Infant Behavior and Development 48 (2017) 114–123

Contents lists available at ScienceDirect

Infant Behavior and Development

journal homepage: www.elsevier.com/locate/inbede

Full length article

Attention focus and self-touch in toddlers: The moderating effect of MARK

attachment security
⁎
Sachiyo Ito-Jägera, Amanda R. Howardb, , Karyn B. Purvisa, David R. Crossa
a
TCU Institute of Child Development, TCU Box 298920, Fort Worth, TX 76129, USA
b
Samford University, Department of Psychology, 800 Lakeshore Dr., Birmingham, AL, 35229, USA

AR TI CLE I NF O AB S T R A CT

Keywords: The superior self-regulation and attention-regulation abilities of securely attached children have
Attachment been repeatedly demonstrated. However, the mechanisms that allow securely attached children
Attention focus to exhibit higher levels of attention focus than insecurely attached (anxious-ambivalent and
Self-touch anxious-avoidant) children need to be explored. One possible mechanism that has been
Attention-regulation
hypothesized to play a role in focusing attention is self-touch. Previous research has shown that
Lateral self-touch
10-year-old children exhibit more bilateral self-touch (i.e., both hands are simultaneously
moving onto each other or on the body, and both hands are in contact with each other or with
the body), but not lateral self-touch (i.e., one hand is moving on the other hand or on the body,
and the hand is in contact with the other hand or with the body), when they focus attention on a
task. Because bilateral coordination is still developing during childhood, we expected that lateral
self-touch, instead of bilateral self-touch, may be associated with attention focus for toddlers. The
objectives of the present study were to examine whether securely attached toddlers exhibit more
self-touch, particularly lateral self-touch, while they focus on a task than while they do not focus
on a task. We expected to find that the association between lateral self-touch and attention focus
is not as strong for insecurely attached toddlers. Data from forty-nine mother-child dyads were
employed for analyses. The attachment classification of the children was determined using the
Strange Situation. The duration of attention focus and self-touch behavior during a reading task
were coded. An association between lateral self-touch and attention focus was found for children
of all attachment classifications. This association was particularly strong for securely attached
children. We discuss the possibility that securely attached toddlers may use lateral self-touch to
regulate attention.

1. Introduction

Attachment refers to “a bond, tie, or enduring relationship between a young child and his mother” (Ainsworth, Blehar,
Waters, & Wall, 1978, p.17; Bowlby, 1969/1982; Bowlby, 1969). Numerous studies that have investigated the relationship between
attachment security and attention-regulation have found that children who were classified as secure in their attachment exhibited
higher levels of attention focus than children who were classified as insecure (anxious-ambivalent and anxious-avoidant) in their
attachment (Ainsworth et al., 1978; Cassidy, 1994; Sroufe, 1995). However, the mechanisms that allow securely attached children to
focus attention better compared to insecurely attached children still need to be examined. Previous research has found that self-touch,
particularly bilateral self-touch (i.e., both hands are simultaneously moving onto each other or on the body, and both hands are in

⁎
Corresponding author at: Samford University, Department of Psychology, 800 Lakeshore Dr Birmingham, AL, 35229, USA.
E-mail addresses: sachiyoito@gmail.com (S. Ito-Jäger), ahoward6@samford.edu (A.R. Howard), k.s.purvis@tcu.edu (K.B. Purvis), d.cross@tcu.edu (D.R. Cross).

http://dx.doi.org/10.1016/j.infbeh.2017.05.001
Received 2 February 2016; Received in revised form 2 May 2017; Accepted 4 May 2017
Available online 01 June 2017
0163-6383/ © 2017 Elsevier Inc. All rights reserved.
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

contact with each other or with the body), is strongly related to attention focus for 10-year-old children (Barroso, Freedman, & Grand,
1980; Barroso, Freedman, Grand, & van Meel, 1978). However, because bilateral coordination is still developing during childhood
(Magalhaes, Koomar, & Cermak, 1989), we expected that lateral self-touch (i.e., one hand is moving on the other hand or on the body,
and the hand is in contact with the other hand or with the body), rather than bilateral self-touch, may be associated with attention
focus for toddlers. In sum, we hypothesized that securely attached toddlers exhibit more self-touch, particularly lateral self-touch,
while they are focusing on a task than while they are not focusing on a task. We expected that this association between lateral self-
touch and attention focus is not as strong for insecurely attached toddlers.

1.1. Attachment security and attention-regulation

The superior self-regulation and attention-regulation abilities of securely attached children have been repeatedly demonstrated
(Cassidy, 1994; Sroufe, 1995). Securely attached children whose caregivers have been consistently responsive and caring learn that
their caregivers would be available at a time of distress (Cassidy, 1994). This history of consistent care during infancy enables
securely attached children to use their caregiver as a secure base, which leads them to explore environments, pursue autonomous
activities, and achieve self-regulation and attention-regulation skills, including attention focus, during toddlerhood (Sroufe, 1995).
Anxious-ambivalent children whose caregivers have been inconsistently available, at times being responsive and at other times
neglectful, during infancy develop to show extreme dependence on their caregiver and seek attention from their caregiver (Ainsworth
et al., 1978; Cassidy, 1994). This intense dependency on their caregiver interferes with the development of autonomy, self-regulation,
and attention-regulation, including attention focus, of these children (Cassidy, 1994).
Children who have a history of consistent rejection from their caregiver in times of need minimize the importance of their
caregiver (Cassidy, 1994). By the end of the first year, children avoid contact with their caregiver when under stress. These anxious-
avoidant children seem to be independent on a behavior level. However, the distress level, measured with cortisol, of anxious-
avoidant children increases after participating in the Strange Situation, whereas it slightly decreases for securely attached children,
suggesting an inappropriate coping strategy of anxious-avoidant children (Spangler & Grossman, 1993).
In fact, both anxious-avoidant children and anxious-ambivalent children are found to have poor self-regulation and attention-
regulation skills, including attention focus. Matas, Arend, and Sroufe (1978) found that, during a problem solving task, 24-month-old
toddlers with insecure attachment styles spend less time focusing on the task, show less positive affect and more negative affect, and
exhibit more frustration behaviors than toddlers with secure attachment style. Main (1983) demonstrated that 21-month-old securely
attached toddlers play with toys longer than insecurely attached toddlers when caregivers are prevented from initiating play.
Securely attached toddlers were also observed to focus their attention on the toys more intensely (e.g., “the face showed intense
monitoring of play”) and fully (e.g., “visual regard was almost constantly directed to the object(s) being manipulated”) than insecure
children (Main, 1983)).
This association between attachment security and attention-regulation has also been observed at other ages. In primary school,
securely attached children show higher levels of attention control (i.e., focusing and shifting attention) than insecurely attached
children (Muris & Dietvorst, 2006). In addition, adults with higher attachment anxiety tend to report having difficulty focusing and
shifting attention than those with lower attachment anxiety (Skowron & Dendy, 2004).
Thus, for securely attached children, their caregiver functions as a secure base, which creates a context in which they can pursue
autonomous activities, engage in challenging tasks, and focus attention (Bowlby, 2005). Lack of this secure base, on the other hand,
may cause more difficulty in attempting to focus attention for anxious-ambivalent and anxious-avoidant attached children. However,
the mechanisms that allow securely attached children to focus attention better compared to insecurely attached children still need to
be explored. One potential mechanism that has been hypothesized to play a role in focusing attention is self-touch. In the next section
(1.2.), we review the importance of touch in attachment theory.

1.2. Attachment and touch

Touch is considered to be one of the core components of attachment theory. By being touched by their caregiver, infants learn that
the caregiver is present and they can be protected in times of threat (Main, 1983). Touch is the largest sensory system of a body and
the first sense to develop (Montague, 1971). Touch has a variety of significant functions, such as enhancing positive emotion and
attention orientation (Clements & Tracy, 1977; Feldman et al., 2003). For instance, mother’s touch elevates the infant’s smiling and
eye contact behaviors (Feldman et al., 2003). Also, adults’ touch has been found to promote attention focus and performance in a
problem-solving task of children (Clements & Tracy, 1977).
Although touch is an important aspect of attachment theory, most studies have focused on its extrinsic forms (i.e., being touched
by a caregiver, and touching a caregiver) and have neglected its intrinsic form (i.e., self-touch). To the best of our knowledge, self-
touch has not been investigated in relation to attachment in toddlers, although this relation would be important because self-touch
has been shown to relate to attention focus.

1.3. Attention focus and two forms of self-touch behavior

Previous studies have shown that the duration of self-touch increases when people focus attention on a task. Barroso et al. (1978)
studied two forms of self-touch: bilateral self-touch and lateral self-touch. Bilateral self-touch refers to types of self-touch, in which
both hands are simultaneously moving onto each other or on the body, and both hands are in contact with each other or with the

115
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

body. Lateral self-touch refers to types of self-touch, in which one hand is moving on the other hand or on the body, and the hand is in
contact with the other hand or with the body. Barroso and his colleagues found that 10-year-old children exhibit more self-touch,
particularly bilateral self-touch, during a task requiring more attention focus (i.e., Stroop Color-Word test) than during a task
requiring less attention focus (i.e., Stroop Color-Naming Test). The authors argued that self-touch behavior could be considered a
manifestation of the extent to which one’s attention was focused. As an alternative conclusion, the authors speculated that self-touch
might function to help regulate attention focus.
In a subsequent study, Barroso et al. (1980) found that longer periods of bilateral self-touch are related to better performance
during three different cognitive tasks: the Stroop test, a memory task, and a reasoning task. Specifically, 10-year-old children who
show larger increases in bilateral self-touch from the Stroop Color-Naming test (i.e., requiring less attention focus) to the Stroop
Color-Word test (i.e., requiring more attention focus) have fewer insertions (i.e., repeating a correct response, correcting of errors in
naming, or repeating previous responses). Longer periods of bilateral self-touch are associated with more numbers of items
remembered during the memory task and fewer requests for information being repeated during the reasoning task. Based on the
assumption that performance in a task reflects attention processes, Barroso and his colleagues argued that self-touch, particularly
bilateral self-touch, is related to improvements in attention focus.

1.4. Lateral and bilateral self-touch: developmental perspectives

Although bilateral self-touch was associated with attention focus for older children (Barroso et al., 1978, 1980), lateral self-touch
may be related to attention focus for younger children. The first movements that infants exhibit are lateral (Cobb,
Goodwin, & Saelens, 1966). Though infants and toddlers begin showing some bilateral movements, such as raising their arms and
clasping their hands at the body midline (White, Castle, & Held, 1964), the development of bilateral motor coordination continues
throughout childhood. For instance, bilateral motor coordination tends to improve with age for children aged between 5 and 9 years
(Magalhaes et al., 1989). These past studies suggest the developmental transition in motor coordination during childhood, changing
from a predominance of lateral movement in toddlerhood to bilateral movement in older children. Therefore, lateral self-touch,
instead of bilateral self-touch, may be related to attention focus for toddlers because bilateral coordination is still developing.

1.5. Present study

Previous studies have investigated the relationship between attachment, attention focus, and (self-)touch. Securely attached
children have superior attention focus abilities compared to insecurely attached children (e.g., Matas et al., 1978). Touch is one of the
core components in attachment theory (e.g., Main, 1983) although the focus has been on its extrinsic forms and has neglected its
intrinsic form. Self-touch, particularly bilateral self-touch, is associated with attention focus for 10-year-old children (Barroso et al.,
1978, 1980). The first movements that infants exhibit, however, are lateral, whereas the development of bilateral motor coordination
continues throughout childhood (Magalhaes et al., 1989).
Based on these previous findings, we examined whether the association between self-touch (combined, lateral, bilateral) and
attention focus is moderated by the attachment classifications. Forty-nine mother-child dyads participated in the Strange Situation to
examine attachment styles of children and in a reading task to assess self-touch and attention focus. We hypothesized that securely
attached toddlers exhibit more self-touch, particularly lateral self-touch, while they focus on a task than while they do not focus on a
task. We expected to find that the association between lateral self-touch and attention focus is not as strong for insecurely attached
toddlers.

2. Method

2.1. Participants

Data of forty-nine mother-child dyads from middle-class families were used for the current study. These families were part of a
longitudinal study conducted in a large metropolitan area in the southwestern United States. All families were recruited from the
Bureau of Vital Statistics birth log including all births from ten county hospitals. Families with a child turning 19 months of age at the
time of the first assessment were recruited by a letter. Criteria for participation included families that were English-speaking, and
having a first-born child with no pre-, peri-, or post-natal complications. Additional criteria included parents having no auditory,
visual, or motor problems. Criteria were assessed by a questionnaire. Of those recruited, 100 families met the criteria. For the current
study, data of forty-nine families who participated in experiments both at 19 months and at 26 months of child’s age and whose
videotapes had sufficient quality for coders to score were used.

2.2. Procedure

The child-mother dyad participated in the Strange Situation to assess attachment style of the child (Ainsworth et al., 1978) at 19
months of age and in a reading task to test self-touch and attention focus behavior of the child at 26 months of age. Before
participation, participants were informed about the course of the study and signed an informed consent form.

116
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

2.2.1. Attachment classifications

The child-mother dyad participated in the Strange Situation paradigm at 19 months of age (Ainsworth et al., 1978). Strange
Situation is most commonly used to assess attachment styles of children between 9 and 18 months. However, it has also been used to
study 19 months old children in previous research (e.g., Hertsgaard, Gunnar, Erickson, & Nachmias, 1995). In the Strange Situation, a
child was placed in an increasingly stressful situation. The Strange Situation was videotaped, and four interactive behaviors of the
child at the reunions with the mother were observed: proximity- and contact-seeking behavior, contact maintaining behavior,
resistant behavior, and avoidant behavior. Based on these behaviors, children were classified into one of the three attachment styles:
secure, anxious-avoidant, and anxious-ambivalent.
Children who actively seek proximity and contact with their mothers, maintain contact with them, and show little or no tendency
to resist or avoid interaction with them are classified as securely attached to their mother. Secure children are thought to exhibit
those behaviors because the history of consistent care during infancy teaches them that their caregivers provide a secure base
(Ainsworth et al., 1978).
Children classified as anxious-ambivalent attachment style also show strong proximity- and contact-seeking behavior and strong
contact maintaining behavior. Anxious-ambivalent children, however, tend to show ambivalent behaviors, such as the mixture of
seeking and resisting contact with the mother (Ainsworth et al., 1978). These behaviors are likely to develop because anxious-
ambivalent children’s caregivers have been inconsistently available during infancy. In consequence, children seek attention from
their caregiver but, at the same time, are anxious that they may lose attention from them any time.
Children with anxious-avoidant attachment style, on the other hand, tend to avoid their mother upon reunion (Ainsworth et al.,
1978). These children tend not to seek proximity or contact with their mother. Children tend to develop these behaviors because a
history of consistent rejection from their caregiver in times of need make them minimize the importance of their caregiver.
Two independent raters, who had been trained to score the Strange Situation based on the Ainsworth coding system (Ainsworth
et al., 1978), scored a sub-sample (20%) of the Strange Situation videotapes. Cohen’s (1960) kappa coefficients were 0.90. A third
rater scored any discrepancies to determine the attachment classification. Out of 49 children, 26 (53%) children were classified with a
secure classification, 14 (29%) children as an anxious-avoidant classification, and 9 children (18%) with an anxious-ambivalent
classification.

2.2.2. Self-touch and attention focus coding

The child-mother interaction during a reading task was videotaped at 26 months of age. Studying attention focus of 26 months old
was appropriate because the higher form of attention system, which enables children to focus attention, gradually develops during
toddlerhood (Ruff & Rothbart, 1996). In a laboratory of the university, child and mother sat next to each other at a table. A camera
had been installed to record the child-mother interaction from the front view, allowing the experimenters to observe and later code
the behaviors. The mother was instructed to “read this book with your child as you would if you were at home.” The experimenter left
the laboratory after he or she explained the task, leaving the child and the mother alone. The videotapes were timed for the duration
of the child’s self-touch behavior and off-task-behavior to the nearest second. Coders used the zoom function and the slow motion
function when necessary in order to make the scoring as accurate as possible. Children were considered to be “in the task/room” from
the time the book was open until it was closed. In order to control for the possibility that children might exhibit self-touch less
frequently while walking, the seconds that children were walking in the room were not used for analyses. The average time (seconds)
that children were in the reading task was 244.49 (SD = 97.63).
Two pairs of coders (four coders in total) independently scored bilateral self-touch, lateral self-touch, and on-task behavior on a
sub-sample (20%) of videotapes, blind to attachment styles of each child. Cohen’s (1960) kappa coefficients were 0.80 for bilateral
self-touch, 0.91 for lateral self-touch, and 0.92 for on-task behavior. Each pair then independently scored behaviors on the rest of the
tapes.

2.2.2.1. Self-touch. Coders scored the duration (units of one second) of a child exhibiting self-touch in the task. Self-touch included
behaviors such as rubbing, squeezing, or touching on any part of the body. Each episode of self-touch was classified as one of two
types based on a coding system adapted from Barroso et al. (1978) and Barroso et al. (1980): bilateral self-touch and lateral self-
touch. Bilateral self-touch refers to types of self-touch, in which both hands are simultaneously moving onto each other or on the
body, and both hands are in contact with each other or with the body. Lateral self-touch refers to types of self-touch, in which one
hand is clearly moving on the other hand or on the body of the child, and the hand is in contact with the other hand or with the body.
First, the two forms of self-touch were combined and analyzed together (i.e., combined self-touch). Then, each form of self-touch was
analyzed separately (i.e., bilateral self-touch, lateral self-touch). Each episode of self-touch was scored whether it was incidental or
non-incidental. Self-touch was considered incidental if the touch was a by-product of another behavior (e.g., a child touches himself
or herself while trying to pick up a book). Only non-incidental touch was included for analysis.

2.2.2.2. Attention focus. Coders scored the duration (units of one second) of a child not being fully engaged with the task of storybook
reading based on the coding system adapted from Ruff and Lawson (1990). Off-task (i.e., not focusing on a task) included episodes of
the child discussing irrelevant topics, such as what they ate for lunch or what time they were leaving. The episodes where children
were in the task/room but not coded as off-task were considered on-task (i.e., focusing on a task).

117
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

Fig. 1. The odds of combined self-touch during on-task and off-task for secure, anxious-avoidant, and anxious-ambivalent children.

3. Results

The R statistical software was used for the analyses (R Development Core Team, 2009). We investigated whether the association
between being on-task (on-task, off-task) and combined self-touch (combined touch, no combined touch), bilateral self-touch
(bilateral touch, no bilateral touch), or lateral self-touch (lateral touch, no lateral touch) was moderated by attachment style (secure,
anxious-avoidant, anxious-ambivalent) using log-linear analyses. For multiple testing, we corrected p-values by applying the
Benjamini-Hochberg procedure to control for false discovery rates (FDR; Benjamini & Hochberg, 1995).

3.1. Combined self-touch and attention focus: the moderating effect of attachment security

Log-linear analyses showed good fit for the 2-way association model, G2(2) = 0.08, p = 0.961, and the comparison between the
independence model and the 2-way association model was statistically significant, G2(5) = 439.50, p < 0.001, suggesting that the
2-way association model is the adequate model. Because the 3-way association model is not the adequate model, it suggests that the
relationship between being on-task and combined self-touch is not moderated by attachment style (see Fig. 1 & Table 1).

3.2. Bilateral self-touch and attention focus: the moderating effect of attachment security

Log-linear analyses showed poor fit for both the independence model, G2(7) = 171.37, p < 0.001, and the 2-way association
model, G2(2) = 6.53, p = 0.004, indicating that the 3-way association model is the unique adequate model. Thus, the relationship
between on-task and bilateral self-touch differs for at least one pair of attachment classifications (see Fig. 2 & Table 2). We therefore

Table 1
Descriptive statistics for combined self-touch when children were on-task versus off-task.

Attachment Combined touch No combined touch Marginal totals Odds Odds ratio
Style

Secure On-task 1451 10,209 11660 0.14 2.03

Off-task 40 570 610 0.07
Column totals 1491 10,779
Odds 36.28 17.91 19.11

Anxious-Avoidant On-task 504 5980 6484 0.08 1.89

Off-task 33 741 774 0.04
Column totals 537 6721
Odds 15.27 8.07 8.38

Anxious-Ambivalent On-task 433 3655 4088 0.12 2

Off-task 27 455 482 0.06
Column totals 460 4110
Odds 16.04 8.03 8.48

Note. Values are frequencies, odds, or odds ratio.

118
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

Fig. 2. The odds of bilateral self-touch during on-task and off-task for secure, anxious-avoidant, and anxious-ambivalent children.

performed log-linear analyses for each pair of attachment classifications to determine which comparisons were statistically
significant.

3.2.1. Secure vs. anxious-avoidant

Secure children are more likely to exhibit bilateral self-touch than anxious-avoidant children both when on-task, OR = 1.25,
G2(1) = 5.93, p = 0.027, and off-task, OR = 3.98, G2(1) = 9.81, p = 0.006. Anxious-avoidant children are likely to exhibit less
bilateral self-touch when off-task than on-task, G2(1) = 14.68, p < 0.001, but there is no difference during on-task and off-task for
secure children, G2(1) = 0.37, p = 0.613. Thus, although significant differences exist both when being on-task and off-task for secure
and anxious-avoidant children, it is the difference in off-task that contributes to the differences in the odds ratio.

3.2.2. Secure vs. anxious-ambivalent

Similar to secure children, anxious-ambivalent children do not differ in the likelihood of exhibiting bilateral self-touch when they
are focusing on a task and when they are not, G2(1) = 2.71, p = 0.150. Also, the likelihood of exhibiting bilateral self-touch does not
differ between secure and anxious-ambivalent children both when they are on-task, OR = 0.86, G2(1) = 2.28, p = 0.168, and off-
task, OR = 1.20, G2(1) = 0.22, p = 0.637.

3.2.3. Anxious-avoidant vs. anxious-ambivalent

Anxious-ambivalent children are more likely to exhibit bilateral self-touch than anxious-avoidant children both when they are on-
task, OR = 1.45, G2(1) = 10.78, p = 0.005, and off-task, OR = 3.33, G2(1) = 6.06, p = 0.027. As indicated above, anxious-avoidant
children, but not anxious-ambivalent children, are more likely to exhibit bilateral self-touch when they are focusing on a task than

Table 2
Descriptive statistics for bilateral self-touch when children were on-task versus off-task.

Attachment Bilateral touch No bilateral touch Marginal totals Odds Odds ratio
Style

Secure On-task 395 5435 5830 0.07 1.16

Off-task 18 287 305 0.06
Column totals 413 5722 0.07
Odds 21.94 18.94 19.11

Anxious-Avoidant On-task 178 3064 3242 0.06 3.69

Off-task 6 381 387 0.02
Column totals 184 3445 0.05
Odds 29.67 8.04 8.38

Anxious-Ambivalent On-task 159 1885 2044 0.08 1.61

Off-task 12 229 241 0.05
Column totals 171 2114 0.08
Odds 13.25 8.23 8.48

Note. Values are frequencies, odds, or odds ratio.

119
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

Fig. 3. The odds of lateral self-touch during on-task and off-task for secure, anxious-avoidant, and anxious-ambivalent children.

when they are not focusing on a task.

3.3. Lateral self-touch and attention focus: the moderating effect of attachment security

Log-linear analyses showed marginally good fit for the 2-way association model, G2(2) = 4.62, p = 0.099, and the model
comparison between the independence model and the 2-way association model was statistically significant, G2(5) = 301.38,
p < 0.001, suggesting that the 2-way association model or 3-way association model is adequate for lateral self-touch. Because the fit
was only marginally good for the 2-way association model, the relationship between lateral self-touch and on-task may be moderated
by attachment style for at least one pair of attachment classifications (see Fig. 3 & Table 3). We therefore performed log-linear
analyses for each pair of attachment classifications to determine which comparisons were statistically significant.

3.3.1. Secure vs. anxious-avoidant

Both secure children, G2(1) = 28.99, p < 0.001, and anxious avoidant children, G2(1) = 4.06, p = 0.066, are more likely to
exhibit lateral self-touch when they are focusing on a task than when they are not focusing on a task though the likelihood was only
marginally significant for anxious-avoidant children. Secure children are more likely to exhibit lateral self-touch than anxious-
avoidant children when they are focused on a task, OR = 1.98, G2(1) = 110.64, p < 0.001. No difference was found between secure
and anxious-avoidant children when they are not focused on a task, OR = 1.04, G2(1) = 0.01, p = 0.904. Thus, locus of moderation
is not in children’s off-task behavior, but in their on-task behavior.

Table 3
Descriptive statistics for lateral self-touch when children were on-task versus off-task.

Attachment Lateral touch No lateral touch Marginal totals Odds Odds ratio
Style

Secure On-task 1056 4774 5830 0.22 2.85

Off-task 22 283 305 0.08
Column totals 1078 5057 0.21
Odds 48.00 16.87 19.12

Anxious-Avoidant On-task 326 2916 3242 0.11 1.49

Off-task 27 360 387 0.08
Column totals 263 3276 0.08
Odds 12.07 8.10 8.38

Anxious-Ambivalent On-task 274 1770 2044 0.16 2.33

Off-task 15 226 241 0.07
Column totals 289 1996 0.15
Odds 18.27 7.83 8.48

Note. Values are frequencies, odds, or odds ratio.

120
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

3.3.2. Secure vs. anxious-ambivalent

Similar to secure and anxious-avoidant children, anxious-ambivalent children are more likely to exhibit lateral self-touch when
they are focusing on a task than when they are not, G2(1) = 11.85, p < 0.001. However, secure children are more likely to exhibit
lateral self-touch than anxious-ambivalent children when they are on task, OR = 1.43, G2(1) = 24.87, p < 0.001. There is no
difference between secure and anxious-ambivalent children when they are not focusing on a task, OR = 1.17, G2(1) = 0.21,
p = 0.799.

3.3.3. Anxious-avoidant vs. anxious-ambivalent

Anxious-ambivalent children are more likely to exhibit lateral self-touch than anxious-avoidant children when they are focusing
on a task, OR = 1.38, G2(1) = 13.75, p < 0.001. In contrast, the odds of lateral self-touch do not differ for anxious-ambivalent and
anxious-avoidant children when they are not focusing on a task, OR = 1.13, G2(1) = 0.14, p = 0.799.

4. Discussion

The hypothesis of the present study was that securely attached toddlers exhibit more self-touch, particularly lateral self-touch,
while they focus on a task than while they do not focus on a task. We expected to find that the association between lateral self-touch
and attention focus is not as strong for insecurely attached toddlers.
Attachment classifications did not contribute to variations in the relationship between combined self-touch and attention focus.
On the other hand, attachment classifications contributed to variations in both the association between bilateral self-touch and
attention focus, as well as the association between lateral self-touch and attention focus. However, the nature of the effect of
attachment classifications was different for the two forms of self-touch.
Regarding bilateral self-touch, anxious-avoidant children are less likely to exhibit it when they are not focusing on a task than
when they are focusing on a task. On the other hand, there is no difference in the likelihood of bilateral self-touch when focusing on a
task and when not focusing on a task for securely attached children and anxious-ambivalent children. The pattern of the anxious-
avoidant children’s bilateral self-touch behavior is the same as that of older children in previous research (Barroso et al., 1978;
Barroso et al., 1980). The data suggest that, only for anxious-avoidant children, bilateral self-touch may function either as a
manifestation of the extent to which attention is focused or to regulate attention focus, as argued by Barroso et al. (1978).
However, closer observations of the children’s behavior revealed that anxious-avoidant children are less likely to exhibit bilateral
self-touch when they are not focusing on a task because they engage in behaviors that prevent them from exhibiting self-touch. More
specifically, when they are not focusing on a task, anxious-avoidant children exhibit behaviors such as throwing away a book and
trying to take a book away from their mother (Bus, Belsky, van IJzendoorn, & Crnic, 1997). In contrast, securely attached children
and anxious-ambivalent children exhibit different types of off-task behaviors, such as talking about another task (e.g., crayon) or
another event (e.g., visiting their father), which allows them to exhibit self-touch. Thus, although anxious-avoidant children are less
likely to exhibit bilateral self-touch when they are not focused, the pattern of this behavior does not necessarily indicate that bilateral
self-touch functions as a manifestation of the extent to which attention is focused or to regulate attention focus in toddlers. As
discussed from the developmental perspectives of lateral and bilateral motor coordination, lateral rather than bilateral self-touch may
be related to attention focus for younger children.
With regard to lateral self-touch, children of all attachment classifications are more likely to exhibit it when they are focusing on a
task than when they are not focusing on a task. The association between lateral self-touch and attention focus is, however, stronger
for securely attached children than for children with either insecure attachment style.
There are two possible reasons why securely attached children are more likely to engage in lateral self-touch when they are
focusing on a task than children with insecure attachment styles. First, as Barroso et al. (1978) discussed in terms of the relationship
between self-touch in general and attention focus, lateral self-touch may be a manifestation of the extent to which children’s attention
is focused. According to this hypothesis, securely attached children are more likely to exhibit lateral self-touch than anxious-
ambivalent or anxious-avoidant children because they are focusing on a task more than the other two groups. Second, as Barroso et al.
argued as an alternative explanation, lateral self-touch may function to self-regulate attention. Thus, one reason why securely attached
children are more likely to focus attention may be because they use lateral self-touch to regulate attention.
Because of the observational nature of the current study, we cannot conclude whether lateral self-touch functions merely as a
manifestation of the extent to which attention is focused, or if, in fact, it functions as a self-regulation tool to focus attention.
However, the strong association between lateral self-touch and attention focus for securely attached children suggests that, at a
minimum, children’s self-touch is an indicator of secure attachment as well as a higher capacity of attention focus in certain
environments; environments such as learning contexts, where focused attention is required. This finding is particularly important
because self-touch is generally perceived negatively because of its association to emotional distress (e.g., Remland, 2008). For
example, infants exhibit more self-touch when they encounter a stranger and when they are left alone in a room compared to being
with or around familiar people (Trevarthen, 1977, 1979). Furthermore, 3-year-olds exhibit more self-touch during storytelling with
an “emotionally evocative picture book” than during storytelling with non-emotionally evocative toys (Landau, Shusel, Eshel, & Ben-
Aaron, 2003), suggesting that self-touch may have the function of soothing.
If self-touch functions as a self-regulation tool, an important question for future research is how securely attached children, in
contrast to insecurely attached children, develop the ability to engage in self-touch in order to focus attention. A secure relationship
with a caregiver may create the context for development of self-touch in relation to attention focus by multiple mechanisms. First,
securely attached children may experience the effect of touch on attention focus through their caregiver’s touch. Adults’ touch has

121
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

been found to promote attention focus and performance in a problem-solving task in children (Clements & Tracy, 1977). Because
mothers of securely attached children tend to be more sensitive with their children than mothers of insecurely attached children
during infancy (De Wolff & van IJzendoorn, 1997), mothers of securely attached children may touch their child more when their
children have difficulty focusing on a task. Additionally, mothers’ unaffectionate touch and aversion to touch with their child
predicted child’s attachment insecurity (Ainsworth et al., 1978), suggesting that securely attached children are, in general, more
experienced than insecurely attached children to receive touch from their caregivers. Therefore, it is possible that, through touch by
their caregivers, securely attached children learn to engage in self-touch when attention focus is required.
Second, securely attached children may learn the relation between self-touch and attention focus by observing their caregivers.
Children can acquire behavior through the observing of and modeling of other individuals (Schunk, 2001), and adults are also more
likely to engage in self-touch when they focus attention than when they do not (Barroso & Feld, 1986). Although no published studies
have examined whether caregivers of securely attached children tend to exhibit more self-touch than those of insecurely attached
children during attention focus, other studies suggest that this may be the case. First, there is a strong relationship between child
attachment style and their parent attachment style (van IJzendoorn, 1995), indicating that parents of securely attached children tend
to be securely attached, whereas parents of insecurely attached children tend to be insecurely attached. Second, securely attached
adults display a greater ability to self-regulate than insecurely attached adults (Bouthillier, Julien, Dubé, Bélanger, & Hamelin, 2002).
These findings suggest that caregivers of securely attached children may exhibit more self-touch than those of insecurely attached
children during attention focus. Therefore, securely attached children may learn the relation between self-touch and attention focus
by observing their caregivers.
Third, securely attached children may learn self-touch as related to attention focus by experiencing themselves that they are better
able to focus attention when they engage in self-touch. In other words, self-touch may be reinforced by successful attention focus, and
consequently children may exhibit self-touch when attention focus is required. Because securely attached children have caregivers
who are sensitive to their needs, they can use their caregiver as a secure base and are better able to develop autonomy than insecurely
attached children (Cassidy, 1994). Consequently, by toddlerhood, securely attached children may be able to engage in self-touch
behavior, which then leads to their own learning of its connection to attention focus.

4.1. Limitations and future research

One limitation of the present study is that attention focus and self-touch behaviors were observed only during toddlerhood. Future
research should examine different age groups, particularly in order to explore the development of lateral and bilateral self-touch in
relation to attention focus. The present study found that lateral self-touch, but not bilateral self-touch, has a strong relation to
attention focus for toddlers, a finding that is contrary to past studies with older children (Barroso et al., 1978; Barroso et al., 1980). As
discussed above, these differences may be derived from the developmental transition in motor coordination during childhood,
changing from a predominance of lateral movement in toddlerhood to bilateral movement in older children. Future research should,
therefore, examine preschoolers in order to clarify developmental trajectories, in which forms of self-touch change in relation to
attention focus and age.
Another limitation of this study is that the behavioral observations were limited to the children. As discussed in the previous
section, self-touch behavior of not only children but also their mothers, both in infancy and toddlerhood, should be examined.
Development of self-regulation has, for example, been studied in relation to mother-child affect synchrony (e.g., Feldman,
Greenbaum, & Yirmiya, 1999). The close investigation of the mother-child interaction may help us clarify how securely attached
toddlers, more than insecurely attached toddlers, develop to engage in self-touch when attention focus is required.
Future research should also investigate whether self-touch is merely a manifestation of the extent to which attention is focused, or
if self-touch functions in regulating attention. Because of the observational nature of the present study, we cannot conclude which
function self-touch has from the present data. Future research should investigate these mechanisms using controlled experiments. For
instance, future researchers may employ an intervention to train children to engage in self-touch in a learning context.

References

Ainsworth, M., Blehar, M., Waters, E., & Wall, S. (1978). Patterns of attachment. Hillsdale, NJ: Erlbaum.
Barroso, F., & Feld, J. (1986). Self-touching and attentional processes: The role of task difficulty, selection stage, and sex differences. Journal of Nonverbal Behavior, 10,
51–64. http://dx.doi.org/10.1007/BF00987205.
Barroso, F., Freedman, N., Grand, S., & van Meel, J. (1978). Evocation of two types of hand movements in information processing. Journal of Experimental Psychology:
Human Perception and Performance, 4, 321–329. http://dx.doi.org/10.1037/0096-1523.4.2.321.
Barroso, F., Freedman, N., & Grand, S. (1980). Self-touching, performance, and attentional processes. Perceptual and Motor Skills, 50, 1083–1089. http://dx.doi.org/10.
2466/pms.1980.50.3c.1083.
Benjamini, Y., & Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. Journal of the Royal Statistical Society
Series B (Methodological), 289–300. http://dx.doi.org/10.2307/2346101.
Bouthillier, D., Julien, D., Dubé, M., Bélanger, I., & Hamelin, M. (2002). Predictive validity of adult attachment measures in relation to emotion regulation behaviors in
marital interactions. Journal of Adult Development, 9, 291–305. http://dx.doi.org/10.1023/A:1020291011587.
Bowlby, J., (1969/1982). Attachment and loss: Vol. 1: Attachment. New York: Basic Books.
Bowlby, J. (2005). A secure base: Clinical applications of attachment theory. Oxon, U.K: Routeldge.
Bus, A., Belsky, J., van IJzendoorn, M., & Crnic, K. (1997). Attachment and bookreading patterns: A study of mothers, fathers and their toddlers. Early Childhood
Research Quarterly, 12, 81–98. http://dx.doi.org/10.1016/S0885-2006(97)90044-2.
Cassidy, J. (1994). Emotion regulation: Influences of attachment relationships. Monographs of the Society for Research in Child Development, 59, 228–283. http://dx.doi.
org/10.1111/j.1540-5834.1994.tb01287.x.
Clements, J., & Tracy, D. (1977). Effects of touch and verbal reinforcement on the classroom behavior of emotionally disturbed boys. Exceptional Children, 43, 453–454.

122
S. Ito-Jäger et al. Infant Behavior and Development 48 (2017) 114–123

http://dx.doi.org/10.1177/001440297704300707.
Cobb, K., Goodwin, R., & Saelens, E. (1966). Spontaneous hand positions of newborn infants. The Journal of Genetic Psychology: Research and Theory of Human
Development, 108, 225–237. http://dx.doi.org/10.1080/00221325.1966.10532781.
Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37–46. http://dx.doi.org/10.1177/
001316446002000104.
De Wolff, M., & van IJzendoorn, M. (1997). Sensitivity and attachment: A meta-analysis on parental antecedents of infant attachment. Child Development, 68, 571–591.
http://dx.doi.org/10.1111/j.1467-8624.1997.tb04218.x.
Feldman, R., Greenbaum, C., & Yirmiya, N. (1999). Mother-infant affect synchrony as an antecedent of the emergence of self-control. Developmental Psychology, 35,
223–231. http://dx.doi.org/10.1037//0012-1649.35.1.223.
Feldman, R., Weller, A., Sirota, L., & Eidelman, A. I. (2003). Testing a family intervention hypothesis: The contribution of mother-infant skin-to-skin contact (Kangaroo
Care) to family interaction, proximity, and touch. Journal of Family Psychology, 17, 94–107. http://dx.doi.org/10.1037/0893-3200.17.1.94.
Hertsgaard, L., Gunnar, M., Erickson, M. F., & Nachmias, M. (1995). Adrenocortical responses to the Strange Situation in infants with disorganized/disoriented
attachment relationships. Child Development, 66, 1100–1106. http://dx.doi.org/10.1111/j.1467-8624.1995.tb00925.x.
Landau, R., Shusel, R., Eshel, Y., & Ben-Aaron, M. (2003). Mother-child and metapelet-child touch behavior with three-year-old Kibbutz children in two contexts. Infant
Mental Health Journal, 24, 529–546. http://dx.doi.org/10.1002/imhj.10069.
Magalhaes, L. C., Koomar, J. A., & Cermak, S. A. (1989). Bilateral motor coordination n 5- to 9-year-old children: A pilot study. The American Journal of Occupational
Therapy, 43, 437–443. http://dx.doi.org/10.5014/ajot.43.7.437.
Main, M. (1983). Exploration, play, and cognitive functioning related to infant-mother attachment. Infant Behavior & Development, 6, 167–174. http://dx.doi.org/10.
1016/S0163-6383(83)80024-1.
Matas, L., Arend, R., & Sroufe, L. (1978). Continuity of adaptation in the second year: The relationship between quality of attachment and later competence. Child
Development, 49, 547–556. http://dx.doi.org/10.1111/j.1467-8624.1978.tb02354.x.
Montague, A. (1971). Touching: The human significance of the skin. New York: Harper & Row.
Muris, P., & Dietvorst, R. (2006). Underlying personality characteristics of behavioral inhibition in children. Child Psychiatry and Human Development, 36, 437–445.
http://dx.doi.org/10.1007/s10578-006-0014-9.
R Development Core Team (2009). R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing [ISBN 3–900051-07-
0. Retrieved, October 20, 2009, from http://www.R-project.org].
Remland, M. (2008). Nonverbal communication in everyday life. Allyn & Bacon: Boston, MA.
Ruff, H., & Lawson, K. (1990). Development of sustained, focused attention in young children during free play. Developmental Psychology, 26, 85–93. http://dx.doi.org/
10.1037/0012-1649.26.1.85.
Ruff, H., & Rothbart, M. (1996). Attention in early development: themes and variations. New York, NY: Oxford Press.
Schunk, D. (2001). Social cognitive theory and self-regulated learning. In B. Zimmerman, & D. Schunk (Eds.), Self-regulated learning and academic achievement:
theoretical perspectives (pp. 125–151). (2nd ed.). Mahwah, NJ: Erlbaum.
Skowron, E., & Dendy, A. (2004). Differentiation of self and attachment in adulthood: Relational correlations of effortful control. Contemporary Family Therapy, 26,
337–357. http://dx.doi.org/10.1023/B:COFT.0000037919.63750.9d.
Spangler, G., & Grossman, K. (1993). Biobehavioral organization in securely and insecurely attached infants. Child Development, 64, 1439–1450. http://dx.doi.org/10.
1111/j.1467-8624.1993.tb02962.x.
Sroufe, A. (1995). Emotion development: The organization of emotional life in the early years. New York: Cambridge Press.
Trevarthen, C. (1977). Descriptive analyses of infant communicative behavior. In R. Schaffer (Ed.), Studies in mother-infant interaction (pp. 227–270). London, U. K:
Academic Press.
Trevarthen, C. (1979). Communication and cooperation in early infancy: A description of primary intersubjectivity. In M. Bullowa (Ed.), Before speech: The beginnings of
human communication (pp. 321–347). London: Cambridge Press.
van IJzendoorn, M. (1995). Adult attachment representations, parental responsiveness, and infant attachment: A meta-analysis on the predictive validity of the adult
attachment interview. Psychological Bulletin, 117, 387–403. http://dx.doi.org/10.1037/0033-2909.117.3.387.
White, B., Castle, P., & Held, R. (1964). Observations on the development of visually-directed reaching. Child Development, 35, 349–364. http://dx.doi.org/10.2307/
1126701.

123