A

COPARENTING INTERACTIONS OBSERVED BY THE PRENATAL LAUSANNE

TRILOGUE PLAY: AN ITALIAN REPLICATION STUDY
ALESSANDRA SIMONELLI, MARA BIGHIN, FRANCESCA DE PALO

University of Padua, Padua, Italy

The infantparent interaction is the focus of interest on early interactive relational models in the triadic perspective. This approach
considers the motherfatherchild interaction as the matrix of early competences, which start to develop well before delivery. This theoretical approach
has generated an observation paradigm, the Lausanne Trilogue Play (LTP; E. Fivaz-Depeursinge & A. Corboz-Warnery, 1999). The aim of the study was
to investigate (a) the psychometric characteristics of coding of the prenatal LTP procedure and their comparison with the Lausanne validation studies
and (b) the characteristics of triadic interactions during pregnancy. Ninety-eight nonreferred, primiparous families were recruited at childbirth courses.
In addition to validated questionnaires, observational data were collected at the seventh month of pregnancy in the prenatal LTP. The collected data show
good reliability of the LTP coding and a consistent factorial structure in line with the Lausanne validation studies (C. Carneiro, A. Corboz-Warnery, &
E. Fivaz-Depeursinge, 2006; N. Favez et al., 2006). The Structure of the Play and the Intuitive Parenting Behaviors Scales seem the most representative
dimensions in the prenatal period. Coparental abilities during pregnancy represent an interactive matrix for the construction of early family relations
and may be considered as protective factors in the childs development of early triadic interactive abilities.

ABSTRACT:

Abstracts translated in Spanish, French, German, and Japanese can be found on the abstract page of each article on Wiley Online Library at
http://wileyonlinelibrary.com/journal/imhj.

* * *
In the field of early family processes, recent theoretical and
empirical approaches have stressed the distinction between the
co-parental couple and the marital couple (Katz & Gottman,
1996; McHale, 2007; McHale & Cowan, 1996; McHale & FivazDepeursinge, 1999; McHale, Kuersten-Hogan, & Rao, 2004). In
particular, in the domain of the research on transition to parenthood and on parenting in the family life cycle, coparenting is a
key notion that describes the coordination between the adults in
their parental roles (Minuchin, 1974). In other words, according
to Minuchins (1974) approach, the coparenting concept refers to
the degree of supportive cooperation between the adults who bear
responsibility for the socialization process of children belonging
to their family system. This notion identifies the presence of a hierarchy within the family, where leadership and decision making
are taken on by the adult members of the family executive subsystem only, as far as child caregiving is concerned (McHale, 2007).
Therefore, the coparenting notion does not only refer to sharing
work and responsibility in child caregiving; rather, it refers to the
coordination and support between adults who are responsible for
childcare and childrearing (McHale, 2007).

Several studies in this area have generally assessed coparenting in the context of family interactions, asking whether parents
support or undermine each other with respect to the child (Belsky
& Kelly, 1994; McHale & Cowan, 1996). The most recent research
has demonstrated links between the marital, the parental, and the
coparental family subsystems, suggesting that the marital relationship has an impact on the quality of parenting. In general, data
have shown that parents with high marital distress tend to show
less adjusted parenting interventions than do parents with low marital distress (Cowan & Cowan, 1992). This research, however, did
not consider some interesting and fundamental aspects and their
role in the parenting processes: The most important, in the authors
view, concerns the definition and the specification of the constructs
of marital, parental, and coparental functions. In fact, the marital
and the parental subsystems are constructs which are neither completely independent nor wholly overlapping in the same way as are
the parental and coparental ones (Margolin, Gordis, & John, 2001;
McHale & Cowan, 1996). The parental function is an individual,
dyadic competence that concerns the adultinfant interaction and
relationships and where each parent actualizes with the infant, independently from the other parent. There is a large area of research
and many studies on the quality of adultinfant dyadic interactions
and on the role of personal (adult personality, attachment experiences, etc.), relational (marital relationship, etc.), and contextual
(social support, social network, etc.) factors in determining the

Direct correspondence to: Alessandra Simonelli, Department of Developmental and Social Psychology, University of Padua, LIRIPAC, via Belzoni, 80
(35131 Padova), Padua, Italy; e-mail: alessandra.simonelli@unipd.it.
INFANT MENTAL HEALTH JOURNAL

C 2012 Michigan Association for Infant Mental Health
View this article online at wileyonlinelibrary.com.
DOI: 10.1002/imhj.21350

A. Simonelli, M. Bighin, and F. DePalo

parenting and the child outcomes (cf. Belsky, 1984). These theoretical and methodological approaches have resulted in data on
the characteristics of the dyadic level of parenting (adultinfant
interaction), but are limited by the absence of consideration of the
coparental level. Coparenting, in fact, is a cooperative function that
involves the two parents and their coordination in the interactive
and representative levels of manifestation of parenting. Consequently, the coparenting function can be defined as one specific
subsystem of family interactions referring to two adults abilities
to coordinate and interact while performing their parental role toward the child. Therefore, at a methodological level, coparenting
can be observed within motherfatherchild family interactions
as one of the levels which structure the overall functioning of the
system as a whole. According to this viewpoint, the coparenting
subsystem is defined and studied according to Fivaz-Depeursinge
and Corboz-Warnerys (1999) approach, which broadened the focus beyond the coparenting subsystem to the family unit of father,
mother, and infant using an assessment tool specifically developed
to study families during the first year of an infants life. The central element of this approach is the use of a seminaturalistic play
situation, the Lausanne Trilogue Play situation (LTP), which involves the two parents and their baby in a cooperative task. In fact,
the goal of trilogue play is a shared experience of positive affects,
regardless of any transitory moments in negative affective states
(e.g., tiredness, frustration, etc.). The capacity to regulate affects as
a group is one of the foundations of family communication. Note
that at the basis of the procedure is the idea that what you want to
assess is the familys best performance during the interaction task.
With this in mind, it is fundamental to put all family members in the
optimal and most functional situation to experience a satisfactory
interaction.
During the LTP, parents and their babies are brought together to
play in four configurations: motherinfant, fatherinfant, mother
father, and a single three-together configuration where all three
partners are active in play. In the two-plus-one configurations, two
partners are active while the third is present as participant-observer.
The LTP not only allows an assessment of coordination between
the two parents as well as the other two dyadic family subunits
(motherchild and fatherchild) but also of the coordination of the
family as a whole.
Fivaz-Depeursinge and Corboz-Warnery (1999) previously
demonstrated that to attain the goal of trilogue play, the partners
fulfill three embedded functions: participation (i.e., all partners are
included), organization (i.e., partners keep to their respective roles
of active vs. participant-observer), and focalization (i.e., partners
share a joint focus). The degree of coordination that they reach
in fulfilling these functions determines their family alliance.
The more coordinated the interactions are, the more functional
the family alliance, and consequently, the partners regularly experience moments of pleasure together. The less coordinated the
interactions are, the more problematic the family alliance, and the
partners remain locked in negative reciprocity loops.
Results of studies using LTP have revealed that the family
alliance is fairly stable over the first year of the childs life (Favez

& Frascarolo, 2002; Fivaz-Depeursinge & Corboz-Warnery, 1999)

and up to 18 months of age (Favez et al., 2006; Weber, 2002). In
other words, some research studies have highlighted how the quality of parents interaction while role-playing in the prenatal LTP
procedure is predictive of the quality of their interactive exchanges
with the child after birth, at least during the first year of life and
until the child is 18 months of age (Favez et al., 2006). Moreover,
they have revealed a link between the quality of family interactions assessed during pregnancy and the first year of the childs
life and the childs developmental outcomes (Favez et al., 2006),
with specific reference to the appearance of symptomatologies at a
psychofunctional level, such as sleep and conduct disorders. These
data emphasize the extent to which parents and the quality of family
interactions can influence the childs developmental path already
during pregnancy and the first years of life. Furthermore, some
research has shown how the quality of family interactions not only
seems to influence the functioning of the family system itself but
also has an effect in other areas such as the affective relational
development of the child. Data on triadic interactions with a low
level of coordination have shown that the child finds him- or herself
involved in a conflict situation in which he or she is unable, due
also to still-incomplete capabilities, to find an adequate position
for him- or herself. Often, the child is seen to be excluded from
the interaction, which can have short- or long-term consequences
for the adequate development of interactive triadic competences.
(Fivaz-Depeursinge, Frascarolo, & Corboz-Warnery, 1998).
While exploring how these alliances set up, it also is helpful to trace their development during the transition to parenthood.
Pregnancy is not only a period of development of the embryo in
the mothers womb but also is a period when parents develop a
psychological attitude toward parenthood and their relationship
with their child. An internal representation is progressively built
in the parents minds, the baby in the head or the imaginary
baby (Stern, 1995): The representation of an imaginary baby
is progressively elaborated during pregnancy, peaking at around
the Month 6 of pregnancy and then declining or becoming less
clear. Stern (1995) postulated that this fading away is a protective
function to make space for any discrepancies in comparison with
the real baby. Most authors have focused on maternal representations of the child-to-be and their influence on the motherchild
relationship after birth (Ammaniti, Candelori, Pola, & Tambelli,
1995; Fava-Vizziello, Antonioli, Cocci, Invernizzi, 1993; Stoleru
& Morales, 1985; Zeanah & Barton, 1989) while few studies have
focused on fathers representations. This research examined the
modifications of the maternal and paternal world representations
during pregnancy and while building the relationship with the baby
in the first year. Remaining critical aspects concern the dyadic vision of the theoretical approach and the methods of observation of
the representations which are not strictly linked to the family level.
Studies in this area, such as the one by Burgin and Von
Klitzing (1995), opened up this field by showing that parents
triangular representations of the family as a threesome during pregnancy predict the place that the parents will afford the child in trilogue interactions, observed during the LTP, 4 months after birth.

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

Coparenting interaction observed by the Prenatal Lausanne Trilogue Play

Similarly, a study by McHale, Carleton, Kazan, and DeCourcey

(2002) demonstrated significant associations between parents prenatal expectations about the future family process and observed
coparental functioning in trilogue interactions after birth. During
interviews, families in which both expectant mothers and fathers
were more pessimistic about their future family process showed
lower family warmth, less cooperation, and more disagreement in
their coparenting than did more optimistic families during the LTP
several months after the childs birth (McHale & Rotman, 2007).
The aforementioned research is limited, however, by the very
association of the findings from evaluations that were carried out
with narrative methods (interviews conducted with parents during pregnancy) with those conducted with observational methods
(mostly used after the birth of the baby). In other words, the studies have shown some aspects of the representational world of the
parents during pregnancy and have connected them with the interactive attitudes shown once the baby is born, but they are not an
observation of the parental behavior during pregnancya time in
which the mental characteristics and behavioral aspects of being
a parent are activated and reorganized. In this sense, a theoretical
and empirical vacuum exists: the possibility of observing during
pregnancy the representational world of the future parents together
with their competences in terms of caring behavior prior to the
actual birth of the baby. The vision of Corboz-Warnery, FivazDepeursinge (2001) introduced into this vacuum highlights how
the quality of parental cooperation is a function which starts with
and evolves from the beginning of pregnancy, a period in which
both mother and father are involved not only in a representative
way but also in an active way, anticipating their future meeting with
their child. In this sense, the proposal consists of the possibility
of observing and evaluating the quality of the agitated representations of the partners before the birth of their baby: The concept
of agitated representation defines a series of anticipatory behaviors which the two partners can put in place in phases preceding
the birth of the baby and which reflects their interactive and cooperative capabilities once they become parents. That is, it refers
to a level intermediate between representation and interaction in
the evaluation of parenthood, which has to do with the institution (thanks to the activation of their representational world during
pregnancy) by future parents of behaviors which in some way prepare and anticipate the interactive coparenting style which will be
manifested in the presence of the baby within a triadic approach to
the study of parenthood.
To date, there have been no observational assessment methods
to study the coparental subsystem in formation at the prenatal stage.
Yet, the importance and the predictive power of observational assessments already has been documented. Comparable assessment
tools for coparenting would allow prenatal detection of postnatal
interactive functioning of the family unit. It is from this perspective
that Carneiro et al. (2006) developed an observational assessment
tool, the prenatal LTP, which is an adaptation of the postnatal LTP,
with its four configurations. Thus, it provides a new perspective,
allowing observation and assessment of coparental coordination as
it relates to the baby-to-be. It is a somewhat unusual situation in

which expectant parents are asked to role-play their first encounter

with their baby-to-be as represented by a doll. The authors hypothesized that having a pregnant couple play with a doll is a window
into the expectant couples emerging coparenting alliance. Beyond
role-playing abilities, the task calls on the parents mutual support
of and cooperation with each other. The main idea at the basis of
this method is that the parents capacity to enact their representations as a threesome will reveal the capacity for coordination
in the emerging coparental subsystem. In addition, knowing that
intuitive parenting behaviors can be activated in play with a doll,
as has been seen with children, we assume that the nature of the
task will trigger the overt manifestation of these behaviors in the
parents-to-be (Papousek & Papousek, 1987).

RESEARCH AIMS

The general goal of this study is a replication, in the Italian context,

of the Carneiro et al. (2006) research of presentation of the new
observational assessment tool, the prenatal LTP. The specific aims
were to investigate the psychometric characteristics of the administration and coding of the prenatal LTP procedure in the Italian
population as well as the comparison with the validation study and
the specific characteristics of triadic interactions during pregnancy
to demonstrate that it can effectively capture significant aspects of
the coparental couples functioning and of the family alliance in
formation.
More specifically, two research aims guide this work. The first
aim is to replicate the Carneiro et al. (2006) study in the Italian
context to verify the psychometric characteristics of the administration and coding of the prenatal LTP procedure; the underlying, working hypothesis is that we should find a unitary factorial structure grouping all coding scales of the procedure. In this
sense, we should find a coding system whose variables all define
and evaluate the same theoretical constructthe quality of coparenting interactionsaccording to different viewpoints which
are, however, linked by the same structure. This replication also
has the aim of producing an intercultural comparison between two
different samples to reflect on the specificity and differences, at
methodological and cultural levels, in the assessment of coparenting through the LTP. Since no other studies have been carried out
using this tool within different populations and cultures, we are
not in a position to formulate a specific hypothesis on possible
differences and/or similarities within the two compared groups.
This aspect of the research has a merely exploratory value and
should be further supported by subsequent, more comprehensive
investigation.
The second aim is to observe whether specific characteristics of the triadic interactions emerge during pregnancy. These
aspects should help in understanding the basic construction of the
coparental interactions on the subsequent motherfatherchild interactions, by providing key information on how the transition to
parenthood will take place and on how new parents will interact
with their child.

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

A. Simonelli, M. Bighin, and F. DePalo

METHOD
Participants

The sample for this study consisted of 98 primiparous families who

volunteered to participate in a large research project on family interactions and child development. They were recruited at childbirth
courses given by the obstetric gynecological clinic of a public Italian hospital. Families were told that they would be participating in
a longitudinal research project on the transition to parenthood and
the development of family communication. They were informed
that they would need to fill out various questionnaires and would
be asked to come to the laboratory to be filmed in various family interaction situations both prenatally and postpartum. For this
specific study, families were seen in Month 7 of pregnancy; data
obtained through questionnaires and interviews administered to the
parents were taken into consideration. Pregnancies and deliveries
were medically uncomplicated, and all infants were in good health.
No parent had a diagnosed psychiatric disorder. Fathers age was
from 28 to 42 years (M = 35.2, SD = 4.15), and mothers age was
from 26 to 41 years (M = 33.2, SD = 3.78). The family socioeconomic level was medium high (dollar amount is $ 3.1610 pro
mounth). The fathers were mainly freelance professionals (38.6%;
e.g., lawyers, medical practitioners, architects, engineers), and the
mothers were mainly employees (48.6%; secretaries, government
employees, teachers, etc.). The educational level was M = 14.38
years for the fathers (SD = 3.51, range = 818 years) and M =
15.26 (SD = 2.84, range = 818 years) for the mothers. The mean
length of the couples relationship was 9 years (including both engagement and marriage years) (SD = 4.80, range = 117 years).
The comparison group was composed of 49 Swiss couples
awaiting their first child who were recruited by the Lausanne Centre
dEtudes de la Famille (Carneiro et al., 2006).
Procedure

Prenatal LTP. Couples visited the laboratory around the 28th week
of pregnancy, which has been found to be a favorable time for evoking images of their child-to-be (Stern, 1995). They were met by
a facilitator, who remained their contact throughout the study. All
facilitators were women. Having discussed their present situation
and their respective family histories, the expectant parents were
interviewed about their representations of their baby-to-be. This
semistructured interview consisted of a total of 21 questions directed toward the father, the mother, or both parents. Questions
posed were, for example, how they imagined the physical appearance and personality of the baby, how they imagined themselves
both one-on-one with the child as well as together as a family,
and who would they like the child to resemble. In addition to
data collection, the interview allowed for rapport-building with the
facilitatoran important prelude to doing the prenatal LTP. The
setting of the prenatal LTP is shown in Figure 1.
Parents-to-be were seated in a triangular configuration, with a
basket. The baby is represented by a neutral doll, with the typical
size and shape of a newborn. The face has features and traits of

FIGURE 1. The four parts of the procedure Lausanne Trilogue Play. From The
Primary Triangle. A Developmental Systems View of Mothers, Fathers and Infants,
by E. Fivaz-Depeursinge and A. Corboz-Warnery, 1999. © 1999 by Basic
Books.

a Caucasian baby, neutral in relation to sex or particular eye,

skin, and hair color. Such neutrality should help the parents-tobe to role-play the situation. Two cameras, coupled by a common
timer, recorded two video screens: a general view (Camera 1)
and a screen with a close-up of the parents faces (Camera 2).
These images were compressed on a single screen to allow for
simultaneous viewing. The facilitator asked the parents to imagine
the moment when the three of them would meet for the first time
after the babys delivery. She explained that the task had four parts:
(a) One of them would play with the baby, (b) then the other, (c)
then the parents would play together with him or her, and finally,
(d) they would let him or her go to sleep and then talk together
about the experience that they just went through. The exercise
took about 5 min. The facilitator helped them warm up by roleplaying the nurse bringing the baby to them. For a more detailed
description of the facilitators role, please refer to Corboz-Warnery
and Fivaz-Depeursinges (2001) specific work on the prenatal LTP
procedure.
Measures

The prenatal coparenting alliance was assessed in the prenatal LTP

situation using five scales on a Likert Scale ranging from 0 to 2.

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

Coparenting interaction observed by the Prenatal Lausanne Trilogue Play

The first three scales were specifically elaborated by Carneiro et al.

(2006) to analyze the prenatal LTP.
Co-Parent Playfulness. This scale assesses the capacity of the couple to create a playful space and to co-construct a game. Couples
receive a score of 2 when both parents affective engagement is
positive and they keep a playful distance from the task. They go
along with the role-play, freely interpreting the instructions. They
show an awareness that this is a game and not reality. There are
several conditions in which couples receive a score of 0: (a) when
one of the parents engages in the play, but the other one struggles,
unable to share in the partners pleasure; (b) the parents adhere
too literally to the instructions, seemingly unaware that they are
engaged in a game; and/or (c) both parents remain constrained or
denigrate the task. A score of 1 is used when parents manage to
play the game only for a moment, but then often lose the thread, or
when parents accomplish the task, but without a playful distance
and without a clear indication of their awareness that this is a game
and not reality.
Structure of the Play. This scale assesses the couples capacity
to structure the four play segments according to the instructions.
Two dimensions are considered: the differentiation of the play
into four discrete segments and the duration of both the entire
play sequence as well as of the four segments. The differentiation
into four discrete play segments is coded at high level when the
coder can easily distinguish the four play segments. In the first two
play segments (Parts 1 and 2), the parent who is playing is easily
identifiable (e.g., leaning forward toward the crib or holding the
baby doll). In Part 3, the parents together address the baby, who
is either held by one of them or in the basket. In Part 4, parents
address each other in adult talk. Durations must be sufficiently
long for interaction to develop, but not too long, leaving time for
each of the partners. Hence, 2 points are allotted when the entire
game lasts about 4 to 5 min, with four distinct components of
about 1 min each. Couples receive a score of 0 when the elapsed
time is too short (2 min) or too long (9 min), or when two or
more elements are omitted. A score of 1 is used when the durations are appropriate, but the differentiation of Parts 1 to 4 is
not clear or when the differentiation is clear, but the durations are
inappropriate.
Intuitive Parenting Behaviors. This scale assesses the parents use
of intuitive parenting behaviors. Six behaviors known from the
literature (Papousek & Papousek, 1987) are coded: holding and
en face orientation, dialogue distance, baby talk and/or smiles
at baby, caresses and/or rocking, exploration of the babys body,
and preoccupation with the babys well-being. These intuitive parenting behaviors are assessed as present or absent separately for
each parent. Coding results in a score for each parent, then these
scores are collapsed together into a global score for the couple.
Thus, if one parents behavior is coded as optimal and the other is
coded as intermediate, the overall couples score will be partially
appropriate. A score of 2 points is given when each parent shows

at least five intuitive parenting behaviors. A score of 0 is given

when one parent shows 2 or less intuitive parenting behaviors. A
score of 1 is given when one parent shows at least five intuitive
parenting behaviors, but the other shows only three or four such
behaviors.
Scales 4 and 5 are based on the Co-Parenting and Family
Scale (McHale, Kuersten-Hogan, & Lauretti, 2001). In the original
version, these scales were used to analyze coparental interactions.
As the scales focus only on the parents and not directly on the child,
they were easily applicable, without modification, to the prenatal
situation, which by definition focuses only on the parents. Coding
is done on the video record of the parents in close-up to better
capture the parents facial expressions.
Couple Cooperation Scale. This scale assesses the degree of active
cooperation between the parents during the play, at a behavioral
level. The absence of antagonism or interference is not sufficient
to attain a high score. A score of 2 is awarded only when there
is a clear indication of active cooperation between the parents,
through the use of gestures and words which facilitate joint play
and mutual support. A score of 0 is given when there are interferences between partners, offending remarks, and mockery about
the partners actions, all of which prevent co-construction of the
play. Finally, couples receive a score of 1 when parents cooperate
only in some parts of the play or when there is a large difference
between the parents, with only one of them engaging in the play
and cooperating.
Family Warmth. This scale captures the affection and humor shared
by the partners during play; namely, whether they manifest affection and tenderness as a couple and toward the baby. Couples
receive a score of 2 points when there are tender words, complicit
smiles, and warm gestures between the parents and toward the
baby. At the other extreme, a score of 0 is given in the absence
of warmth and the expression of negative affect. Sometimes, it
is toward the doll baby that the parents are unable to express affection; at other times, it is toward each other that the partners
show coldness, disdain, or even contempt. A score of 1 is given
when the parents show warmth only some of the time (e.g., when
playing with the baby in Parts 1 and 2, but show distance and
coldness in the two last segments when they interact with each
other).
The scores of the five scales are added to obtain a global score
between 0 and 10. The higher the score, the more the prenatal
alliance is considered to be functional. Two independent observers
coded all sessions to assess interrater reliability. Interrater reliability was assessed using Cohens index. Reliability of the two raters
on the total prenatal score was = .87. The Inter Class Correlation
index was .85. When each individual scale of the coding system
was taken into consideration, judges agreement appeared as follows: Co-Parent Playfulness ( = .76), Structure of the Play ( =
.74), Intuitive Parenting Behaviors ( = .74), Couple Cooperation
( = .72), and Family Warmth ( = .72).

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

A. Simonelli, M. Bighin, and F. DePalo

TABLE 1. Descriptive Statistics of the Scores in the Scales and Total

Prenatal Lausanne Trilogue Play (N = 98) and Differences Between the
Two Groups
Padova
(N = 98)

Coparent Playfulness
Structure of Play
Intuitive Behavior
Couple Cooperation
Family Warmth
Total Family Alliance

Losanna
(N = 49)

MannWithney

SD

SD

Sign

1.08
1.42
1.28
1.47
1.34
6.58

.67
.67
.89
.68
.70
2.58

1.49
1.18
1.12
1.14
1.27
6.20

.54
.75
.66
.76
.75
2.52

3.50
1.81
1.54
2.50
.48
.98

.00
.07
.12
.01
.63
.33

RESULTS
Descriptive Results

Table 1 shows the means and standard deviations of the prenatal assessment scales of our studied sample and of the Carneiro
et al. (2006) research control. There were no significant differences

between the scale scores of our samples and those of Carneiro et al.
(2006); the only difference is related to the Structure of Play scale,
which has a higher mean in the Italian sample than it does in the
Swiss one.
Figures 2 and 3 represent the scores distribution curves of
the two samples (Italian and Swiss); as can be observed, these
distributions are still homogeneous with respect to the normal one:
Results show that the distribution of the Carneiro et al. (2006)
sample has asymmetric and kurtosis values in line with normal
distribution of observed data (asymmetry = .321, SE = .34,
kurtosis = .65, SE = .668). Our sample also has a distribution
that does not differ significantly from the normal one (asymmetry =
.61, SE = .24, kurtosis = .44, SE = .48).
The degree of reliability of the coding system is assessed
by the Cronbachs coefficient: Internal consistency of the five
prenatal scales was good (Cronbachs = .78). In the Carneiro
et al. (2006) study, the scales of the LTP prenatal coding system
also have good internal consistency (Cronbachs = .79). Table 2
shows the results of consistency statistics applied to the single
scales of the prenatal LTP coding system. The results relative to
the reliability statistics also applied to the individual scales of the
coding system for both the Italian research and the Swiss group.

FIGURE 2.

Distribution of scores by the Italian group on the scales in the coding system by Carneiro et al. (2006).

FIGURE 3.

Distribution of scores by the Swiss group on the scales of the coding system by Carneiro et al. (2006).

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

Coparenting interaction observed by the Prenatal Lausanne Trilogue Play

TABLE 2. Reliability of Scores in the Italian Group on the Scales of the

Prenatal Lausanne Trilogue Play Coding System (Carneiro,
Corboz-Warnery, & Fivaz-Depeursinge, 2006) in the Italian and Swiss
Groups
Padova (N = 98)

Losanna (N = 49)

Total
Cronbachs
Total
Cronbachs
Correlation
Excluded
Correlation
Excluded
Correct Scale
Scale
Correct Scale
Scale
Coparent Playfulness
Structure of Play
Intuitive Behavior
Couple Cooperation
Family Warmth

.67
.21
.54
.64
.59

.66
.80
.71
.67
.69

.48
.57
.51
.53
.60

.74
.71
.73
.73
.69

The alpha levels remain at a similarly high level if individual

scales are systematically excluded, as do the overall correlations
between the scales. These data show good reliability of the prenatal LTP coding system, both in the general sense and for each
individual scale. In fact, if the last column of Table 2 is observed,
eliminating the contribution given by each scale, the alpha values
are reduced with respect to the reference value obtained by =
.78. The only exception is given by the Structure of Play scale:
This represents a correlation with rather low total points (R =
.21), and if eliminated from the analysis, the index of reliability
for the method increases.
Given the reassuring results obtained by the verification of the
reliability of the coding method, the basis of the LTP technique, further analysis examined the factor structure underlying the coding
system. To investigate this aspect, a factor analysis of the principal components using varimax rotation of the results obtained on
the evaluation scale of the LTP was used. Varimax rotation was
chosen because it is an orthogonal rotation which maximizes the

sum of the variants maintaining factors not correlated with each

other. As such, this places a greater emphasis on the simplification
of the structure of the factors in terms of variability and not vice
versa, allowing for the amplification of the high correlations of
each factor and the reduction of those which were low, thus aiding
interpretation. No confirmative factorial analysis was applied. In
fact, since this was the first factorial study on this procedure, we
did not have any factorial group to confirm. It was only possible to
explore groupingif anyamong variables.
As reported in Table 3, the results showed the presence of a single factor which explained 52.87% of the total variation (The correlation matrix underlying the applied factorial analysis is included in
the Appendix.) Furthermore, the KaiserMeyerOlkin Test (.78)
and the Bartlett sphericity test, 2 (N = 98, 10) = 134,02, p <
.001, illustrate the satisfactory application of the factorial model.
This result shows how all the evaluation scales of the prenatal LTP
procedure lead homogeneously and uniformly to a single base construct, as identified by the emerging factors, which is defined by
the authors as the quality of the coparenthood interaction when
role-playing. In other words, the diverse aspects observed through
the scales lead to the description of a single function linked to the
parenting competences which is described in the area of evaluation observed by the variables of the coding system used. This
analysis also highlights the presence of another factor with an
eigenvalue of .98; nevertheless, we deleted this second solution
while there is a consistent difference in the values between the two
factors.
Further verification of the internal consistency of the prenatal
LTP coding system was carried out applying Pearson correlational
analysis to the results obtained from the families in each of the five
evaluation scales of the method and between these and the overall
result (Table 4).
As seen in Table 4, the scales correlate positively with the
Family Warmth Scale and with the overall results. The only scale

TABLE 3. Factorial Analysis Applied to the Prenatal Lausanne Trilogue Play Procedure
1 Component
Coparent Playfulness
Couple Cooperation
Family Warmth
Intuitive Behavior
Structure of Play
A single factor explains 52.82% of the total variance.

.83
.81
.78
.75
.34

Initial Values

Rotated Factors

Component

Total

%Variance

%Cumulative

Total

%Variance

%Cumulative

1
2
3
4
5

2.64
.98
.56
.45
.37

52.82
19.52
11.18
9.06
7.42

52.82
72.34
83.52
92.58
100.00

2.64

52.82

52.82

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

A. Simonelli, M. Bighin, and F. DePalo

TABLE 4. Correlation Between Scores Obtained by the Italian Group in the Scales and Total Prenatal Lausanne Trilogue Play (N = 98)

Coparent Playfulness
Structure of Play
Intuitive Behavior
Couple Cooperation
Family Warmth
p

Coparent
Playfulness

Structure
of Play

.24

Intuitive
Behavior
.55
.09

Couple
Cooperation
.57
.27
.43

Family
Warmth
.52
.09
.51
.55

Total
Alliance
.80
.45
.77
.78
.75

< .01. *p < .05.

TABLE 5. Correlation Between Scores Obtained by the Swiss Group in the Scales and Total Prenatal Lausanne Trilogue Play (N = 49)

Coparent Playfulness
Structure of Play
Intuitive Behavior
Couple Cooperation
Family Warmth
p

Coparent
Playfulness

Structure
of Play

Intuitive
Behavior

Couple
Cooperation

Family
Warmth

Total
Alliance

.35

.30
.39

.25
.49
.41

.54
.42
.42
.41

.63
.74
.68
.75
.77

< .01.

that does not show this type of correlation is that which describes
the structure of the play, in terms of organization and of duration, which shows significant statistical correlations only with the
Couple Cooperation Scale and with the overall result. These data
deviate in part from that observed in the study by Carneiro et al.
(2006), where all scales of the coding system showed significant correlation between themselves and with the overall score
(Table 5). Also in this study, however, most correlation is shown
between the different scales and that of the Family Warmth Scale
and the overall score. In that sense, these two dimensions seem to
be those which confer the highest level of reliability to the coding
procedures used.
To make a comparison between the scores obtained by the
two groups on the evaluation scales of the prenatal LTP coding
system, a MannWhitney test was performed. The results shown
in Table 1 do not show any significant differences for most of
the scales, with the exception of the Couple Cooperation and CoParent Playfulness Scales. This result highlights small differences
between the two groups relative to the interactive mode despite
the application of the MannWhitney test to the distribution of
the overall scores transformed in rows giving data which do not
show any significant differences between the two groups regarding
the level of total prenatal alliance, Z = .981, p = .326, n.s.
The choice to use the MannWhitney test to make a comparison
between the two groups derived from the difference of the two
samples, which required the application of nonparametric, rather
than parametric, statistics.
Given the results obtained relative to the two groups, but more
important, given the low correlation of the Structure of Play Scale
with respect to the other scales obtained by the Italian group, it was

deemed necessary to study in more depth the analysis relative to

the scores obtained on that scale to carry out a comparison between
the group studied and that of the Swiss. The analysis applied to
the distribution of scores obtained by the families showed how the
Italian group devoted more time to carrying out the triadic play than
did the Swiss group, and therefore in consideration of the criteria
for the coding system, achieved scores substantially different from
those of the Swiss (Figure 4).
As can be seen in Figure 4, the duration of the procedure is
rather long, with an average time of 7 min 13 s (M = 401.99,
SD = 175.08). This may be so for two principal reasons: The first
reason regards the composition of the Structure of Play Scale.
This is composed of two subscales: Organization and Length
of Play. For the organizational part, the presence or absence of
all four parts of the procedure are evaluated, gradually penalizing the family for the parts which are not undertaken. For duration of play, the score is given according to a temporal subdivision related to the total duration of the play (Table 6). The
second reason regards the assignment given to the parents. Before the start of the play, the future parents were given a time
(5 min) indicative of the duration of the play, but the partners were free to choose when to start and when to finish the
play.
Observing Table 6 relative to the temporal subdivision during the play proposed by the authors Del the LTP procedure the
distribution of time devoted by the Italian group in finishing the
procedure (Table 7), future proposals could reconsider the time
used by the Italian families during the undertaking of the triadic
play and propose a new table based on the distribution of times of
the Italian families.

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

Coparenting interaction observed by the Prenatal Lausanne Trilogue Play

FIGURE 4.

inappropriate
partially appropriate
appropriate
partially appropriate
inappropriate

TABLE 7. Distribution of the Time Trend in the Swiss Group

<4 min
4 to 5 min 18 s
5 min 18 s to 6 min 28 s
6 min 28 s to 8 min 18 s
>8 min 18 s

25 th percentile
50 th
70 th percentile

Trend in the total time in the prenatal Lausanne Trilogue Play.

TABLE 6. Distribution of the Time Trend in the Swiss Group

<2 min
2 min to 2 min 30 s
2 min 30 s to 6 min 30 s
6 min 30 s to 9 min
>9 min

inappropriate
partially appropriate
percentile appropriate
partially appropriate
inappropriate

DISCUSSION

Dynamic and evolutionary psychological research has always highlighted the need to be equipped with methods and tools that are suitable for the evaluation of complex constructs which characterize
the subject of interest, require precise definition at the theoretical
level, and result in observable and therefore analyzable behaviors.
In particular, the evaluation of parenthood and of parental competences poses a series of construct problems and problems relative
to the choices of variable to be measured. For this reason, this research proposed the application in an Italian context of a particular
methodology which supplied a specific definition of parenthood in
terms of the quality of couple cooperation, and which identified
a specific situation and a series of behaviors in which the observations were considered by the authors to be an evaluation of the

parenting competences shown by the couple before the birth of the

child. An earlier study along the same lines that was conducted
by the authors was limited by the small number of families and
therefore the scope of the data collected, and the findings were
very limited (Simonelli, Fava Vizziello, Bighin, & De Palo, 2009).
In fact, we know that the application and verification of new tools
necessitate numerous trials having a significant sample size because of the difficulty of testing the tools and of understanding
their limitations and the resources needed.
The aim of the work was the verification of the applicability
of the prenatal LTP in an Italian context using a coding system
designed by the authors; that is, whether the scale defined for the
evaluation of the quality of the interaction of the couple in the
play of parenthood could be used in an Italian context obtaining
results analogous to or at least comparable to those obtained in the
original studies by Carneiro et al. (2006). The statistical analysis
used in our study has a good level of internal reliability within the
coding system, which shows a normal distribution and significant
correlations between these scores and overall. Furthermore, the
factor analysis applied to the scales identifies them as belonging
to a single base construct, identified by an emerging factor defined
by the authors as the quality of the interaction of the couple in
undertaking the role-play. In other words, the diverse aspects observed through the scales lead to the description of a single function
linked to the parenting competences, which is described in the area
of evaluation observed by the variables of the coding system used.
At the same time, no significant differences were seen between the
data obtained by the authors and that obtained by the reference
study of Carneiro et al. (2002); nevertheless, the distributions of
the two groups showed a high degree of homogeneity. The only
exceptions were with respect to the findings on the Couple Cooperation Scale, where the score obtained by the Italian group was
higher than that of the Swiss group, and the Co-Parent Playfulness
Scale, where the Italian group gained lower points than did the

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

10

A. Simonelli, M. Bighin, and F. DePalo

Swiss group. In general, therefore, we can conclude by identifying

the prenatal LTP method as one which, wherever applied, allows
the evaluation of a particular ability of the couple to operate parenting competences within a fictitious environment (role-play). This
seems adequately useable and coding in different cultural contexts
without the emergence of errors linked to the application of the
process and/or codification.
The only critical aspect which in part seems to counter the
coherence of the data regards the Scale of the Structure of Play,
which differentiates the Italian group from that of the Swiss and
does not correlate with the other scales or with the overall score.
In general, this scale is made up of various variables which concern the organization and duration of the play of the couple when
role-playing with the doll. The duration seems to be the differentiating characteristic between the two groups examined. In fact, the
analysis undertaken shows a greater duration of the time of play
by the Italian couples, a characteristic which in the coding system
receives a specific score. In other words, the authors described the
duration of the play identifying an average time suitable to the
proposed situation (time range: 29 min): Couples who exceed or,
alternatively, who work too quickly are penalized by the scores that
they receive. For this reason, the lack of correlation found using
this scale seems to be linked to this aspect. The longer duration
of the play with respect to the duration of the variable, considered
to be suitable in the coding system, means that the Italian couples
receive overall scores on the scale which deviate from those expected and which consequently render the scale of the Structure of
Play unreliable.
Seen in a broader perspective, this may be considered a typical characteristic of the population observed, and an option may
be to make some changes to the coding system scale regarding
the duration of the game. This would seem to be a necessary step
also because for this variable, the couple does not receive any particular instructions from the researchers, but is left to play for a
time which they deem suitable and to signal the end of the play
when desired. This aspect therefore appears to be that which is
most affected by components that can be defined as belonging to
a cultural matrix and which intervene to influence the course of
the procedure. In fact, this aspect could be considered a typical
feature of the observed population, and a comparison with other
Italian groups who also have used this procedure is currently being developed to investigate the reasons for the differences found
with respect to the Swiss group, thus evaluating the possibility of
creating an Italian distribution of the duration of play variable and
therefore identifying an appropriate range of its assessment and a
range of related scores more suited to the sample population.
Finally, there are some limitations of our study that future
studies should try to overcome. The first limitation concerns
participant-selection criteria. In the present work, participant families spontaneously offered to take part in the project after hearing
about it during birth preparation courses; however, no mention is
made of family participation rate as compared with the total number of families. Moreover, the sampling procedure did not foresee
different groups according to the participants sociodemographic

characteristics. These couples spontaneously decided to participate in the research, and their sociocultural and economic level
was high. It would certainly be preferable to build a sample group
including families from different sociocultural levels to make it
more homogeneous with the general population. An additional,
critical aspect refers to the procedure itself, which might prove to
be artificial: the quality of its coding system as far as setting
and required task are concerned. It would be necessary to check if
and to what extent the participants feel they are facing a strange,
unnatural condition, which would imply a low ecological validity of the observed interactions. The authors already explored this
issue with reference to the postnatal LTP procedure. In fact, after role-playing with the child, parents were invited to answer a
questionnaire investigating these themes. Reported data indicate
that parents experienced their playing situation at three, together
with the child, as being very similar to moments of family interactions during their daily life, which confirmed the ecological
features of the procedure (Favez, Lavanchy-Scaiola, Tissot, Darwiche, & Frascarolo, 2011). In the same way, a similar tool could
be conceived and used for the prenatal LTP procedure to investigate and assess any possible limitation in the role-playing situation
and whether such a condition is really suitable to observe parents
behavior before the childs birth.
A final remark concerns the manifestation of the acted-out
parental representations which constitute, in the authors opinion,
the focus of the observation and evaluation proposed by the procedure: the ability to assess these behaviors during pregnancy through
a role-play situation which involves parents in a pleasant way, for
a limited time, in a not-too-intrusive and not-anxiety-inducing situation is, in our opinion, the power of this methodology and of
any future applications. In this sense, the shift to the measurement
of parenting before the birth of the child, the interface between
behavior and narrative representation, may contribute to the understanding of complex dynamics which affect the transition to
parenthood of both groups of nonclinical participants (e.g., those
in this work) and of couples who have either individual pathological conditions or fragility related to the evolutionary phase which
they are about to face. Future clinical and empirical applications
will allow for the investigation of the various aspects related to this
assessment and whose use in a clinical setting will certainly open
the possibility of increased communication between such settings
and research.
REFERENCES
Ammaniti, M., Candelori, C., Pola, M., & Tambelli, R. (1995). Maternity
and Pregnancy Milan: Raffaello Cortina.
Belsky, J. (1984). The determinants of parenting: A process model. Child
Development, 55, 8398.
Belsky, J., & Kelly, J. (1994). The transition to parenthood: How a first
changes a marriage. Why some couples grow closer and others apart.
New York: Dell.
Burgin, D., & von Klitzing, K. (1995). Prenatal representations and postnatal interactions of a threesome (mother, father and baby). In J. Bitzer

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.

Coparenting interaction observed by the Prenatal Lausanne Trilogue Play

& M. Stauber (Eds.), Psychosomatic obstetrics and gynaecology (pp.

185192). Bologna, Italy: Monduzzi.
Carneiro, C., Corboz-Warnery, A., & Fivaz-Depeursinge, E. (2006). Prenatal coparenting and postnatal family alliance. Infant Mental Health
Journal, 27(2), 207228.
Corboz-Warnery, A., & Fivaz-Depeursinge, E. (2001). By the Couple to
the Family: the Coparenting Prenatal Alliance in the transition to
Parenthood. Cahiers Critiques de Therapie Familiale et de Pratique
de Reseaux, 27(2), 1734.
Cowan, C.P., & Cowan, P.A. (1992). When partners became parents. New
York: Basic Books.
Fava Vizziello, G.M., Antonioli, M.A., Cocci, V., & Invernizzi, R. (1993).
From pregnancy to motherhood: The structure of representative and
narrative change. Infant Mental Health Journal, 14, 416.
Favez, N., & Frascarolo, A. (2002, July, 1620). The nascent family alliance. Paper presented at the 8th Congress of the World Association
of Infant Mental Health, Amsterdam.
Favez, N., Frascarolo, F., Carneiro, C., Montfort, V., Corboz-Warnery,
A., & Fivaz-Depeursinge, E. (2006). The development of the family
alliance from pregnancy to toddlerhood and children outcomes at
18 months. Infant and Child Development, 15, 59
73.

11

McHale, J., Carleton, M., Kazan, I., & DeCourcey, W. (2002, July, 1620).
Do prebirth expectancies about co-parenting foreshadow postbirth
co-parenting dynamics? Paper presented at the 8th Congress of the
World Association of Infant Mental Health, Amsterdam.
McHale, J., & Cowan, P. (1996). New directions in the study
family-level dynamics during infancy and early childhood. In
J. McHale & P. Cowan (Eds.), Understanding how family-level
dynamics affect childrens development. Studies of two parents families. New Directions for Child Development, 74, 5
26.
McHale, J., & Fivaz-Depeursinge, E. (1999). Understanding triadic and
family group interactions during infancy and toddlerhood. Clinical
Child and Family Psychology Review, 2, 107127.
McHale, J., Kuersten-Hogan, R., & Lauretti, A. (2001). Evaluating
co-parenting and family-level dynamics during infancy and early
childhood: The co-parenting and family rating scale. In P.K.
King & K.M. Lindhal (Eds.), Family observation coding systems. Resources for systemic research (pp. 151170). London:
Erlbaum.
McHale, J., Kuersten-Hogan, R., & Rao, N. (2004). Growing points for
coparenting therapy and research. Journal of Adult Development,
11(3), 221234.

Favez, N., Lavancy-Scaiola, C., Tissot, E., Darwiche, J., & Frascarolo, F.
(2011). The Family Alliance Assessment Scale: Steps toward Validation and Reliability of an Observational Assessment Tool for Early
family Interactions. Journal of Child and Family studies, 20 (1), 2337.
Fivaz-Depeursinge, E., & Corboz-Warnery, A. (1999). The primary triangle. A developmental systems view of mothers, fathers and infants.
New York: Basic Books [Trad. It. Il triangolo primario. Milano:
Raffaello Cortina, 2000].

McHale, J., & Rotman, T. (2007). Is seeing believing? Expectant parents outlooks on coparenting and later coparenting solidarity. Infant
Behavior & Development, 30, 6381.

Fivaz-Depeursinge, E., Frascarolo, F., & Corboz-Warnery, A. (1998).

Evaluation of Triadic Alliance during mother-father-baby game. Devenir, 10(4), 79104.
Katz, L.F., & Gottman, J.M. (1996). Spillover effect of marital conflict:
In search of parenting and co-parenting mechanisms. In J. McHale
& P. Cowan (Eds.), Understanding how family-level dynamics affect
childrens development: Studies of two-parents families (pp. 5776).
San Francisco: Jossey-Bass.
Margolin, G., Gordis, E.B., & John, R.S. (2001). Co-parenting: A link
between marital conflict and parenting in two-parent families. Journal
of Family Psychology, 15, 221.

Simonelli, A., Fava Vizziello, G.M., Bighin, M., & De Palo, F. (2009).
Preliminary Validation Analysis of Prenatal Lausanne Trilogue Play
Procedure to study Coparenting Competences in Pregnancy. Et`a Evolutiva, 92, 517.

McHale, J. (2007). Charting the bumpty of coparenthood: Understanding

the challenges of family life. Washington, DC: ZERO TO THREE
Press.

Minuchin, S. (1974). Families and family therapy. Cambridge, MA: Harvard University Press.
Papousek, H., & Papousek, N. (1987). Intuitive parenting: A dialectic
counterpart to the infants integrative competence. In J.D. Ososky
(Ed.), Handbook of infant development (2nd ed., pp. 669720). New
York: Wiley.

Stern, D.N. (1995). The motherhood constellation. New York: Basic

Books.
Stoleru, S., & Morales, M. (1985). From the ghostly baby in pregnancy to
the mother-infant interaction. Psychiatrie de lEnfant, 28, 441484.
Weber, K. (2002). Family Alliance: continuity between the first and second year of a childs life. Faculte de Psychologie et de Science de
lEducation, travail de DESS Gen`eve, Universite de Gen`eve.
Zeanah, C.H., & Barton, M.L. (1989). Introduction: Internal representations and parentinfant relationship. Infant Mental Health Journal,
10, 135141.

Infant Mental Health Journal DOI 10.1002/imhj. Published on behalf of the Michigan Association for Infant Mental Health.