The data used in these analyses comes from the Early Experiences Lab BABY Affect and Behavior (BABY 1.0) study at the University of Utah, overseen by Dr. Lee Raby.

These analyses are likely to be used in my Honors thesis, which will be completed in spring of 2025. My complete Honors thesis proposal is linked here.

Let’s load the dataset.

library(haven)
baby <- read_sav("theadata_baby1.0.sav")

Next, let’s get an overview of the data.

head(baby)
## # A tibble: 6 × 7
##   FamilyID DERStotmA FreePlay_Toy_OverallParentingQu…¹ ITSEAdysTmD1 ITSEAextTmD1
##      <dbl>     <dbl>                             <dbl>        <dbl>        <dbl>
## 1        4        55                                NA           48           57
## 2        6       101                                NA           38           52
## 3        9        66                                NA           38           70
## 4       16        52                                NA           NA           NA
## 5       17        50                                NA           NA           NA
## 6       18       104                                NA           NA           NA
## # ℹ abbreviated name: ¹​FreePlay_Toy_OverallParentingQuality
## # ℹ 2 more variables: ITSEAintTmD1 <dbl>, ITSEAcompTmD1 <dbl>
str(baby)
## tibble [162 × 7] (S3: tbl_df/tbl/data.frame)
##  $ FamilyID                            : num [1:162] 4 6 9 16 17 18 22 23 27 28 ...
##   ..- attr(*, "label")= chr "Family unit ID"
##   ..- attr(*, "format.spss")= chr "F8.0"
##   ..- attr(*, "display_width")= int 13
##  $ DERStotmA                           : num [1:162] 55 101 66 52 50 104 82 45 110 49 ...
##   ..- attr(*, "label")= chr "prenatal mom DERStot DERS total score (range 36-180; higher score indicates greater difficulties with regulating emotions)"
##   ..- attr(*, "format.spss")= chr "F8.2"
##   ..- attr(*, "display_width")= int 11
##  $ FreePlay_Toy_OverallParentingQuality: num [1:162] NA NA NA NA NA NA NA NA NA NA ...
##   ..- attr(*, "format.spss")= chr "F8.2"
##   ..- attr(*, "display_width")= int 34
##  $ ITSEAdysTmD1                        : num [1:162] 48 38 38 NA NA NA 69 NA 36 NA ...
##   ..- attr(*, "label")= chr "18-month mom ITSEA Dysregulattion T score."
##   ..- attr(*, "format.spss")= chr "F8.2"
##   ..- attr(*, "display_width")= int 22
##  $ ITSEAextTmD1                        : num [1:162] 57 52 70 NA NA NA 71 NA 51 NA ...
##   ..- attr(*, "label")= chr "18-month mom ITSEA Externalizing T score."
##   ..- attr(*, "format.spss")= chr "F8.2"
##   ..- attr(*, "display_width")= int 22
##  $ ITSEAintTmD1                        : num [1:162] 52 52 68 NA NA NA 46 NA 58 NA ...
##   ..- attr(*, "label")= chr "18-month mom ITSEA Internalizing T score."
##   ..- attr(*, "format.spss")= chr "F8.2"
##   ..- attr(*, "display_width")= int 22
##  $ ITSEAcompTmD1                       : num [1:162] 47 57 49 NA NA NA 51 NA 39 NA ...
##   ..- attr(*, "label")= chr "18-month mom ITSEA Competence T score."
##   ..- attr(*, "format.spss")= chr "F8.2"
##   ..- attr(*, "display_width")= int 19
summary(baby)
##     FamilyID        DERStotmA      FreePlay_Toy_OverallParentingQuality
##  Min.   :  4.00   Min.   : 36.00   Min.   :1.750                       
##  1st Qu.: 79.25   1st Qu.: 59.75   1st Qu.:2.771                       
##  Median :177.50   Median : 73.00   Median :3.281                       
##  Mean   :206.88   Mean   : 79.93   Mean   :3.313                       
##  3rd Qu.:294.75   3rd Qu.: 98.25   3rd Qu.:3.875                       
##  Max.   :621.00   Max.   :155.00   Max.   :4.812                       
##                   NA's   :2        NA's   :60                          
##   ITSEAdysTmD1    ITSEAextTmD1    ITSEAintTmD1   ITSEAcompTmD1  
##  Min.   :23.00   Min.   :33.00   Min.   :31.00   Min.   :19.00  
##  1st Qu.:34.50   1st Qu.:44.00   1st Qu.:41.50   1st Qu.:43.00  
##  Median :41.00   Median :49.00   Median :49.00   Median :50.00  
##  Mean   :44.15   Mean   :50.59   Mean   :49.27   Mean   :49.28  
##  3rd Qu.:51.00   3rd Qu.:55.00   3rd Qu.:56.50   3rd Qu.:55.00  
##  Max.   :95.00   Max.   :86.00   Max.   :76.00   Max.   :70.00  
##  NA's   :51      NA's   :51      NA's   :51      NA's   :52

The column names are a bit messy, so let’s rename them using dplyr.

library(dplyr)
## 
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
## 
##     filter, lag
## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union
baby <- baby %>%
  rename(DERS = DERStotmA,
         Sensitivity_to_ND = FreePlay_Toy_OverallParentingQuality,
         ITSEA_dysreg = ITSEAdysTmD1,
         ITSEA_exter = ITSEAextTmD1,
         ITSEA_inter = ITSEAintTmD1,
         ITSEA_comp = ITSEAcompTmD1)

Let’s look at the distribution of DERS scores, just for fun.

library(ggplot2)
baby %>%
  ggplot(aes(x = DERS)) +
  geom_histogram(binwidth = 5, fill = "skyblue", color = "black") +
  labs(x = "DERS Scores", y = "Frequency", 
       title = "Distribution of Maternal Prenatal DERS Scores")
## Warning: Removed 2 rows containing non-finite outside the scale range
## (`stat_bin()`).

We can see that this sample is skewed a bit to the right, which is to be expected, given the difficulties surrounding sampling a high DERS population. However, this study made a deliberate effort to sample high DERS participants.

Let’s also look at the distributions of our ITSEA scores in each of the four domains.

library(tidyr)
baby %>%
  pivot_longer(cols = starts_with("ITSEA"),
               names_to = "Domain", values_to = "Score") %>%
  ggplot(aes(x = Score)) +
  geom_histogram(binwidth = 5, fill = "skyblue", color = "black") +
  labs(x = "", y = "Frequency") +
  facet_wrap(~ Domain, scales = "free", ncol = 2) +
  theme_minimal() +
  theme(strip.text.x = element_text(size = 12, face = "bold"))
## Warning: Removed 205 rows containing non-finite outside the scale range
## (`stat_bin()`).

Now, let’s move on to some analyses. The hypotheses that will be used to guide these analyses are as follows:

Let’s begin by looking at Hypothesis 1.

In order to do so, we will test the correlation between maternal prenatal DERS scores and 18-month toddler socioemotional outcomes by looking at ITSEA scores in each of the four domains: Dysregulation, Externalizing, Internalizing, and Competence.

Maternal DERS and Toddler Dysregulation

cor.test(baby$DERS, baby$ITSEA_dysreg)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$DERS and baby$ITSEA_dysreg
## t = 2.3001, df = 109, p-value = 0.02335
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.02995503 0.38605796
## sample estimates:
##       cor 
## 0.2151467
plot(baby$DERS, baby$ITSEA_dysreg, 
     xlab = "Prenatal Maternal DERS", 
     ylab = "18-month Toddler ITSEA Dysregulation", 
     main = "Maternal DERS and Child Dysregulation")
  fit <- lm(ITSEA_dysreg ~ DERS, data = baby)
  abline(fit, col = "red")

Given that p = 0.0234, this correlation is SIGNIFICANT, and we can reject the null hypothesis. With cor = 0.215, we can say that there is weak positive correlation between prenatal maternal DERS score and 18-month toddler ITSEA Dysregulation score.

Maternal DERS and Toddler Externalizing

cor.test(baby$DERS, baby$ITSEA_exter)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$DERS and baby$ITSEA_exter
## t = 3.4049, df = 109, p-value = 0.0009272
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.1312547 0.4693306
## sample estimates:
##       cor 
## 0.3100619
plot(baby$DERS, baby$ITSEA_exter,
     xlab = "Prenatal Maternal DERS", 
     ylab = "18-month Toddler ITSEA Externalizing", 
     main = "Maternal DERS and Child Externalizing")
  fit <- lm(ITSEA_exter ~ DERS, data = baby)
  abline(fit, col = "red")

Given that p = 0.001, this correlation is SIGNIFICANT, and we can reject the null hypothesis. With cor = 0.310, we can say that there is weak positive correlation between prenatal maternal DERS score and 18-month toddler ITSEA Dysregulation score.

Maternal DERS and Toddler Internalizing

cor.test(baby$DERS, baby$ITSEA_inter)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$DERS and baby$ITSEA_inter
## t = 2.1855, df = 109, p-value = 0.03099
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.01923035 0.37688845
## sample estimates:
##       cor 
## 0.2048888
plot(baby$DERS, baby$ITSEA_inter,
     xlab = "Prenatal Maternal DERS", 
     ylab = "18-month Toddler ITSEA Internalizing", 
     main = "Maternal DERS and Child Internalizing")
  fit <- lm(ITSEA_inter ~ DERS, data = baby)
  abline(fit, col = "red")

Given that p = 0.031, this correlation is SIGNIFICANT, and we can reject the null hypothesis. With cor = 0.205, we can say that there is weak positive correlation between prenatal maternal DERS score and 18-month toddler ITSEA Internalizing score.

Maternal DERS and Toddler Competence

cor.test(baby$DERS, baby$ITSEA_comp)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$DERS and baby$ITSEA_comp
## t = 0.88946, df = 108, p-value = 0.3757
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.1036198  0.2682347
## sample estimates:
##        cor 
## 0.08527618
plot(baby$DERS, baby$ITSEA_comp,
     xlab = "Prenatal Maternal DERS", 
     ylab = "18-month Toddler ITSEA Competence", 
     main = "Maternal DERS and Child Competence")
  fit <- lm(ITSEA_comp ~ DERS, data = baby)
  abline(fit, col = "red")

Since p = 0.376, this correlation is NOT SIGNIFICANT, and we fail to reject the null hypothesis. The correlation between prenatal maternal DERS score and 18-month toddler ITSEA Competence score is also close to zero.

Next, let’s look at Hypothesis 2a.

In order to do so, we will test the correlation between maternal prenatal DERS scores and maternal sensitivity to 7-month infant non-distress.

cor.test(baby$DERS, baby$Sensitivity_to_ND)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$DERS and baby$Sensitivity_to_ND
## t = -2.0094, df = 98, p-value = 0.04724
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.380479312 -0.002611586
## sample estimates:
##        cor 
## -0.1989277
plot(baby$DERS, baby$Sensitivity_to_ND,
     xlab = "Prenatal Maternal DERS", 
     ylab = "Maternal Sensitivity to Infant Non-Distress at 7-months",
     main = "Maternal DERS and Sensitivity to Infant Non-Distress")
fit <- lm(Sensitivity_to_ND ~ DERS, data = baby)
abline(fit, col = "red")

Given that p = 0.047, these results are barely significant, and we can reject the null hypothesis. With cor = -0.199, we can say that there is a weak negative correlation between maternal DERS score and 7-month maternal sensitivity to infant non-distress.

Finally, let’s look at Hypothesis 2b. We will test the correlation between maternal sensitivity to 7-month infant non-distress and 18-month toddler socioemotional outcomes by looking at ITSEA scores in each of the four domains: Dysregulation, Externalizing, Internalizing, and Competence.

Maternal Sensitivity and Toddler Dysregulation

cor.test(baby$Sensitivity_to_ND, baby$ITSEA_dysreg)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$Sensitivity_to_ND and baby$ITSEA_dysreg
## t = -1.3673, df = 84, p-value = 0.1752
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.34853439  0.06639565
## sample estimates:
##        cor 
## -0.1475554
# p = 0.1752, cor = -0.1475554

plot(baby$Sensitivity_to_ND, baby$ITSEA_dysreg,
     xlab = "Maternal Sensitivity to Infant Non-Distress",
     ylab = "18-month Toddler ITSEA Dysregulation",
     main = "Maternal Sensitivity and Toddler Dysregulation")
fit <- lm(baby$ITSEA_dysreg ~ Sensitivity_to_ND, data = baby)
abline(fit, col = "red")

Given that p = 0.175, this correlation is not significant, and we cannot reject the null hypothesis.

Maternal Sensitivity and Toddler Externalizing

cor.test(baby$Sensitivity_to_ND, baby$ITSEA_exter)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$Sensitivity_to_ND and baby$ITSEA_exter
## t = -2.017, df = 84, p-value = 0.04689
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.40821863 -0.00320377
## sample estimates:
##        cor 
## -0.2149332
# p = 0.04689, cor = -0.2149332
    # BARELY SIGNIFICANT

plot(baby$Sensitivity_to_ND, baby$ITSEA_exter,
     xlab = "Maternal Sensitivity to Infant Non-Distress",
     ylab = "18-month Toddler ITSEA Externalizing",
     main = "Maternal Sensitivity and Toddler Externalizing")
fit <- lm(baby$ITSEA_exter ~ Sensitivity_to_ND, data = baby)
abline(fit, col = "red")

Given that p = 0.0.0469, this correlation is barely significant, and we can reject the null hypothesis. With cor = -0.215, we can say that there is a weak negative correlation between maternal sensitivity to infant non-distress and 18-month toddler externalizing.

Maternal Sensitivity and Toddler Internalizing

cor.test(baby$Sensitivity_to_ND, baby$ITSEA_inter)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$Sensitivity_to_ND and baby$ITSEA_inter
## t = -0.060575, df = 84, p-value = 0.9518
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.2181789  0.2055540
## sample estimates:
##          cor 
## -0.006609142
# p = 0.9518, cor = -0.006609142

plot(baby$Sensitivity_to_ND, baby$ITSEA_inter,
     xlab = "Maternal Sensitivity to Infant Non-Distress",
     ylab = "18-month Toddler ITSEA Internalizing",
     main = "Maternal Sensitivity and Toddler Internalizing")
fit <- lm(baby$ITSEA_inter ~ Sensitivity_to_ND, data = baby)
abline(fit, col = "red")

Given that p = 0.952, this correlation is not significant, and we cannot reject the null hypothesis.

Maternal Sensitivity and Toddler Competence

cor.test(baby$Sensitivity_to_ND, baby$ITSEA_comp)
## 
##  Pearson's product-moment correlation
## 
## data:  baby$Sensitivity_to_ND and baby$ITSEA_comp
## t = 1.0536, df = 83, p-value = 0.2951
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.1007127  0.3201627
## sample estimates:
##       cor 
## 0.1148771
# p = 0.2951, cor = 0.1148771

plot(baby$Sensitivity_to_ND, baby$ITSEA_comp,
     xlab = "Maternal Sensitivity to Infant Non-Distress",
     ylab = "18-month Toddler ITSEA Competence",
     main = "Maternal Sensitivity and Toddler Competence")
fit <- lm(baby$ITSEA_comp ~ Sensitivity_to_ND, data = baby)
abline(fit, col = "red")

Given that p = 0.295, this correlation is not significant, and we cannot reject the null hypothesis.

The Findings (In Short)

Hypothesis 1: We hypothesize that women who report higher prenatal emotion dysregulation will have toddlers with greater socioemotional difficulties.

Analyses suggest that women who reported higher prenatal emotion dysregulation had toddlers with greater socioemotional difficulties in three of the four ITSEA domains: Dysregulation, Externalizing, and Internalizing. However, there was no significant correlation found prenatal maternal DERS score and 18-month toddler ITSEA Competence.

Hypothesis 2a: We hypothesize that mothers with higher prenatal emotion dysregulation will be less sensitive during non-distress mother-infant interactions.

Analyses suggest that women with higher prenatal emotion dysregulation were less sensitive during non-distress mother-infant interactions at 7-months. The negative correlation shows that mothers with lower DERS scores were generally more sensitive to their infant’s needs, while mothers with higher DERS scores were generally less sensitive, as hypothesized.

Hypothesis 2b: We also hypothesize that mothers who are less sensitive to infants will have toddlers with greater socioemotional difficulties.

Analyses suggest that there is not a significant correlation between maternal sensitivity during non-distress at 7-months and 18-month socioemotional outcomes in three of the four ITSEA domains: Dysregulation, Internalizing, and Competence. However, there is a significant correlation between maternal sensitivity during non-distress at 7-months and 18-month socioemotional outcomes in the Externalizing domain.

Thea Wulff

Simon Brewer

GEOG 5680 - Intro to R

University of Utah

June 21, 2024