Predictors of infant jealousy
Jiska Westen (0922277)
Supervisors: Prof. dr. Lenneke Alink, Dr. Szilvia Biro
February, 2015
Centre for Child and Family Studies
Leiden University
Contents Abstract 3 Introduction 4 Jealousy 5 Maternal sensitivity 10 Infant temperament 13 Differential susceptibility 14 Present study 16 Method 17 Participants 17 Procedure 17 Measures 18 Analytic plan 21 Results 22 Data inspection 22 Preliminary analyses 22
Predictors of infant jealousy 25
Differential susceptibility hypothesis 25
Discussion 26
Implications 28
Limitations 29
Recommendations for future research 30
Conclusion 30
Abstract
The goal of the present study was to investigate whether maternal sensitivity and infant temperament were predictors of infant jealousy. The differential susceptibility hypothesis was also tested. It was investigated if infant temperament could moderate the relationship between maternal sensitivity and infant jealousy. Forty-eight 10-month-olds and their mothers
participated in the study. Jealousy was measured during a 10 minutes episode of free-play between mother and infant during which mothers were instructed to care for a life-like doll. The doll cried for 5 minutes of the total episode. Infants were able to play with a large variety of toys during the jealousy evoking situation. Maternal sensitivity was measured during three additional short episodes (break, free-play without toys, free-play with toys). Infant
temperament was measured by the Infant Characteristic Questionnaire (ICQ). The results showed that maternal sensitivity and infant temperament were not predictors of infant jealousy within the present study. Also, no proof was found for the differential susceptibility hypothesis. Infant temperament was not a significant predictor in the relationship between maternal sensitivity and infant temperament. The results of the present study indicate that the relationships between maternal sensitivity and infant jealousy and between infant
temperament and infant jealousy might not be as strong as suggested by previous studies.
Key words: infant jealousy, maternal sensitivity, infant temperament, differential susceptibility hypothesis
Introduction
All people experience the feeling of jealousy at some point in their lives. Siblings can experience jealousy because another sibling receives more attention from their parents and romantic partners can be jealous of potential rivals. Many different definitions of jealousy have been formulated. One of the most well-known definitions is written by Parrot (1991). In his view jealousy is an emotion that a person experiences when he or she is afraid to lose an important relationship to another person, also called the rival. Research has shown that not only adults are able to feel jealous, but children can experience feelings of jealousy as well. Several studies have found that older siblings can experience feelings of jealousy, for
example, when a new-born sibling arrives (Dunn, Kendrick & MacNamee, 1981; Kendrick & Dunn, 1982; Field & Reite, 1984). However, older siblings are toddlers by the time their younger sibling is born. This raised the question whether even younger children are able to feel jealous. Therefore several studies have investigated whether infants can experience feelings of jealousy using life-like dolls to resemble social rivals. These studies found that infants are capable of an early form of jealousy, which is often called jealousy protest (Hart, Field & Del Valle, 1998a; Hart, Field, Letourneau & Del Valle, 1998b; Hart & Carrington, 2002; Hart, Carrington, Tronick & Caroll, 2004; Mize & Jones, 2012).
Although it is known that infants can show an early form of jealousy, not much is known about the predictors of infant jealousy. Only a few studies focused on which constructs are related to infant jealousy. For example, Hart and Behrens (2013) found that
temperamental emotionality and maternal sensitivity were related to infant jealousy. More specifically, they found that infants who scored high on positive emotionality showed more distress in the jealousy evoking condition. It was also found that infants with more sensitive mothers displayed less distress during the eliciting condition. Mize and Jones (2012) also found that several aspects of infant temperament were related to infant jealousy. For example, it was found that infants who smiled and laughed a lot during the eliciting condition displayed less jealousy than infants who laughed and smiled less.
The present study will focus on two predictors of jealousy in 10-month-olds. Since previous studies have found that maternal sensitivity and infant temperament are related to jealousy (Hart & Behrens, 2013; Mize & Jones, 2012), the first goal of the present study is to investigate whether maternal sensitivity and infant temperament are also predictors of infant jealousy within the present study. The second goal of the study is to determine which of the two constructs is the strongest predictor, if both constructs are found to be related to infant jealousy. A life-like doll will be used as a social rival to study infant jealousy. A unique
aspect of the present study is that this doll not only resembles a 6-week-old infant in height and weight, but it can also cry like a new-born. Infant jealousy will be measured when the doll cries, but also when the doll is quiet. An advantage of using a doll that is able to cry is that it might be more representative of a real social rival or family situation. Another advantage is that it allows the field to investigate infant jealousy, while mother is trying to soothe the doll. This might be more stressful for infants and could result in more jealousy. The third goal of the study is to investigate whether temperament can moderate the relationship between jealousy and sensitivity. It will be tested whether the relationship between sensitivity and jealousy is stronger for infants with a difficult temperament. This is also called testing the differential susceptibility hypothesis and is introduced by Belsky (Belsky,1997; 2005). The present study will, to my knowledge, be the first study that tests the differential susceptibility hypothesis in the infant jealousy field.
Jealousy
What is jealousy?
Jealousy is described as an emotion that a person experiences when he or she is afraid to lose an important relationship to a rival (Parrot, 1991). Jealousy is considered to be a complex and mixed emotion. It is thought to be a mix of several more basic emotions. Research has shown that jealous individuals can react with anger, sadness, and fear but jealousy is also associated with love and affection (Hart, 2010). Although jealousy is a well-known emotion, it is often confused with the emotion envy. The emotion envy refers to feelings of dissatisfaction a person can experience regarding his or her own fate and a desire for the attributes, reputation or possessions of another person (Bryson, 1977 as described in Salovey & Rodin, 1984). Both emotions describe the disturbed feelings that people can have towards other people’s
possessions. However, there is a clear difference between jealousy and envy. When feeling envious, a person desires to have something that is in possession of another person. When feeling jealous, a person is afraid to lose something (a relationship) that he or she already possesses. Envy can only occur in a two-party relationship, namely between the envious person and the person who possesses something that the other person desires. Jealousy, on the other hand, always occurs within a party relationship (Ben-Ze’ev, 2010). This three-party relationship is also known as the ‘social triangle’, which is a triangle of relationships (Parrott, 1991). This triangle involves four relationships, namely the relationship all three involved parties have and three dyadic relationships. The first dyadic relationship in the triangle is the relationship between the jealous person and the desired person, which is also
called the primary or jealous relationship. The second dyadic relationship involves the desired person and the rival, which is also called the secondary or rival relationship. The last dyadic relationship within the triangle concerns the relationship between the jealous person and the rival, also called the adverse relationship. Before jealousy actually occurs, there are four criteria that need to be met. The first criterion concerns the issue that there needs to be a social triangle in order for jealousy to occur. The second criterion is that the jealous individual should assess his or her relationship with the desired person as a very important relationship. Jealousy will not occur if the relationship is not considered to be of great importance. The third criterion concerns the loss of the relationship. A person will only become jealous when there is a real or a perceived chance that he or she will lose an important relationship to a rival. For example, the feelings that a person experiences after losing a love-relationship due to separation or death of the partner is not considered to be jealousy. Finally, jealousy is not only the result of losing someone’s love. If this was the case, jealousy could only have existed in romantic relationships, which is not the case (White & Mullen, 1989 as described in
Volling, McElwain & Miller, 2002).
Sibling jealousy in early childhood
Most researchers who investigate jealousy in childhood, focus on jealousy between siblings. Sibling jealousy refers to the conflict between siblings for the love and attention of their parents. Children can react with anger and sadness when they fear that the relationship that they had with their parents is lost to the sibling rival (Volling, Kennedy & Jackey, 2010).
Research has shown that the birth of a sibling is one of the first events where sibling jealousy occurs. It has been found that the behavior of the older sibling can change after the birth of a sibling. Older children can show many different behaviors after the arrival of a sibling. They are found to react by trying to disturb or protest the interaction between mother and new-born infant (Kendrick & Dunn, 1982), show more withdrawal and also show more tearful behavior (Dunn et al., 1981). Furthermore, parents reported changes in eating patterns, toilet habits and sleeping patterns of their oldest children. Some parents also reported that their oldest child imitated their youngest, for instance by drinking out of a bottle, lying on the changing mat of their younger sibling and imitating their younger siblings behavior (e.g., crying and noises) (Dunn et al., 1981). However, there is some debate whether the above described behaviors of the older sibling are actually the result of jealousy. The arrival of a second child causes a lot of changes within a family. The problematic behaviors shown by the older sibling can be a result of these changes, instead of jealousy (Volling et al., 2010).
Although it is not clear whether the changes in behavior after the arrival of a sibling can always be considered jealousy, it is clear that sibling jealousy occurs in early childhood. Several studies have focused on sibling jealousy in the triadic context (one of the parents and both siblings). Miller, Volling & McElwain (2000) conducted a study in which they included 16-month-old toddlers, their older siblings (between 2 and 6 years of age) and one of their parents in their sample. Both parents were observed in a jealousy-evoking condition, in which the parent was instructed to play with one child while the other child had to play on its own with other (less attractive) toys. After 3 minutes the roles were reversed and the other sibling had to play alone. The results of the study revealed that the older siblings showed more sadness when they had to play alone. The younger siblings showed more distress during this challenging condition. These findings indicate that toddlers and pre-schoolers are capable of demonstrating jealous behavior. Volling et al. (2010) also conducted a study in which they investigated sibling jealousy in early childhood. Both parents were included in this study. At the time the study was conducted, the younger sibling was approximately 2 years of age (toddler) and the older sibling was between 4 and 6 years of age (pre-schooler). The procedures were similar to the study conducted by Miller and colleagues and were adapted from the procedures used by Teti & Ablard (Teti & Ablard, 1989 as described in Volling, Kennedy & Jackey, 2010). Parents were instructed to play with one child and had to tell the other child to play with other toys that were present in the room. Volling et al. (2010) state that this procedure provokes jealous feelings in young children, because the children need to control their emotions and behaviors, while their parent is only paying attention to the other sibling (or the rival). Results of the study revealed that both the younger and the older sibling showed jealous behavior, when were instructed to play alone. During the mother session the younger sibling showed less happiness and more sadness, whereas the older sibling only showed a significant increase in sadness. During the father session the younger sibling showed more anger, more sadness and less happiness. The older siblings showed more sadness and less happiness when they had to play alone. These findings indicate that within this study both the older and the younger sibling showed jealous behavior.
Jealousy in infancy
The literature on sibling jealousy has shown that toddlers are capable of expressing jealousy, however researchers started to wonder if even younger children were also able to experience feelings of jealousy. Studies on sibling jealousy most often focus on toddlers and their older siblings or on jealous toddlers and their new-born siblings. Researchers started to include
younger children in their samples, to investigate the emergence of jealousy. One of the first studies that investigated emotional behavior of infants was conducted by Bridges (1932). She observed the emotional behavior of a group of infants (ranging between 1-month-olds to toddlers of over 2 years) for a couple of months. She found that toddlers between 15 to 18-months of age started to show signs of jealous behavior. These children seemed to be very upset when a well-known adult shifted his/her attention to other children in the group. She also observed that some jealous children reacted aggressively when another child received more attention. Another study that gave insight in infant jealousy was conducted by Masciuch and Kienapple (1993). They conducted a study in which they included children between 4 months and 7 years of age. Their study consisted of two jealousy evoking conditions. In the first jealousy-evoking condition the mother had to cuddle the control child and ignore her own child. During the second jealousy-evoking condition mother was instructed to place the
control child on her lap and read a story to this child, while ignoring her own child. The outcome of the study was that all children showed signs of jealousy, however children aged between 1.1 and 4.5 years of age showed significantly more jealous behavior than children between 4.5 and 8.4 months of age. Another finding was that the jealous behaviors shown by children aged between 1.1 and 2.3 years were more intense, than the expressions of jealousy of the younger children in the sample. The findings of the studies conducted by Bridges (1932) and Masciuch and Kienapple (1993) suggest that jealousy did not emerge until toddlerhood.
However, there are also studies that did conclude that young infants are capable of experiencing jealousy. One of these studies was conducted by Hart et al. (1998a). They found that 12-month-olds were capable of a primitive form of jealousy, which is often called
jealousy protest. In this study mothers were instructed to either hold a storybook or a life-like doll, the latter represents a social rival. While holding these objects, mothers had to ignore their infants. During the condition in which the mother was instructed to hold to doll, infants showed increased protest, played less and demonstrated more negative vocalizations. Another study conducted by Hart and Carrington (2002) showed that even younger infants are capable of experiencing jealousy. They found that 6-month-olds showed more negative facial affect during the condition in which the mother was instructed to hold the social rival (doll
condition), than during the book condition. Subsequent research showed that 6-month-olds responded to the jealousy-evoking condition with increased anger, reactions of intense negative emotionality, and decreased demonstrations of joy. It was also found that when infant’s reactions to the jealousy-evoking situation were compared to the reactions of the
infants to the still-face procedure, infants showed more sadness during the jealousy-evoking situation (Hart, Carrington, Tronick & Carroll, 2004).
All of these findings by Hart and her colleagues are in contradiction with the findings of Bridges (1932) and Masciuich and Kienapple (1993). When comparing these two studies with the studies conducted by Hart and colleagues, some differences become apparent. The first difference concerns the methods used to measure jealousy. Bridges used naturalistic observations in a hospital to measure emotions in infancy (including jealousy), whereas Hart and Masciuch and Kienapple (1993) used experimental settings. Bridges (1993) measured jealousy while a group of 15- to 24-month-old children were playing together, while Hart and Masciuch and Kienapple (1993) both used an experiment that was designed to measure child jealousy. The second difference that might have resulted in opposite findings concerns the social rivals used in the studies. Hart and colleagues used life-like dolls to represent the social rival and these dolls were always younger than the jealous infant. In the study conducted by Masciuch and Kienapple (1993) the social rival was represented by other children that participated in their study. In one of their experimental conditions the social rival was represented by a younger child than the jealous infant. However, in their other experimental condition a child of the same age and sex as the jealous infant represented the social rival. These differences in how these researchers measured jealousy might have resulted in opposite findings.
Another important finding in the field of infant jealousy was that infants reacted to the jealousy-evoking condition with more approach-style behaviors (Mize & Jones, 2012; Hart et al., 2004), including an increase in maternal gaze, an increase in interest and limited
distancing from mother during the jealousy evoking condition (Hart et al., 2004). Mize and Jones (2012) conducted a study in infant jealousy, in which they included a physiological measure (EEG) and maternal reported temperament. The goal of the study was to replicate and extent the findings of Hart and colleagues, using another sample. Especially, the researchers wanted to extent to findings of approach-style behaviors. The
jealousy-inducement condition was similar to the procedures of Hart and colleagues. A life-like doll represented the social rival and in the other condition mothers were instructed to read a book. During both conditions mothers had to ignore their infants. The results of the study replicated findings by Hart and colleagues. Infants showed more negativity and sadness during the social-rival condition, than during the book-condition. Another finding of the study was that infants responded with more approach style behaviors (more mother-directed gaze, proximity and touch) to the jealousy inducement condition, as compared to the book-condition. This
means that infants responded to the potential loss of attention from their mothers with behaviors that seek their mothers’ proximity. It seems only logical that infants respond with approach behaviors, since infants rely on their caregivers for survival and emotional support. These approach behaviors are used to ensure that the attention of their caregiver is not lost to a social rival (Mize & Jones, 2012). Also, these approach behaviors suggest that infants are jealous of the social rival during the jealousy inducement condition. Infants show these behaviors in order to regain their mother’s exclusive attention. Mize and Jones (2012) also found higher reactivity levels, increased arousal levels, more aggression and more anxiety during the jealousy-inducement condition, than during the book-condition. Another finding of the study was that there was a significant association between approach-style behaviors shown in the eliciting condition and greater relative left frontal tonic EEG activity. Finally, a significant relationship between jealousy profile (high versus low jealousy) and tonic frontal left EEG activity was also discovered.
Maternal sensitivity
One of the constructs that has been marked as a possible predictor of infant jealousy is maternal sensitivity. Mary Ainsworth is the founder of the concept maternal sensitivity. Maternal sensitivity refers to the ability of the mother to observe the signals that her child sends, give a correct interpretation to these signals, and give an adequate and prompt response to them (Ainsworth, 1974 as described in Mesman & Emmen, 2013). Sensitivity is considered as the most important predictor of a secure attachment relationship. This means that children of more sensitive mothers more often have a secure relationship with their mother. A meta-analysis conducted by de Wolff & van IJzendoorn (1997) confirmed that sensitivity is indeed the most important predictor of sensitivity. Another finding from this meta-analysis was that the strength of the relationship between sensitivity and attachment security was only modest, which is not in line with the findings reported by Ainsworth and her colleagues. They
reported a relatively strong relationship between the two constructs (Ainsworth, Blehar, Waters & Wall, 1978 as described in McElwain & Booth- LaForce, 2006). Despite this difference in the strength of the relationship, sensitivity is seen as the most important predictor of attachment security.
Maternal sensitivity and infant jealousy
There are only a few studies that investigated whether maternal sensitivity is related to infant jealousy. Hart et al. (2004) found that 6-month-old infants of more sensitive mothers showed
more anger during the jealousy evoking situation. Another study found that 10-month-old infants of more sensitive mothers showed less distress during the jealousy evoking situation, than infants of less sensitive mothers (Hart & Behrens, 2013). These two findings seem contradictory. According to Hart and Behrens (2013) a possible explanation can be found in a difference in physical closeness to mother during the eliciting condition. Infants in the study conducted by Hart and Behrens (2013) were able to crawl during the eliciting condition and were able to be physically close to their mothers. This was not the case for the infants of the study conducted by Hart et al. (2004). These infants were not mobile yet and placed in a high chair during the eliciting context, enabling them to be physically close to their mothers. Hart and Behrens state that this is an important difference. Previous studies showed that infants, who were able to have physical contact with their mothers, were less distressed by the emotional unavailability of their mothers (Stack & Muir, 1990 as described by Hart & Behrens, 2013). Therefore more research is needed to confirm whether the difference in physical contact can explain the contradicting findings.
Another indication that maternal sensitivity is related to infant jealousy is provided by studies which include depressed mothers in their sample. Since maternal depression is known to be related to low maternal sensitivity (Campbell, Matestic, von Stauffenberg, Mohan & Kirchner, 2007; Milgrom, Westley & Gemmill, 2004; Murray, Fiori-Cowley, Hooper & Cooper, 1996), these studies can provide insight in the relationship between maternal sensitivity and infant jealousy. Hart et al. (1998b) found that 1-year-old infants of depressed mothers, in comparison to infants of non-depressed mothers, showed less protest in the jealousy evoking condition. These infants also touched mother less often, showed less
proximity to mother and gazed less often towards their mother. This may indicate that infants of less sensitive mothers (or depressed mothers) showed less approach behaviors. These approach behaviors have proven to be related to infant jealousy (Mize & Jones, 2012; Hart et al., 2004). The findings of the study by Hart et al. (1998b) might suggest that infants of less sensitive mothers show less jealousy protest, however there are also findings that contradict this suggestion. For example, the finding of Hart et al. (2004) that 6-month-old infants of less sensitive mothers showed more anger during the evoking condition. More research is needed to determine whether infants of less sensitive mothers indeed experience less jealousy. At this moment, there is not enough information to draw a valid conclusion regarding this issue.
Finally, there are studies that focus on the differences in jealousy expressions of infants that have mothers with depression that show withdrawn behavior, and the jealousy expressions of infants with depressed mothers that show intrusive behavior. These studies also
help to demonstrate that maternal sensitivity is an important predictor of infant jealousy. Besides emphasizing the importance of maternal sensitivity, these studies also provide insight in how different types of insensitive behavior (intrusive of withdrawn behavior) can lead to different responses in the jealousy evoking condition. Hart, Aaron Jones, Field & Lundy (1999) investigated the differences in jealousy responses in infants of intrusive or withdrawn mothers. They found that, when compared to the responses of infants of intrusive mothers, infants of withdrawn mothers showed more toy manipulation. They also found that infants of intrusive mothers showed more negative and positive affect. Hart et al. (2003) also
investigated the differences in jealousy responses of infants of intrusive and withdrawn mothers. This study found that infants of intrusive mothers responded to the jealousy evoking condition by less often touching their mother and also less often looked at their mother. They also found that infants of intrusive mothers played more during the jealousy evoking
condition. Another finding of the study was that infants of withdrawn mothers, as compared to infants of intrusive mothers, touched the stranger more often during the jealousy evoking condition. When taking the responses from both groups of infants into account, it becomes clear that there are large differences in the behaviors shown by the infants. The group of infants of withdrawn mothers show approach behaviors, however these behaviors are directed to the stranger instead of the mother. Whereas the group of infants of intrusive mothers showed behaviors that are associated with avoidance. This is in line with previous studies that suggested that avoidance is a characteristic of infants of intrusive mothers (Cicchetti & Toth, 1998; Hart, Aaron Jones & Field, 2003).
When comparing the behaviors shown by infants of intrusive and withdrawn mothers it also becomes apparent that there is a contradiction in the findings of Hart et al. (1999) and Hart et al. (2003). Whereas Hart et al. (2003) found that infants of intrusive mothers played more, Hart et al. (1999) found that infants of withdrawn mothers showed more toy
manipulation. This means that infants of mothers with depression that show withdrawn and intrusive behavior respond differently to the jealousy inducement situation. Their reactions might suggest that these infants show less jealousy protest, since some previous studies with infants of non-depressed mothers have found that jealousy protest is mainly characterized by approach-style behaviors directed at mother (Mize & Jones, 2012). However more
information is needed in order to draw valid conclusions. Future studies should focus more on the differences in jealous behavior of infants of depressed mothers that show intrusive and withdrawn behavior. Also, future research should focus on the contradiction in findings in this field. This goes beyond the scope of the present study.
Infant temperament
Infant temperament has received some attention within the field of infant jealousy research. It has been suggested that infant temperament might be an important predictor of infant jealousy protest. Infant temperament is a construct that has received a lot of attention from researchers for many years now. Although infant temperament is often studied, there is still some
discussion about the correct definition of the construct. However, most researchers do agree that variations in temperament are differences in styles of behavior. These styles of behavior become apparent early in life and are a product of neurological factors. These behavioral styles are also known to reflect a predictable pattern of response (Rutter, 1987; Thomas & Chess, 1977 as described in Fox & Henderson, 1999). It is also suggested that variations in infant temperament form the heart of a child’s identity (Fox & Henderson, 1999).
Infant difficultness
There are different ways of determining an infant’s temperament. One often-used way of measuring infant temperament is determining infant difficultness. Infants’ behavior is rated (often by their own parents) on several domains and the position of the infant on all of these domains combined, determines an infant’s difficultness. Two researchers that used this way of measuring infant temperament were Thomas and Chess (Goldsmith, Buss, Plomin, Rothbart, Thomas, Chess, Hinde & McCall, 1987). They proposed nine dimensions, which in their view, were needed to determine infant difficultness. These dimensions were activity, rhythmicity, approach-withdrawal, adaptability, intensity of reaction, threshold of response, quality of mood, distractibility and attention-span persistence (Thomas, Chess & Birch, 1968; Thomas, Chess, Birch, Hertzig & Korn, 1963, as described in Rowe & Plomin, 1977).
Thomas and Chess noticed that there were three combinations of scores on the nine
dimensions that occurred often. They introduced three types of infant temperament to describe these often occurring combinations. The names of these three types are easy, difficult and slow-to-warm-up temperament. Infants with an easy temperament do not have problems with adapting, approach novelty and score high on regularity. Difficult temperament is
characterized by body functions that are irregular, withdraw in response to novelty, show extreme responses to stimulation, show a lot of negative affect and have problems with adjusting to changes in the surroundings. Infants are considered to have a slow-to-warm-up temperament if they need some time to adjust, if they withdraw themselves from novelty, if they have a low activity level and if they have a mood that is somewhat negative. These three temperamental types are often used in research, especially the difficult type (Fox &
Henderson, 1999). This type has received more attention from researchers, because it has been proposed that infants with a more difficult temperament are at risk for developing behavior problems later in their childhood (Bates, Bennet Freeland & Lounsbury, 1979).
Temperament and jealousy
There are only a few studies that investigated whether temperament can be a predictor of jealousy. Bauminger, Chomsky-Smolkin & Orbach-Caspi (2008) conducted a study to investigate jealousy in typically developing pre-schoolers and pre-schoolers with autism spectrum disorder (ASD). They found that typically developing pre-schoolers who received a high score on emotional responsiveness showed more jealous behaviors within the jealousy evoking condition, than typically developing pre-schoolers that had lower scores on emotional responsiveness. Volling, McElwain & Miller (2002) also found that temperament can be a predictor of jealousy. Their study focused on sibling jealousy and the relationships this construct has with several characteristics of the family and the child. The study revealed a positive relationship between temperamental anger and jealousy. However, this relationship between anger and jealousy was only found for the younger sibling in the mother-session. Another study that investigated whether temperament is related to jealousy is conducted by Mize and Jones (2012). They found that several components of infant temperament, as rated by mother, were related to jealousy. They found a positively moderate relationship between jealousy and activity levels and also between jealousy and distress to limitations. The smiling and laughter index was found to be related to infant jealousy in the opposite direction, namely a negatively moderate relationship was found. Finally, Hart and Behrens (2013) found that infants who scored low on positive emotionality showed more distress during the jealousy evoking condition, than infants who received a high score on positive emotionality.
Differential susceptibility
The term differential susceptibility refers to the idea that children differ in their susceptibility to rearing experiences. Not all children within one family are equally susceptible to beneficial and unfavourable parenting influences. It has been proposed that this difference in
susceptibility is the result of evolution (Belsky, 1997; Belsky, 2005). Evolution theory states that humans need to have different characteristics in order to ensure a person’s success in reproduction (Gilissen, Bakermans-Kranenburg, van IJzendoorn & van der Veer, 2007). Since the future is uncertain, parents cannot know which styles of parenting would benefit the survival of their offspring the most. Raising children that differ in their susceptibility to
parenting would result in the survival of at least some of the offspring (Belsky, 1997; Belsky, 2005). In other words, the fact that parents would benefit from raising children with
differences in susceptibility to parenting is a result of natural selection (Belksy, 2005).
Differential susceptibility to unfavorable and beneficial parenting experiences and temperament
Several studies have focused on temperament as a marker of differential susceptibility. Many studies have provided evidence for the differential susceptibility hypothesis. Belsky (2005) reviewed all relevant studies conducted at that point. He found that the relationship between parenting influences and several behavioral outcomes were stronger for children with a more difficult temperament. There are several studies that confirm this finding, especially for children with more difficult temperament. One of these studies is conducted by Bradley and Corwyn (2008). They used the data of the longitudinal NICHD study and found evidence for the differential susceptibility hypothesis. Within in their study there was a stronger
relationship between parenting and externalizing problem behavior in first grade for children with a more difficult temperament, than for children with an easy or average temperament. The study also revealed that for children with an easy temperament there was hardly any relationship between the three parenting variables (harsh parenting, maternal sensitivity and productive activity) and the level of externalizing problem behavior these children showed in the classroom. Dopkins-Stright, Cranley-Gallager and Kelley (2008) also used the data of the NICHD study to test the differential susceptibility hypothesis. The results of the study showed a stronger relationship between maternal parenting and the adjustment of the children in first grade, for children with a more difficult temperament. This relationship was less strong for children with a more easy temperament. This indicates that this study found proof for the differential susceptibility hypothesis. Another study that found proof for this hypothesis was conducted by Klein Velderman, Bakermans-Kranenburg, Juffer & van IJzendoorn (2006). They conducted an intervention-study and the goal of this study was to increase maternal sensitivity. The results of the study showed that infants with a more difficult temperament (highly reactive) and their mothers benefited most from the intervention. Mothers of infants with a more difficult temperament showed a higher increase in maternal sensitivity than mothers of infants with an easy temperament (less reactive). Furthermore, it was found that there was only a significant association between attachment security and mother’s increase in sensitivity (from pre-test to post-test) for mothers of infants with a more difficult
receptive to their mothers’ change (or increase) in sensitivity, than infants with a more easy temperament.
Differential susceptibility and jealousy
Although there are some studies that found associations between sensitivity and jealousy and between temperament and jealousy, there are no studies that investigated the differential susceptibility hypothesis. The present study will, to my knowledge, be the first study that will test whether temperament is a moderator in the relationship between parenting and jealousy. More specifically, it will be tested whether temperament moderates the relationship between maternal sensitivity and infant jealousy protest. If the association between maternal sensitivity and infant jealousy protest is stronger for infants with difficult temperament, this is proof for the differential susceptibility hypothesis.
Present study
The aim of the current study is three-folded. The first aim of the study is to investigate whether maternal sensitivity and infant temperament are predictors of infant jealousy. Previous studies (Mize & Jones, 2012; Hart & Behrens, 2013) suggested that these two constructs can predict infant jealousy. The second aim of the study is determine which of the two constructs is the strongest predictor. The third aim of the study concerns testing the differential susceptibility hypothesis. It will be tested whether infant temperament can
moderate the relationship between maternal sensitivity and infant jealousy protest. This means that it will be investigated if the relationship between maternal sensitivity and infant jealousy is stronger for infants with a more difficult temperament.
The first hypothesis of the present study is that infant temperament and maternal sensitivity are both predictors of infant jealousy protest. This hypothesis is based on previous findings (Mize & Jones, 2012; Hart & Behrens, 2013). Regarding the question which of the two predictors is the strongest, no clear hypothesis was formulated. The reason for this decision is that there are no studies that have focused specifically on these two predictors. Next, the present study will, to my knowledge, be the first study to test whether infant temperament can moderate the relationship between maternal sensitivity and infant jealousy protest. However, several studies have found proof for the differential susceptibility
hypothesis, with temperament as the moderator variable (Belsky, 2005; Bradley & Corwyn, 2008; Dopkins-Stright et al., 2008; Klein Velderman et al., 2006). Therefore the final hypothesis of the present study is that infant temperament will moderate the relationship
between maternal sensitivity and infant jealousy protest.
Method Participants
The sample for the current study consisted of forty-eight 10-month-old infants (24 boys, 24 girls) and their mothers. Infants were on average 44 weeks old (SD = 0.75) at the date of testing. Most children had the Dutch nationality (83%), as had most of their mothers (88%). Maternal ages ranged between 25 and 42 years, with an average age of 32 years at the testing date (M = 32.04, SD = 3.94).
The mother-infant pairs were participating in a short-term longitudinal study designed to investigate social information processing in infancy. The study consisted of three visits. Mother-infant pairs were invited to come to the laboratory when the infants were
approximately 7, 10 and 12 months old. The focus of the present paper will be on the 10-month visit. Participants were recruited by sending letters to all families with young infants that lived in Leiden. Families who were willing to participate in research could register themselves. All registered families were entered in a data-base. When the infants were approximately 7 months of age, their mothers received a phone call to invite them for the study. Visits were scheduled at convenient times for the mother-infant pairs.
Procedure
Mother-infant pairs were invited to come to the laboratory when the infants were
approximately 10 months old. After their arrival, mothers received instructions about the procedures and were asked to give their informed consent. Infants first participated in a short EEG-experiment, which was not used in the current study and will therefore not be explained in further detail. Next, the mother-infant pairs had a short break. This break took place in a room which contained a chair, an infant seat and a small table. Two cameras, positioned in the diagonal of the room, videotaped the break. After approximately 5 minutes, the female
experimenter entered the room and gave instructions to the mother about the free-play
episodes. The free play episodes consisted of 10 minutes free play. During the first 5 minutes mothers were instructed to play with their infants without the presence of any toys and during the last 5 minutes the mother-infant pairs could play with a large variety of toys. Next, the jealousy-evoking situation occurred, which consisted of three episodes. The experimenter entered the room, holding a car-seat with a life-like doll in it. This doll was a RealCare Baby Infant simulator and resembled a 6-week-old infant. Mothers were instructed to care for the
doll for 10 minutes, with their own infants and the toys still present in the room. Mothers were also told that the doll had a very sensitive neck and, like a real infant, the doll could start crying. The experimenter gently handed over the doll to the mother and left the room. This is the start of the first of the three episodes, which is called the pre-cry episode. This episode ends when the doll starts to cry. The second episode (cry-episode) begins when the doll starts to cry and ends when the doll stops crying. The doll was programmed to cry for 5 minutes. The doll did not cry continuously for 5 minutes. There were short breaks in between the cries of the doll, meaning that the doll was quiet for a few seconds and then started crying again. This sequence repeated itself until the doll had cried for five minutes. After the doll stops crying the last episode starts, which is called the post-cry episode. This final episodes ends when the experimenter enters the room, which was the end of the eliciting condition. After the jealousy evoking condition the second visit ends. Infants received a diploma and a small gift for their participation. A few days after their visit, mothers received an email with a link to several online questionnaires including questionnaires about their background, infant
temperament and maternal depression. Two months later, when the infants were 12 months of age, the third visit took place.
Measures
Infant temperament
The Infant Characteristics Questionnaire (ICQ) was used to measure infant temperament (Bates et al., 1979). Mothers completed a Dutch version of the ICQ (Kohnstamm, 1984). This version contains 33 items, which is one item more than the original version designed by Bates and colleagues. The additional item concerns overall difficultness of the infant, as perceived by parents. This item was rated on a 7-point scale, ranging from 0 very easy to 6 very difficult. There is a second item that is rated on the same 7-point scale, this item concerns overall difficultness of the infant as perceived by other parents. The remaining 31 items are rated on a 5-point scale, ranging from 0 not true to 4 true.
The Dutch version of the ICQ consists of four scales, namely crying- easy/difficult (7 items), persistence (3 items), adaptability (4 items) and cuddly (3 items). The present study will only use the difficultness scale. An example of an item that belongs to this scale concerns the reaction of the infant to getting dressed by mother. Also, items concerning how often an infant gets agitated or cries and the reaction of an infant to new toys belong to this scale (Kohnstamm, 1984). Before Cronbach’s alpha of the difficultness-scale could be computed, several items needed to be reversed (items 1,2,5,6,9,10,13,14,15,17,23,25,27,29). Next, the 7
items belonging to the difficultness-scale (items 5,12,20,22,23,31,32,33) were summed. Finally, Cronbach’s alpha was computed and was found to be satisfactory (α = .76).
Maternal sensitivity
Maternal sensitivity was coded during the break and both free-play episodes. The episodes were coded by three coders. The Ainsworth scale sensitivity was used to code maternal sensitivity (Ainsworth, Bell & Stayton, 1974, as described in Mesman & Emmen, 2013). In this coding system only the odd numbers are defined. The highest score of the sensitivity scale that can be assigned to the mother is 9, which means that she is highly sensitive. The lowest score that a mother can receive is 1, meaning that she is highly insensitive.
In order to determine the inter-rater reliability, three coders independently scored 20 cases. All coders scored both maternal sensitivity for each of the three episodes (break, play without toys, play with toys). Besides the scores for the episodes, an overall score and an average score for both constructs were computed. The present study will only use the overall sensitivity score. Inter-rater reliability was computed for each pair of coders separately. The inter-rater reliability score for the overall maternal sensitivity was 0.81.
Jealousy
An adapted version of the coding system developed by Hart and Behrens (2013) was used to code jealousy. Within the present study jealousy was measured during a free play situation. During this situation mothers had to take care of a doll, while her own infant was also present. The original coding system was designed by Hart and Behrens (2013) to code jealousy in toddlerhood. The coding system needed to be adapted to the repertoire of jealous behaviors that 10-month-olds can show. In the adapted version of the coding system, the maternal and infant behaviors were coded by two separate coders. All behaviors were coded every 15 seconds and behaviors could be either present or not present.
Maternal behaviors
Three maternal behaviors were coded in the adapted version of the coding system. Maternal
attention was the first maternal behavior that was coded. Four different categories could be
chosen, namely attention to the own infant, attention to the doll, attention to both the infant and the doll and attention to neither of the two. The categories were mutually exclusive, meaning that only one category could be chosen per 15-second episode. Mothers needed to show at least two occasional behaviors to the infant or the doll in order to code this category.
Mothers could also show continuous behavior towards the doll or the infant and this would also lead to coding one of the four categories. If mother did not show two occasional behaviors to the infant, the doll or to both, the category neither was coded. The second maternal behavior that was coded was whether mothers were trying to draw their infant’s
attention towards to the doll. Mothers could verbally and physically draw their infant’s
attention towards the doll. Verbal behaviors that were coded as drawing attention to the doll, were comments that for example included an invitation for the infant to touch the doll. Mothers could also physically draw their infant behavior towards the doll, by taking their infants hand and forcing the infant to touch the doll. The last maternal behavior that was coded, was criticizing the infant’s behavior towards the doll. This included comments such as “careful", "do not poke the baby" or "be gentle".
Infant behaviors
There were two main categories of infant behavior that were coded, namely doll-directed
behavior and externalizing/under-regulated behavior. Three main categories were formed for
the doll directed-behavior category, namely positive, negative and exploratory behavior. Infants could show more than one of these behaviors in one 15-second episode. However, there can also be infants who do not show any behavior towards the doll. These infants did not receive any scores in the doll-directed-behavior category. In the externalizing/under-regulated category, five different behaviors were coded, namely aggression towards mother,
aggression towards self, acute agitation, destructive play and attention seeking. These
behaviors could all be present or not present during an episode. Aggression towards mother concerned all the hostile behaviors towards mother, including hitting, biting and throwing toys. Aggression towards self was coded when infants attacked or tried to attack itself. Infants showed acute agitation if they screamed, had temper tantrums or threw themselves on the floor. Destructive play was coded when infants were throwing or banging toys during the jealousy-evoking situation. Attention seeking was the last infant behavior that was included.
Attention seeking is characterized by behaviors aiming at regaining mother’s attention. This
included infants showing toys to mother, making loud noises or whining.
In order to determine the inter-rater reliability for the maternal and the infant behaviors, both coders scored 15 cases. For reliability an additional coder scored both maternal and infant behaviors, whereas the other two coders only coded one of the two categories. The inter-rater reliability scores for the infant behaviors were sufficient. For externalizing behavior the score was 0.82 (ICC, single measures, absolute agreement). The
reliability score of the infant’s exploratory behavior towards the doll was 0.90. Reliability scores for the maternal behaviors were also computed. For the maternal attention variable the reliability scores were satisfactory (Attention to own infant: 0.73, attention to the doll: 0.75 and attention to both the infant and the doll: 0.77). For mother’s attempts to draw the infant’s attention towards the doll, the score was 0.79. Finally, the score for the variable criticizing infant’s behavior towards the doll was also sufficient (inter-rater reliability: 0.89).
For the further analyses not all maternal and infant behaviors were used. For the maternal behaviors, only the maternal attention variable (attention to child, doll, both or neither) was used. Of all infant behaviors, only the behaviors included in the
externalizing/under-regulated category were used. These behaviors were summed into one
externalizing category. Also, separate sum-scores were created in order to determine the amount of externalizing behaviour before the doll started crying, while the doll was crying and after the doll started crying. Sum-scores were also created for the overall amount of jealous behavior, the jealous behavior shown in the cry and post-cry episodes combined and the average amount of jealous behavior of the entire episode (pre-cry, cry and post-cry).
Analytic plan
In order to investigate whether infant temperament and maternal sensitivity are predictors of infant jealousy protest, several steps need to be taken. First, it will be investigated whether infant temperament and maternal sensitivity are related to infant jealousy. Correlations between maternal sensitivity and jealousy and between infant temperament and jealousy will be used for this purpose. Next, regression analyses will be computed to investigate the
multivariate association between infant temperament and maternal sensitivity on the one hand and infant jealousy on the other.
In order to investigate whether infant temperament is a moderator in the relationship between maternal sensitivity and infant jealousy, a hierarchical regression analysis with two blocks will be used. Infant jealousy will be the dependent variable in this analysis. Before actually computing the hierarchical regression analysis there are three variables that need to be centered. Centering means that the variables will be transformed into deviations from their own mean. It serves as a way to prevent multicollinearity (Frazier, Tix & Barron, 2004). The variables that need to be centered are the independent variable (maternal sensitivity) and the moderator (infant temperament). After centering, the first step of the hierarchical regression can be computed. In this first block, only the centered independent variable and the centered moderator are included. These are the maternal sensitivity and the infant temperament score.
The second and last block does not only include the centered maternal sensitivity and infant temperament score, but also includes the interaction. If this interaction is significant, it means that infant temperament serves as a moderator in the relationship between maternal sensitivity and infant jealousy.
Results
Data inspection
Data inspection was performed to have a better overview of the data. Within the sample of 48 mother-infant pairs, no missing scores were found. However, for three participants the
temperament scores of the last visit of the study (12 months) were used. These participants were missing the temperament scores at 10 months, but had available temperament-data of the last visit (12 months). All analyses were also computed without these three participants, but no differences in results were found. Therefore it was decided to only include the 12-month data for these participants.
Furthermore, it was investigated if there were outliers for one of the predictors or the outcome. In order to detect outliers, all values of the predictors and the outcome were transformed into z-scores. All z-scores below -3.29 and above 3.29 were considered outliers. No outliers were found for the maternal sensitivity variable and the difficultness-scale. All the sum-scores for the jealousy variable contained one outlier. These outlying scores were
winsorized. This means that the participants with the outlying scores received a score of one value above or below the score of the participant with the second highest or lowest score. An advantage of this procedure is that the order of scores is still intact.
Preliminary analyses
Infant jealousy
First, the distributions of all the infant jealousy scores were checked. None of the sum-scores had a normal distribution, due to the low amount of externalizing behavior shown by the infants. Therefore a logarithmic transformation was applied to the externalizing scores during the cry-episode, since this episode was considered to be the most important (see next page). After the transformation, the externalizing data still did not show a normal distribution. Based on this finding, it was decided to use the original externalizing scores of the
cry-episode.
Next, the relationships between the sum-scores were investigated using correlations. A significant correlation was found between externalizing behavior in the cry episode and
overall externalizing behavior (r = .93, p < .001). There was also a significant correlation between the sum-score of both the cry and post-cry episode and overall externalizing behavior (r =.99, p < .001). A significant but somewhat lower correlation was found between
externalizing behavior shown after the doll stopped crying and overall externalizing behavior (r =.79, p < .001).
Differences in externalizing behavior between the three episodes (pre-cry, cry and post-cry) were also investigated. A repeated measures analysis of variance (ANOVA) was used for this purpose. The results show that there was a significant difference in externalizing behavior between the three episodes (F(1.76,82.62) = 11.20, p <.001). Post-hoc testing showed that there was a significant difference in externalizing behavior before the doll cried,
M = 0.40, SD = 0.12, and when the doll was crying, M = 2.40, SD = 0.50, (p < .001). There
was also a difference in externalizing behavior between when the doll was crying and after the doll stopped crying, M = 1.23, SD = 0.32, (p = .01). Finally, there was a difference in
externalizing behavior between before the doll was crying and after the doll stopped crying (p = .02). This means that there was a significant increase in externalizing behavior as the doll started crying and a significant decrease in externalizing behavior after the doll stopped crying, which can be seen in Figure 1.
Figure 1. Amount of externalizing behavior before, during and after the doll was crying
It was decided to only use the externalizing behavior that the infants showed while the doll was crying (cry-episode) for further analyses. This decision was based on the finding that the episode in which the doll cried was the most stressful episode for the infant. During this
episode mother was trying to sooth the doll and the infant received less attention than before the doll started crying. This elicits the most externalizing behavior, which can also be seen in Figure 1. Also, there was only a small difference between the correlation of the externalizing behavior shown in both the cry and post-cry episode and overall externalizing behavior (r = .93, p < .001) and the correlation between the correlation between the externalizing behavior shown in the cry-episode and the overall externalizing behavior (r =.99, p < .001).
Finally, it was investigated if mother’s behavior during the jealousy episode was related to the jealous behavior of the infant. This was investigated in order to determine whether mother’s behavior during the jealousy episode could explain a possible association between the child’s jealousy behavior and maternal sensitivity or child temperament. A significant relationship was found between externalizing behavior of the infant and the amount of attention mother gives to both the child and the doll (r = .49, p < .001). It was also investigated if this maternal variable was related to the predictors (sensitivity and
temperament). If this was the case, then the maternal variable should be used as a covariate in further analyses. No significant relationships between the maternal variable and the predictors were found and therefore the maternal variable was not used as a covariate in the upcoming analyses. In addition, no associations between the other maternal variables (draws attention to doll, criticizing child’s behavior towards doll) and any infant jealousy sum-scores were found.
Maternal sensitivity
The data inspection revealed that there were no outliers within the maternal sensitivity variable. The values of the Skewness and Kurtosis were examined and found to be sufficient (Skewness: -0.45, Kurtosis: -0.74). Since the Skewness and Kurtosis were satisfactory; it was decided to use the maternal sensitivity without performing any transformations.
Infant temperament
The difficultness-scale of the ICQ was used in the present study. No outliers were found for the difficultness-scale. Next, a closer look at the distribution of the difficultness-scale was taken. The values of the Skewness (0.20) and Kurtosis (-0.65) were satisfactory and there was no need to use any transformations.
Gender and siblings
In order to investigate if there were differences in externalizing behavior between boys and girls, t-tests were used. The results revealed that there were no differences in externalizing
behavior between boys (M = 2.79, SD = 3.56) and girls (M = 2.00, SD = 3.23), while the doll was crying (t(46) = .79, p = .44). There were also no differences in the overall externalizing behavior (t(46) = -.14, p = .89) between boys (M = 3.92, SD = 4.81) and girls (M = 4.13, SD = 5.68). This indicates that there is no main effect of gender on infant jealousy.
Also, differences in externalizing behavior between infants who did have siblings and did not have siblings at the testing date were examined using a t-test. The results showed that there were no difference in externalizing behavior between children with and without siblings while the doll was crying (t(46) = -.23, p = .82). Also, there were no differences in overall externalizing behavior between infants with and without siblings (t(46) = -.33, p = .74). It can be concluded that there was no main effect of having siblings on infant jealousy.
Maternal sensitivity and infant temperament as predictors of infant jealousy
No significant associations were found between maternal sensitivity and infant jealousy (r = .01, p = .95), and between infant temperament and infant jealousy (r = -.05, p = .72). Next, a regression analysis was performed to test multivariate associations. The results of the
regression analysis showed that maternal sensitivity and infant temperament were not related to infant jealousy, which can be seen in Table 1.
Table 1. Regression table: Dependent variable: infant jealousy during cry episode (N = 48)
Differential susceptibility hypothesis
Finally, the differential susceptibility hypothesis was tested. A block-wise regression analysis was used for this purpose. In the first block only the centered predictors were entered. None of the predictors were significant (Maternal sensitivity: β = .01, t(45) = -.08, p = .94, Infant temperament: β = -.05, t(45) = -.26, p = .72). In the second block the interaction was added to the model. As can be seen in Table 2, this interaction was not significant. This means that
Unstandardized Coeffients Standardized Coeffients t p
Model B Std. Error β
Constant 2.45 4.75 .52 .61
Maternal sensitivity .05 .66 .01 .08 .94
infant temperament was not a moderator in the relationship between maternal sensitivity and infant jealousy and there was no proof of the differential susceptibility hypothesis.
Table 2. Regression table: Dependent variable: infant jealousy during cry episode (N= 48)
Discussion
The present study investigated whether maternal sensitivity and infant temperament could univariately and multivariately predict jealousy in 10-month-old infants. Also, the differential susceptibility hypothesis was tested. This means that it was investigated whether infant temperament could moderate the relationship between maternal sensitivity and infant
jealousy. Results of the study revealed that maternal sensitivity and infant temperament were not related to the jealousy response of 10-month-olds. It was also found that infant
temperament did not moderate the relationship between maternal sensitivity and infant temperament. This indicated that there was no proof for the differential susceptibility hypothesis within the present study.
As mentioned above, the present study did not find significant evidence for any of the research questions. A possible explanation for the lack of significant findings can be the minimal amount of jealous behavior shown by the infants. Because of the minimal amount of jealous behavior, the original coding system designed by Hart and colleagues to measure jealousy in toddlers, needed to be adapted. This coding system consisted of two parts, namely an internalizing and an externalizing part. Because the infants did not show any internalizing behaviors, this part was removed from the adapted version of the coding system. Infants also did not show many externalizing behaviors. This jealousy variable showed skewness within the cry-episode and therefore a logarithmic transformation was applied. However this transformation did not result in a normal distribution. Therefore it was decided to use the
Model B Std. Error t p
Block 2
Constant 2.41 .52 4.68 <.001
Maternal sensitivity (centered) -.05 .68 -.07 .95
Infant temperament (centered) -.04 .09 -.48 .64
Maternal sensitivity x Infant temperament -.08 .11 -.72 .48 -.01
-.07
-.11
Unstandardized Coeffients Standardized Coefficients
original (skewed) jealousy scores. However, a disadvantage of using variables with a skewed distribution is that most statistical tests assume that variables have a normal distribution. If a variable does not have a normal distribution, the results of the statistical tests are less reliable. This also applies to the present study. The results of the present study might have been
significant, when the infants showed more jealous behavior. More jealous behavior might have resulted in a normal distribution. A possible solution that may increase variability in jealous responses in 10-month-olds is to also focus on the intensity of the jealous response instead of only looking at the presence of jealousy.
No association between maternal sensitivity and jealousy in 10-month-olds was found. This might suggest that maternal sensitivity does not have an influence on jealousy in 10-month-olds. Jealousy in infancy might develop disregarding how sensitive mothers respond to their infants signals. It might develop independently of mother’s sensitivity in order to ensure that mother’s exclusive attention is always regained. Infants depend on their mothers for survival and receiving mothers’ exclusive attention is crucial. Therefore jealousy might develop disregarding how sensitive a mother is to ensure that she will care for the infant. Also, there might be other factors that contribute to the development of infant jealousy. A possible explanation for lack of significant findings in the present study can be found in the instructions that mothers received about how they should behave during the jealousy evoking condition. Previous studies (Hart & Behrens, 2013; Hart et al., 2004; Hart et al., 1998b; Mize& Jones, 2012) instructed mothers to ignore their own infants during the jealousy inducement condition, whereas mothers in the present study did not receive this instruction and were allowed to respond to their infants’ behavior. Studies that instructed mothers to ignore their infants during the jealousy evoking condition did find significant relationships between maternal sensitivity and infant jealousy (Hart & Behrens, 2013; Hart et al., 2004; Hart et al., 1998b; Mize & Jones, 2012). Not responding to the infants during the eliciting condition might be more stressful for the infants, which might have thus resulted in a more explicit jealous behavior. However, it might be more realistic not to ignore the own infant during the eliciting condition. Within families with more than one child it seems unnatural to completely ignore one child, when soothing the other. When compared to previous studies that were able to find a significant association between maternal sensitivity and infant jealousy the present study might have been more realistic and this might have caused the differences in findings.
No evidence was found for the hypothesis that infant temperament is related to jealousy protest in 10-month-old infants. This hypothesis was based on findings of previous
studies (Bauminger et al., 2008; Volling et al., 2002; Mize & Jones, 2012; Hart & Behrens, 2013) and therefore it was expected to find comparable results. However, within the present study most infants scored on the lower end of the difficultness-scale. There was not a wide variety in temperament-scores. More variability in the scores could have led to different results regarding infant jealousy. A possible explanation for the lack of a significant
relationship between infant temperament and infant jealousy can be found in using only one scale of the ICQ. Within the present study only the difficultness-scale was used. Thus, including all scales of the ICQ might have led to different results. Finally, two of the studies that found support for the hypothesis that there is a significant relationship between
temperament and jealousy used observational instruments to measure temperament (Bauminger et al., 2008; Hart & Behrens, 2013). Hart and Behrens (2013) used different aspects (Big Simon and Puppet Game) of an observational measure called Laboratory Assessment Battery (Lab-TAB) to measure positive emotionality in 45-week-old infants. Bauminger et al. (2008) used the hurt-knee experimental scenario to measure emotional responsiveness in pre-schoolers. During this scenario parents pretended to bang their knee into a table. Their child’s reaction was evaluated and scored on emotional responsiveness. Using observational instruments to measure infant temperament might lead to a more realistic view of the child’s temperament. It can overcome possible bias that mothers have regarding their child’s temperament. This could also apply to the present study, since mothers reported on their infant’s temperament. Replacing the ICQ by an observational measure within the present study might lead to different results.
The present study expected to find support for the differential susceptibility
hypothesis. However, infant temperament was not a moderator in the relationship between maternal sensitivity and infant jealousy. It is difficult to determine whether the lack of significant findings are due to aspects of the present study or if the differential susceptibility hypothesis does not apply to infant jealousy at 10 months of age. The lack of findings could also be due to the small sample size or inadequate measures of the predictors or the outcome. More information is needed in order to draw a valid conclusion about the differential
susceptibility hypothesis within the infant jealousy field.
Implications
The lack of significant findings within the present study has some practical and theoretical implications. Previous studies did found associations between maternal sensitivity and infant jealousy (Hart & Behrens, 2013; Hart et al., 2004; Hart et al., 1998b; Mize & Jones, 2012)
and between infant temperament and infant jealousy (Bauminger et al., 2008; Mize & Jones, 2012; Hart & Behrens, 2013; Volling et al., 2002). The lack of significant findings within the present study could indicate that the associations with maternal sensitivity and infant
temperament are not be as strong as assumed. Especially, because the present study had more realistic procedures and the doll rival used in the present study more closely resembles a social rival. Therefore the present study does provide new insights in the infant jealousy field, despite the lack of significant findings.
Also, the results of the present study suggest that interventions designed to enhance maternal sensitivity might not result in a decrease of jealous behavior in infants. If infant jealousy develops disregarding maternal sensitivity, increasing maternal behavior would not result in a change in jealous behavior. Also, it was found that infant difficult temperament was not related to the amount of jealous behavior shown by infants. However, it must be noted that infants in the present study scored relatively low on the difficultness-scale. This means that infants within the sample that are considered to have a difficult temperament did not score in the highest ranges of the difficultness-scale.
Limitations
Several limitations of the present study should be mentioned. First, an important limitation of the present study concerns the small sample size. The sample of the present study only included 48 infants. It is difficult to find significant results within studies with a small sample size, due to the small amount of power. This problem could also apply to the present study.
Second, the minimal amount of externalizing behavior shown by the infants is
considered to be a limitation. This resulted in skewness within the jealousy scores during the episode in which the doll was crying. A logarithmic transformation did not result in a normal distribution and therefore the original scores were used for further analyses. However, it might also be the case that the minimal amount of jealous behavior found within the present study is a more realistic reproduction of reality. Within the present study mothers were allowed to respond to their infants during the eliciting condition. This might be a more realistic situation for the infant, than the procedures in previous studies. It might be the case that these procedures used in previous studies caused more jealous behavior than the infants would normally show.
Finally, only the presence of jealousy was measured and not the intensity of the jealous response. This results in the loss of possibly important information about the jealous behavior. Some infants might have shown jealous behavior at a much higher intensity, but are
