Analysis Results
Linh Nguyen
2022-07-13
I. Meta-data
# list of loaded packages and versions
si <- devtools::session_info()[[2]]
rownames(si) <- NULL
si %>%
select(package, loadedversion, date, source) %>%
#red bold the called packages
mutate(package = cell_spec(package,
color = ifelse(package %in% libraries, "red", "black"),
bold = ifelse(package %in% libraries, TRUE, FALSE))) %>%
knitr::kable(escape = F, caption = "All loaded packages. Bolded in red are those loaded explicitly with <code>library()</code>") %>%
kable_styling() %>%
scroll_box(height = "300px")| package | loadedversion | date | source |
|---|---|---|---|
| abind | 1.4-5 | 2016-07-21 | CRAN (R 4.2.0) |
| arm | 1.12-2 | 2021-10-15 | CRAN (R 4.2.0) |
| assertthat | 0.2.1 | 2019-03-21 | CRAN (R 4.2.0) |
| backports | 1.4.1 | 2021-12-13 | CRAN (R 4.2.0) |
| base64enc | 0.1-3 | 2015-07-28 | CRAN (R 4.2.0) |
| bayestestR | 0.12.1 | 2022-05-02 | CRAN (R 4.2.0) |
| boot | 1.3-28 | 2021-05-03 | CRAN (R 4.2.1) |
| brio | 1.1.3 | 2021-11-30 | CRAN (R 4.2.0) |
| broom | 0.8.0 | 2022-04-13 | CRAN (R 4.2.0) |
| bslib | 0.3.1 | 2021-10-06 | CRAN (R 4.2.0) |
| cachem | 1.0.6 | 2021-08-19 | CRAN (R 4.2.0) |
| callr | 3.7.0 | 2021-04-20 | CRAN (R 4.2.0) |
| carData | 3.0-5 | 2022-01-06 | CRAN (R 4.2.0) |
| cellranger | 1.1.0 | 2016-07-27 | CRAN (R 4.2.0) |
| checkmate | 2.1.0 | 2022-04-21 | CRAN (R 4.2.0) |
| cli | 3.3.0 | 2022-04-25 | CRAN (R 4.2.0) |
| cluster | 2.1.3 | 2022-03-28 | CRAN (R 4.2.1) |
| coda | 0.19-4 | 2020-09-30 | CRAN (R 4.2.0) |
| codetools | 0.2-18 | 2020-11-04 | CRAN (R 4.2.1) |
| colorspace | 2.0-3 | 2022-02-21 | CRAN (R 4.2.0) |
| corpcor | 1.6.10 | 2021-09-16 | CRAN (R 4.2.0) |
| cowplot | 1.1.1 | 2020-12-30 | CRAN (R 4.2.0) |
| crayon | 1.5.1 | 2022-03-26 | CRAN (R 4.2.0) |
| curl | 4.3.2 | 2021-06-23 | CRAN (R 4.2.0) |
| data.table | 1.14.2 | 2021-09-27 | CRAN (R 4.2.0) |
| datawizard | 0.4.1 | 2022-05-16 | CRAN (R 4.2.0) |
| DBI | 1.1.2 | 2021-12-20 | CRAN (R 4.2.0) |
| dbplyr | 2.2.0 | 2022-06-05 | CRAN (R 4.2.0) |
| desc | 1.4.1 | 2022-03-06 | CRAN (R 4.2.0) |
| devtools | 2.4.3 | 2021-11-30 | CRAN (R 4.2.0) |
| digest | 0.6.29 | 2021-12-01 | CRAN (R 4.2.0) |
| dplyr | 1.0.9 | 2022-04-28 | CRAN (R 4.2.0) |
| effectsize | 0.7.0 | 2022-05-26 | CRAN (R 4.2.0) |
| ellipsis | 0.3.2 | 2021-04-29 | CRAN (R 4.2.0) |
| emmeans | 1.7.4-1 | 2022-05-15 | CRAN (R 4.2.0) |
| estimability | 1.3 | 2018-02-11 | CRAN (R 4.2.0) |
| evaluate | 0.15 | 2022-02-18 | CRAN (R 4.2.0) |
| fansi | 1.0.3 | 2022-03-24 | CRAN (R 4.2.0) |
| fastmap | 1.1.0 | 2021-01-25 | CRAN (R 4.2.0) |
| fdrtool | 1.2.17 | 2021-11-13 | CRAN (R 4.2.0) |
| forcats | 0.5.1 | 2021-01-27 | CRAN (R 4.2.0) |
| foreign | 0.8-82 | 2022-01-16 | CRAN (R 4.2.1) |
| Formula | 1.2-4 | 2020-10-16 | CRAN (R 4.2.0) |
| fs | 1.5.2 | 2021-12-08 | CRAN (R 4.2.0) |
| generics | 0.1.2 | 2022-01-31 | CRAN (R 4.2.0) |
| ggeffects | 1.1.2 | 2022-04-10 | CRAN (R 4.2.0) |
| ggplot2 | 3.3.6 | 2022-05-03 | CRAN (R 4.2.0) |
| glasso | 1.11 | 2019-10-01 | CRAN (R 4.2.0) |
| glue | 1.6.2 | 2022-02-24 | CRAN (R 4.2.0) |
| gridExtra | 2.3 | 2017-09-09 | CRAN (R 4.2.0) |
| gtable | 0.3.0 | 2019-03-25 | CRAN (R 4.2.0) |
| gtools | 3.9.2.1 | 2022-05-23 | CRAN (R 4.2.0) |
| haven | 2.5.0 | 2022-04-15 | CRAN (R 4.2.0) |
| Hmisc | 4.7-0 | 2022-04-19 | CRAN (R 4.2.0) |
| hms | 1.1.1 | 2021-09-26 | CRAN (R 4.2.0) |
| htmlTable | 2.4.0 | 2022-01-04 | CRAN (R 4.2.0) |
| htmltools | 0.5.2 | 2021-08-25 | CRAN (R 4.2.0) |
| htmlwidgets | 1.5.4 | 2021-09-08 | CRAN (R 4.2.0) |
| httr | 1.4.3 | 2022-05-04 | CRAN (R 4.2.0) |
| igraph | 1.3.1 | 2022-04-20 | CRAN (R 4.2.0) |
| insight | 0.17.1 | 2022-05-13 | CRAN (R 4.2.0) |
| jpeg | 0.1-9 | 2021-07-24 | CRAN (R 4.2.0) |
| jquerylib | 0.1.4 | 2021-04-26 | CRAN (R 4.2.0) |
| jsonlite | 1.8.0 | 2022-02-22 | CRAN (R 4.2.0) |
| kableExtra | 1.3.4 | 2021-02-20 | CRAN (R 4.2.0) |
| knitr | 1.39 | 2022-04-26 | CRAN (R 4.2.0) |
| kutils | 1.70 | 2020-04-29 | CRAN (R 4.2.0) |
| labelled | 2.9.1 | 2022-05-05 | CRAN (R 4.2.0) |
| lattice | 0.20-45 | 2021-09-22 | CRAN (R 4.2.1) |
| latticeExtra | 0.6-29 | 2019-12-19 | CRAN (R 4.2.0) |
| lavaan | 0.6-11 | 2022-03-31 | CRAN (R 4.2.0) |
| lifecycle | 1.0.1 | 2021-09-24 | CRAN (R 4.2.0) |
| likert | 1.3.5 | 2016-12-31 | CRAN (R 4.2.0) |
| lisrelToR | 0.1.5 | 2022-05-09 | CRAN (R 4.2.0) |
| lme4 | 1.1-29 | 2022-04-07 | CRAN (R 4.2.0) |
| lubridate | 1.8.0 | 2021-10-07 | CRAN (R 4.2.0) |
| magick | 2.7.3 | 2021-08-18 | CRAN (R 4.2.0) |
| magrittr | 2.0.3 | 2022-03-30 | CRAN (R 4.2.0) |
| MASS | 7.3-57 | 2022-04-22 | CRAN (R 4.2.1) |
| Matrix | 1.4-1 | 2022-03-23 | CRAN (R 4.2.1) |
| matrixStats | 0.62.0 | 2022-04-19 | CRAN (R 4.2.0) |
| memoise | 2.0.1 | 2021-11-26 | CRAN (R 4.2.0) |
| mi | 1.1 | 2022-06-06 | CRAN (R 4.2.0) |
| minqa | 1.2.4 | 2014-10-09 | CRAN (R 4.2.0) |
| mnormt | 2.1.0 | 2022-06-07 | CRAN (R 4.2.0) |
| modelr | 0.1.8 | 2020-05-19 | CRAN (R 4.2.0) |
| munsell | 0.5.0 | 2018-06-12 | CRAN (R 4.2.0) |
| mvtnorm | 1.1-3 | 2021-10-08 | CRAN (R 4.2.0) |
| nlme | 3.1-157 | 2022-03-25 | CRAN (R 4.2.1) |
| nloptr | 2.0.3 | 2022-05-26 | CRAN (R 4.2.0) |
| nnet | 7.3-17 | 2022-01-16 | CRAN (R 4.2.1) |
| OpenMx | 2.20.6 | 2022-03-09 | CRAN (R 4.2.0) |
| openxlsx | 4.2.5 | 2021-12-14 | CRAN (R 4.2.0) |
| pander | 0.6.5 | 2022-03-18 | CRAN (R 4.2.0) |
| parameters | 0.18.1 | 2022-05-29 | CRAN (R 4.2.0) |
| pbapply | 1.5-0 | 2021-09-16 | CRAN (R 4.2.0) |
| pbivnorm | 0.6.0 | 2015-01-23 | CRAN (R 4.2.0) |
| performance | 0.9.0 | 2022-03-30 | CRAN (R 4.2.0) |
| pillar | 1.7.0 | 2022-02-01 | CRAN (R 4.2.0) |
| pkgbuild | 1.3.1 | 2021-12-20 | CRAN (R 4.2.0) |
| pkgconfig | 2.0.3 | 2019-09-22 | CRAN (R 4.2.0) |
| pkgload | 1.2.4 | 2021-11-30 | CRAN (R 4.2.0) |
| plyr | 1.8.7 | 2022-03-24 | CRAN (R 4.2.0) |
| png | 0.1-7 | 2013-12-03 | CRAN (R 4.2.0) |
| prettyunits | 1.1.1 | 2020-01-24 | CRAN (R 4.2.0) |
| processx | 3.6.0 | 2022-06-10 | CRAN (R 4.2.0) |
| pryr | 0.1.5 | 2021-07-26 | CRAN (R 4.2.0) |
| ps | 1.7.0 | 2022-04-23 | CRAN (R 4.2.0) |
| psych | 2.2.5 | 2022-05-10 | CRAN (R 4.2.0) |
| purrr | 0.3.4 | 2020-04-17 | CRAN (R 4.2.0) |
| qgraph | 1.9.2 | 2022-03-04 | CRAN (R 4.2.0) |
| R6 | 2.5.1 | 2021-08-19 | CRAN (R 4.2.0) |
| rapportools | 1.1 | 2022-03-22 | CRAN (R 4.2.0) |
| RColorBrewer | 1.1-3 | 2022-04-03 | CRAN (R 4.2.0) |
| Rcpp | 1.0.8.3 | 2022-03-17 | CRAN (R 4.2.0) |
| RcppParallel | 5.1.5 | 2022-01-05 | CRAN (R 4.2.0) |
| readr | 2.1.2 | 2022-01-30 | CRAN (R 4.2.0) |
| readxl | 1.4.0 | 2022-03-28 | CRAN (R 4.2.0) |
| remotes | 2.4.2 | 2021-11-30 | CRAN (R 4.2.0) |
| reprex | 2.0.1 | 2021-08-05 | CRAN (R 4.2.0) |
| reshape2 | 1.4.4 | 2020-04-09 | CRAN (R 4.2.0) |
| rio | 0.5.29 | 2021-11-22 | CRAN (R 4.2.0) |
| rlang | 1.0.3 | 2022-06-27 | CRAN (R 4.2.0) |
| rmarkdown | 2.14 | 2022-04-25 | CRAN (R 4.2.0) |
| rockchalk | 1.8.152 | 2022-05-20 | CRAN (R 4.2.0) |
| rpart | 4.1.16 | 2022-01-24 | CRAN (R 4.2.1) |
| rprojroot | 2.0.3 | 2022-04-02 | CRAN (R 4.2.0) |
| rstudioapi | 0.13 | 2020-11-12 | CRAN (R 4.2.0) |
| rvest | 1.0.2 | 2021-10-16 | CRAN (R 4.2.0) |
| sass | 0.4.1 | 2022-03-23 | CRAN (R 4.2.0) |
| scales | 1.2.0 | 2022-04-13 | CRAN (R 4.2.0) |
| sem | 3.1-15 | 2022-04-10 | CRAN (R 4.2.0) |
| semPlot | 1.1.5 | 2022-03-08 | CRAN (R 4.2.0) |
| sessioninfo | 1.2.2 | 2021-12-06 | CRAN (R 4.2.0) |
| sjlabelled | 1.2.0 | 2022-04-10 | CRAN (R 4.2.0) |
| sjmisc | 2.8.9 | 2021-12-03 | CRAN (R 4.2.0) |
| sjPlot | 2.8.10 | 2021-11-26 | CRAN (R 4.2.0) |
| sjstats | 0.18.1 | 2021-01-09 | CRAN (R 4.2.0) |
| stringi | 1.7.6 | 2021-11-29 | CRAN (R 4.2.0) |
| stringr | 1.4.0 | 2019-02-10 | CRAN (R 4.2.0) |
| summarytools | 1.0.1 | 2022-05-20 | CRAN (R 4.2.0) |
| survival | 3.3-1 | 2022-03-03 | CRAN (R 4.2.1) |
| svglite | 2.1.0 | 2022-02-03 | CRAN (R 4.2.0) |
| systemfonts | 1.0.4 | 2022-02-11 | CRAN (R 4.2.0) |
| testthat | 3.1.4 | 2022-04-26 | CRAN (R 4.2.0) |
| tibble | 3.1.7 | 2022-05-03 | CRAN (R 4.2.0) |
| tidyr | 1.2.0 | 2022-02-01 | CRAN (R 4.2.0) |
| tidyselect | 1.1.2 | 2022-02-21 | CRAN (R 4.2.0) |
| tidyverse | 1.3.1 | 2021-04-15 | CRAN (R 4.2.0) |
| tzdb | 0.3.0 | 2022-03-28 | CRAN (R 4.2.0) |
| usethis | 2.1.6 | 2022-05-25 | CRAN (R 4.2.0) |
| utf8 | 1.2.2 | 2021-07-24 | CRAN (R 4.2.0) |
| vctrs | 0.4.1 | 2022-04-13 | CRAN (R 4.2.0) |
| viridisLite | 0.4.0 | 2021-04-13 | CRAN (R 4.2.0) |
| webshot | 0.5.3 | 2022-04-14 | CRAN (R 4.2.0) |
| withr | 2.5.0 | 2022-03-03 | CRAN (R 4.2.0) |
| xfun | 0.31 | 2022-05-10 | CRAN (R 4.2.0) |
| XML | 3.99-0.10 | 2022-06-09 | CRAN (R 4.2.0) |
| xml2 | 1.3.3 | 2021-11-30 | CRAN (R 4.2.0) |
| xtable | 1.8-4 | 2019-04-21 | CRAN (R 4.2.0) |
| yaml | 2.3.5 | 2022-02-21 | CRAN (R 4.2.0) |
| zip | 2.2.0 | 2021-05-31 | CRAN (R 4.2.0) |
II. MPQ cleaning and descriptives
#load data
mpq <- rio::import(file = "./Data/SIBS_LONGITUDINAL_PERSONALITY_20210413.csv") %>%
select(-1)
rel_IN <- rio::import(file = "./Data/rel_IN.sav")
rel_FU1 <- rio::import(file = "./Data/rel_FU1.sav")
lei <- rio::import(file = './Data/lei.sav')
lei[is.na(lei)] <- 0
#aggression items
mpq$AG_IN <- mpq %>%
select(starts_with("AG") & ends_with("IN")) %>%
rowMeans(na.rm = T)
mpq$AG_FU1 <- mpq %>%
select(starts_with("AG") & ends_with("FU1")) %>%
rowMeans(na.rm = T)
mpq$AG_FU2 <- mpq %>%
select(starts_with("AG") & ends_with("FU2")) %>%
rowMeans(na.rm = T)
mpq$AG_FU3 <- mpq %>%
select(starts_with("AG") & ends_with("FU3")) %>%
rowMeans(na.rm = T)
#control items
mpq$CON_IN <- mpq %>%
select(starts_with("CON") & ends_with("IN")) %>%
rowMeans(na.rm = T)
mpq$CON_FU1 <- mpq %>%
select(starts_with("CON") & ends_with("FU1")) %>%
rowMeans(na.rm = T)
mpq$CON_FU2 <- mpq %>%
select(starts_with("CON") & ends_with("FU2")) %>%
rowMeans(na.rm = T)
mpq$CON_FU3 <- mpq %>%
select(starts_with("CON") & ends_with("FU3")) %>%
rowMeans(na.rm = T)
#harm avoidance items
mpq$HA_IN <- mpq %>%
select(starts_with("HA") & ends_with("IN")) %>%
rowMeans(na.rm = T)
mpq$HA_FU1 <- mpq %>%
select(starts_with("HA") & ends_with("FU1")) %>%
rowMeans(na.rm = T)
mpq$HA_FU2 <- mpq %>%
select(starts_with("HA") & ends_with("FU2")) %>%
rowMeans(na.rm = T)
mpq$HA_FU3 <- mpq %>%
select(starts_with("HA") & ends_with("FU3")) %>%
rowMeans(na.rm = T)
#clean up NaN
mpq$AG_IN[is.nan(mpq$AG_IN)]<-NA
mpq$AG_FU1[is.nan(mpq$AG_FU1)]<-NA
mpq$AG_FU2[is.nan(mpq$AG_FU2)]<-NA
mpq$AG_FU3[is.nan(mpq$AG_FU3)]<-NA
mpq$CON_IN[is.nan(mpq$CON_IN)]<-NA
mpq$CON_FU1[is.nan(mpq$CON_FU1)]<-NA
mpq$CON_FU2[is.nan(mpq$CON_FU2)]<-NA
mpq$CON_FU3[is.nan(mpq$CON_FU3)]<-NA
mpq$HA_IN[is.nan(mpq$HA_IN)]<-NA
mpq$HA_FU1[is.nan(mpq$HA_FU1)]<-NA
mpq$HA_FU2[is.nan(mpq$HA_FU2)]<-NA
mpq$HA_FU3[is.nan(mpq$HA_FU3)]<-NA
# participants with relationship data at FU1 but not IN
FU1_ID <- rel_FU1[which(!rel_FU1$ID %in% rel_IN$ID), "ID"]
rel_FU1 <- rel_FU1 %>%
filter(ID %in% FU1_ID)
# merge relationship data (IN for all, FU1 for those without IN data)
mpq_IN <- mpq %>% filter(!ID %in% FU1_ID)
mpq_FU1 <- mpq %>% filter(ID %in% FU1_ID)
mpq_IN <- merge(mpq_IN, rel_IN, all.x = TRUE)
mpq_FU1 <- merge(mpq_FU1, rel_FU1, all.x = TRUE)
mpq <- rbind(mpq_IN, mpq_FU1)
# merge life events data
mpq <- merge(mpq, lei, all.x = TRUE)
# reverse code some relationship items (into new items _R)
#mpq <- mpq %>%
# mutate(S_ENURY_R = 6 - S_ENURY,
# S_YADME_R = 6 - S_YADME,
# S_EADMY_R = 6 - S_EADMY,
# S_EDOMY_R = 6 - S_EDOMY,
# S_AFFECT_R = 6 - S_AFFECT)
#create wide dyad dataframe
mpq <- mpq[order(mpq$ID),]
mpq$P <- rep(c(1:2), length.out = nrow(mpq)) #person dummy var
yo <- mpq %>% select(IDYRFAM, BDAY, P) %>%
mutate(BDAY = as.character(BDAY),
P = as.character(P))
#wide data with 2 siblings per row
yo <- reshape2::dcast(yo, IDYRFAM ~ P, value.var = "BDAY")
yo <- yo %>% mutate(`1` = as.Date(`1`),
`2` = as.Date(`2`))
yo <- yo %>% mutate(older = ifelse(`1`<`2`, "P1", "P2")) %>%
select(IDYRFAM, older)
mpq <- merge(mpq, yo)
#y = younger, o = older
mpq <- mpq %>%
mutate(yo = ifelse(P == 1 & older == "P2" |
P == 2 & older == "P1", "y", "o")) %>%
select(-P, -older)
#data with only younger sib
young <- mpq %>%
filter(yo == "y")
names(young) <- paste0(names(young), "_y", sep = "")
young <- young %>% rename(IDYRFAM = IDYRFAM_y)
#data with only older sib
old <- mpq %>%
filter(yo == "o")
names(old) <- paste0(names(old), "_o", sep = "")
old <- old %>% rename(IDYRFAM = IDYRFAM_o)
yo <- merge(young, old)
rm(young, old)Demographics
#descriptives
mpq %>%
select(IDSEX:FU3_AGE) %>%
descr(stats = "common", order = "p")Descriptive Statistics
mpq
N: 1231
| IDSEX | IDAB | IDFAMAB | IDSC | IN_AGE | FU1_AGE | FU2_AGE | FU3_AGE | |
|---|---|---|---|---|---|---|---|---|
| Mean | 1.55 | 1.44 | 1.74 | 3.12 | 14.92 | 18.05 | 20.10 | 31.92 |
| Std.Dev | 0.50 | 0.50 | 0.77 | 2.26 | 1.93 | 1.98 | 1.08 | 2.71 |
| Min | 1.00 | 1.00 | 1.00 | 0.00 | 10.73 | 13.67 | 18.95 | 25.66 |
| Median | 2.00 | 1.00 | 2.00 | 3.00 | 14.98 | 18.05 | 19.97 | 31.76 |
| Max | 2.00 | 2.00 | 3.00 | 7.00 | 20.98 | 25.34 | 23.09 | 40.53 |
| N.Valid | 1231.00 | 1231.00 | 1231.00 | 1231.00 | 1221.00 | 1065.00 | 132.00 | 769.00 |
| Pct.Valid | 100.00 | 100.00 | 100.00 | 100.00 | 99.19 | 86.52 | 10.72 | 62.47 |
# graphs: continuous
ggplot(data = yo, aes(IN_AGE_y)) +
geom_histogram(binwidth = 0.2, fill = "#710c0c") +
geom_vline(xintercept = mean(yo$IN_AGE_y, na.rm = T),
color = "#f1c232", size = 1) +
annotate("text",
label = paste0("Mean = ", round(mean(yo$IN_AGE_y, na.rm = T), 2),
", SD = ",round(sd(yo$IN_AGE_y, na.rm = T), 2)),
x = 17, y = 25) +
labs(
title = "Distribution of younger sibling's age at intake",
x = "Age (in years)",
y = NULL
) +
theme_classic()
ggplot(data = yo, aes(IN_AGE_o)) +
geom_histogram(binwidth = 0.2, fill = "#710c0c") +
geom_vline(xintercept = mean(yo$IN_AGE_o, na.rm = T),
color = "#f1c232", size = 1) +
annotate("text",
label = paste0("Mean = ", round(mean(yo$IN_AGE_o, na.rm = T), 2),
", SD = ",round(sd(yo$IN_AGE_o, na.rm = T), 2)),
x = 18, y = 35) +
labs(
title = "Distribution of older sibling's age at intake",
x = "Age (in years)",
y = NULL
) +
theme_classic()
# graphs: discrete
ggplot(mpq,
aes(factor(mpq$IDAB,
levels = c(1, 2),
labels = c("Adoptive", "Non-Adoptive")))) +
geom_bar(fill = c("#710c0c", "#f1c232")) +
geom_text(stat='count', aes(label=..count..), vjust = -1) +
labs(
title = "Count of adoption status",
x = NULL
) +
ylim(0, 700) +
theme_classic()
ggplot(mpq,
aes(factor(mpq$IDSEX,
levels = c(1, 2),
labels = c("Male", "Female")))) +
geom_bar(fill = c("#f1c232", "#710c0c")) +
geom_text(stat='count', aes(label=..count..), vjust = -1) +
labs(
title = "Count of sex",
x = NULL
) +
ylim(0, 700) +
theme_classic()
The average time between assessments is:
- Between intake and follow-up 1: 3.306 years.
- Between follow-up 1 and follow-up 2: 4.283 years.
- Between follow-up 3 and follow-up 2: 9.393 years.
Scale scores
#descriptives
mpq %>% select(AG_IN:HA_FU3) %>% descr(stats = "common", order = "p")Descriptive Statistics
mpq
N: 1231
| AG_IN | AG_FU1 | AG_FU2 | AG_FU3 | CON_IN | CON_FU1 | CON_FU2 | CON_FU3 | |
|---|---|---|---|---|---|---|---|---|
| Mean | 2.16 | 2.06 | 1.89 | 1.71 | 2.58 | 2.63 | 2.79 | 2.98 |
| Std.Dev | 0.58 | 0.56 | 0.49 | 0.44 | 0.47 | 0.48 | 0.47 | 0.45 |
| Min | 1.00 | 1.00 | 1.00 | 1.00 | 1.17 | 1.07 | 1.67 | 1.25 |
| Median | 2.11 | 2.00 | 1.89 | 1.67 | 2.61 | 2.67 | 2.83 | 3.00 |
| Max | 4.00 | 4.00 | 3.22 | 3.75 | 4.00 | 3.94 | 3.83 | 4.00 |
| N.Valid | 1221.00 | 1061.00 | 132.00 | 748.00 | 1221.00 | 1061.00 | 132.00 | 748.00 |
| Pct.Valid | 99.19 | 86.19 | 10.72 | 60.76 | 99.19 | 86.19 | 10.72 | 60.76 |
| HA_IN | HA_FU1 | HA_FU2 | HA_FU3 | |
|---|---|---|---|---|
| Mean | 2.66 | 2.60 | 2.63 | 3.03 |
| Std.Dev | 0.60 | 0.60 | 0.66 | 0.57 |
| Min | 1.00 | 1.00 | 1.11 | 1.33 |
| Median | 2.69 | 2.61 | 2.75 | 3.08 |
| Max | 4.00 | 4.00 | 3.94 | 4.00 |
| N.Valid | 1221.00 | 1061.00 | 132.00 | 748.00 |
| Pct.Valid | 99.19 | 86.19 | 10.72 | 60.76 |
#random sample for plots
sample = sample(mpq$ID, size = 100)
sample <- mpq %>% filter(ID %in% c(sample))
sample <- reshape(sample, direction = "long",
varying = list(c("IN_AGE", "FU1_AGE", "FU2_AGE", "FU3_AGE"),
c("AG_IN", "AG_FU1", "AG_FU2", "AG_FU3"),
c("CON_IN", "CON_FU1", "CON_FU2", "CON_FU3"),
c("HA_IN", "HA_FU1", "HA_FU2", "HA_FU3")),
timevar = "time",
times = c(0,1,2,3),
v.names = c("age","AG","CON","HA"),
idvar = c("ID"))
row.names(sample) <- 1:nrow(sample)
#plot by timepoints
pAG <- ggplot(data = sample,
aes(x = time, y = AG, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$AG,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
pCON <- ggplot(data = sample,
aes(x = time, y = CON, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$CON,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
pHA <- ggplot(data = sample,
aes(x = time, y = HA, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.5) +
theme_bw () +
stat_summary(aes(data=sample$HA,group=1),fun=mean,geom="line",lwd = 1.5, color= "red")
#plot by age
paAG <- ggplot(data = sample,
aes(x = age, y = AG, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$AG,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
paCON <- ggplot(data = sample,
aes(x = age, y = CON, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$CON,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
paHA <- ggplot(data = sample,
aes(x = age, y = HA, group = ID)) +
geom_line (linetype = "dashed")+
geom_point(size = 0.25) +
theme_bw () +
stat_smooth(aes(data=sample$HA,group=1),method="lm",formula=y ~ poly(x, 2),lwd = 1.5, color= "red")
cowplot::plot_grid(pAG, paAG,
pCON, paCON,
pHA, paHA,
nrow = 3, ncol = 2)
rm(sample, pAG, paAG, pCON, paCON, pHA, paHA)There are 1221 participants with complete responses at intake, 1061 at follow-up 1, 132 at follow-up 2, and 748 at follow-up 3.
There are 615 families with at least 1 sibling at intake, 569 at follow-up 1, 123 at follow-up 2, and 495 at follow-up 3.
There are 606 families with both siblings at intake, 492 at follow-up 1, 9 at follow-up 2, and 253 at follow-up 3.
III. Analyses with 3 waves
These analyses do NOT use data points at wave 3 (FU2) due to very high and non-random missingness of the study design. The supplementary analyses (section IV.) includes these data points for robustness checks. Missingness was too severe such that when grouping variables (e.g., adoption status) were added, some variables only have one observations across older siblings and as a result the models could not run.
Random parcel sampling
par <- c(1:18)
par1 <- sample(par, size = 6) #parcel 1
par2 <- sample(par[which(!par %in% par1)], size = 6) #parcel 2
par3 <- par[which(!par %in% par1 & !par %in% par2)] #parcel 3
# aggression AG
yo <- yo %>%
mutate(# older sibling
AG_IN_o_p1 = rowMeans(select(yo, AG12_IN_o, AG18_IN_o, AG4_IN_o, AG15_IN_o, AG6_IN_o, AG2_IN_o), na.rm = T),
AG_IN_o_p2 = rowMeans(select(yo, AG10_IN_o, AG17_IN_o, AG7_IN_o, AG13_IN_o, AG1_IN_o, AG3_IN_o), na.rm = T),
AG_IN_o_p3 = rowMeans(select(yo, AG5_IN_o, AG8_IN_o, AG9_IN_o, AG11_IN_o, AG14_IN_o, AG16_IN_o), na.rm = T),
AG_FU1_o_p1 = rowMeans(select(yo, AG12_FU1_o, AG18_FU1_o, AG4_FU1_o, AG15_FU1_o, AG6_FU1_o, AG2_FU1_o), na.rm = T),
AG_FU1_o_p2 = rowMeans(select(yo, AG10_FU1_o, AG17_FU1_o, AG7_FU1_o, AG13_FU1_o, AG1_FU1_o, AG3_FU1_o), na.rm = T),
AG_FU1_o_p3 = rowMeans(select(yo, AG5_FU1_o, AG8_FU1_o, AG9_FU1_o, AG11_FU1_o, AG14_FU1_o, AG16_FU1_o), na.rm = T),
AG_FU2_o_p1 = rowMeans(select(yo, AG12_FU2_o, AG18_FU2_o, AG4_FU2_o, AG15_FU2_o, AG6_FU2_o, AG2_FU2_o), na.rm = T),
AG_FU2_o_p2 = rowMeans(select(yo, AG10_FU2_o, AG17_FU2_o, AG7_FU2_o, AG13_FU2_o, AG1_FU2_o, AG3_FU2_o), na.rm = T),
AG_FU2_o_p3 = rowMeans(select(yo, AG5_FU2_o, AG8_FU2_o, AG9_FU2_o, AG11_FU2_o, AG14_FU2_o, AG16_FU2_o), na.rm = T),
AG_FU3_o_p1 = rowMeans(select(yo, AG12_FU3_o, AG18_FU3_o, AG4_FU3_o, AG6_FU3_o, AG2_FU3_o), na.rm = T),
AG_FU3_o_p2 = rowMeans(select(yo, AG10_FU3_o, AG7_FU3_o), na.rm = T),
AG_FU3_o_p3 = rowMeans(select(yo, AG5_FU3_o, AG8_FU3_o, AG11_FU3_o, AG14_FU3_o, AG16_FU3_o), na.rm = T),
# younger sibling
AG_IN_y_p1 = rowMeans(select(yo, AG12_IN_y, AG18_IN_y, AG4_IN_y, AG15_IN_y, AG6_IN_y, AG2_IN_y), na.rm = T),
AG_IN_y_p2 = rowMeans(select(yo, AG10_IN_y, AG17_IN_y, AG7_IN_y, AG13_IN_y, AG1_IN_y, AG3_IN_y), na.rm = T),
AG_IN_y_p3 = rowMeans(select(yo, AG5_IN_y, AG8_IN_y, AG9_IN_y, AG11_IN_y, AG14_IN_y, AG16_IN_y), na.rm = T),
AG_FU1_y_p1 = rowMeans(select(yo, AG12_FU1_y, AG18_FU1_y, AG4_FU1_y, AG15_FU1_y, AG6_FU1_y, AG2_FU1_y), na.rm = T),
AG_FU1_y_p2 = rowMeans(select(yo, AG10_FU1_y, AG17_FU1_y, AG7_FU1_y, AG13_FU1_y, AG1_FU1_y, AG3_FU1_y), na.rm = T),
AG_FU1_y_p3 = rowMeans(select(yo, AG5_FU1_y, AG8_FU1_y, AG9_FU1_y, AG11_FU1_y, AG14_FU1_y, AG16_FU1_y), na.rm = T),
AG_FU2_y_p1 = rowMeans(select(yo, AG12_FU2_y, AG18_FU2_y, AG4_FU2_y, AG15_FU2_y, AG6_FU2_y, AG2_FU2_y), na.rm = T),
AG_FU2_y_p2 = rowMeans(select(yo, AG10_FU2_y, AG17_FU2_y, AG7_FU2_y, AG13_FU2_y, AG1_FU2_y, AG3_FU2_y), na.rm = T),
AG_FU2_y_p3 = rowMeans(select(yo, AG5_FU2_y, AG8_FU2_y, AG9_FU2_y, AG11_FU2_y, AG14_FU2_y, AG16_FU2_y), na.rm = T),
AG_FU3_y_p1 = rowMeans(select(yo, AG12_FU3_y, AG18_FU3_y, AG4_FU3_y, AG6_FU3_y, AG2_FU3_y), na.rm = T),
AG_FU3_y_p2 = rowMeans(select(yo, AG10_FU3_y, AG7_FU3_y), na.rm = T),
AG_FU3_y_p3 = rowMeans(select(yo, AG5_FU3_y, AG8_FU3_y, AG11_FU3_y, AG14_FU3_y, AG16_FU3_y), na.rm = T))
# control CON
yo <- yo %>%
mutate(# older sibling
CON_IN_o_p1 = rowMeans(select(yo, CON12_IN_o, CON18_IN_o, CON4_IN_o, CON15_IN_o, CON6_IN_o, CON2_IN_o), na.rm = T),
CON_IN_o_p2 = rowMeans(select(yo, CON10_IN_o, CON17_IN_o, CON7_IN_o, CON13_IN_o, CON1_IN_o, CON3_IN_o), na.rm = T),
CON_IN_o_p3 = rowMeans(select(yo, CON5_IN_o, CON8_IN_o, CON9_IN_o, CON11_IN_o, CON14_IN_o, CON16_IN_o), na.rm = T),
CON_FU1_o_p1 = rowMeans(select(yo, CON12_FU1_o, CON18_FU1_o, CON4_FU1_o, CON15_FU1_o, CON6_FU1_o, CON2_FU1_o), na.rm = T),
CON_FU1_o_p2 = rowMeans(select(yo, CON10_FU1_o, CON17_FU1_o, CON7_FU1_o, CON13_FU1_o, CON1_FU1_o, CON3_FU1_o), na.rm = T),
CON_FU1_o_p3 = rowMeans(select(yo, CON5_FU1_o, CON8_FU1_o, CON9_FU1_o, CON11_FU1_o, CON14_FU1_o, CON16_FU1_o), na.rm = T),
CON_FU2_o_p1 = rowMeans(select(yo, CON12_FU2_o, CON18_FU2_o, CON4_FU2_o, CON15_FU2_o, CON6_FU2_o, CON2_FU2_o), na.rm = T),
CON_FU2_o_p2 = rowMeans(select(yo, CON10_FU2_o, CON17_FU2_o, CON7_FU2_o, CON13_FU2_o, CON1_FU2_o, CON3_FU2_o), na.rm = T),
CON_FU2_o_p3 = rowMeans(select(yo, CON5_FU2_o, CON8_FU2_o, CON9_FU2_o, CON11_FU2_o, CON14_FU2_o, CON16_FU2_o), na.rm = T),
CON_FU3_o_p1 = rowMeans(select(yo, CON4_FU3_o, CON15_FU3_o, CON6_FU3_o, CON2_FU3_o), na.rm = T),
CON_FU3_o_p2 = rowMeans(select(yo, CON10_FU3_o, CON17_FU3_o, CON7_FU3_o), na.rm = T),
CON_FU3_o_p3 = rowMeans(select(yo, CON5_FU3_o, CON8_FU3_o, CON9_FU3_o, CON16_FU3_o), na.rm = T),
# younger sibling
CON_IN_y_p1 = rowMeans(select(yo, CON12_IN_y, CON18_IN_y, CON4_IN_y, CON15_IN_y, CON6_IN_y, CON2_IN_y), na.rm = T),
CON_IN_y_p2 = rowMeans(select(yo, CON10_IN_y, CON17_IN_y, CON7_IN_y, CON13_IN_y, CON1_IN_y, CON3_IN_y), na.rm = T),
CON_IN_y_p3 = rowMeans(select(yo, CON5_IN_y, CON8_IN_y, CON9_IN_y, CON11_IN_y, CON14_IN_y, CON16_IN_y), na.rm = T),
CON_FU1_y_p1 = rowMeans(select(yo, CON12_FU1_y, CON18_FU1_y, CON4_FU1_y, CON15_FU1_y, CON6_FU1_y, CON2_FU1_y), na.rm = T),
CON_FU1_y_p2 = rowMeans(select(yo, CON10_FU1_y, CON17_FU1_y, CON7_FU1_y, CON13_FU1_y, CON1_FU1_y, CON3_FU1_y), na.rm = T),
CON_FU1_y_p3 = rowMeans(select(yo, CON5_FU1_y, CON8_FU1_y, CON9_FU1_y, CON11_FU1_y, CON14_FU1_y, CON16_FU1_y), na.rm = T),
CON_FU2_y_p1 = rowMeans(select(yo, CON12_FU2_y, CON18_FU2_y, CON4_FU2_y, CON15_FU2_y, CON6_FU2_y, CON2_FU2_y), na.rm = T),
CON_FU2_y_p2 = rowMeans(select(yo, CON10_FU2_y, CON17_FU2_y, CON7_FU2_y, CON13_FU2_y, CON1_FU2_y, CON3_FU2_y), na.rm = T),
CON_FU2_y_p3 = rowMeans(select(yo, CON5_FU2_y, CON8_FU2_y, CON9_FU2_y, CON11_FU2_y, CON14_FU2_y, CON16_FU2_y), na.rm = T),
CON_FU3_y_p1 = rowMeans(select(yo, CON4_FU3_y, CON15_FU3_y, CON6_FU3_y, CON2_FU3_y), na.rm = T),
CON_FU3_y_p2 = rowMeans(select(yo, CON10_FU3_y, CON17_FU3_y, CON7_FU3_y), na.rm = T),
CON_FU3_y_p3 = rowMeans(select(yo, CON5_FU3_y, CON8_FU3_y, CON9_FU3_y, CON16_FU3_y), na.rm = T))
# harm avoidance HA
yo <- yo %>%
mutate(# older sibling
HA_IN_o_p1 = rowMeans(select(yo, HA12_IN_o, HA18_IN_o, HA4_IN_o, HA15_IN_o, HA6_IN_o, HA2_IN_o), na.rm = T),
HA_IN_o_p2 = rowMeans(select(yo, HA10_IN_o, HA17_IN_o, HA7_IN_o, HA13_IN_o, HA1_IN_o, HA3_IN_o), na.rm = T),
HA_IN_o_p3 = rowMeans(select(yo, HA5_IN_o, HA8_IN_o, HA9_IN_o, HA11_IN_o, HA14_IN_o, HA16_IN_o), na.rm = T),
HA_FU1_o_p1 = rowMeans(select(yo, HA12_FU1_o, HA18_FU1_o, HA4_FU1_o, HA15_FU1_o, HA6_FU1_o, HA2_FU1_o), na.rm = T),
HA_FU1_o_p2 = rowMeans(select(yo, HA10_FU1_o, HA17_FU1_o, HA7_FU1_o, HA13_FU1_o, HA1_FU1_o, HA3_FU1_o), na.rm = T),
HA_FU1_o_p3 = rowMeans(select(yo, HA5_FU1_o, HA8_FU1_o, HA9_FU1_o, HA11_FU1_o, HA14_FU1_o, HA16_FU1_o), na.rm = T),
HA_FU2_o_p1 = rowMeans(select(yo, HA12_FU2_o, HA18_FU2_o, HA4_FU2_o, HA15_FU2_o, HA6_FU2_o, HA2_FU2_o), na.rm = T),
HA_FU2_o_p2 = rowMeans(select(yo, HA10_FU2_o, HA17_FU2_o, HA7_FU2_o, HA13_FU2_o, HA1_FU2_o, HA3_FU2_o), na.rm = T),
HA_FU2_o_p3 = rowMeans(select(yo, HA5_FU2_o, HA8_FU2_o, HA9_FU2_o, HA11_FU2_o, HA14_FU2_o, HA16_FU2_o), na.rm = T),
HA_FU3_o_p1 = rowMeans(select(yo, HA18_FU3_o, HA4_FU3_o, HA6_FU3_o, HA2_FU3_o), na.rm = T),
HA_FU3_o_p2 = rowMeans(select(yo, HA10_FU3_o, HA7_FU3_o, HA13_FU3_o, HA1_FU3_o, HA3_FU3_o), na.rm = T),
HA_FU3_o_p3 = rowMeans(select(yo, HA8_FU3_o, HA9_FU3_o, HA16_FU3_o), na.rm = T),
# younger sibling
HA_IN_y_p1 = rowMeans(select(yo, HA12_IN_y, HA18_IN_y, HA4_IN_y, HA15_IN_y, HA6_IN_y, HA2_IN_y), na.rm = T),
HA_IN_y_p2 = rowMeans(select(yo, HA10_IN_y, HA17_IN_y, HA7_IN_y, HA13_IN_y, HA1_IN_y, HA3_IN_y), na.rm = T),
HA_IN_y_p3 = rowMeans(select(yo, HA5_IN_y, HA8_IN_y, HA9_IN_y, HA11_IN_y, HA14_IN_y, HA16_IN_y), na.rm = T),
HA_FU1_y_p1 = rowMeans(select(yo, HA12_FU1_y, HA18_FU1_y, HA4_FU1_y, HA15_FU1_y, HA6_FU1_y, HA2_FU1_y), na.rm = T),
HA_FU1_y_p2 = rowMeans(select(yo, HA10_FU1_y, HA17_FU1_y, HA7_FU1_y, HA13_FU1_y, HA1_FU1_y, HA3_FU1_y), na.rm = T),
HA_FU1_y_p3 = rowMeans(select(yo, HA5_FU1_y, HA8_FU1_y, HA9_FU1_y, HA11_FU1_y, HA14_FU1_y, HA16_FU1_y), na.rm = T),
HA_FU2_y_p1 = rowMeans(select(yo, HA12_FU2_y, HA18_FU2_y, HA4_FU2_y, HA15_FU2_y, HA6_FU2_y, HA2_FU2_y), na.rm = T),
HA_FU2_y_p2 = rowMeans(select(yo, HA10_FU2_y, HA17_FU2_y, HA7_FU2_y, HA13_FU2_y, HA1_FU2_y, HA3_FU2_y), na.rm = T),
HA_FU2_y_p3 = rowMeans(select(yo, HA5_FU2_y, HA8_FU2_y, HA9_FU2_y, HA11_FU2_y, HA14_FU2_y, HA16_FU2_y), na.rm = T),
HA_FU3_y_p1 = rowMeans(select(yo, HA18_FU3_y, HA4_FU3_y, HA6_FU3_y, HA2_FU3_y), na.rm = T),
HA_FU3_y_p2 = rowMeans(select(yo, HA10_FU3_y, HA7_FU3_y, HA13_FU3_y, HA1_FU3_y, HA3_FU3_y), na.rm = T),
HA_FU3_y_p3 = rowMeans(select(yo, HA8_FU3_y, HA9_FU3_y, HA16_FU3_y), na.rm = T))
# clean up NaN
for (row in seq(nrow(yo))) {
for (col in seq(ncol(yo))) {
if (is.nan(yo[row, col])) {
yo[row, col] <- NA
}}} # END for row/col LOOPFor SEM analyses of each personality variable, the three parcels were randomized to contain items numbered:
- Parcel 1: 2, 4, 6, 7, 14, 15
- Parcel 2: 3, 8, 9, 12, 13, 17
- Parcel 3: 1, 5, 10, 11, 16, 18
General change trajectories and correlated change
Aggression AG
lgmAG <- '
#younger sibling
AG0y =~ AG_IN_y_p1 + p2 * AG_IN_y_p2 + p3 * AG_IN_y_p3
AG3y =~ AG_FU1_y_p1 + p2 * AG_FU1_y_p2 + p3 * AG_FU1_y_p3
AG17y =~ AG_FU3_y_p1 + p2 * AG_FU3_y_p2 + p3 * AG_FU3_y_p3
#older sibling
AG0o =~ AG_IN_o_p1 + p2 * AG_IN_o_p2 + p3 * AG_IN_o_p3
AG3o =~ AG_FU1_o_p1 + p2 * AG_FU1_o_p2 + p3 * AG_FU1_o_p3
AG17o =~ AG_FU3_o_p1 + p2 * AG_FU3_o_p2 + p3 * AG_FU3_o_p3
#latent intercept and slopes
iy =~ 1*AG0y + 1*AG3y + 1*AG17y
io =~ 1*AG0o + 1*AG3o + 1*AG17o
sy =~ 0*AG0y + 3.3*AG3y + 17*AG17y
so =~ 0*AG0o + 3.3*AG3o + 17*AG17o
iy ~ 1
io ~ 1
sy ~ 1
so ~ 1
#fix zero latent intercepts
AG0y ~ 0*1
AG3y ~ 0*1
AG17y ~ 0*1
AG0o ~ 0*1
AG3o ~ 0*1
AG17o ~ 0*1
#fix zero/equal item intercepts
AG_IN_y_p1 ~ 0*1
AG_FU1_y_p1 ~ 0*1
AG_FU3_y_p1 ~ 0*1
AG_IN_o_p1 ~ 0*1
AG_FU1_o_p1 ~ 0*1
AG_FU3_o_p1 ~ 0*1
AG_IN_y_p2 ~ par2*1
AG_FU1_y_p2 ~ par2*1
AG_FU3_y_p2 ~ par2*1
AG_IN_o_p2 ~ par2*1
AG_FU1_o_p2 ~ par2*1
AG_FU3_o_p2 ~ par2*1
AG_IN_y_p3 ~ par3*1
AG_FU1_y_p3 ~ par3*1
AG_FU3_y_p3 ~ par3*1
AG_IN_o_p3 ~ par3*1
AG_FU1_o_p3 ~ par3*1
AG_FU3_o_p3 ~ par3*1
#residual covariances
AG_IN_y_p1 ~~ c(cov11)*AG_FU1_y_p1 + c(cov13)*AG_FU3_y_p1
AG_FU1_y_p1 ~~ c(cov12)*AG_FU3_y_p1
AG_IN_y_p2 ~~ c(cov21)*AG_FU1_y_p2 + c(cov23)*AG_FU3_y_p2
AG_FU1_y_p2 ~~ c(cov22)*AG_FU3_y_p2
AG_IN_y_p3 ~~ c(cov31)*AG_FU1_y_p3 + c(cov33)*AG_FU3_y_p3
AG_FU1_y_p3 ~~ c(cov32)*AG_FU3_y_p3
AG_IN_o_p1 ~~ c(cov11)*AG_FU1_o_p1 + c(cov13)*AG_FU3_o_p1
AG_FU1_o_p1 ~~ c(cov12)*AG_FU3_o_p1
AG_IN_o_p2 ~~ c(cov21)*AG_FU1_o_p2 + c(cov23)*AG_FU3_o_p2
AG_FU1_o_p2 ~~ c(cov22)*AG_FU3_o_p2
AG_IN_o_p3 ~~ c(cov31)*AG_FU1_o_p3 + c(cov33)*AG_FU3_o_p3
AG_FU1_o_p3 ~~ c(cov32)*AG_FU3_o_p3
'
#ad_sibAG <- sem(sibAG, data = yo[which(yo$adopt == 1),], missing = "FIML")
#summary(ad_sibAG, fit.measures = T, standardized = T)
#bio_sibAG <- sem(sibAG, data = yo[which(yo$adopt == 0),], missing = "FIML")
#summary(bio_sibAG, fit.measures = T, standardized = T)
#g_sibAG <- sem(sibAG, data = yo, missing = "FIML", group = "adopt", em.iter.max = 20000)
#summary(g_sibAG, fit.measures = T, standardized = T, ci = T)
sibAG <- sem(lgmAG, data = yo, missing = "FIML")
fitMeasures(sibAG, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 272.327
## Degrees of freedom 138
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.979
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.040
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.052as.data.frame(unclass(standardizedSolution(sibAG))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|
| iy | 4.063 | 0.180 | 22.56 | 0e+00 | 3.71 | 4.416 |
| io | 3.876 | 0.165 | 23.54 | 0e+00 | 3.55 | 4.199 |
| sy | -1.080 | 0.238 | -4.53 | 6e-06 | -1.55 | -0.613 |
| so | -0.831 | 0.141 | -5.90 | 0e+00 | -1.11 | -0.555 |
| Note: | ||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
as.data.frame(unclass(standardizedSolution(sibAG))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
select("Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| iy | io | 0.311 | 0.0518 | 6.02 | 0.000000 | 0.2099 | 0.413 |
| iy | so | -0.251 | 0.0733 | -3.42 | 0.000616 | -0.3949 | -0.107 |
| sy | so | 0.229 | 0.1010 | 2.27 | 0.023180 | 0.0314 | 0.427 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
semPaths(sibAG, what = "col", whatLabels = "est", intercepts = T)
Control CON
lgmCON <- '
#younger sibling
CON0y =~ CON_IN_y_p1 + p2 * CON_IN_y_p2 + p3 * CON_IN_y_p3
CON3y =~ CON_FU1_y_p1 + p2 * CON_FU1_y_p2 + p3 * CON_FU1_y_p3
CON17y =~ CON_FU3_y_p1 + p2 * CON_FU3_y_p2 + p3 * CON_FU3_y_p3
#older sibling
CON0o =~ CON_IN_o_p1 + p2 * CON_IN_o_p2 + p3 * CON_IN_o_p3
CON3o =~ CON_FU1_o_p1 + p2 * CON_FU1_o_p2 + p3 * CON_FU1_o_p3
CON17o =~ CON_FU3_o_p1 + p2 * CON_FU3_o_p2 + p3 * CON_FU3_o_p3
#latent intercept and slopes
iy =~ 1*CON0y + 1*CON3y + 1*CON17y
io =~ 1*CON0o + 1*CON3o + 1*CON17o
sy =~ 0*CON0y + 3.3*CON3y + 17*CON17y
so =~ 0*CON0o + 3.3*CON3o + 17*CON17o
iy ~ 1
io ~ 1
sy ~ 1
so ~ 1
#fix zero latent intercepts
CON0y ~ 0*1
CON3y ~ 0*1
CON17y ~ 0*1
CON0o ~ 0*1
CON3o ~ 0*1
CON17o ~ 0*1
#fix zero/equal item intercepts
CON_IN_y_p1 ~ 0*1
CON_FU1_y_p1 ~ 0*1
CON_FU3_y_p1 ~ 0*1
CON_IN_o_p1 ~ 0*1
CON_FU1_o_p1 ~ 0*1
CON_FU3_o_p1 ~ 0*1
CON_IN_y_p2 ~ par2*1
CON_FU1_y_p2 ~ par2*1
CON_FU3_y_p2 ~ par2*1
CON_IN_o_p2 ~ par2*1
CON_FU1_o_p2 ~ par2*1
CON_FU3_o_p2 ~ par2*1
CON_IN_y_p3 ~ par3*1
CON_FU1_y_p3 ~ par3*1
CON_FU3_y_p3 ~ par3*1
CON_IN_o_p3 ~ par3*1
CON_FU1_o_p3 ~ par3*1
CON_FU3_o_p3 ~ par3*1
#residual covariances
CON_IN_y_p1 ~~ c(cov11)*CON_FU1_y_p1 + c(cov13)*CON_FU3_y_p1
CON_FU1_y_p1 ~~ c(cov12)*CON_FU3_y_p1
CON_IN_y_p2 ~~ c(cov21)*CON_FU1_y_p2 + c(cov23)*CON_FU3_y_p2
CON_FU1_y_p2 ~~ c(cov22)*CON_FU3_y_p2
CON_IN_y_p3 ~~ c(cov31)*CON_FU1_y_p3 + c(cov33)*CON_FU3_y_p3
CON_FU1_y_p3 ~~ c(cov32)*CON_FU3_y_p3
CON_IN_o_p1 ~~ c(cov11)*CON_FU1_o_p1 + c(cov13)*CON_FU3_o_p1
CON_FU1_o_p1 ~~ c(cov12)*CON_FU3_o_p1
CON_IN_o_p2 ~~ c(cov21)*CON_FU1_o_p2 + c(cov23)*CON_FU3_o_p2
CON_FU1_o_p2 ~~ c(cov22)*CON_FU3_o_p2
CON_IN_o_p3 ~~ c(cov31)*CON_FU1_o_p3 + c(cov33)*CON_FU3_o_p3
CON_FU1_o_p3 ~~ c(cov32)*CON_FU3_o_p3
'
sibCON <- sem(lgmCON, data = yo, missing = "FIML")
fitMeasures(sibCON, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 265.283
## Degrees of freedom 138
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.975
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.039
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.049as.data.frame(unclass(standardizedSolution(sibCON))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|
| iy | 7.97 | 0.445 | 17.922 | 0.000 | 7.095 | 8.84 |
| io | 6.67 | 0.312 | 21.362 | 0.000 | 6.057 | 7.28 |
| sy | 3.54 | 7.185 | 0.493 | 0.622 | -10.541 | 17.62 |
| so | 1.49 | 0.579 | 2.575 | 0.010 | 0.356 | 2.62 |
| Note: | ||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
as.data.frame(unclass(standardizedSolution(sibCON))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io") & z < 1.324) %>%
select("Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| iy | io | -0.00339 | 0.0637 | -0.0533 | 0.958 | -0.128 | 0.122 |
| iy | so | 0.14394 | 0.1209 | 1.1907 | 0.234 | -0.093 | 0.381 |
| sy | so | -0.67387 | 1.4293 | -0.4715 | 0.637 | -3.475 | 2.127 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
semPaths(sibCON, what = "col", whatLabels = "est", intercepts = T)
Harm Avoidance HA
lgmHA <- '
#younger sibling
HA0y =~ HA_IN_y_p1 + p2 * HA_IN_y_p2 + p3 * HA_IN_y_p3
HA3y =~ HA_FU1_y_p1 + p2 * HA_FU1_y_p2 + p3 * HA_FU1_y_p3
HA17y =~ HA_FU3_y_p1 + p2 * HA_FU3_y_p2 + p3 * HA_FU3_y_p3
#older sibling
HA0o =~ HA_IN_o_p1 + p2 * HA_IN_o_p2 + p3 * HA_IN_o_p3
HA3o =~ HA_FU1_o_p1 + p2 * HA_FU1_o_p2 + p3 * HA_FU1_o_p3
HA17o =~ HA_FU3_o_p1 + p2 * HA_FU3_o_p2 + p3 * HA_FU3_o_p3
#latent intercept and slopes
iy =~ 1*HA0y + 1*HA3y + 1*HA17y
io =~ 1*HA0o + 1*HA3o + 1*HA17o
sy =~ 0*HA0y + 3.3*HA3y + 17*HA17y
so =~ 0*HA0o + 3.3*HA3o + 17*HA17o
iy ~ 1
io ~ 1
sy ~ 1
so ~ 1
#fix zero latent intercepts
HA0y ~ 0*1
HA3y ~ 0*1
HA17y ~ 0*1
HA0o ~ 0*1
HA3o ~ 0*1
HA17o ~ 0*1
#fix zero/equal item intercepts
HA_IN_y_p1 ~ 0*1
HA_FU1_y_p1 ~ 0*1
HA_FU3_y_p1 ~ 0*1
HA_IN_o_p1 ~ 0*1
HA_FU1_o_p1 ~ 0*1
HA_FU3_o_p1 ~ 0*1
HA_IN_y_p2 ~ par2*1
HA_FU1_y_p2 ~ par2*1
HA_FU3_y_p2 ~ par2*1
HA_IN_o_p2 ~ par2*1
HA_FU1_o_p2 ~ par2*1
HA_FU3_o_p2 ~ par2*1
HA_IN_y_p3 ~ par3*1
HA_FU1_y_p3 ~ par3*1
HA_FU3_y_p3 ~ par3*1
HA_IN_o_p3 ~ par3*1
HA_FU1_o_p3 ~ par3*1
HA_FU3_o_p3 ~ par3*1
#residual covariances
HA_IN_y_p1 ~~ c(cov11)*HA_FU1_y_p1 + c(cov13)*HA_FU3_y_p1
HA_FU1_y_p1 ~~ c(cov12)*HA_FU3_y_p1
HA_IN_y_p2 ~~ c(cov21)*HA_FU1_y_p2 + c(cov23)*HA_FU3_y_p2
HA_FU1_y_p2 ~~ + c(cov22)*HA_FU3_y_p2
HA_IN_y_p3 ~~ c(cov31)*HA_FU1_y_p3 + c(cov33)*HA_FU3_y_p3
HA_FU1_y_p3 ~~ c(cov32)*HA_FU3_y_p3
HA_IN_o_p1 ~~ c(cov11)*HA_FU1_o_p1 + c(cov13)*HA_FU3_o_p1
HA_FU1_o_p1 ~~ c(cov12)*HA_FU3_o_p1
HA_IN_o_p2 ~~ c(cov21)*HA_FU1_o_p2 + c(cov23)*HA_FU3_o_p2
HA_FU1_o_p2 ~~ c(cov22)*HA_FU3_o_p2
HA_IN_o_p3 ~~ c(cov31)*HA_FU1_o_p3 + c(cov33)*HA_FU3_o_p3
HA_FU1_o_p3 ~~ c(cov32)*HA_FU3_o_p3
'
sibHA <- sem(lgmHA, data = yo, missing = "FIML")
fitMeasures(sibHA, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 502.965
## Degrees of freedom 138
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.918
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.066
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.064as.data.frame(unclass(standardizedSolution(sibHA))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|
| iy | 6.17 | 0.372 | 16.56 | 0.000000 | 5.436 | 6.89 |
| io | 4.91 | 0.221 | 22.22 | 0.000000 | 4.480 | 5.35 |
| sy | 1.21 | 0.718 | 1.69 | 0.091608 | -0.196 | 2.62 |
| so | 1.08 | 0.296 | 3.64 | 0.000278 | 0.495 | 1.65 |
| Note: | ||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
as.data.frame(unclass(standardizedSolution(sibHA))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
select("Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| iy | io | 0.326 | 0.0617 | 5.291 | 0.000 | 0.205 | 0.4471 |
| iy | so | -0.135 | 0.1017 | -1.329 | 0.184 | -0.335 | 0.0642 |
| sy | so | 0.127 | 0.1773 | 0.717 | 0.473 | -0.220 | 0.4746 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
semPaths(sibHA, what = "col", whatLabels = "est", intercepts = T)
Adopted vs. Non-adopted
Here we compare between-sibling correlations in adopted vs. non-adopted sibling pairs.
Correlations are considered significantly different from one another if the confidence interval of one correlation does not include the point estimate for the other correlation.
# create adopted/bio variable
yo <- yo %>%
mutate(adopt = ifelse(IDAB_y == 1 | IDAB_o == 1, 1, 0))
# histogram
ggplot(yo,
aes(factor(yo$adopt,
levels = c(0, 1),
labels = c("Non-Adoptive", "Adoptive")))) +
geom_bar(fill = c("#710c0c", "#f1c232")) +
geom_text(stat='count', aes(label=..count..), vjust = -1) +
labs(
title = "Count of adoption status in sibling pairs",
x = NULL
) +
ylim(0, 450) +
theme_classic()
There are 406 adopted pairs and 208 non-adopted sibling pairs.
Adoption status - Aggression AG
adoptAG <- sem(lgmAG, data = yo, missing = "FIML",
group = "adopt")
fitMeasures(adoptAG, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 484.633
## Degrees of freedom 289
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.969
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.047
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.061# intercept and slope
as.data.frame(unclass(standardizedSolution(adoptAG))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Adopted" = group, "Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(Adopted = ifelse(Adopted == 1, "adopt", "bio")) %>%
arrange(Parameters) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Adopted | Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| adopt | io | 3.780 | 0.189 | 20.007 | 0.000000 | 3.41 | 4.150 |
| bio | io | 4.073 | 0.323 | 12.607 | 0.000000 | 3.44 | 4.706 |
| adopt | iy | 4.152 | 0.234 | 17.763 | 0.000000 | 3.69 | 4.610 |
| bio | iy | 3.921 | 0.281 | 13.949 | 0.000000 | 3.37 | 4.472 |
| adopt | so | -0.757 | 0.167 | -4.527 | 0.000006 | -1.08 | -0.429 |
| bio | so | -0.969 | 0.255 | -3.806 | 0.000141 | -1.47 | -0.470 |
| adopt | sy | -0.777 | 0.150 | -5.185 | 0.000000 | -1.07 | -0.483 |
| bio | sy | -3.263 | 6.766 | -0.482 | 0.629576 | -16.52 | 9.998 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
# correlated intercepts and slopes
as.data.frame(unclass(standardizedSolution(adoptAG))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
arrange(lhs) %>%
select("Adopted" = group, "Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(Adopted = ifelse(Adopted == 1, "adopt", "bio")) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Adopted | Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|---|
| adopt | iy | io | 0.235 | 0.0650 | 3.612 | 0.000303 | 0.1073 | 0.3620 |
| adopt | iy | so | -0.259 | 0.0963 | -2.690 | 0.007155 | -0.4479 | -0.0703 |
| bio | iy | io | 0.470 | 0.0833 | 5.636 | 0.000000 | 0.3063 | 0.6330 |
| bio | iy | so | -0.231 | 0.1089 | -2.117 | 0.034255 | -0.4439 | -0.0171 |
| adopt | sy | so | 0.158 | 0.0965 | 1.637 | 0.101665 | -0.0312 | 0.3472 |
| bio | sy | so | 0.668 | 1.4185 | 0.471 | 0.637643 | -2.1121 | 3.4484 |
| Note: | ||||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
Adoption status - Control CON
adoptCON <- sem(lgmCON, data = yo, missing = "FIML",
group = "adopt")
fitMeasures(adoptCON, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 426.770
## Degrees of freedom 289
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.973
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.039
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.059# intercept and slope
as.data.frame(unclass(standardizedSolution(adoptCON))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Adopted" = group, "Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(Adopted = ifelse(Adopted == 1, "adopt", "bio")) %>%
arrange(Parameters) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Adopted | Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| adopt | io | 6.04 | 0.314 | 19.22 | 0.000000 | 5.426 | 6.66 |
| bio | io | 8.48 | 0.888 | 9.54 | 0.000000 | 6.735 | 10.22 |
| adopt | iy | 7.81 | 0.504 | 15.50 | 0.000000 | 6.824 | 8.80 |
| bio | iy | 8.24 | 0.872 | 9.46 | 0.000000 | 6.535 | 9.95 |
| adopt | so | 1.05 | 0.267 | 3.91 | 0.000091 | 0.523 | 1.57 |
| bio | so | NA | NA | NA | NA | NA | NA |
| adopt | sy | 1.65 | 0.993 | 1.66 | 0.096409 | -0.295 | 3.60 |
| bio | sy | NA | NA | NA | NA | NA | NA |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
# correlated intercepts and slopes
as.data.frame(unclass(standardizedSolution(adoptCON))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
arrange(lhs) %>%
select("Adopted" = group, "Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(Adopted = ifelse(Adopted == 1, "adopt", "bio")) %>%
slice(-c(2,4)) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Adopted | Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|---|
| adopt | iy | io | -0.0239 | 0.072 | -0.331 | 0.741 | -0.165 | 0.117 |
| bio | iy | io | 0.0401 | 0.135 | 0.297 | 0.766 | -0.224 | 0.305 |
| adopt | sy | so | -0.2106 | 0.221 | -0.952 | 0.341 | -0.644 | 0.223 |
| bio | sy | so | -1.4592 | 10.540 | -0.138 | 0.890 | -22.117 | 19.199 |
| Note: | ||||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
Adoption status - Harm Avoidance HA
adoptHA <- sem(lgmHA, data = yo, missing = "FIML",
group = "adopt")
fitMeasures(adoptHA, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 666.446
## Degrees of freedom 289
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.915
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.065
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.072# intercept and slope
as.data.frame(unclass(standardizedSolution(adoptHA))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Adopted" = group, "Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(Adopted = ifelse(Adopted == 1, "adopt", "bio")) %>%
arrange(Parameters) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Adopted | Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| adopt | io | 4.968 | 0.275 | 18.081 | 0.00000 | 4.429 | 5.51 |
| bio | io | 4.817 | 0.360 | 13.376 | 0.00000 | 4.111 | 5.52 |
| adopt | iy | 6.889 | 0.574 | 11.997 | 0.00000 | 5.763 | 8.01 |
| bio | iy | 5.409 | 0.491 | 11.015 | 0.00000 | 4.447 | 6.37 |
| adopt | so | 1.078 | 0.407 | 2.651 | 0.00802 | 0.281 | 1.88 |
| bio | so | 1.071 | 0.414 | 2.585 | 0.00975 | 0.259 | 1.88 |
| adopt | sy | 1.857 | 3.267 | 0.568 | 0.56971 | -4.546 | 8.26 |
| bio | sy | 0.988 | 0.595 | 1.661 | 0.09680 | -0.178 | 2.15 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
# correlated intercepts and slopes
as.data.frame(unclass(standardizedSolution(adoptHA))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
arrange(lhs) %>%
select("Adopted" = group, "Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(Adopted = ifelse(Adopted == 1, "adopt", "bio")) %>%
slice(-c(2,4)) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Adopted | Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|---|
| adopt | iy | io | 0.2123 | 0.0826 | 2.570 | 0.0102 | 0.0504 | 0.374 |
| bio | iy | io | 0.5213 | 0.0911 | 5.725 | 0.0000 | 0.3429 | 0.700 |
| adopt | sy | so | 0.2617 | 0.5674 | 0.461 | 0.6446 | -0.8504 | 1.374 |
| bio | sy | so | 0.0523 | 0.2016 | 0.259 | 0.7955 | -0.3430 | 0.447 |
| Note: | ||||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
Same-sex vs. Opposite-sex
Here we compare between-sibling correlations in same-sex and opposite-sex pairs.
Correlations are considered significantly different from one another if the confidence interval of one correlation does not include the point estimate for the other correlation.
# create same/opposite sex variable
yo <- yo %>%
mutate(samesex = as.numeric(IDSEX_y == IDSEX_o))
# histogram
ggplot(yo,
aes(factor(yo$samesex,
levels = c(0, 1),
labels = c("Different-Sex", "Same-Sex")))) +
geom_bar(fill = c("#f1c232", "#710c0c")) +
geom_text(stat='count', aes(label=..count..), vjust = -1) +
labs(
title = "Count of sex similarity in sibling pairs",
x = NULL
) +
ylim(0, 450) +
theme_classic()
There are 374 same-sex and 240 opposite-sex sibling pairs.
Sex similarity - Aggression AG
sexAG <- sem(lgmAG, data = yo, missing = "FIML",
group = "samesex")
fitMeasures(sexAG, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 423.557
## Degrees of freedom 289
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.979
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.039
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.062# intercept and slope
as.data.frame(unclass(standardizedSolution(sexAG))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("SameSex" = group, "Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(SameSex = ifelse(SameSex == 1, "same-sex", "opposite-sex")) %>%
arrange(Parameters) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| SameSex | Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| same-sex | io | 3.787 | 0.203 | 18.65 | 0.000000 | 3.389 | 4.185 |
| opposite-sex | io | 4.036 | 0.279 | 14.49 | 0.000000 | 3.490 | 4.582 |
| same-sex | iy | 3.954 | 0.216 | 18.34 | 0.000000 | 3.531 | 4.376 |
| opposite-sex | iy | 4.228 | 0.314 | 13.47 | 0.000000 | 3.613 | 4.843 |
| same-sex | so | -0.909 | 0.216 | -4.21 | 0.000026 | -1.333 | -0.485 |
| opposite-sex | so | -0.671 | 0.141 | -4.75 | 0.000002 | -0.948 | -0.394 |
| same-sex | sy | -1.172 | 0.319 | -3.68 | 0.000235 | -1.797 | -0.548 |
| opposite-sex | sy | -1.016 | 0.370 | -2.75 | 0.006028 | -1.741 | -0.291 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
# correlated intercepts and slopes
as.data.frame(unclass(standardizedSolution(sexAG))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
arrange(lhs) %>%
select("SameSex" = group, "Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(SameSex = ifelse(SameSex == 1, "same-sex", "opposite-sex")) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| SameSex | Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|---|
| same-sex | iy | io | 0.4462 | 0.0609 | 7.328 | 0.00000 | 0.3268 | 0.5655 |
| same-sex | iy | so | -0.3383 | 0.1123 | -3.013 | 0.00258 | -0.5584 | -0.1183 |
| opposite-sex | iy | io | 0.0776 | 0.0904 | 0.858 | 0.39064 | -0.0995 | 0.2547 |
| opposite-sex | iy | so | -0.1510 | 0.0886 | -1.704 | 0.08844 | -0.3248 | 0.0227 |
| same-sex | sy | so | 0.1918 | 0.1359 | 1.411 | 0.15818 | -0.0746 | 0.4581 |
| opposite-sex | sy | so | 0.2805 | 0.1506 | 1.862 | 0.06263 | -0.0148 | 0.5757 |
| Note: | ||||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
Sex similarity - Control CON
sexCON <- sem(lgmCON, data = yo, missing = "FIML",
group = "samesex")
fitMeasures(sexCON, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 441.077
## Degrees of freedom 289
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.970
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.041
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.063# intercept and slope
as.data.frame(unclass(standardizedSolution(sexCON))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("SameSex" = group, "Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(SameSex = ifelse(SameSex == 1, "same-sex", "opposite-sex")) %>%
arrange(Parameters) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| SameSex | Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| same-sex | io | 6.97 | 0.437 | 15.947 | 0.00000 | 6.11 | 7.83 |
| opposite-sex | io | 6.29 | 0.436 | 14.439 | 0.00000 | 5.44 | 7.14 |
| same-sex | iy | 8.21 | 0.618 | 13.293 | 0.00000 | 7.00 | 9.42 |
| opposite-sex | iy | 7.69 | 0.630 | 12.193 | 0.00000 | 6.45 | 8.92 |
| same-sex | so | 2.04 | 1.987 | 1.028 | 0.30393 | -1.85 | 5.94 |
| opposite-sex | so | 1.15 | 0.407 | 2.819 | 0.00482 | 0.35 | 1.95 |
| same-sex | sy | NA | NA | NA | NA | NA | NA |
| opposite-sex | sy | 2.54 | 4.344 | 0.584 | 0.55892 | -5.97 | 11.05 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
# correlated intercepts and slopes
as.data.frame(unclass(standardizedSolution(sexCON))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
arrange(lhs) %>%
select("SameSex" = group, "Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(SameSex = ifelse(SameSex == 1, "same-sex", "opposite-sex")) %>%
slice(-c(2,4)) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| SameSex | Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|---|
| same-sex | iy | io | -0.0317 | 0.0863 | -0.368 | 0.713 | -0.201 | 0.137 |
| opposite-sex | iy | io | 0.0297 | 0.0953 | 0.311 | 0.755 | -0.157 | 0.216 |
| same-sex | sy | so | -1.1208 | 3.7820 | -0.296 | 0.767 | -8.533 | 6.292 |
| opposite-sex | sy | so | -0.4284 | 0.8126 | -0.527 | 0.598 | -2.021 | 1.164 |
| Note: | ||||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
Sex similarity - Harm Avoidance HA
sexHA <- sem(lgmHA, data = yo, missing = "FIML",
group = "samesex")
fitMeasures(sexHA, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 669.213
## Degrees of freedom 289
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.915
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.065
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.073# intercept and slope
as.data.frame(unclass(standardizedSolution(sexHA))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("SameSex" = group, "Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(SameSex = ifelse(SameSex == 1, "same-sex", "opposite-sex")) %>%
arrange(Parameters) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| SameSex | Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| same-sex | io | 4.954 | 0.283 | 17.50 | 0.000000 | 4.399 | 5.51 |
| opposite-sex | io | 4.876 | 0.342 | 14.27 | 0.000000 | 4.206 | 5.55 |
| same-sex | iy | 6.316 | 0.521 | 12.12 | 0.000000 | 5.294 | 7.34 |
| opposite-sex | iy | 5.950 | 0.512 | 11.62 | 0.000000 | 4.946 | 6.95 |
| same-sex | so | 1.396 | 0.802 | 1.74 | 0.081786 | -0.176 | 2.97 |
| opposite-sex | so | 0.865 | 0.271 | 3.19 | 0.001433 | 0.333 | 1.40 |
| same-sex | sy | 0.751 | 0.223 | 3.36 | 0.000769 | 0.313 | 1.19 |
| opposite-sex | sy | NA | NA | NA | NA | NA | NA |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
# correlated intercepts and slopes
as.data.frame(unclass(standardizedSolution(sexHA))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
arrange(lhs) %>%
select("SameSex" = group, "Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
mutate(SameSex = ifelse(SameSex == 1, "same-sex", "opposite-sex")) %>%
slice(-c(2,4)) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| SameSex | Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|---|
| same-sex | iy | io | 0.433 | 0.0787 | 5.497 | 0.0000 | 0.27836 | 0.587 |
| opposite-sex | iy | io | 0.190 | 0.0991 | 1.915 | 0.0555 | -0.00447 | 0.384 |
| same-sex | sy | so | 0.302 | 0.2588 | 1.169 | 0.2424 | -0.20469 | 0.810 |
| opposite-sex | sy | so | -0.123 | 0.1847 | -0.664 | 0.5067 | -0.48469 | 0.239 |
| Note: | ||||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
IV. Analyses with full data
General change trajectories and correlated change with 4 waves
These analyses analyzed the general change pattern using all 4 waves of data: IN, FU1, FU2, FU3. This is to compare with analyses in section III, which only included 3 waves, excluding FU2 for high missingness. Because these analyses made use of the maximum amount of data, they will be presented in reports and presentations for general change patterns.
Aggression AG
lgmAG <- '
#younger sibling
AG0y =~ AG_IN_y_p1 + p2 * AG_IN_y_p2 + p3 * AG_IN_y_p3
AG3y =~ AG_FU1_y_p1 + p2 * AG_FU1_y_p2 + p3 * AG_FU1_y_p3
AG7y =~ AG_FU2_y_p1 + p2 * AG_FU2_y_p2 + p3 * AG_FU2_y_p3
AG17y =~ AG_FU3_y_p1 + p2 * AG_FU3_y_p2 + p3 * AG_FU3_y_p3
#older sibling
AG0o =~ AG_IN_o_p1 + p2 * AG_IN_o_p2 + p3 * AG_IN_o_p3
AG3o =~ AG_FU1_o_p1 + p2 * AG_FU1_o_p2 + p3 * AG_FU1_o_p3
AG7o =~ AG_FU2_o_p1 + p2 * AG_FU2_o_p2 + p3 * AG_FU2_o_p3
AG17o =~ AG_FU3_o_p1 + p2 * AG_FU3_o_p2 + p3 * AG_FU3_o_p3
#latent intercept and slopes
iy =~ 1*AG0y + 1*AG3y + 1*AG7y + 1*AG17y
io =~ 1*AG0o + 1*AG3o + 1*AG7o + 1*AG17o
sy =~ 0*AG0y + 3.3*AG3y + 7.6*AG7y + 17*AG17y
so =~ 0*AG0o + 3.3*AG3o + 7.6*AG7o + 17*AG17o
iy ~ 1
io ~ 1
sy ~ 1
so ~ 1
#fix zero latent intercepts
AG0y ~ 0*1
AG3y ~ 0*1
AG7y ~ 0*1
AG17y ~ 0*1
AG0o ~ 0*1
AG3o ~ 0*1
AG7o ~ 0*1
AG17o ~ 0*1
#fix zero/equal item intercepts
AG_IN_y_p1 ~ 0*1
AG_FU1_y_p1 ~ 0*1
AG_FU2_y_p1 ~ 0*1
AG_FU3_y_p1 ~ 0*1
AG_IN_o_p1 ~ 0*1
AG_FU1_o_p1 ~ 0*1
AG_FU2_o_p1 ~ 0*1
AG_FU3_o_p1 ~ 0*1
AG_IN_y_p2 ~ par2*1
AG_FU1_y_p2 ~ par2*1
AG_FU2_y_p2 ~ par2*1
AG_FU3_y_p2 ~ par2*1
AG_IN_o_p2 ~ par2*1
AG_FU1_o_p2 ~ par2*1
AG_FU2_o_p2 ~ par2*1
AG_FU3_o_p2 ~ par2*1
AG_IN_y_p3 ~ par3*1
AG_FU1_y_p3 ~ par3*1
AG_FU2_y_p3 ~ par3*1
AG_FU3_y_p3 ~ par3*1
AG_IN_o_p3 ~ par3*1
AG_FU1_o_p3 ~ par3*1
AG_FU2_o_p3 ~ par3*1
AG_FU3_o_p3 ~ par3*1
#residual covariances
AG_IN_y_p1 ~~ c(cov11)*AG_FU1_y_p1 + c(cov12)*AG_FU2_y_p1 + c(cov13)*AG_FU3_y_p1
AG_FU1_y_p1 ~~ c(cov11)*AG_FU2_y_p1 + c(cov12)*AG_FU3_y_p1
AG_FU2_y_p1 ~~ c(cov11)*AG_FU3_y_p1
AG_IN_y_p2 ~~ c(cov21)*AG_FU1_y_p2 + c(cov22)*AG_FU2_y_p2 + c(cov23)*AG_FU3_y_p2
AG_FU1_y_p2 ~~ c(cov21)*AG_FU2_y_p2 + c(cov22)*AG_FU3_y_p2
AG_FU2_y_p2 ~~ c(cov21)*AG_FU3_y_p2
AG_IN_y_p3 ~~ c(cov31)*AG_FU1_y_p3 + c(cov32)*AG_FU2_y_p3 + c(cov33)*AG_FU3_y_p3
AG_FU1_y_p3 ~~ c(cov31)*AG_FU2_y_p3 + c(cov32)*AG_FU3_y_p3
AG_FU2_y_p3 ~~ c(cov31)*AG_FU3_y_p3
AG_IN_o_p1 ~~ c(cov11)*AG_FU1_o_p1 + c(cov12)*AG_FU2_o_p1 + c(cov13)*AG_FU3_o_p1
AG_FU1_o_p1 ~~ c(cov11)*AG_FU2_o_p1 + c(cov12)*AG_FU3_o_p1
AG_FU2_o_p1 ~~ c(cov11)*AG_FU3_o_p1
AG_IN_o_p2 ~~ c(cov21)*AG_FU1_o_p2 + c(cov22)*AG_FU2_o_p2 + c(cov23)*AG_FU3_o_p2
AG_FU1_o_p2 ~~ c(cov21)*AG_FU2_o_p2 + c(cov22)*AG_FU3_o_p2
AG_FU2_o_p2 ~~ c(cov21)*AG_FU3_o_p2
AG_IN_o_p3 ~~ c(cov31)*AG_FU1_o_p3 + c(cov32)*AG_FU2_o_p3 + c(cov33)*AG_FU3_o_p3
AG_FU1_o_p3 ~~ c(cov31)*AG_FU2_o_p3 + c(cov32)*AG_FU3_o_p3
AG_FU2_o_p3 ~~ c(cov31)*AG_FU3_o_p3
'
#ad_sibAG <- sem(sibAG, data = yo[which(yo$adopt == 1),], missing = "FIML")
#summary(ad_sibAG, fit.measures = T, standardized = T)
#bio_sibAG <- sem(sibAG, data = yo[which(yo$adopt == 0),], missing = "FIML")
#summary(bio_sibAG, fit.measures = T, standardized = T)
#g_sibAG <- sem(sibAG, data = yo, missing = "FIML", group = "adopt", em.iter.max = 20000)
#summary(g_sibAG, fit.measures = T, standardized = T, ci = T)
sibAG <- sem(lgmAG, data = yo, missing = "FIML")
fitMeasures(sibAG, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 455.575
## Degrees of freedom 265
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.971
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.034
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.137as.data.frame(unclass(standardizedSolution(sibAG))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|
| iy | 4.152 | 0.185 | 22.46 | 0 | 3.79 | 4.514 |
| io | 3.885 | 0.165 | 23.59 | 0 | 3.56 | 4.207 |
| sy | -1.043 | 0.132 | -7.88 | 0 | -1.30 | -0.783 |
| so | -0.852 | 0.127 | -6.68 | 0 | -1.10 | -0.602 |
| Note: | ||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
as.data.frame(unclass(standardizedSolution(sibAG))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
select("Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| iy | io | 0.314 | 0.0522 | 6.02 | 0.000000 | 0.2121 | 0.417 |
| iy | so | -0.252 | 0.0729 | -3.46 | 0.000536 | -0.3954 | -0.110 |
| sy | so | 0.215 | 0.0893 | 2.41 | 0.015864 | 0.0404 | 0.390 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
semPaths(sibAG, what = "col", whatLabels = "est", intercepts = T)
Control CON
lgmCON <- '
#younger sibling
CON0y =~ CON_IN_y_p1 + p2 * CON_IN_y_p2 + p3 * CON_IN_y_p3
CON3y =~ CON_FU1_y_p1 + p2 * CON_FU1_y_p2 + p3 * CON_FU1_y_p3
CON7y =~ CON_FU2_y_p1 + p2 * CON_FU2_y_p2 + p3 * CON_FU2_y_p3
CON17y =~ CON_FU3_y_p1 + p2 * CON_FU3_y_p2 + p3 * CON_FU3_y_p3
#older sibling
CON0o =~ CON_IN_o_p1 + p2 * CON_IN_o_p2 + p3 * CON_IN_o_p3
CON3o =~ CON_FU1_o_p1 + p2 * CON_FU1_o_p2 + p3 * CON_FU1_o_p3
CON7o =~ CON_FU2_o_p1 + p2 * CON_FU2_o_p2 + p3 * CON_FU2_o_p3
CON17o =~ CON_FU3_o_p1 + p2 * CON_FU3_o_p2 + p3 * CON_FU3_o_p3
#latent intercept and slopes
iy =~ 1*CON0y + 1*CON3y + 1*CON7y + 1*CON17y
io =~ 1*CON0o + 1*CON3o + 1*CON7o + 1*CON17o
sy =~ 0*CON0y + 3.3*CON3y + 7.6*CON7y + 17*CON17y
so =~ 0*CON0o + 3.3*CON3o + 7.6*CON7o + 17*CON17o
iy ~ 1
io ~ 1
sy ~ 1
so ~ 1
#fix zero latent intercepts
CON0y ~ 0*1
CON3y ~ 0*1
CON7y ~ 0*1
CON17y ~ 0*1
CON0o ~ 0*1
CON3o ~ 0*1
CON7o ~ 0*1
CON17o ~ 0*1
#fix zero/equal item intercepts
CON_IN_y_p1 ~ 0*1
CON_FU1_y_p1 ~ 0*1
CON_FU2_y_p1 ~ 0*1
CON_FU3_y_p1 ~ 0*1
CON_IN_o_p1 ~ 0*1
CON_FU1_o_p1 ~ 0*1
CON_FU2_o_p1 ~ 0*1
CON_FU3_o_p1 ~ 0*1
CON_IN_y_p2 ~ par2*1
CON_FU1_y_p2 ~ par2*1
CON_FU2_y_p2 ~ par2*1
CON_FU3_y_p2 ~ par2*1
CON_IN_o_p2 ~ par2*1
CON_FU1_o_p2 ~ par2*1
CON_FU2_o_p2 ~ par2*1
CON_FU3_o_p2 ~ par2*1
CON_IN_y_p3 ~ par3*1
CON_FU1_y_p3 ~ par3*1
CON_FU2_y_p3 ~ par3*1
CON_FU3_y_p3 ~ par3*1
CON_IN_o_p3 ~ par3*1
CON_FU1_o_p3 ~ par3*1
CON_FU2_o_p3 ~ par3*1
CON_FU3_o_p3 ~ par3*1
#residual covariances
CON_IN_y_p1 ~~ c(cov11)*CON_FU1_y_p1 + c(cov12)*CON_FU2_y_p1 + c(cov13)*CON_FU3_y_p1
CON_FU1_y_p1 ~~ c(cov11)*CON_FU2_y_p1 + c(cov12)*CON_FU3_y_p1
CON_FU2_y_p1 ~~ c(cov11)*CON_FU3_y_p1
CON_IN_y_p2 ~~ c(cov21)*CON_FU1_y_p2 + c(cov22)*CON_FU2_y_p2 + c(cov23)*CON_FU3_y_p2
CON_FU1_y_p2 ~~ c(cov21)*CON_FU2_y_p2 + c(cov22)*CON_FU3_y_p2
CON_FU2_y_p2 ~~ c(cov21)*CON_FU3_y_p2
CON_IN_y_p3 ~~ c(cov31)*CON_FU1_y_p3 + c(cov32)*CON_FU2_y_p3 + c(cov33)*CON_FU3_y_p3
CON_FU1_y_p3 ~~ c(cov31)*CON_FU2_y_p3 + c(cov32)*CON_FU3_y_p3
CON_FU2_y_p3 ~~ c(cov31)*CON_FU3_y_p3
CON_IN_o_p1 ~~ c(cov11)*CON_FU1_o_p1 + c(cov12)*CON_FU2_o_p1 + c(cov13)*CON_FU3_o_p1
CON_FU1_o_p1 ~~ c(cov11)*CON_FU2_o_p1 + c(cov12)*CON_FU3_o_p1
CON_FU2_o_p1 ~~ c(cov11)*CON_FU3_o_p1
CON_IN_o_p2 ~~ c(cov21)*CON_FU1_o_p2 + c(cov22)*CON_FU2_o_p2 + c(cov23)*CON_FU3_o_p2
CON_FU1_o_p2 ~~ c(cov21)*CON_FU2_o_p2 + c(cov22)*CON_FU3_o_p2
CON_FU2_o_p2 ~~ c(cov21)*CON_FU3_o_p2
CON_IN_o_p3 ~~ c(cov31)*CON_FU1_o_p3 + c(cov32)*CON_FU2_o_p3 + c(cov33)*CON_FU3_o_p3
CON_FU1_o_p3 ~~ c(cov31)*CON_FU2_o_p3 + c(cov32)*CON_FU3_o_p3
CON_FU2_o_p3 ~~ c(cov31)*CON_FU3_o_p3
'
sibCON <- sem(lgmCON, data = yo, missing = "FIML")
fitMeasures(sibCON, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 514.279
## Degrees of freedom 265
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.955
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.039
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.163as.data.frame(unclass(standardizedSolution(sibCON))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|
| iy | 8.01 | 0.437 | 18.31 | 0.000000 | 7.152 | 8.87 |
| io | 6.67 | 0.310 | 21.52 | 0.000000 | 6.064 | 7.28 |
| sy | 1.68 | 0.496 | 3.38 | 0.000724 | 0.705 | 2.65 |
| so | 1.60 | 0.681 | 2.35 | 0.018653 | 0.267 | 2.94 |
| Note: | ||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
as.data.frame(unclass(standardizedSolution(sibCON))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io") & z < 1.324) %>%
select("Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| iy | io | -0.00522 | 0.0634 | -0.0823 | 0.934 | -0.129 | 0.119 |
| iy | so | 0.15649 | 0.1328 | 1.1786 | 0.239 | -0.104 | 0.417 |
| sy | so | -0.35653 | 0.2607 | -1.3676 | 0.171 | -0.867 | 0.154 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
semPaths(sibCON, what = "col", whatLabels = "est", intercepts = T)
Harm Avoidance HA
lgmHA <- '
#younger sibling
HA0y =~ HA_IN_y_p1 + p2 * HA_IN_y_p2 + p3 * HA_IN_y_p3
HA3y =~ HA_FU1_y_p1 + p2 * HA_FU1_y_p2 + p3 * HA_FU1_y_p3
HA7y =~ HA_FU2_y_p1 + p2 * HA_FU2_y_p2 + p3 * HA_FU2_y_p3
HA17y =~ HA_FU3_y_p1 + p2 * HA_FU3_y_p2 + p3 * HA_FU3_y_p3
#older sibling
HA0o =~ HA_IN_o_p1 + p2 * HA_IN_o_p2 + p3 * HA_IN_o_p3
HA3o =~ HA_FU1_o_p1 + p2 * HA_FU1_o_p2 + p3 * HA_FU1_o_p3
HA7o =~ HA_FU2_o_p1 + p2 * HA_FU2_o_p2 + p3 * HA_FU2_o_p3
HA17o =~ HA_FU3_o_p1 + p2 * HA_FU3_o_p2 + p3 * HA_FU3_o_p3
#latent intercept and slopes
iy =~ 1*HA0y + 1*HA3y + 1*HA7y + 1*HA17y
io =~ 1*HA0o + 1*HA3o + 1*HA7o + 1*HA17o
sy =~ 0*HA0y + 3.3*HA3y + 7.6*HA7y + 17*HA17y
so =~ 0*HA0o + 3.3*HA3o + 7.6*HA7o + 17*HA17o
iy ~ 1
io ~ 1
sy ~ 1
so ~ 1
#fix zero latent intercepts
HA0y ~ 0*1
HA3y ~ 0*1
HA7y ~ 0*1
HA17y ~ 0*1
HA0o ~ 0*1
HA3o ~ 0*1
HA7o ~ 0*1
HA17o ~ 0*1
#fix zero/equal item intercepts
HA_IN_y_p1 ~ 0*1
HA_FU1_y_p1 ~ 0*1
HA_FU2_y_p1 ~ 0*1
HA_FU3_y_p1 ~ 0*1
HA_IN_o_p1 ~ 0*1
HA_FU1_o_p1 ~ 0*1
HA_FU2_o_p1 ~ 0*1
HA_FU3_o_p1 ~ 0*1
HA_IN_y_p2 ~ par2*1
HA_FU1_y_p2 ~ par2*1
HA_FU2_y_p2 ~ par2*1
HA_FU3_y_p2 ~ par2*1
HA_IN_o_p2 ~ par2*1
HA_FU1_o_p2 ~ par2*1
HA_FU2_o_p2 ~ par2*1
HA_FU3_o_p2 ~ par2*1
HA_IN_y_p3 ~ par3*1
HA_FU1_y_p3 ~ par3*1
HA_FU2_y_p3 ~ par3*1
HA_FU3_y_p3 ~ par3*1
HA_IN_o_p3 ~ par3*1
HA_FU1_o_p3 ~ par3*1
HA_FU2_o_p3 ~ par3*1
HA_FU3_o_p3 ~ par3*1
#residual covariances
HA_IN_y_p1 ~~ c(cov11)*HA_FU1_y_p1 + c(cov12)*HA_FU2_y_p1 + c(cov13)*HA_FU3_y_p1
HA_FU1_y_p1 ~~ c(cov11)*HA_FU2_y_p1 + c(cov12)*HA_FU3_y_p1
HA_FU2_y_p1 ~~ c(cov11)*HA_FU3_y_p1
HA_IN_y_p2 ~~ c(cov21)*HA_FU1_y_p2 + c(cov22)*HA_FU2_y_p2 + c(cov23)*HA_FU3_y_p2
HA_FU1_y_p2 ~~ c(cov21)*HA_FU2_y_p2 + c(cov22)*HA_FU3_y_p2
HA_FU2_y_p2 ~~ c(cov21)*HA_FU3_y_p2
HA_IN_y_p3 ~~ c(cov31)*HA_FU1_y_p3 + c(cov32)*HA_FU2_y_p3 + c(cov33)*HA_FU3_y_p3
HA_FU1_y_p3 ~~ c(cov31)*HA_FU2_y_p3 + c(cov32)*HA_FU3_y_p3
HA_FU2_y_p3 ~~ c(cov31)*HA_FU3_y_p3
HA_IN_o_p1 ~~ c(cov11)*HA_FU1_o_p1 + c(cov12)*HA_FU2_o_p1 + c(cov13)*HA_FU3_o_p1
HA_FU1_o_p1 ~~ c(cov11)*HA_FU2_o_p1 + c(cov12)*HA_FU3_o_p1
HA_FU2_o_p1 ~~ c(cov11)*HA_FU3_o_p1
HA_IN_o_p2 ~~ c(cov21)*HA_FU1_o_p2 + c(cov22)*HA_FU2_o_p2 + c(cov23)*HA_FU3_o_p2
HA_FU1_o_p2 ~~ c(cov21)*HA_FU2_o_p2 + c(cov22)*HA_FU3_o_p2
HA_FU2_o_p2 ~~ c(cov21)*HA_FU3_o_p2
HA_IN_o_p3 ~~ c(cov31)*HA_FU1_o_p3 + c(cov32)*HA_FU2_o_p3 + c(cov33)*HA_FU3_o_p3
HA_FU1_o_p3 ~~ c(cov31)*HA_FU2_o_p3 + c(cov32)*HA_FU3_o_p3
HA_FU2_o_p3 ~~ c(cov31)*HA_FU3_o_p3
'
sibHA <- sem(lgmHA, data = yo, missing = "FIML")
fitMeasures(sibHA, c("chisq", "df", "pvalue", "cfi", "rmsea", "srmr"),
output = "text")##
## Model Test User Model:
##
## Test statistic 757.383
## Degrees of freedom 265
## P-value 0.000
##
## User Model versus Baseline Model:
##
## Comparative Fit Index (CFI) 0.900
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.055
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.163as.data.frame(unclass(standardizedSolution(sibHA))) %>%
filter(op == "~1" & lhs %in% c("iy", "io", "sy", "so")) %>%
select("Parameters" = lhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Intercept and Slope", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Parameters | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|
| iy | 6.162 | 0.361 | 17.07 | 0.000000 | 5.455 | 6.87 |
| io | 4.942 | 0.222 | 22.26 | 0.000000 | 4.506 | 5.38 |
| sy | 0.884 | 0.158 | 5.59 | 0.000000 | 0.574 | 1.19 |
| so | 1.060 | 0.284 | 3.73 | 0.000193 | 0.503 | 1.62 |
| Note: | ||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
as.data.frame(unclass(standardizedSolution(sibHA))) %>%
filter(op == "~~" & lhs %in% c("sy", "iy") & rhs %in% c("so", "io")) %>%
select("Younger sibling" = lhs, "Older sibling" = rhs, Beta = est.std, SE = se, Z = z, "p-value" = pvalue, "CI lower" = ci.lower, "CI upper" = ci.upper) %>%
knitr::kable(caption = "Correlated intercepts and slopes", digits = 6) %>%
kable_styling() %>%
footnote(general = "i = intercept, s = slope, o = older sibling, y = younger sibling")| Younger sibling | Older sibling | Beta | SE | Z | p-value | CI lower | CI upper |
|---|---|---|---|---|---|---|---|
| iy | io | 0.3276 | 0.0613 | 5.347 | 0.000 | 0.207 | 0.4476 |
| iy | so | -0.1295 | 0.0996 | -1.300 | 0.194 | -0.325 | 0.0657 |
| sy | so | 0.0782 | 0.1094 | 0.715 | 0.475 | -0.136 | 0.2926 |
| Note: | |||||||
| i = intercept, s = slope, o = older sibling, y = younger sibling |
semPaths(sibHA, what = "col", whatLabels = "est", intercepts = T)
Adoption status
Adoption status - Aggression AG
# extract latent intercepts and slopes
adoptAG <- lavPredict(sibAG) %>%
as.data.frame() %>%
select(iy, io, sy, so)
# scale all variables
adoptAG <- scale(adoptAG, scale = F) %>% as.data.frame()
# add adoption variable
adoptAG$adopt <- factor(yo$adopt, levels = c(0,1),
labels = c("non-adoptive", "adoptive"))
# moderation
summary(lm(iy ~ io * adopt, data = adoptAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.056 | 0.026 | 2.18 | 0.029 |
| io | 0.501 | 0.059 | 8.47 | 0.000 |
| adoptadoptive | -0.082 | 0.032 | -2.59 | 0.010 |
| io:adoptadoptive | -0.216 | 0.072 | -3.01 | 0.003 |
summary(lm(sy ~ so * adopt, data = adoptAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.002 | 0.001 | -1.750 | 0.081 |
| so | 0.292 | 0.057 | 5.154 | 0.000 |
| adoptadoptive | 0.003 | 0.001 | 2.194 | 0.029 |
| so:adoptadoptive | -0.048 | 0.069 | -0.705 | 0.481 |
# interaction plot
plot_model(lm(iy ~ io * adopt, data = adoptAG),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Aggression intercept using Adoption status") +
theme_classic()
Adoption status - Control CON
# extract latent intercepts and slopes
adoptCON <- lavPredict(sibCON) %>%
as.data.frame() %>%
select(iy, io, sy, so)
# scale all variables
adoptCON <- scale(adoptCON, scale = F) %>% as.data.frame()
# add adoption variable
adoptCON$adopt <- factor(yo$adopt, levels = c(0,1),
labels = c("non-adoptive", "adoptive"))
# moderation
summary(lm(iy ~ io * adopt, data = adoptCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.013 | 0.018 | -0.684 | 0.494 |
| io | 0.059 | 0.058 | 1.012 | 0.312 |
| adoptadoptive | 0.019 | 0.023 | 0.821 | 0.412 |
| io:adoptadoptive | -0.074 | 0.069 | -1.062 | 0.289 |
summary(lm(sy ~ so * adopt, data = adoptCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.000 | 0.730 | 0.466 |
| so | -0.502 | 0.056 | -8.957 | 0.000 |
| adoptadoptive | 0.000 | 0.001 | -0.879 | 0.380 |
| so:adoptadoptive | 0.077 | 0.067 | 1.151 | 0.250 |
Adoption status - Harm Avoidance HA
# extract latent intercepts and slopes
adoptHA <- lavPredict(sibHA) %>%
as.data.frame() %>%
select(iy, io, sy, so)
# scale all variables
adoptHA <- scale(adoptHA, scale = F) %>% as.data.frame()
# add adoption variable
adoptHA$adopt <- factor(yo$adopt, levels = c(0,1),
labels = c("non-adoptive", "adoptive"))
# moderation
summary(lm(iy ~ io * adopt, data = adoptHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.068 | 0.021 | -3.25 | 0.001 |
| io | 0.443 | 0.045 | 9.80 | 0.000 |
| adoptadoptive | 0.106 | 0.026 | 4.09 | 0.000 |
| io:adoptadoptive | -0.201 | 0.056 | -3.60 | 0.000 |
summary(lm(sy ~ so * adopt, data = adoptHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance slope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.001 | 0.001 | -0.558 | 0.577 |
| so | 0.096 | 0.068 | 1.417 | 0.157 |
| adoptadoptive | 0.001 | 0.001 | 0.705 | 0.481 |
| so:adoptadoptive | 0.042 | 0.086 | 0.492 | 0.623 |
# interaction plot
plot_model(lm(iy ~ io * adopt, data = adoptHA),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Harm Avoidance intercept using Adoption status") +
theme_classic()
Sex similarity
Sex similarity - Aggression AG
# extract latent intercepts and slopes
sexAG <- lavPredict(sibAG) %>%
as.data.frame() %>%
select(iy, io, sy, so)
# scale all variables
sexAG <- scale(sexAG, scale = F) %>% as.data.frame()
# add adoption variable
sexAG$samesex <- factor(yo$samesex, levels = c(0,1),
labels = c("different-sex", "same-sex"))
# moderation
summary(lm(iy ~ io * samesex, data = sexAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.010 | 0.024 | 0.396 | 0.692 |
| io | 0.187 | 0.056 | 3.357 | 0.001 |
| samesexsame-sex | -0.013 | 0.031 | -0.407 | 0.684 |
| io:samesexsame-sex | 0.258 | 0.070 | 3.693 | 0.000 |
summary(lm(sy ~ so * samesex, data = sexAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.325 | 0.745 |
| so | 0.300 | 0.054 | 5.515 | 0.000 |
| samesexsame-sex | -0.001 | 0.001 | -0.421 | 0.674 |
| so:samesexsame-sex | -0.055 | 0.067 | -0.811 | 0.418 |
# interaction plot
plot_model(lm(iy ~ io * samesex, data = sexAG),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Aggression intercept using Sex similarity") +
theme_classic()
Sex similarity - Control CON
# extract latent intercepts and slopes
sexCON <- lavPredict(sibCON) %>%
as.data.frame() %>%
select(iy, io, sy, so)
# scale all variables
sexCON <- scale(sexCON, scale = F) %>% as.data.frame()
# add adoption variable
sexCON$samesex <- factor(yo$samesex, levels = c(0,1),
labels = c("different-sex", "same-sex"))
# moderation
summary(lm(iy ~ io * samesex, data = sexCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.018 | 0.017 | -1.044 | 0.297 |
| io | 0.029 | 0.049 | 0.594 | 0.553 |
| samesexsame-sex | 0.030 | 0.022 | 1.348 | 0.178 |
| io:samesexsame-sex | -0.040 | 0.064 | -0.630 | 0.529 |
summary(lm(sy ~ so * samesex, data = sexCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.000 | -0.234 | 0.815 |
| so | -0.456 | 0.046 | -9.838 | 0.000 |
| samesexsame-sex | 0.000 | 0.001 | 0.303 | 0.762 |
| so:samesexsame-sex | 0.015 | 0.061 | 0.253 | 0.800 |
Sex similarity - Harm Avoidance
# extract latent intercepts and slopes
sexHA <- lavPredict(sibHA) %>%
as.data.frame() %>%
select(iy, io, sy, so)
# scale all variables
sexHA <- scale(sexHA, scale = F) %>% as.data.frame()
# add adoption variable
sexHA$samesex <- factor(yo$samesex, levels = c(0,1),
labels = c("different-sex", "same-sex"))
# moderation
summary(lm(iy ~ io * samesex, data = sexHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.031 | 0.020 | 1.55 | 0.122 |
| io | 0.278 | 0.044 | 6.32 | 0.000 |
| samesexsame-sex | -0.053 | 0.026 | -2.07 | 0.039 |
| io:samesexsame-sex | 0.069 | 0.056 | 1.23 | 0.219 |
summary(lm(sy ~ so * samesex, data = sexHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance slope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.001 | 0.001 | 0.534 | 0.593 |
| so | 0.011 | 0.061 | 0.182 | 0.856 |
| samesexsame-sex | -0.001 | 0.001 | -0.692 | 0.489 |
| so:samesexsame-sex | 0.206 | 0.083 | 2.469 | 0.014 |
# interaction plot
plot_model(lm(sy ~ so * samesex, data = sexHA),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Harm Avoidance slope using Sex similarity") +
theme_classic()
Age gap
Here, we examined whether a smaller age gap between siblings would result in higher correlated change patterns compared to siblings who are further apart in age.
In order to fit the latent interaction model, factor scores for the latent intercepts and slopes are extracted from the LGM of general change using full data. All variables are then mean-centered.
# descriptives
yo$gap <- yo$IN_AGE_o - yo$IN_AGE_y
yo$gap %>%
descr(stats = "common", order = "p")## Descriptive Statistics
## yo$gap
## N: 614
##
## gap
## --------------- --------
## Mean 2.34
## Std.Dev 0.89
## Min 0.04
## Median 2.26
## Max 4.93
## N.Valid 606.00
## Pct.Valid 98.70# histogram
ggplot(data = yo, aes(gap)) +
geom_histogram(binwidth = 0.1, fill = "#710c0c") +
geom_vline(xintercept = mean(yo$gap, na.rm = T),
color = "#f1c232", size = 1) +
annotate("text",
label = paste0("Mean = ", round(mean(yo$gap, na.rm = T), 2),
", SD = ",round(sd(yo$gap, na.rm = T), 2)),
x = 4, y = 20) +
labs(
title = "Distribution of Between-Sibling Age Gap",
x = "Age Gap (in years)",
y = NULL
) +
theme_classic()
Age gap - Aggression AG
Age gap did not moderate the relationship between sibling Aggression scores or change over time.
# extract latent intercepts and slopes
ageAG <- lavPredict(sibAG) %>%
as.data.frame() %>%
select(iy, io, sy, so)
ageAG$gap <- yo$gap
# scale all variables
ageAG <- scale(ageAG, scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * gap, data = ageAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.001 | 0.015 | -0.082 | 0.935 |
| io | 0.351 | 0.034 | 10.289 | 0.000 |
| gap | -0.026 | 0.017 | -1.516 | 0.130 |
| io:gap | -0.034 | 0.038 | -0.888 | 0.375 |
summary(lm(sy ~ so * gap, data = ageAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.108 | 0.914 |
| so | 0.268 | 0.033 | 8.209 | 0.000 |
| gap | 0.001 | 0.001 | 1.220 | 0.223 |
| so:gap | 0.004 | 0.036 | 0.107 | 0.914 |
Age gap - Control CON
Age gap did not moderate the relationship between sibling Control scores, but it did moderate their correlated change over time \((B = -0.075, SE = 0.035, t = -2.16, p = 0.031)\). Siblings who are further apart in age showed more strongly negative correlated change in Control over time.
# extract latent intercepts and slopes
ageCON <- lavPredict(sibCON) %>%
as.data.frame() %>%
select(iy, io, sy, so)
ageCON$gap <- yo$gap
# scale all variables
ageCON <- scale(ageCON, scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * gap, data = ageCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.011 | 0.014 | 0.989 |
| io | 0.011 | 0.031 | 0.364 | 0.716 |
| gap | 0.009 | 0.012 | 0.786 | 0.432 |
| io:gap | -0.007 | 0.035 | -0.198 | 0.843 |
summary(lm(sy ~ so * gap, data = ageCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.000 | -0.094 | 0.925 |
| so | -0.447 | 0.030 | -14.786 | 0.000 |
| gap | 0.000 | 0.000 | -0.927 | 0.354 |
| so:gap | -0.075 | 0.035 | -2.160 | 0.031 |
# interaction plot at minimum and maximum age gap
plot_model(lm(sy ~ so * gap, data = ageCON),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Change in Control over time") +
theme_classic()
Age gap - Harm Avoidance HA
Age gap did not moderate the relationship between sibling Harm Avoidance scores or change over time.
Interestingly, age gap was a significant predictor of Harm Avoidance score such that younger siblings had higher average HA scores with higher between-sibling age gaps \((B = 0.038, SE = 0.014, t = 2.70, p = 0.007)\).
# extract latent intercepts and slopes
ageHA <- lavPredict(sibHA) %>%
as.data.frame() %>%
select(iy, io, sy, so)
ageHA$gap <- yo$gap
# scale all variables
ageHA <- scale(ageHA, scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * gap, data = ageHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.001 | 0.013 | 0.047 | 0.962 |
| io | 0.310 | 0.027 | 11.384 | 0.000 |
| gap | 0.038 | 0.014 | 2.696 | 0.007 |
| io:gap | -0.046 | 0.029 | -1.576 | 0.115 |
summary(lm(sy ~ so * gap, data = ageHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance sllope ", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.013 | 0.989 |
| so | 0.124 | 0.042 | 2.936 | 0.003 |
| gap | 0.001 | 0.001 | 0.721 | 0.471 |
| so:gap | -0.069 | 0.052 | -1.314 | 0.189 |
Relationship quality
Sibling relationship was measured by the Sibling Relationship Questionnaire (SRQ; Furman, & Buhrmester, 1985). The scale has 48 items. We excluded items 2, 18, 34 (subscale maternal partiality), and items 7, 23, 39 (subscale paternal partiality) during data collection. This was measured at the first two waves: IN and FU1. There are a total of 14 subscales, each with 3 items, and they can be organized into 3 (overlapping) factors. The original paper included a 4th factor, Rivalry, which consists of Competition and Parental partiality (which we did not measure).
- Warmth/Closeness:
- Initimacy
- Prosocial behavior
- Companionship
- Similarity
- Nurturance of sibling
- Nurturance by sibling
- Admiration of sibling
- Admiration by sibling
- Affection
- Relative status/power:
- Nurturance by sibling (R)
- Nurturance of sibling
- Admiration by sibling
- Admiration of sibling (R)
- Dominance by sibling (R)
- Dominance over sibling
- Conflict:
- Admiration by sibling (R)
- Affection (R)
- Dominance by sibling
- Dominance over sibling
- Quarreling
- Antagonism
- Competition
# descriptives
mpq %>%
select(S_PROSOC:S_QUARR) %>%
descr(stats = "common", order = "p")## Descriptive Statistics
## mpq
## N: 1231
##
## S_PROSOC S_YNURE S_ENURY S_YDOME S_EDOMY S_AFFECT S_COMPAN S_ANTAG
## --------------- ---------- --------- --------- --------- --------- ---------- ---------- ---------
## Mean 3.06 2.93 2.83 2.33 2.30 3.83 2.94 2.72
## Std.Dev 0.70 0.77 0.84 0.86 0.84 0.89 0.87 0.86
## Min 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00
## Median 3.00 3.00 3.00 2.33 2.33 4.00 3.00 2.67
## Max 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00
## N.Valid 1219.00 1219.00 1219.00 1219.00 1219.00 1217.00 1217.00 1217.00
## Pct.Valid 99.03 99.03 99.03 99.03 99.03 98.86 98.86 98.86
##
## Table: Table continues below
##
##
##
## S_SIMLR S_INTIM S_COMPET S_YADME S_EADMY S_QUARR
## --------------- --------- --------- ---------- --------- --------- ---------
## Mean 2.85 2.38 2.49 3.45 3.17 2.81
## Std.Dev 0.88 1.05 0.93 0.91 0.90 0.92
## Min 1.00 1.00 1.00 1.00 1.00 1.00
## Median 3.00 2.33 2.33 3.67 3.00 3.00
## Max 5.00 5.00 5.00 5.00 5.00 5.00
## N.Valid 1217.00 1217.00 1217.00 1217.00 1215.00 1216.00
## Pct.Valid 98.86 98.86 98.86 98.86 98.70 98.78We will use relationship quality data at intake for all participants. However, there are 26 participants who did not have relationship data at IN but had FU1 data that will be used instead.
Explanatory factor analysis
We ran an explanatory factor analysis to examine the factor structure and extract factor scores.
This resulted in two main factors, similarly for younger and older siblings:
Factor 1 resembles the original paper’s Warmth/Closeness; Factor 2 resembles Conflict.
# younger siblings
efa_y <- yo %>%
select(IDYRFAM, starts_with("S_")) %>%
select(IDYRFAM, ends_with("_y")) %>%
na.omit()
scree(efa_y[, -1], pc = FALSE, main = "Scree plot for younger siblings")
efa_fit_y <- factanal(efa_y[, -1], 2, rotation = "promax",
score = "Bartlett")
print(efa_fit_y$loadings, digits = 3, cutoff = 0.3, sort = T)##
## Loadings:
## Factor1 Factor2
## S_PROSOC_y 0.836
## S_YNURE_y 0.644
## S_ENURY_y 0.755
## S_AFFECT_y 0.737
## S_COMPAN_y 0.815
## S_SIMLR_y 0.689
## S_INTIM_y 0.749
## S_YADME_y 0.749
## S_EADMY_y 0.751
## S_YDOME_y 0.505
## S_EDOMY_y 0.611
## S_ANTAG_y 0.846
## S_COMPET_y 0.512
## S_QUARR_y 0.835
##
## Factor1 Factor2
## SS loadings 5.220 2.408
## Proportion Var 0.373 0.172
## Cumulative Var 0.373 0.545# older siblings
efa_o <- yo %>%
select(IDYRFAM, starts_with("S_")) %>%
select(IDYRFAM, ends_with("_o")) %>%
na.omit()
scree(efa_o[, -1], pc = FALSE, main = "Scree plot for older siblings")
efa_fit_o <- factanal(efa_o[, -1], 2, rotation = "promax",
score = "Bartlett")
print(efa_fit_o$loadings, digits = 3, cutoff = 0.3, sort = T)##
## Loadings:
## Factor1 Factor2
## S_PROSOC_o 0.850
## S_YNURE_o 0.748
## S_ENURY_o 0.745
## S_AFFECT_o 0.769
## S_COMPAN_o 0.820
## S_SIMLR_o 0.729
## S_INTIM_o 0.721
## S_YADME_o 0.764
## S_EADMY_o 0.736
## S_YDOME_o 0.656
## S_ANTAG_o 0.828
## S_COMPET_o 0.604
## S_QUARR_o 0.840
## S_EDOMY_o 0.434
##
## Factor1 Factor2
## SS loadings 5.47 2.464
## Proportion Var 0.39 0.176
## Cumulative Var 0.39 0.566Relationship - Aggression AG
1. Using younger sibling’s relationship ratings:
There were significant interaction effects between relationship quality and sibling similarity in Aggression, but not in changes in Aggression over time.
In particular, there was a stronger between-sibling correlation in Aggression when the younger sibling indicated higher Warmth/Closeness scores, \(B = 0.081, SE = 0.030, t = 2.724, p = 0.007\). Interestingly, between-sibling correlation in Aggression was also stronger when the younger sibling indicated higher Conflict scores, \(B = 0.057, SE = 0.028, t = 2.057, p = 0.040\).
Further, younger sibling’s Conflict scores were associated with their own higher Aggression intercept \((B = 0.088, SE = 0.014, t = 6.234, p < 0.001)\) and slower decrease in Aggression over time \((B = -0.003, SE = =.001, t = -5.097, p < 0.001)\).
#### younger siblings
# merge rela scores with latent intercepts and slopes
relaAG_y <- cbind(ageAG, yo$IDYRFAM)
efa_y <- cbind(efa_y$IDYRFAM, efa_fit_y$scores)
colnames(efa_y)[1] <- colnames(relaAG_y)[6] <- "IDYRFAM"
relaAG_y <- merge(relaAG_y, efa_y)
# scale all variables
relaAG_y <- scale(relaAG_y[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * Factor1, data = relaAG_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept using younger sibling's Warmth ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.007 | 0.015 | 0.447 | 0.655 |
| io | 0.344 | 0.035 | 9.898 | 0.000 |
| Factor1 | -0.026 | 0.015 | -1.708 | 0.088 |
| io:Factor1 | 0.081 | 0.030 | 2.724 | 0.007 |
summary(lm(iy ~ io * Factor2, data = relaAG_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept using younger sibling's Conflict ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.004 | 0.015 | -0.276 | 0.783 |
| io | 0.321 | 0.034 | 9.571 | 0.000 |
| Factor2 | 0.088 | 0.014 | 6.234 | 0.000 |
| io:Factor2 | 0.057 | 0.028 | 2.057 | 0.040 |
summary(lm(sy ~ so * Factor1, data = relaAG_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope using younger sibling's Warmth ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | -0.149 | 0.882 |
| so | 0.263 | 0.032 | 8.103 | 0.000 |
| Factor1 | 0.001 | 0.001 | 1.027 | 0.305 |
| so:Factor1 | 0.045 | 0.029 | 1.533 | 0.126 |
summary(lm(sy ~ so * Factor2, data = relaAG_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope using younger sibling's Conflict ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.029 | 0.977 |
| so | 0.249 | 0.032 | 7.792 | 0.000 |
| Factor2 | -0.003 | 0.001 | -5.097 | 0.000 |
| so:Factor2 | 0.010 | 0.029 | 0.338 | 0.735 |
# interaction plot at minimum and maximum relationship ratings
plot_model(lm(iy ~ io * Factor1, data = relaAG_y),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Aggression intercept using younger sibling's Warmth ratings") +
theme_classic()
plot_model(lm(iy ~ io * Factor2, data = relaAG_y),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Aggression intercept using younger sibling's Conflict ratings") +
theme_classic()
2. Using older sibling’s relationship ratings:
Results were different when we used older sibling’s relationship ratings. The only significant interaction effect was to predict Aggression intercept. In particular, there was a higher between-sibling correlation in Aggression when the older sibling provided higher Warmth scores, \(B = 0.104, SE = 0.031, t = 3.303, p = 0.001\). Older siblings’ ratings of Warmth and Conflict did not have any association with correlated change in Aggression.
#### older siblings
# merge rela scores with latent intercepts and slopes
relaAG_o <- cbind(ageAG, yo$IDYRFAM)
efa_o <- cbind(efa_o$IDYRFAM, efa_fit_o$scores)
colnames(efa_o)[1] <- colnames(relaAG_o)[6] <- "IDYRFAM"
relaAG_o <- merge(relaAG_o, efa_o)
# scale all variables
relaAG_o <- scale(relaAG_o[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * Factor1, data = relaAG_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept using older sibling's Warmth ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.012 | 0.016 | 0.765 | 0.445 |
| io | 0.362 | 0.035 | 10.301 | 0.000 |
| Factor1 | 0.006 | 0.015 | 0.409 | 0.683 |
| io:Factor1 | 0.104 | 0.031 | 3.303 | 0.001 |
summary(lm(iy ~ io * Factor2, data = relaAG_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept using older sibling's Conflict ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.004 | 0.016 | -0.273 | 0.785 |
| io | 0.330 | 0.037 | 8.898 | 0.000 |
| Factor2 | 0.022 | 0.016 | 1.398 | 0.163 |
| io:Factor2 | 0.025 | 0.029 | 0.834 | 0.405 |
summary(lm(sy ~ so * Factor1, data = relaAG_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope using older sibling's Warmth ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | -0.244 | 0.807 |
| so | 0.278 | 0.033 | 8.492 | 0.000 |
| Factor1 | -0.002 | 0.001 | -2.381 | 0.018 |
| so:Factor1 | 0.043 | 0.030 | 1.402 | 0.161 |
summary(lm(sy ~ so * Factor2, data = relaAG_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope using older sibling's Conflict ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.186 | 0.852 |
| so | 0.242 | 0.033 | 7.283 | 0.000 |
| Factor2 | -0.002 | 0.001 | -2.572 | 0.010 |
| so:Factor2 | 0.027 | 0.031 | 0.877 | 0.381 |
# interaction plot at minimum and maximum relationship ratings
plot_model(lm(iy ~ io * Factor1, data = relaAG_o),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Aggression intercept using older sibling's Warmth ratings") +
theme_classic()
Relationship - Control CON
1. Using younger sibling’s relationship ratings:
Relationship quality, as rated by the younger siblings, had no association with between-sibling correlation in Control scores or changes in Control over time.
Younger siblings’ ratings of Warmth was associated with higher intercepts of their own Control scores \((B = 0.036, SE = 0.010, t = 3.478, p = 0.001)\) but also a slower increase in their Control scores over time \((B = -0.001, SE = 0.00, t = -2.036, p = 0.042)\).
#### younger siblings
# merge rela scores with latent intercepts and slopes
relaCON_y <- cbind(ageCON, yo$IDYRFAM)
colnames(relaCON_y)[6] <- "IDYRFAM"
relaCON_y <- merge(relaCON_y, efa_y)
# scale all variables
relaCON_y <- scale(relaCON_y[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * Factor1, data = relaCON_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept using younger sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.001 | 0.011 | -0.058 | 0.953 |
| io | -0.008 | 0.032 | -0.256 | 0.798 |
| Factor1 | 0.036 | 0.010 | 3.478 | 0.001 |
| io:Factor1 | 0.017 | 0.029 | 0.590 | 0.556 |
summary(lm(iy ~ io * Factor2, data = relaCON_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept using younger sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.001 | 0.011 | -0.130 | 0.897 |
| io | -0.018 | 0.031 | -0.567 | 0.571 |
| Factor2 | -0.057 | 0.010 | -5.731 | 0.000 |
| io:Factor2 | -0.040 | 0.027 | -1.492 | 0.136 |
summary(lm(sy ~ so * Factor1, data = relaCON_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope using younger sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.00 | 0.033 | 0.974 |
| so | -0.454 | 0.03 | -14.944 | 0.000 |
| Factor1 | -0.001 | 0.00 | -2.036 | 0.042 |
| so:Factor1 | 0.016 | 0.03 | 0.525 | 0.600 |
summary(lm(sy ~ so * Factor2, data = relaCON_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope using younger sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.000 | -0.004 | 0.997 |
| so | -0.449 | 0.030 | -14.932 | 0.000 |
| Factor2 | 0.001 | 0.000 | 3.793 | 0.000 |
| so:Factor2 | -0.016 | 0.028 | -0.590 | 0.555 |
2. Using older sibling’s relationship ratings:
Similarly to younger siblings’ ratings, older siblings’ ratings of relationship quality had no association with between-sibling correlation in Control scores or changes in Control over time.
#### older siblings
# merge rela scores with latent intercepts and slopes
relaCON_o <- cbind(ageCON, yo$IDYRFAM)
colnames(relaCON_o)[6] <- "IDYRFAM"
relaCON_o <- merge(relaCON_o, efa_o)
# scale all variables
relaCON_o <- scale(relaCON_o[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * Factor1, data = relaCON_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.001 | 0.011 | -0.128 | 0.899 |
| io | -0.004 | 0.032 | -0.134 | 0.893 |
| Factor1 | 0.018 | 0.011 | 1.633 | 0.103 |
| io:Factor1 | 0.022 | 0.028 | 0.784 | 0.434 |
summary(lm(iy ~ io * Factor2, data = relaCON_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.002 | 0.011 | -0.156 | 0.876 |
| io | -0.013 | 0.032 | -0.402 | 0.688 |
| Factor2 | -0.032 | 0.010 | -3.063 | 0.002 |
| io:Factor2 | -0.025 | 0.028 | -0.867 | 0.387 |
summary(lm(sy ~ so * Factor1, data = relaCON_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.000 | 0.021 | 0.983 |
| so | -0.453 | 0.031 | -14.622 | 0.000 |
| Factor1 | 0.000 | 0.000 | -0.452 | 0.651 |
| so:Factor1 | 0.004 | 0.028 | 0.147 | 0.883 |
summary(lm(sy ~ so * Factor2, data = relaCON_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.000 | -0.049 | 0.961 |
| so | -0.461 | 0.030 | -15.223 | 0.000 |
| Factor2 | 0.001 | 0.000 | 4.156 | 0.000 |
| so:Factor2 | 0.016 | 0.028 | 0.571 | 0.568 |
Relationship - Harm Avoidance HA
There was absolutely no main or interaction effects involving relationship quality in predicting Harm Avoidance intercepts or slopes.
#### younger siblings
# merge rela scores with latent intercepts and slopes
relaHA_y <- cbind(ageHA, yo$IDYRFAM)
colnames(relaHA_y)[6] <- "IDYRFAM"
relaHA_y <- merge(relaHA_y, efa_y)
# scale all variables
relaHA_y <- scale(relaHA_y[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * Factor1, data = relaHA_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.001 | 0.013 | -0.084 | 0.933 |
| io | 0.312 | 0.028 | 11.275 | 0.000 |
| Factor1 | 0.012 | 0.012 | 0.944 | 0.345 |
| io:Factor1 | 0.018 | 0.026 | 0.696 | 0.486 |
summary(lm(iy ~ io * Factor2, data = relaHA_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.013 | -0.034 | 0.973 |
| io | 0.313 | 0.027 | 11.382 | 0.000 |
| Factor2 | -0.007 | 0.012 | -0.585 | 0.559 |
| io:Factor2 | -0.014 | 0.026 | -0.550 | 0.582 |
summary(lm(sy ~ so * Factor1, data = relaHA_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance slope using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.061 | 0.951 |
| so | 0.123 | 0.042 | 2.914 | 0.004 |
| Factor1 | 0.000 | 0.001 | 0.335 | 0.738 |
| so:Factor1 | 0.049 | 0.041 | 1.193 | 0.233 |
summary(lm(sy ~ so * Factor2, data = relaHA_y))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance slope using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | -0.002 | 0.998 |
| so | 0.119 | 0.042 | 2.824 | 0.005 |
| Factor2 | 0.000 | 0.001 | 0.313 | 0.754 |
| so:Factor2 | -0.007 | 0.042 | -0.165 | 0.869 |
#### older siblings
# merge rela scores with latent intercepts and slopes
relaHA_o <- cbind(ageHA, yo$IDYRFAM)
colnames(relaHA_o)[6] <- "IDYRFAM"
relaHA_o <- merge(relaHA_o, efa_o)
# scale all variables
relaHA_o <- scale(relaHA_o[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * Factor1, data = relaHA_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.003 | 0.013 | -0.237 | 0.812 |
| io | 0.314 | 0.028 | 11.228 | 0.000 |
| Factor1 | -0.006 | 0.012 | -0.443 | 0.658 |
| io:Factor1 | 0.036 | 0.025 | 1.450 | 0.148 |
summary(lm(iy ~ io * Factor2, data = relaHA_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.013 | -0.005 | 0.996 |
| io | 0.311 | 0.028 | 11.284 | 0.000 |
| Factor2 | 0.013 | 0.012 | 1.121 | 0.263 |
| io:Factor2 | -0.002 | 0.024 | -0.090 | 0.928 |
summary(lm(sy ~ so * Factor1, data = relaHA_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance slope using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.011 | 0.991 |
| so | 0.116 | 0.042 | 2.733 | 0.006 |
| Factor1 | 0.000 | 0.001 | 0.269 | 0.788 |
| so:Factor1 | 0.006 | 0.041 | 0.156 | 0.876 |
summary(lm(sy ~ so * Factor2, data = relaHA_o))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance slope using older sibling's relationship ratings", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | -0.006 | 0.995 |
| so | 0.115 | 0.042 | 2.723 | 0.007 |
| Factor2 | 0.000 | 0.001 | 0.632 | 0.527 |
| so:Factor2 | -0.016 | 0.042 | -0.386 | 0.700 |
Life events
Life events data were collected from a Life Event Interview at intake, follow-up 1, and follow-up 2. These events were categorized into 3 groups following previous research byBillig et al., 1996:
# discrepancy count between siblings
fam_list <- c("Q1B","Q3A","Q10","Q11A","Q17","Q17A","Q18B",
"Q19AP","Q20AP","Q21A","Q21B","Q15","Q22S","Q35",
"Q41D","Q49A","Q49B","Q41C","Q22P","Q43F","Q46P")
inf_list <- c("Q7","Q8","Q9","Q24","Q25","Q26","Q27",
"Q29E","Q51B")
ninf_list <- c("Q4A","Q5","Q6A","Q14B","Q20B","Q28B1","Q29B",
"Q29C","Q29D","Q38B","Q38H","Q38C","Q13","Q30A",
"Q31","Q16F","Q16D","Q16B")
count <- vector(mode = "numeric", length = nrow(yo))
for(i in c(inf_list, ninf_list)){
current <- yo %>% select(starts_with(i))
count <- count + as.numeric(current[,1] == current[,2])
}
yo$lei_shared <- count
# descriptives
fam <- lei %>%
select(all_of(fam_list))
inf <- lei %>%
select(all_of(inf_list))
ninf <- lei %>%
select(all_of(ninf_list))
colnames(fam) <- var_label(fam)
colnames(inf) <- var_label(inf)
colnames(ninf) <- var_label(ninf)
fam <- fam %>%
mutate_all(.funs = function(x){
factor(x, levels = c(0:1),
labels = c("No", "Yes"))
})
inf <- inf %>%
mutate_all(.funs = function(x){
factor(x, levels = c(0:1),
labels = c("No", "Yes"))
})
ninf <- ninf %>%
mutate_all(.funs = function(x){
factor(x, levels = c(0:1),
labels = c("No", "Yes"))
})- Family related events (FAM): events that happened to the family or to individuals within the family
ggplot(data = lei, aes(FAM)) +
geom_histogram(binwidth = 1, fill = "#710c0c") +
geom_vline(xintercept = mean(lei$FAM, na.rm = T),
color = "#f1c232", size = 1) +
annotate("text",
label = paste0("Mean = ", round(mean(lei$FAM, na.rm = T), 2),
", SD = ",round(sd(lei$FAM, na.rm = T), 2)),
x = 10, y = 150) +
labs(
title = "Distribution of Family-Related Events",
x = "Count of events",
y = NULL
) +
theme_classic()
plot(likert(fam))
- Independent non-family events (INF): non-family events that were independent of the respondent’s behavior
ggplot(data = lei, aes(INF)) +
geom_histogram(binwidth = 1, fill = "#710c0c") +
geom_vline(xintercept = mean(lei$INF, na.rm = T),
color = "#f1c232", size = 1) +
annotate("text",
label = paste0("Mean = ", round(mean(lei$INF, na.rm = T), 2),
", SD = ",round(sd(lei$INF, na.rm = T), 2)),
x = 7.5, y = 150) +
labs(
title = "Distribution of Independent Non-Family Events",
x = "Count of events",
y = NULL
) +
theme_classic()
plot(likert(inf))
- Non-independent non-family events (NINF): non-family events in which the respondent’s behavior may have influenced whether or not the event occurred
ggplot(data = lei, aes(NINF)) +
geom_histogram(binwidth = 1, fill = "#710c0c") +
geom_vline(xintercept = mean(lei$NINF, na.rm = T),
color = "#f1c232", size = 1) +
annotate("text",
label = paste0("Mean = ", round(mean(lei$NINF, na.rm = T), 2),
", SD = ",round(sd(lei$NINF, na.rm = T), 2)),
x = 10, y = 150) +
labs(
title = "Distribution of Non-Independent Non-Family Events",
x = "Count of events",
y = NULL
) +
theme_classic()
plot(likert(ninf))
In the following analyses, we used the number of family-related events (as reported by the older siblings), as well as the count of similar non-family events that the siblings both experienced.
ggplot(data = yo, aes(lei_shared)) +
geom_histogram(binwidth = 1, fill = "#710c0c") +
geom_vline(xintercept = mean(yo$lei_shared, na.rm = T),
color = "#f1c232", size = 1) +
annotate("text",
label = paste0("Mean = ", round(mean(yo$lei_shared, na.rm = T), 2),
", SD = ",round(sd(yo$lei_shared, na.rm = T), 2)),
x = 15, y = 50) +
labs(
title = "Distribution of Similar Non-Family Events",
x = "Count of events",
y = NULL
) +
theme_classic()
Life events - Aggression AG
There were significant interaction effects between family-related events and sibling similarity in changes in Aggression over time.
In particular, there was a stronger between-sibling correlation in Aggression development when there were more family-related events reported, \(B = 0.031, SE = 0.012, t = 2.584, p = 0.010\).
Further, interestingly, there were significant main effects between life events and Aggression (in younger siblings) such that:
- Family-related events were associated with higher Aggression intercept \((B = 0.033, SE = 0.006, t = 5.730, p < 0.001)\) and faster decrease in Aggression \((B = -0.001, SE = 0.000, t = -2.750, p = 0.006)\)
- Similarity in non-family events were associated with lower Aggression intercept \((B = -0.025, SE = 0.005 t = -4.696, p < 0.001)\) and slower decrease in Aggression \((B = 0.000, SE = 0.000, t = -2.001, p = 0.046)\)
# merge lei counts with latent intercepts and slopes
leiAG <- cbind(select(ageAG, c(io,iy,so,sy)), yo$IDYRFAM)
colnames(leiAG)[5] <- "IDYRFAM"
leiAG <- merge(leiAG, select(yo, c(IDYRFAM, FAM_o, lei_shared)))
# scale all variables
leiAG <- scale(leiAG[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * FAM_o, data = leiAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept using count of family-related events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.001 | 0.015 | -0.060 | 0.953 |
| io | 0.334 | 0.033 | 10.042 | 0.000 |
| FAM_o | 0.033 | 0.006 | 5.730 | 0.000 |
| io:FAM_o | 0.010 | 0.012 | 0.764 | 0.445 |
summary(lm(iy ~ io * lei_shared, data = leiAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression intercept using count of similar non-family events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.015 | 0.020 | 0.984 |
| io | 0.318 | 0.034 | 9.238 | 0.000 |
| lei_shared | -0.025 | 0.005 | -4.696 | 0.000 |
| io:lei_shared | 0.002 | 0.012 | 0.134 | 0.893 |
summary(lm(sy ~ so * FAM_o, data = leiAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope using count of family-related events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.224 | 0.823 |
| so | 0.254 | 0.032 | 7.977 | 0.000 |
| FAM_o | -0.001 | 0.000 | -2.750 | 0.006 |
| so:FAM_o | 0.031 | 0.012 | 2.584 | 0.010 |
summary(lm(sy ~ so * lei_shared, data = leiAG))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Aggression slope using count of similar non-family events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.087 | 0.930 |
| so | 0.255 | 0.032 | 7.891 | 0.000 |
| lei_shared | 0.000 | 0.000 | 2.001 | 0.046 |
| so:lei_shared | -0.007 | 0.010 | -0.723 | 0.470 |
# interaction plot at minimum and maximum family events
plot_model(lm(sy ~ so * FAM_o, data = leiAG),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Aggression slope using count of family-related events") +
theme_classic()
Life events - Control CON
There were no interaction effects between life events and between-sibling correlations in Control or changes in Control over time.
Similarly to Aggression, however, there were significant main effects of life events on Control such that:
- Family-related events were associated with lower Control intercept \((B = -0.018, SE = 0.004, t = -4.323, p < 0.001)\) and faster increase in Control \((B = 0.000, SE = 0.000, t = 3.173, p = 0.002)\)
- Similarity in non-family events were associated with higher Control intercept \((B = 0.015, SE = 0.004, t = 3.937, p < 0.001)\) and slower increase in Control \((B = 0.000, SE = 0.000, t = -4.049, p < 0.001)\).
# merge lei counts with latent intercepts and slopes
leiCON <- cbind(select(ageCON, c(io,iy,so,sy)), yo$IDYRFAM)
colnames(leiCON)[5] <- "IDYRFAM"
leiCON <- merge(leiCON, select(yo, c(IDYRFAM, FAM_o, lei_shared)))
# scale all variables
leiCON <- scale(leiCON[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * FAM_o, data = leiCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept using count of family-related events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.011 | 0.005 | 0.996 |
| io | 0.001 | 0.031 | 0.019 | 0.985 |
| FAM_o | -0.018 | 0.004 | -4.324 | 0.000 |
| io:FAM_o | 0.001 | 0.012 | 0.103 | 0.918 |
summary(lm(iy ~ io * lei_shared, data = leiCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control intercept using count of similar non-family events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.001 | 0.011 | -0.086 | 0.931 |
| io | -0.016 | 0.032 | -0.506 | 0.613 |
| lei_shared | 0.015 | 0.004 | 3.937 | 0.000 |
| io:lei_shared | 0.007 | 0.011 | 0.617 | 0.537 |
summary(lm(sy ~ so * FAM_o, data = leiCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope using count of family-related events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.000 | 0.005 | 0.996 |
| so | -0.446 | 0.030 | -14.834 | 0.000 |
| FAM_o | 0.000 | 0.000 | 3.173 | 0.002 |
| so:FAM_o | 0.003 | 0.012 | 0.238 | 0.812 |
summary(lm(sy ~ so * lei_shared, data = leiCON))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Control slope using count of similar non-family events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.000 | -0.088 | 0.930 |
| so | -0.459 | 0.030 | -15.280 | 0.000 |
| lei_shared | 0.000 | 0.000 | -4.049 | 0.000 |
| so:lei_shared | -0.007 | 0.011 | -0.622 | 0.534 |
Life events - Harm Avoidance HA
Similarity in non-family events significantly moderated between-sibling correlation in both intercepts and slopes of Harm Avoidance.
In particular, siblings with more similar non-family events were more similar in Harm Avoidance intercept \((B = 0.024, SE = 0.010, t = 2.494, p = 0.013)\) and slope \((B = 0.030, SE = 0.015, t = 2.016, p = 0.044)\).
There were significant main effects of life events on changes in Harm Avoidance overtime such that:
- Family-related events were associated with faster increase in Harm Avoidance Control \((B = 0.001, SE = 0.000, t = 2.408, p = 0.016)\)
- Similarity in non-family events were associated with slower increase in Harm Avoidance \((B = -0.001, SE = 0.000, t = -2.682, p = 0.008)\).
# merge lei counts with latent intercepts and slopes
leiHA <- cbind(select(ageHA, c(io,iy,so,sy)), yo$IDYRFAM)
colnames(leiHA)[5] <- "IDYRFAM"
leiHA <- merge(leiHA, select(yo, c(IDYRFAM, FAM_o, lei_shared)))
# scale all variables
leiHA <- scale(leiHA[, -1], scale = F) %>% as.data.frame()
# moderation
summary(lm(iy ~ io * FAM_o, data = leiHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept using count of family-related events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.012 | 0.006 | 0.995 |
| io | 0.314 | 0.027 | 11.568 | 0.000 |
| FAM_o | -0.007 | 0.005 | -1.368 | 0.172 |
| io:FAM_o | 0.009 | 0.010 | 0.828 | 0.408 |
summary(lm(iy ~ io * lei_shared, data = leiHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance intercept using count of similar non-family events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | -0.003 | 0.013 | -0.248 | 0.804 |
| io | 0.314 | 0.027 | 11.472 | 0.000 |
| lei_shared | 0.006 | 0.004 | 1.336 | 0.182 |
| io:lei_shared | 0.024 | 0.010 | 2.494 | 0.013 |
summary(lm(sy ~ so * FAM_o, data = leiHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance slope using count of family-related events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.012 | 0.990 |
| so | 0.117 | 0.042 | 2.797 | 0.005 |
| FAM_o | 0.001 | 0.000 | 2.408 | 0.016 |
| so:FAM_o | -0.004 | 0.015 | -0.280 | 0.779 |
summary(lm(sy ~ so * lei_shared, data = leiHA))$coefficients %>%
knitr::kable(caption = "Predicting younger siblings's Harm Avoidance slope using count of similar non-family events", digits = 3) %>%
kable_styling()| Estimate | Std. Error | t value | Pr(>|t|) | |
|---|---|---|---|---|
| (Intercept) | 0.000 | 0.001 | 0.082 | 0.935 |
| so | 0.121 | 0.042 | 2.901 | 0.004 |
| lei_shared | -0.001 | 0.000 | -2.682 | 0.008 |
| so:lei_shared | 0.030 | 0.015 | 2.016 | 0.044 |
# interaction plot at minimum and maximum family events
plot_model(lm(iy ~ io * lei_shared, data = leiHA),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Harm Avoidance intercept using count of similar non-family events") +
theme_classic()
plot_model(lm(sy ~ so * lei_shared, data = leiHA),
type = "int",
axis.title = c("Older sibling", "Younger sibling"),
title = "Harm Avoidance slope using count of similar non-family events") +
theme_classic()