)## lavaan 0.6-3 ended normally after 9 iterations
##
## Optimization method NLMINB
## Number of free parameters 4
##
## Number of observations 10940
##
## Estimator ML
## Model Fit Test Statistic NA
## Degrees of freedom -1
## Minimum Function Value 0.0000000000000
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## g_factor_2007_young =~
## raven_2007_yng 0.792 NA
## math_2007_yong 0.608 NA
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .raven_2007_yng 0.372 NA
## .math_2007_yong 0.631 NA
## g_fctr_2007_yn 1.000nomiss$g_factor_2007_young = predict(cfa_g)[,1]
qplot(nomiss$g_factor_2007_young)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
nomiss = nomiss %>% select(pidlink, g_factor_2007_young)
iq = left_join(iq, nomiss, by = "pidlink")Personality
Personality based on Big Five 15
### Personality
##Rearrange personality data so that every individual has only one row
pers = spread(b3b_psn, psntype, psn01)
##name columns
colnames(pers) <- c("hhid14_9", "pid14", "hhid14", "pidlink", "version", "module", "e1", "c1", "o1", "e2r", "n1r", "a1", "n2", "o2", "c2r", "o3", "a2", "c3", "e3", "a3r", "n3")
pers = pers %>% select(hhid14_9, pidlink, e1, c1, o1, e2r, n1r, a1, n2, o2, c2r, o3, a2, c3, e3, a3r, n3)
##Extraversion
pers$e2r_reversed = 6 - pers$e2r
extraversion = pers %>% select(e1, e2r_reversed, e3)
psych::alpha(data.frame(extraversion), check.keys = T)##
## Reliability analysis
## Call: psych::alpha(x = data.frame(extraversion), check.keys = T)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.37 0.35 0.28 0.15 0.54 0.0056 3.4 0.67 0.14
##
## lower alpha upper 95% confidence boundaries
## 0.36 0.37 0.38
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## e1 0.072 0.081 0.042 0.042 0.088 0.0092 NA 0.042
## e2r_reversed 0.218 0.246 0.140 0.140 0.326 0.0077 NA 0.140
## e3 0.429 0.429 0.273 0.273 0.753 0.0064 NA 0.273
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## e1 31446 0.77 0.71 0.48 0.30 3.1 1.14
## e2r_reversed 31446 0.73 0.67 0.37 0.24 3.0 1.12
## e3 31446 0.44 0.60 0.19 0.11 4.2 0.67
##
## Non missing response frequency for each item
## 1 2 3 4 5 miss
## e1 0.03 0.39 0.08 0.39 0.11 0
## e2r_reversed 0.07 0.36 0.09 0.43 0.05 0
## e3 0.00 0.03 0.05 0.64 0.28 0pers$big5_ext = rowMeans(extraversion)
qplot(pers$big5_ext)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
##conscientiousness
pers$c2r_reversed = 6 - pers$c2r
conscientiousness = pers %>% select(c1, c2r_reversed, c3)
psych::alpha(data.frame(conscientiousness), check.keys = T)##
## Reliability analysis
## Call: psych::alpha(x = data.frame(conscientiousness), check.keys = T)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.29 0.31 0.24 0.13 0.45 0.007 3.8 0.55 0.12
##
## lower alpha upper 95% confidence boundaries
## 0.27 0.29 0.3
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## c1 0.10 0.10 0.054 0.054 0.11 0.0101 NA 0.054
## c2r_reversed 0.35 0.36 0.219 0.219 0.56 0.0071 NA 0.219
## c3 0.20 0.21 0.116 0.116 0.26 0.0087 NA 0.116
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## c1 31446 0.61 0.69 0.42 0.23 4.1 0.71
## c2r_reversed 31446 0.65 0.60 0.19 0.10 3.6 0.95
## c3 31446 0.67 0.65 0.34 0.17 3.8 0.90
##
## Non missing response frequency for each item
## 1 2 3 4 5 miss
## c1 0.01 0.04 0.04 0.66 0.25 0
## c2r_reversed 0.03 0.18 0.07 0.65 0.07 0
## c3 0.01 0.12 0.08 0.63 0.15 0pers$big5_con = rowMeans(conscientiousness)
qplot(pers$big5_con)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
##Openness
openness = pers %>% select(o1, o2, o3)
psych::alpha(data.frame(openness), check.keys = T)##
## Reliability analysis
## Call: psych::alpha(x = data.frame(openness), check.keys = T)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.45 0.45 0.35 0.21 0.81 0.0054 3.7 0.67 0.22
##
## lower alpha upper 95% confidence boundaries
## 0.43 0.45 0.46
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## o1 0.30 0.30 0.18 0.18 0.43 0.0078 NA 0.18
## o2 0.39 0.39 0.25 0.25 0.65 0.0068 NA 0.25
## o3 0.35 0.36 0.22 0.22 0.55 0.0073 NA 0.22
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## o1 31446 0.71 0.71 0.45 0.30 3.7 0.98
## o2 31446 0.70 0.67 0.37 0.25 3.5 1.05
## o3 31446 0.65 0.69 0.41 0.27 4.0 0.88
##
## Non missing response frequency for each item
## 1 2 3 4 5 miss
## o1 0.02 0.18 0.08 0.58 0.14 0
## o2 0.03 0.22 0.09 0.54 0.12 0
## o3 0.01 0.10 0.05 0.61 0.23 0pers$big5_open = rowMeans(openness)
qplot(pers$big5_open)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
## Neuroticism
pers$n1r_reversed = 6 - pers$n1r
neuroticism = pers %>% select(n1r_reversed, n2, n3)
pers$big5_neu = rowMeans(neuroticism)
qplot(pers$big5_neu)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
##Agreeableness
pers$a3r_reversed = 6- pers$a3r
agreeableness= pers %>% select(a1, a2, a3r_reversed)
pers$big5_agree = rowMeans(agreeableness)
qplot(pers$big5_agree)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
Risk aversion
Risk aversion based on adaptive lottery task
###Risktaking
risk = b3a_si %>% select(hhid14_9, pidlink, random_si, si01, si02, si03, si04, si05, si11, si12, si13, si14, si15)
## 8 means they didnt know which answer they would choose
risk$si01[ risk$si01 == 8] = NA
risk$si02[ risk$si02 == 8] = NA
risk$si03[ risk$si03 == 8] = NA
risk$si04[ risk$si04 == 8] = NA
risk$si05[ risk$si05 == 8] = NA
risk$si11[ risk$si11 == 8] = NA
risk$si12[ risk$si12 == 8] = NA
risk$si13[ risk$si13 == 8] = NA
risk$si14[ risk$si14 == 8] = NA
risk$si15[ risk$si15 == 8] = NA
## calculate a risk score for risk game A
# (5 = gamble averse, Ordinalskala : 1 = risk loving, 4 = risk averse)
risk$riskA = ifelse(risk$si01 == 1 & risk$si02 == 1, 5,
ifelse(risk$si01 == 2 & risk$si03 == 1 & risk$si04 == 1, 4,
ifelse(risk$si01 == 2 & risk$si03 == 1 & risk$si04 == 2, 3,
ifelse(risk$si01 == 2 & risk$si03 == 2 & risk$si05 == 1, 2,
ifelse(risk$si01 == 2 & risk$si03 == 2 & risk$si05 == 2, 1,
NA)))))
## calculate a risk score for risk game B
# (5 = gamble loving, Ordinalskala : 1 = risk loving, 4 = risk averse)
risk$riskB = ifelse(risk$si11 == 2 & risk$si12 == 1, 5,
ifelse(risk$si11 == 1 & risk$si13 == 1 & risk$si14 == 1, 4,
ifelse(risk$si11 == 1 & risk$si13 == 1 & risk$si14 == 2, 3,
ifelse(risk$si11 == 1 & risk$si13 == 2 & risk$si15 == 1, 2,
ifelse(risk$si11 == 1 & risk$si13 == 2 & risk$si15 == 2, 1,
NA)))))
# recode risk variable
risk = risk %>%
mutate(riskB = ifelse(riskB == 5, 0, riskB),
riskB = riskB + 1)Educational Attainment
# Select, rename and mutate data
ea = b3a_dl1 %>%
select(pidlink, dl04, dl06, dl07, dl07a, dl07aa) %>%
rename(pidlink = pidlink, attended_school = dl04, highest_education = dl06,
highest_grade=dl07, currently_attending_school = dl07a, hours_in_class = dl07aa) %>%
mutate(attended_school = as.factor(ifelse(attended_school == 8, NA,
ifelse(attended_school == 1, "yes",
ifelse(attended_school == 3, "no",
attended_school)))),
highest_education = as.factor(ifelse(highest_education == 2 | highest_education == 72,
"Elementary",
ifelse(highest_education == 3 |
highest_education == 4 |
highest_education == 73,
"Junior High",
ifelse(highest_education == 5 |
highest_education == 6 |
highest_education == 74,
"Senior High",
ifelse(highest_education == 60 |
highest_education == 61 |
highest_education == 62 |
highest_education == 63 |
highest_education == 13,
"University", NA))))),#"other" = NA
highest_grade = ifelse(highest_grade == 98, NA,
ifelse(highest_grade == 99, NA, highest_grade)),
currently_attending_school = as.factor(ifelse(currently_attending_school == 1, "yes",
ifelse(currently_attending_school == 3, "no", NA))),
hours_in_class = ifelse(hours_in_class == 99, NA,
ifelse(hours_in_class == 98, NA, hours_in_class)))
## Create variable that includes years of education (highest_education.highest_grade) as a numeric variable
# People who started an education level, but dropped out within the first year, are coded with half a year of schooling
ea = ea %>%
mutate(years_of_education_factor = as.factor(str_c(highest_education, ".", highest_grade)),
years_of_education = as.numeric(ifelse(attended_school == "no", 0,
ifelse(years_of_education_factor == "Elementary.0", 0.5,
ifelse(years_of_education_factor == "Elementary.1", 1,
ifelse(years_of_education_factor == "Elementary.2", 2,
ifelse(years_of_education_factor == "Elementary.3", 3,
ifelse(years_of_education_factor == "Elementary.4", 4,
ifelse(years_of_education_factor == "Elementary.5", 5,
ifelse(years_of_education_factor == "Elementary.6", NA,
ifelse(years_of_education_factor == "Elementary.7", 6,
ifelse(years_of_education_factor == "Junior High.0", 6.5,
ifelse(years_of_education_factor == "Junior High.1", 7,
ifelse(years_of_education_factor == "Junior High.2", 8,
ifelse(years_of_education_factor == "Junior High.3", NA,
ifelse(years_of_education_factor == "Junior High.7", 9,
ifelse(years_of_education_factor == "Senior High.0", 9.5,
ifelse(years_of_education_factor == "Senior High.1", 10,
ifelse(years_of_education_factor == "Senior High.2", 11,
ifelse(years_of_education_factor == "Senior High.3", NA,
ifelse(years_of_education_factor == "Senior High.7", 12,
ifelse(years_of_education_factor == "University.0", 12.5,
ifelse(years_of_education_factor == "University.1", 13,
ifelse(years_of_education_factor == "University.2", 14,
ifelse(years_of_education_factor == "University.3", 15,
ifelse(years_of_education_factor == "University.4", 16,
ifelse(years_of_education_factor == "University.5", 17,
ifelse(years_of_education_factor == "University.6", NA,
ifelse(years_of_education_factor == "University.7", 18,
NA)))))))))))))))))))))))))))))
qplot(ea$years_of_education)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
EBTANAS/UAN/UN Score
EBTANAS_long = b3a_dl3 %>%
select(pidlink, "Level_of_schooling" = dl3type, "Type_of_test" = dl16c1, "Indonesia_score" = dl16db,
"English_score" = dl16dc, "Math_score" = dl16dd, "Total_score" = dl16e) %>%
mutate(Level_of_schooling = ifelse(Level_of_schooling == 1, "Elementary",
ifelse(Level_of_schooling == 2, "Junior High",
ifelse(Level_of_schooling == 3, "Senior High", NA))),
Type_of_test = ifelse(Type_of_test == 1, "EBTANAS",
ifelse(Type_of_test == 2, "UAN/UN", NA))) %>%
filter(!is.na(Indonesia_score) | !is.na(English_score) | !is.na(Math_score) | !is.na(Total_score))
# remove people with no EBTANAS Information
#Wrangle Data
EBTANAS_Elemenatry = EBTANAS_long %>%
filter(Level_of_schooling == "Elementary") %>%
rename("Indonesia_score_elementary" = Indonesia_score, "English_score_elementary" = English_score,
"Math_score_elemenatry" = Math_score, "Total_score_elemenatry" = Total_score,
"Type_of_test_elementary" = Type_of_test) %>%
select(-Level_of_schooling)
EBTANAS_Junior_High = EBTANAS_long %>%
filter(Level_of_schooling == "Junior High") %>%
rename("Indonesia_score_Junior_High" = Indonesia_score, "English_score_Junior_High" = English_score,
"Math_score_Junior_High" = Math_score, "Total_score_Junior_High" = Total_score,
"Type_of_test_Junior_High" = Type_of_test)%>%
select(-Level_of_schooling)
EBTANAS_Senior_High = EBTANAS_long %>%
filter(Level_of_schooling == "Senior High") %>%
rename("Indonesia_score_Senior_High" = Indonesia_score, "English_score_Senior_High" = English_score,
"Math_score_Senior_High" = Math_score, "Total_score_Senior_High" = Total_score,
"Type_of_test_Senior_High" = Type_of_test)%>%
select(-Level_of_schooling)
EBTANAS = full_join(EBTANAS_Elemenatry, EBTANAS_Junior_High, by = "pidlink")
EBTANAS = full_join(EBTANAS, EBTANAS_Senior_High, by = "pidlink")
table(is.na(EBTANAS$Total_score_elemenatry))| FALSE | TRUE |
|---|---|
| 4275 | 2002 |
table(is.na(EBTANAS$Total_score_Junior_High))| FALSE | TRUE |
|---|---|
| 3911 | 2366 |
table(is.na(EBTANAS$Total_score_Senior_High))| FALSE | TRUE |
|---|---|
| 2459 | 3818 |
EBTANAS = EBTANAS %>%
mutate(Total_score_highest = Total_score_elemenatry,
Total_score_highest = ifelse(!is.na(Total_score_Senior_High), Total_score_Senior_High,
ifelse(!is.na(Total_score_Junior_High), Total_score_Junior_High,
Total_score_highest)),
Total_score_highest_type = ifelse(!is.na(Total_score_Senior_High), "Senior High",
ifelse(!is.na(Total_score_Junior_High), "Junior High",
ifelse(!is.na(Total_score_elemenatry), "Elementary",
NA))),
Math_score_highest = Math_score_elemenatry,
Math_score_highest = ifelse(!is.na(Math_score_Senior_High), Math_score_Senior_High,
ifelse(!is.na(Math_score_Junior_High), Math_score_Junior_High,
Math_score_highest)),
Math_score_highest_type = ifelse(!is.na(Math_score_Senior_High), "Senior High",
ifelse(!is.na(Math_score_Junior_High), "Junior High",
ifelse(!is.na(Math_score_elemenatry), "Elementary",
NA))))
qplot(EBTANAS$Total_score_highest)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
qplot(EBTANAS$Math_score_highest)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
Child Labour
## Working while still in school
child_labour = b3a_dl4 %>%
select(pidlink, "Level_of_schooling" = dl4type, "Worked" = dl15, "Missed_school" = dl14a) %>%
mutate(Level_of_schooling = ifelse(Level_of_schooling == 1, "Elementary",
ifelse(Level_of_schooling == 2, "JuniorHigh",
ifelse(Level_of_schooling == 3, "SeniorHigh",
ifelse(Level_of_schooling == 4, "University", NA)))),
Worked = ifelse(Worked == 1, 1,
ifelse(Worked == 3, 0, NA)),
Missed_school = ifelse(Missed_school == 1, 1,
ifelse(Missed_school == 3, 0, NA)))
# Wrangle Data
child_labour_work = child_labour %>%
select(pidlink, Level_of_schooling, Worked) %>%
spread(., Level_of_schooling, Worked) %>%
rename("Elementary_worked" = Elementary, "Junior_high_worked" = JuniorHigh,
"Senior_high_worked" = SeniorHigh, "University_worked" = University) %>%
mutate(total_worked = ifelse(Elementary_worked == 1 | Junior_high_worked == 1 |
Senior_high_worked == 1 | University_worked == 1, 1, 0),
total_worked = ifelse(is.na(total_worked), 0, total_worked))
## Missed school
child_labour_missed = child_labour %>%
select(pidlink, Level_of_schooling, Missed_school) %>%
spread(., Level_of_schooling, Missed_school) %>%
rename("Elementary_missed" = Elementary, "Junior_high_missed" = JuniorHigh,
"Senior_high_missed" = SeniorHigh, "University_missed" = University) %>%
mutate(total_missed = ifelse(Elementary_missed == 1 | Junior_high_missed == 1 |
Senior_high_missed == 1 | University_missed == 1, 1, 0))
child_labour = full_join(child_labour_work, child_labour_missed, by = "pidlink")
table(child_labour$total_missed)| 0 | 1 |
|---|---|
| 2608 | 1017 |
table(child_labour$total_worked)| 0 | 1 |
|---|---|
| 9379 | 3498 |
Job Information
job = b3a_tk2 %>% select(pidlink, "Category" = tk24a, "Sector" = tk19ab, "wage_last_month" = tk25a1,
"wage_last_year" = tk25a2) %>%
mutate(Category = ifelse(Category == 1 | Category == 2 | Category == 3, "Self-employed",
ifelse(Category == 4, "Government worker",
ifelse(Category == 5, "Private worker",
ifelse(Category == 7, "Casual worker in agriculture",
ifelse(Category == 8, "Casual worker not in agriculture",
ifelse(Category == 6, "Unpaid family worker",
NA)))))),
Self_employed = ifelse(Category == "Self-employed", 1,
ifelse(is.na(Category), NA, 0)),
Sector = ifelse(Sector == 1, "Agriculture, forestry, fishing and hunting",
ifelse(Sector == 2, "Mining and quarrying",
ifelse(Sector == 3, "Manufacturing",
ifelse(Sector == 4, "Electricity, gas, water",
ifelse(Sector == 5, "Construction",
ifelse(Sector == 6, "Wholesale, retail, restaurants and hotels",
ifelse(Sector == 7, "Transportation, storage and communications",
ifelse(Sector == 8, "Finance, insurance, real estate and business services",
ifelse(Sector == 9, "Social services",
ifelse(Sector == 10, "Activities that cannot be classified", NA)))))))))),
wage_last_month = ifelse(wage_last_month == 8, NA, wage_last_month),
wage_last_year = ifelse(wage_last_year == 8, NA, wage_last_year))
table(job$Category)| Casual worker in agriculture | Casual worker not in agriculture | Government worker | Private worker | Self-employed | Unpaid family worker |
|---|---|---|---|---|---|
| 830 | 1888 | 1616 | 8364 | 9194 | 2589 |
table(job$Sector)| Agriculture, forestry, fishing and hunting | Construction | Electricity, gas, water | Finance, insurance, real estate and business services | Manufacturing | Mining and quarrying | Social services | Transportation, storage and communications | Wholesale, retail, restaurants and hotels |
|---|---|---|---|---|---|---|---|---|
| 5940 | 29 | 394 | 5003 | 5469 | 839 | 1510 | 787 | 3754 |
qplot(job$wage_last_month)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
summary(job$wage_last_month)| Min. | 1st Qu. | Median | Mean | 3rd Qu. | Max. | NA’s |
|---|---|---|---|---|---|---|
| 0 | 550000 | 1250000 | 1923793 | 2400000 | 999999997 | 11970 |
x = job %>% filter(wage_last_month < 5000000)
qplot(x$wage_last_month)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
qplot(job$wage_last_year)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
summary(job$wage_last_year)| Min. | 1st Qu. | Median | Mean | 3rd Qu. | Max. | NA’s |
|---|---|---|---|---|---|---|
| 0 | 3960000 | 12000000 | 20680463 | 27000000 | 999999997 | 12064 |
x = job %>% filter(wage_last_year < 50000000)
qplot(x$wage_last_year)## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.
Health Information - Smoking Behavior
smoking = b3b_km %>%
select(pidlink, "ever_smoked" = km01a, "still_smoking" = km04, "amount" = km08,
"money_spent_smoking" = km09, "age_first_smoke" = km10) %>%
mutate(ever_smoked = ifelse(ever_smoked == 1, 1, 0),
still_smoking = ifelse(still_smoking == 1, 1, 0),
amount = ifelse(amount == 8, NA,
ifelse(amount == 98, NA, amount)),
amount_still_smokers = ifelse(still_smoking == 1, amount, NA))Merge data
### Merge all data for people with birthorder informations
alldata_birthorder = left_join(alldata_pregnancy, iq, by = "pidlink")
alldata_birthorder = left_join(alldata_birthorder, pers, by = "pidlink")
alldata_birthorder = left_join(alldata_birthorder, risk, by = "pidlink")
alldata_birthorder = left_join(alldata_birthorder, ea, by = "pidlink")
alldata_birthorder = left_join(alldata_birthorder, EBTANAS, by = "pidlink")
alldata_birthorder = left_join(alldata_birthorder, child_labour, by = "pidlink")
alldata_birthorder = left_join(alldata_birthorder, job, by = "pidlink")
alldata_birthorder = left_join(alldata_birthorder, smoking, by = "pidlink")
alldata = left_join(individuals_unchanged, iq, by = "pidlink")
alldata = left_join(alldata, pers, by = "pidlink")
alldata = left_join(alldata, risk, by = "pidlink")
alldata = left_join(alldata, ea, by = "pidlink")
alldata = left_join(alldata, EBTANAS, by = "pidlink")
alldata = left_join(alldata, child_labour, by = "pidlink")
alldata = left_join(alldata, job, by = "pidlink")
alldata = left_join(alldata, smoking, by = "pidlink")Fix types
alldata_birthorder <- alldata_birthorder %>%
mutate_at(vars(age_2015_old, age_2015_young, alive, death_yr,
English_score_elementary, English_score_Junior_High, English_score_Senior_High,
Indonesia_score_elementary, Indonesia_score_Junior_High, Indonesia_score_Senior_High,
Math_score_elemenatry, Math_score_Junior_High, Math_score_Senior_High,
e1, c1, o1, e2r, n1r, a1, n2, o2, c2r, o3, a2, c3, e3, a3r, n3, e2r_reversed, c2r_reversed,
n1r_reversed, a3r_reversed,
si01 ,si02, si03, si04, si05, si11, si12, si13, si14, si15
), funs(as.numeric)) %>%
mutate_at(vars(Math_score_highest_type, Total_score_highest_type, mother_pidlink, father_pidlink, marriage_id,
pidlink, random_si, Sector, Type_of_test_elementary, Type_of_test_Junior_High,
Type_of_test_Senior_High, lifebirths, Category
), funs(as.factor)) %>%
mutate(male = if_else(sex == 1, 1L, 0L),
wage_last_month_log = log(wage_last_month+1),
wage_last_year_log = log(wage_last_year+1),
money_spent_smoking_log = log(money_spent_smoking+1)) %>%
mutate_at(vars(alive, Elementary_missed, Elementary_worked, ever_smoked, Junior_high_missed, Junior_high_worked,
Self_employed, total_missed, Senior_high_missed, Senior_high_worked, still_smoking, University_missed,
University_worked, wave), funs(as.integer)) %>%
mutate_at(vars(birthdate), funs(as.Date)) %>%
select(-motherID, -fatherID, -mother_birthorder, -mother_birthdate, -month, -end, -start, -gender, -sex, -sex_2015_old, -sex_2015_young,
-relation_to_HH_head, -years_of_education_factor, -birthdate_duped_in_earlier_wave, -birthorder_duped_in_earlier_wave, -birthorder_naive_ind,
-order_marriage,
-ar08day, -ar08mth, -ar08yr, -birth_day, -birth_month, -birth_year, -highest_grade, -wage_last_month, -wage_last_year, -status,
-money_spent_smoking, -starts_with("hhid14_"), sc05, province, sc01_14_14)Save data
for future analyses
alldata_birthorder = alldata_birthorder %>% ungroup()
alldata = alldata %>% ungroup()
pregnancy = pregnancy %>% ungroup()
iq = iq %>% ungroup()
saveRDS(alldata_birthorder, file = "data/alldata_birthorder.rds")
saveRDS(alldata, file = "data/alldata.rds")
saveRDS(pregnancy, file = "data/pregnancy.rds")
saveRDS(iq, file = "data/iq.rds")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