A longitudinal perspective: Do the pointing abilities of children relate to
temperament?
Master’s thesis Orthopedagogics Graduate School of Childhood Development and Education
University of Amsterdam Name student: A.M. Sweers Thesis supervisor: drs. E. A. Salvadori (MSc) Second thesis supervisor: mw. dr. C. Colonnesi Amsterdam 3 December 2019
Abstract
A crucial pre-verbal ability used to communicate with others is the declarative pointing gesture, described as an extension of the arm and index finger/hand to share interest in something. Because of its important role for the socio-communicative development, it seems paramount to investigate what factors relate to such referential abilities, for example temperament. Temperament is the biological disposition of individual differences in reactivity and self-regulation. This research was first to examine the relation between the declarative pointing production and comprehension of 60 children, as experimentally observed and measured via maternal and paternal parent-reports, and their temperament, equally measured via parent-reports, in children at 12 and 15 months. Correlation and regression analyses with maternal reports yielded that extraverted children pointed more and longer during positive and during all trials at 12 months. Based on paternal measurement, they pointed more and longer during neutral trials at 15 months. Effortful controlled children, defined by fathers, pointed more and longer during positive trials at 12 months. No relation between declarative
comprehension and temperament and between negative affective children and the declarative production was found. In conclusion, this research with its promising results provides reason to continue exploring the relation between pointing abilities and temperament.
Samenvatting
Een cruciaal pre-verbaal communicatief vermogen is het declaratieve wijzen, omschreven als het uitstrekken van de arm en wijsvinger/hand om interesses te delen. Vanwege de cruciale rol voor de sociaal-communicatieve ontwikkeling, is het belangrijk om te onderzoeken welke factoren gerelateerd zijn aan deze referentiële vaardigheid, zoals temperament. Temperament is een biologische dispositie van individuele verschillen in reactiviteit en zelfregulatie. Dit onderzoek was de eerste in het onderzoeken van de relatie tussen het declaratieve wijzen (zowel productie als begrip) van 60 experimenteel
geobserveerde kinderen en hun temperament, gemeten door moeder- en vaderlijke
ouderrapportages, bij kinderen van 12 en 15 maanden. Correlatie- en regressieanalyses met moederlijke rapportages toonden aan dat extraverte kinderen meer en langer wezen tijdens de positieve trial en alle trials bij 12 maanden. Gebaseerd op de vaderlijke meting, wezen zij meer en langer tijdens de neutrale trials bij 15 maanden. Effortful control kinderen,
gedefinieerd door vaders, wezen meer en langer tijdens de positieve trials bij 12 maanden. Er was geen verband tussen het begrip van wijzen en temperament en tussen negative affective kinderen en het declaratieve wijzen. Concluderend toont dit onderzoek veelbelovende resultaten en biedt het reden voor verder onderzoek om de relatie tussen wijzen en temperament te blijven onderzoeken.
A longitudinal perspective: Do the pointing abilities of children relate to temperament? The socio-communicative development of humans starts long before children express themselves verbally. Pre-verbal infants in their first months of life already engage in social interaction with nonverbal communicative modalities. A child first starts communicating with adults via gaze, vocalizations, and facial expressions when they are just few months old (i.e., emotional communication; Colonnesi, Zijlstra, van der Zande, & Bögels, 2012; Vaughan et al., 2003). Later, typically from 9 months onwards, their nonverbal communicative abilities expand with the use of referential communication, and particularly deictic gestures, such as pointing, showing, offering and giving (Camaioni, Perucchini, Muratori, Parrini, & Cesari, 2003).
One of its most studied modalities is the pointing gesture, defined as the simultaneous extension of the arm and the index finger (i.e., finger pointing; Colonnesi, Rieffe, Koops, & Perucchini, 2008; Tomasello, Carpenter, & Liszkowski, 2007) or the extension of the whole hand (i.e., hand pointing; Cochet & Vauclair, 2010) towards a third entity (e.g., object, person, event). More general, the pointing gesture can be conceptualized as an extension of joint attention (Salo, Rowe, & Reeb‐Sutherland, 2018): the coordination of attention between the child and a social partner towards a third entity (object, event, person; Nichols, Fox, & Mundy, 2005). Specifically, pointing is conceptualized as a higher level of joint attention (Cassel, Messinger, Ibanez, Haltigan, Acosta & Buchman, 2007; Mundy et al., 2003). Higher level behaviors follow later developmental stages compared to the lower level behaviors (such as eye contact or gaze alternation). Moreover, higher level behaviors have communicative intentions. For example, when walking outside, a child can point to the ice cream vendor and can thereby intentionally communicate something to the adult: the child may point to the ice cream vendor because s/he wants an ice cream, or just because the ice cream car is of a peculiar color and the child wants to share this beautiful experience with the adult.
Just as the example above shows, communication via pointing underlies an intention. There exist two different intentional forms and two intentional modalities: imperative or declarative and comprehension or production, respectively (Colonnesi et al., 2008). Imperative pointing refers to when the pointing gesture is used to request an object or an action. Differently, pointing with the purpose to share interest in an object (or event) is known as declarative pointing. Each intention, whether imperative or declarative, has two different modalities: comprehension and production. More specifically, comprehension refers to episodes involving the child’s understanding of the other person’s pointing, while
the different intentions and modalities leads to four subtypes of pointing: imperative
comprehension, imperative production, declarative comprehension and declarative production (Colonnesi et al., 2008).
Different social-cognitive mechanisms underlie these four subtypes of pointing (Camaioni et al., 2003). The declarative abilities require the most social-cognitive awareness, as the declarative pointing gesture implies the understanding of other person’s mental states and intentions (e.g. their attentional state). In this case, the other person is seen as an
intentional or mental agent. By using the declarative pointing gesture the child will influence the attentional state of the other person. Indeed, before doing so the child first needs be aware that s/he can influence the attentional state (Camaioni et al., 2003). The declarative pointing abilities are therefore of interest for this research, with particular attention to the latency as research shows that children point longer, but not more frequent, in the case of declarative pointing compared to imperative pointing (Cochet & Vauclair, 2016). Occasionally, latency is measured by pedagogical researches such as Meltzoff (1995) with his study on understanding of intentions, measuring the time from the moment the toy was presented until the child first touched the toy (latency of a nonverbal reaction). More often, latency is used in linguistic research, especially for children with a more problematic language development (e.g. children with dyslexia or children with ASD) to account for the reaction time between the eye gaze and a target behavior (e.g. reading a sentence) or for a neural reaction in the brain (Bucci, Brémond-Gignac, & Kapoula, 2008; Roberts et al., 2011). Following the linguistic research an innovative perspective to the measurement of preverbal pointing abilities were proposed by adding explorative variables about latency.
The pointing gesture is an important milestone for the further socio-communicative development. The development of the pointing abilities has a positive influence to later language development (Colonnesi et al., 2010), theory of mind and understanding of intentionality (Colonnesi et al., 2008), and self-recognition (Nichols et al., 2005).
Furthermore, when the pointing abilities are impaired, children can be classified as at-risk for developing psychopathologies, such as autism. Research showed that children with Autism Spectrum Disorder (ASD) are impaired in their declarative pointing (Zwaigenbaum et al., 2005).
It is difficult to verify when the development of the pointing abilities exactly happens, since a lot of variety has been observed between children’s onset age of pointing abilities, and the velocity of its development (Camaioni, Perucchini, Bellagamba, & Colonnesi, 2004).
Children first develop the pointing abilities within their own visual range (e.g. around 12 months), followed by the pointing abilities outside of their own visual range (e.g. around 15 months; Butterworth & Grover, 1988, 1990). In addition, the declarative pointing abilities are developed at a later age, as a consequence of the more psychologically complex capacity that is needed fort his specific pointing (Camaioni et al., 2004). Generally, according to Colonnesi, Stams, Koster, and Noom’s (2010) meta-analysis, children typically start pointing at around 11 or 12 months of age, ranging between 7 and 15 months, depending on the child’s
individual dispositions, such as temperament.
It seems crucial to examine which factors contribute to such individual variability in the range of the children’s onset age of the pointing abilities. Temperament, in particular, is an individual disposition that might influence the child’s socio-cognitive development (Rothbart et al., 1994) and more specifically the pointing abilities. Temperament refers to the “constitutionally based individual differences in reactivity and self-regulation” (Rothbart, Ahadi & Hershey, 1994, p. 22). In other words, it is an underlying biological construct, shaped by heredity, maturation and experience that creates differences in arousability of affect, motor activity, related responses, attention, approach-withdrawal, behavioral inhibition and self-soothing (Rothbart et al. 1994). According to Gartstein and Rothbart (2003), three temperament patterns can be distinguished among children: 1) Extraverted children are characterized by their impulsivity, smiling, laughter and positive anticipation. They are in general not shy (Gartstein & Rothbart, 2003), tend to engage in positive affect with others (Vaughan et al., 2003) and like to explore new environments (Schoeps et al., 2018). 2) The negative affective children often experience sadness, discomfort and fear. They are easily scared by new stimuli (Rothbart et al., 1994) and easily upset (Schoeps et al., 2018). When being distressed, they have difficulties with soothing themselves. Moreover, negative affective children are more likely to develop (social) anxiety (Gulley, Hanking, & Young, 2016). 3) The effortful controlled children stand out in there attentional focus and perceptual sensitivity. These children have no problem in focusing on a book or toy for a longer period of time (Schoeps et al., 2018). Their inhibitory control is well developed and they enjoy low intensity activities. In new situations or when meeting new persons, they usually react with distress and/or avoidance (Paulus, Backes, Sander, Weber, & von Gontard, 2015), but they are easily soothed again (Schoeps et al., 2018). On average, boys tend to be more extraverted and girls more effortful controlled (Else-Quest, Hyde, Goldsmith, & Van Hulle, 2006). On the
contrary, no gender effect was ever reported for negative affect(Else-Quest et al., 2006; Farkas & Valloton, 2016).
Previous research reported that several temperamental factors (such as emotional reactivity, self-regulation, agreeableness/sociability and inhibitory control) play a role in the socio-communicative development of children (Vaughan et al., 2003, 2007). The only thing that is known so far is that positive emotional reactivity, which underlines an extraverted temperament (Ripper, Boyes, Clarke, & Hasking, 2018), relates to initiating joint attention episodes at nine months, whereas negative emotional reactivity relates to initiating joint attention episodes three months later, at 12 months (Vaughan et al., 2003). This shows that temperament already influences the broader socio-communicative construct joint attention around the same time children develop the socio-communicative pointing gesture (i.e. around 12 months). Therefore, this research measures temperament at 12 months, and three months later, when the pointing abilities are even more established. This research aims to extend previous literature by investigating for the first time the relation between temperament and pointing gesture development, by longitudinally investigating declarative comprehension and production abilities at 12 and 15 months of age. Specifically, in order to capture a
comprehensive overview of their pointing production abilities, both parental reports
describing the use of children’s pointing in naturalistic contexts were collected, and children were observed with experiment design with different contextual cues. Three different
emotion-inducing stimuli where presented namely presenting a positive, a neutral and
negative stimulus. The present study has implications for further research in the pedagogical, developmental, cognitive and linguistic fields. By examining the influence of temperament in the pointing abilities of children, this research explores a factor (i.e. temperament) that may explain individual variability within the population of children with ASD. As explained in the article of Schwartz and colleagues (2009), the biggest obstacle for researchers and clinicians working with children with ASD is the enormous variability within the population of children with ASD, thereby complicating the development of effective interventions that are adaptive for every individual. It is important to identify what contributes to the wide variability within this population to understand what impacts the course of the psychopathology and the
individual’s response to interventions. If an influence of temperament is proven, children’s temperament factors could be assessed in the future in order to determine which type of screening and/or intervention techniques and settings would be the most effective for every child.
With this in mind, the present research explores the relation between the declarative pointing abilities (i.e. declarative comprehension and declarative production) of 12 and 15-month-old children and their temperament (i.e. extraversion, negative affectivity, effortful control), examining the pointing abilities also in relation to the children's gender. Quite innovatively, this research also adds an explorative variable to the measurement of preverbal pointing abilities, such as latency. Consequently, the pointing abilities are not only measured as frequency (e.g. Camaioni et al., 2004; Colonnesi et al., 2008), but also as latency in terms of its duration, i.e. duration of the pointing gesture (Cochet & Vauclair, 2016), and reaction time, i.e. time needed to start pointing (Meltzoff,1995). Due to the numerous amount of variables already presented within this thesis, the measurement and results of the reaction time are presented separately in the Appendix A. See Figure 1 for an overview of the research questions.
Figure 1. An overview of the research questions.
Firstly, a positive relation between extraverted children and their declarative comprehension and production during all trials was expected, both at 12 and 15 months (hypothesis 1). These children tend to show positive affect towards others (Vaughan et al., 2003), which can be attributed to the development of joint attention and, more specifically, to the development of pointing comprehension. At the same time, a negative relation between negative affective children and their declarative comprehension and production during all trials was predicted, both at 12 and 15 months (hypothesis 2). Negative affective children are more scared for novel stimuli, especially those that provoke fear (Rothbart et al., 1994), and have more often (social) anxiety (Gulley et al., 2016). They have probably less social interactions with others and are therefore likely to be slower in their social development. Furthermore, a negative relation between effortful controlled children and their declarative comprehension and production during all trials was expected, both at 12 and 15 months (hypothesis 3). They usually react with distress and/or avoidance to new situations or persons (Paulus et al., 2015), thereby probably having less opportunities to develop their pointing
abilities. For these three hypotheses, more intense emotional reactions were expected during the positive and negative trials, therefore a stronger relation during these trials were predicted. In contrast, since the neutral stimulus was expected to elicit no specific emotional reaction, as it is a familiar everyday object, a less strong relation was predicted between the different temperament factors and the pointing production during the negative trials. Furthermore, a significant effect of gender was expected, especially for extraverted and effortful controlled children (hypothesis 4). Girls score on average higher on effortful control and boys score on average higher on extraversion (Else-Quest et al., 2006). Lastly, the latency of the children’s pointing comprehension and production was related to their own temperament (hypothesis 5). Because of the explorative nature of this hypothesis, there are no further expectations
regarding the three different types of temperament. Method Participants
The current research was part of an ongoing longitudinal study from the Child Development Department of the University of Amsterdam, called My-BEST (Babies Emotional and Social Trajectories). The sample (N = 60; n = 27 female ) participated in a structured lab-visit at 12 and 15 months of age (first visit: Mage = 12.3 months, SD = 0.53; second visit: Mage =15.2 months, SD = 0.39). All parents were married or lived together at the
moment of the data collection. The vast majority of the parents where highly educated: 57 mothers and 43 fathers followed higher professional education (called HBO in Dutch) or university (WO). At time of the research, most mothers were working part-time (53%), whereas most fathers worked fulltime (63%). Moreover, 73,3% of the mothers and 68,3% of the fathers were born in the Netherlands. All children had at least one older sibling and the child was biologically related to at least one of the primary caretakers. Families were recruited in Amsterdam and surrounding areas, via professionals who work with children (e.g.
daycares, schools, children cafés, etc.) and via social media (Facebook, the My-BEST website and Instagram). In order to participate, parents were asked to fill in a short screening
questionnaire to see if their children met the inclusion criteria. For each lab visit, parents were thanked with a gift card of 15 euros and a small toy for the child. Moreover, travel costs were reimbursed. Parents were allowed to withdraw participation from the study at any time and anonymity was guaranteed. The research was approved by the ethical committee of the Research Institute Child Development and Education of the University of Amsterdam, project number: 2016-CDE-7403.
Research Design and Procedure
The research consisted of a) an online questionnaire for the parents about their
children’s temperament and pointing in naturalistic settings and b) one lab visit at the Family Lab of the University of Amsterdam to observe pointing abilities under experimentally manipulated conditions.
After singing up, the research assistant contacted the families to invite the child together with one of the parents for a lab visit of approximately 30 minutes. During this observation, the child was presented with a large battery of tasks, among which the pointing gestures, central for this study. While the child engaged with the experiment, the parent was invited to fill out the online questionnaire and was instructed not to interact in the experiment unless the child became fussy. The other parent, not present during the lab visit, was asked to fill in the questionnaire at home. Before the lab visit, in an earlier measurement at the
beginning of the study, both parents signed an informed consent form to authorize the child’s participation in the research. During the lab visit, for the pointing tasks the experimenter and child were sitting at a table across each other (for a picture of the setting, see Appendix B, figure B1 and B2). The child was sitting in a high-chair and the parent closely behind him, but out of the child’s sight. By knowing that the parent was close, the child could feel safe, but at the same time experience the least possible distraction. A comprehensive view of the
experimenter, the child and the different stimuli was captured with two cameras. Three experimenters were involved in each lab visit. All experimenters were female. At the
beginning of the lab visit the child played with one of the experimenter for two minutes, so to familiarize to the new environment. After the observation, parents signed a data management form to use the observation for this research and possibly also for Open Science1.
Measurements
Pointing Abilities. To measure the child’s declarative pointing abilities two out of the original four pointing gesture subtasks by Camaioni and colleagues (2004) were used. The two following tasks were presented after each other, starting with the comprehension task. Afterwards, the recorded observations of the comprehension and production tasks were offline micro-coded using the Observer XT, version 13 (Noldus, Trienes, Hendriksen, Jansen, & Jansen, 2000). All the behaviors relevant for this research were coded as state events, which are events measured in duration.
1 Open Science is a scientific global platform where data and videos are saved, managed and shared
Declarative pointing comprehension task. In the declarative comprehension task, the
ability of the child to understand other people’s use of the pointing to share something was measured, by engaging a bike light flashed from ¾ of the child’s view that was directed from the dark room. The task included two trials, whereby the experimenter for each trial for 15 seconds pointed at the stimulus (i.e. the light) to direct the child’s attention. Although in the original task three trials are carried out, in the present study two trials were chosen to avoid any possibility of distress of the child due to the length of the experiment. While doing so, the experimenter shifted her gaze between the child and the stimulus, but was instructed not to use any verbal cue to direct the child’s attention. Declarative pointing comprehension was seen as the child first looking at the experimenter or the experimenter’s pointing gesture and then looking at the stimulus, as this indicated that the child successfully followed the pointing gesture.
In order to measure this target behavior, the trials and the gaze of the child were micro-coded. The time span between when the experimenter started pointing and the moment she stopped pointing was coded and called a trial for the comprehension task. The relevant child’s gaze directions for this research were the pointing gesture of the experimenter, the experimenter’s body or face, and the stimuli. In case the child’s eyes were not visible, it was coded as gaze direction not visible.
After coding the data, a state lag sequential analysis (see Figures C1 and C2 in Appendix C for the data profile) was used to measure the target behavior, i.e. switching gaze from the experimenter, or her pointing gesture, to the stimulus. This analysis can only calculate the results measured in frequency of the declarative comprehension, which is the most used sort of variable in the literature of the pointing abilities (Camaioni et al., 2004; Colonnesi et al., 2008). Due to the broad individual variability in length of the trials (range: 13,52- 48,08 seconds), caused by differences in child’s individual responses (e.g. there was a lot of variation in when the child saw the pointing gesture of the experimenter) and child’s fussiness, corrected values of the frequency variables were calculated. Specifically, for the declarative comprehension task there was one outcome variables: the corrected frequencies.
Declarative pointing production task. The declarative pointing production task used
in the present study is an adapted and extended version of the declarative pointing production task developed by Camaioni and colleagues (2004). Indeed, in the current research, different valences of stimuli were presented to the child. Specifically, besides presenting positive stimuli, negative and neutral stimuli were included as well (see Appendix B, figure B2). For
the positive stimulus, the experimenters used an object that creates an optical effect, with the rationale that the child would be more inclined to react to the stimulus. For the neutral stimuli, they used objects that the child was expected to be familiar with, with the idea that this would elicit no specific interest in the child. Finally, for the negative stimuli, the experimenters used a spider and a snake, based on previous research that these specific animals created a natural fear response in children as young as 6-month-old (Hoehl, Hellmer, Johansson, & Gredebäck, 2017), therefore plausibly eliciting a negative fear reaction as well in the children tested during this study. Order of the stimuli presentation was fully counterbalanced within children’s gender.
Stimuli were presented to children one by one from behind the curtains on the
experimenter’s back, who pretended not to be aware of the presence of it. Every stimulus had two trials of 15 seconds each: it appeared on each side of the experimenter, always starting on the right side from the child’s point of view. Experimenters were trained to provide
communicative cues to elicit children’s communicative behavior when looking at the stimulus (e.g. the spider). If the child pointed towards the stimulus, the experimenter turned around and gave a reaction appropriate to the valence of the stimuli (e.g. a positive reaction for the
positive stimulus).
For this production task, the target behavior was the pointing gesture of the child. In order to measure this target behavior, the trials and the pointing gesture of the child were micro-coded. For the production task, the time span between when the stimulus appeared from the curtain and the moment the stimulus disappeared from the curtain was coded and defined as a trial. The pointing gesture of the child was coded from the moment the child had extended his/her arm and the index finger or whole hand towards the stimulus. Coding was stopped when the child lowered his/her arm or his/her index finger/hand was not pointing anymore to the stimulus. For each pointing gesture that was coded, the sort of pointing (hand or index finger, with an without arm extension) and the direction (target stimulus or non-target stimulus) were specified. The code non-non-target stimulus was used when the child pointed to the place where there used to be a stimulus in the first trial (e.g. positive stimulus), but in the meantime the stimulus changed position for the second trial. All these coded behaviors were used in the present research.
For the declarative production task, the pointing duration and the pointing frequencies of the pointing per child within the two positive, two negative, and two neutral trials, and overall, were exported (see Appendix D, figures D1, D2 and D3). Similarly to the
comprehension task, there was a broad individual variability in the length of the trials for the production task, due to differences in child’s individual responses (e.g. some children saw the pointing gesture of the experimenter later than other children) and child’s fussiness (range: 34,20 - 176,84 seconds). Therefore corrected values of the pointing duration and the pointing frequency were calculated and used in the final analyses. For the production task there were two outcome variables: the corrected frequency of pointing and the proportional duration of pointing.
In addition for every observation of both tasks a selection of ‘yes’ and ‘no’ was coded. Yes was coded when the data was good enough to be included in the data analyses. No was coded for parts of the observation that were excluded for the data analyses. Parts of the data were excluded due to the fussiness of the child (except when it was a reaction to the stimulus; N= 35), wrong position of the child (e.g. child is trying to stand up and needs both hands to find balance making it impossible for him/ her to point; N= 7), lack of attention (e.g. the child did not see the pointing gesture of the experimenter or the stimulus; N= 5) and related too many trials (N= 4), procedural errors (N= 4).
To assess the inter-rater reliability, a random 18% of the total observations were double coded. The average of Cohen’s kappa was consistently high and comparable to the kappa’s reported in Camaioni and colleagues (2004). For declarative comprehension, kappa was kdc = .97 (SD = 0.08) and for declarative production equally kdp = .97 (SD = 0.05).
QPOINT. In addition to the experimental measurement of the pointing production via
the lab observations, the pointing production in a natural setting was measured via the subscale declarative pointing of the QPOINT, based on the parent’s view. This subscale has eight items, e.g. “How often does your child point to an image, picture or illustration?”. The items measure if and how often the child shows the declarative pointing production (but not the comprehension) in a domestic setting on a 3-point scale (never, once or twice, more than two times; Colonnesi et al., 2008; Camaioni et al., 2004). Good reliability and sensitivity to individual variation were reported for the QPOINT (Camaioni et al., 2004). Table 1 displays the parent’s reliability of the QPOINT as measured in the current research. Good reliability was found: Cronbach’s alpha ranged from .80 to .85, in line with reported alpha of Camaioni and colleagues (2004).
Mothers’ and fathers’ QPOINT Internal Consistency (Cronbach’s alpha) and their Pearson Correlation α r QPOINT 12 months Mothers (n =47) ,800 Fathers (n =41) .799 Correlations (n = 36) .429** QPOINT 15 months Mothers (n =44) .848 Fathers (n =35) .805 Correlations (n =31) .386*
Note. The calculated correlation is the correlation between the mothers and the fathers for the declarative production subscale.
* p < .05. ** p < .001.
To see if the questionnaires of mothers and fathers for the QPOINT could be merged being both parental view on child’s temperament, Pearson correlations were calculated. There was a significant correlation between mothers and fathers at 12 months (r = .429, p < .05) and at 15 months (r = .3869, p < .05). However, since it was not possible to merge the
temperament questionnaires (see the explanation at page 16), a decision has been made to not merge the QPOINT in order to keep the structure of the variables consistent.
Temperament. In order to measure temperament, both parents’ reports on their experiences with the children’s behavior were collected, with the Infant Behavior
Questionnaire Revised very short form (IBQ-R) at 12 months, and with the Early Childhood Behavior Questionnaire very short form (ECBQ) at 15 months. These are two separate and equivalent instruments to measure temperament and its development over time. Specifically, the ECBQ is described as an “upward extension” of the revised version of the IBQ-R. For both questionnaires the same three-factors structure is found: extraversion, negative affectivity and effortful control (in IBQ-R described as Orienting/Regulation; Gartstein & Rothbart, 2003; Putnam, Gartstein, & Rothbart, 2006). All three factors were tested in this study and were included in the final analyses.
The IBQ-R very short form has 37 items, e.g. “When being dressed or undressed during the last week, how often did the baby squirm and/or try to roll away?”. The items measure the different temperament dimensions with an 8-point scale (range between 1 = never and 7 = always and a “Does Not Apply” option). Items rated as “Does Not Apply” were treated as missing (Putman, Helbig, Gartstein, Rothbart, & Leerkes, 2014). The items measure three general factors: positive affectivity/surgency, negative emotionality and
orienting/regulatory capacity (Putman et al., 2014). The IBQ-R has been validated for American children between three and 12 months old (Enlow, White, Hails, Cabrera, &
Wright, 2016; Putman et al., 2014) and has a good internal consistency, reliability and validity (Cronbach’s alpha range between 0.70 and 0.90; Enlow et al., 2016). The Cronbach’s alpha of the current research ranged from 0.53 to 0.68 (see Table 2), consistently with the reliability of the study of Farkas and Vallotton (2016) on another American sample. For each subscale, one item has been deleted to create a more acceptable reliability. For more detailed information about the deleted items, please see Table E2 in the Appendix E.
The ECBQ very short form has 36 items, e.g. “When approached by an unfamiliar person in a public place (for example, the grocery store), how often did your child cling to a parent?”. They equally measure the different temperament dimension with an 8-point scale (range between 1= never and 7= always and a “Does Not Apply” option), where the “Does Not Apply” option is treated as missing (Putman, Jacobs, Gartstein, & Rothbart, 2010). The ECBQ is developed for children between 18 and 32 months old (Putman et al., 2010).
However, for children in the age gap between 12 and 18 months, the ECBQ is still advised as the most appropriate temperament measurement2. The items measure three scales similar to the IBQ-R: extraversion, negative emotionality and effortful control. Effortful control is comparable to the scale orienting/regulatory capacity of the IBQ-R very short form. In this research the term effortful control will be used for both the IBQ-R and ECBQ for clarity reasons. Also for the ECBQ, good internal consistency, reliability and validity are reported (Cronbach’s alpha between 0.60 and 0.90; Putnam et al., 2006). The current research found Cronbach’s alpha range between 0.54 and 0.71 with an exception for Negative Affectivity scale of fathers, which had a Cronbach’s alpha of 0.22 (see Table 2). Due to such low reliability, negative affectivity was left out of the data analyses. In order to gain a more
acceptable reliability, one item per subscale of the ECBQ was deleted. For more detailed information about the deleted items, please see Table E2 in the Appendix E.
To see if the questionnaires of mothers and fathers for the IBQ-R and ECBQ could be merged being both parental view on child’s temperament, Pearson correlations were
calculated. Since there was not a significant correlation for every subscale, the questionnaires have not been merged.
Table 2
Mothers’ and fathers’ IBQ-R and ECBQ Internal Consistency (Cronbach’s alpha), original and after item deletion, and their Pearson Correlation
IBQ-R Mothers (n =47) IBQ-R Fathers (n =41) Correlati ons of IBQ (n = 36) ECBQ Mothers (n = 42) ECBQ Fathers (n =34) Correlati ons of ECBQ (n =31) α α r α α r Extraversion 0.443 0.509 0.697 0.537 Extraversion modified 0.526 0.540 .455* 0.711 0.577 .439* Negative Affectivity 0.643 0.616 0.493 0.102a Negative Affectivity modified 0.663 0.684 .373* 0.544 0.224 a .148 Effortful Control 0.573 0.571 0.572 0.568 Effortful Control modified 0.640 0.646 .355* 0.600 0.595 .316
Note. The calculated correlation is the correlation between the mothers and the fathers per subscale.
a subscale has been excluded from the analyses due to the low reliability.
*p < .05.
Statistical Analyses
The data of the Observer and the data of the questionnaires in Qualtrics were exported to SPSS IBM Statistics, version 24. The missing data is presented in Table 3.
Table 3
Of these missing data, observations of 44 children were missing due to no lab visit (12 months: n= 18; 15 months: n = 26). Furthermore, two observations of the declarative comprehension task at 15 months and two observations of the declarative production task (one from each time point) were excluded, because the child was too fussy during the whole task. Finally, one observation of the declarative production task was excluded, because of too much involvement of the parent during the task and thereby biasing the reaction of the child. The data collection for the 15 months is not yet completed, which explains partly the higher rate of missing data at 15 months. However, in order to not lose any informative data, a decision to include all the available data has been made. According to the not statistically significant MCAR-Tests, it can be concluded that the missing data was completely at random (χ2 = 12.731, p > .05 for the pointing variables from the observer, χ2 = 70.639, p > .05 for the temperament questionnaires, and χ2 = 61.764, p > .05 for the pointing questionnaires).
The differences among children in pointing duration and frequency towards the different stimuli (positive, negative and neutral) were tested by running a Friedman test. In case a difference was found, post hoc analyses with Wilcoxon signed-rank tests were conducted to further examine the difference across the trials. A Wilcoxon signed-rank test was as well calculated in order to check for developmental differences for all the variables. Mann-Whitney U tests were run to control for differences between the various orders of the stimulus (positive-negative-neutral and positive-neutral-negative) and to determine the role of gender in the declarative pointing abilities at both ages.
All the variables were checked for outliers via the use of boxplot. In addition, investigations of normality were carried out with the Shapiro-Wilk test, indicating that the IBQ, ECBQ subscales ( except for the Effortful Control for mothers), the QPOINT data at 12 months, and the frequency of the comprehension task at 15 months were normally distributed. Consequently, to examine the relation between the declarative pointing production as
Questionnaires 12 months 15 months
Temperament mothers 21,7% 30,0%
Temperament fathers 31,7% 43,3%
QPoint mothers 21,7% 26,7%
QPoint fathers 31,7% 41,7%
Labvisit 12 months 15 months
Pointing comprehension 23,3% 43,3%
described by both parents and the three temperament factors at 12 months Pearson’s correlation was used. Differently, the ECBQ maternal subscale Effortful Controlled, the QPOINT data at 15 months, the frequency of the comprehension task at 12 months, and all the observed variables (i.e. frequency and duration) of the production task at both time points were not normally distributed, with a general tendency to be positively skewed. Therefore, a Spearman rank-order correlation was calculated to examine the relation between a) the declarative comprehension and the three temperament factors at both ages, b) between the observed declarative pointing production and the three temperament factors at both ages and c) between the declarative pointing production questionnaires and the three temperament factors.
Results Preliminary Analyses
Table 4 shows the descriptive statistics of all the data. Declarative pointing comprehension increased across trials in frequency at 15 months when compared to 12 months. Declarative pointing production increased as well, but only in duration of pointing during the negative trials. This means that the development of pointing abilities improved over time, which is confirmed by the significant increase of the parental measurement of the pointing production. Please see Table F1 in the Appendix F for more statistical information.
When looking at the relation between the declarative comprehension task and the pointing production reported by mothers and fathers, no significant results were found at 12 months (mothers: rs = .231, p > .05; fathers: rs = -.125, p > .05), nor at 15 months (mothers: rs
= .193, p > .05; fathers: rs = -.153, p > .05 ). This means that the pointing production
described by parents does not relate to the child’s understanding of the pointing gesture made by the experimenter.
For the production task, some extra preliminary analyses were carried out to check for the different methods (observation versus questionnaires) and the different stimuli presented to the child (positive, negative and neutral). When comparing declarative production between the two measurements (experimental observation and questionnaires capturing the children’s use of pointing in naturalistic settings), different patterns of results were found for maternal and paternal reports. The pointing production described by mothers were positively correlated with the observed frequency of the pointing production during all trials at 12 months (rs =
.344, p < .05), and more specifically the observed pointing frequency during the negative trials at 12 months (rs = .341, p < .05). This was not found for the other trials described by
mothers at 12 months (rs = .202 for the positive trials, rs = for the neutral trials, p > .05),
neither for the observed duration (range: rs =.171 - .306, p > .05), nor at 15 months for all the
variables (range: rs = .095 - .249, p > .05). In addition, no significant correlation was found
between the pointing production observed in the lab and the questionnaires fathers filled in about the pointing production, neither at 12, nor at 15 months (range: rs = -.207- .337, p >
.05). In other words, when a child showed more pointing, either across trials or within the negative trials during the observation at 12 months, mothers reported equally more pointing for this child. Nonetheless, when a child showed more or longer pointing this was not related to the father’s view on the pointing production of that child at the same point of time. Finally, declarative pointing production reported by mothers were positively correlated with the declarative pointing production reported by fathers both at 12 months (r = .429, p < .05) and at 15 months (rs = .386, p < .05), indicating a consistency in parents’ perception of the child’s
pointing production.
There were significant results for the Friedman test regarding the production task at 12 months, indicating that children showed significant differences in pointing duration between the three sort of trials (positive, negative and neutral), χ2 = 7.786, p < .05 and χ2 = 6.222, p < .05. To examine between which trials the differences occurred, post hoc analyses with
Wilcoxon signed-rank tests were conducted. A significance level of p < .017 was set, because of a Bonferroni correction. Children pointed longer during the positive trial than to the
negative (Z = -2.540, p < .017) and neutral trial (Z = -2.391, p < .017). Between the negative and neutral trial there was not a significant difference in pointing duration, Z = -.432, p > .017. The frequency of pointing across the different sort of trials was not significant, χ2 = 4.786, p > .05. In short, these results show that children pointed the longest during the positive trials compared to the negative and neutral trials at 12 months, but there was no difference in duration of pointing between the negative and neutral trials. Moreover, no difference was found concerning how often children point across the different sort of trials. Table 4
12 months 15 months
Questionnaires data M SD Range M SD Range
Temperament mothers
Extraversion a,b 5.36 0.66 3.80 - 6.58 4.88 0.82 2.70 – 6.86
Negative Affectivity a,b 4.11 0.89 2.18 – 5.90 2.79 0.63 1.55 – 4.18
Effortful Control a,b 5.28 0.63 3.82 – 6.45 4.58 0.63 3.27 – 5.55
Temperament fathers Extraversion c, d 5.02 0.60 3.55 - 6.18 4.76 0.82 2.71 – 7.00 Negative Affectivity c, d 3.91 0.79 2.20 – 5.63 2.65 0.51 1.78 – 3.80 Effortful Control c, d 5.05 0.71 3.60 – 6.55 4.33 0.73 2.89 – 6.00 Q Point Declarative Production Mothers a, e 14.77 4.72 3.00 – 24.00 19.74 4.27 10.00 – 24.00 Fathers c, f 14.46 4.33 7.00 – 24.00 18.54 3.80 10.00 – 24.00
Observed pointing abilities M SD Range M SD Range
Pointing comprehension
Frequency 1.51 1.66 0 – 6.21 2.64 1.50 0 – 5.58
Pointing production: all trials Frequency e, h 2.41 2.56 0.00 - 8.61 3.08 3.15 0.00 - 9.91 Duration e, h 6.90 8.29 0.00 - 29.00 6.07 6.58 0.00 -22.12 Pointing production: positive trials Frequency e, h 0.92 1.05 0.00 - 3.40 1.18 1.09 0.00 - 3.46 Duration e, h 0.10 0.12 0.00 - 0.40 0.08 0.08 0.00 – 0.28 Pointing production: negative trials Frequency g, h 0.70 0.94 0.00 - 3.00 1.03 1.24 0.00 - 3.90 Duration g, h 0.04 0.06 0.00 - 0.18 0.08 0.10 0.00 – 0.32 Pointing production: neutral trials Frequency e, i 0.80 1.09 0.00 - 3.80 0.94 1.33 0.00 - 4.28 Duration e, i 0.05 0.07 0.00 - 0.20 0.04 0.06 0.00 – 0.17
Note. For the QPOINT, pointing production and comprehension variables only the declarative parts were taken into account. a n = 47. b n =42. c n = 4. d n = 34. e n = 44. f n = 35. g n = 43. h n = 38. i n = 37.
At 15 months, the Friedman test showed a significant difference in duration of
pointing between the three sort of trials (positive, negative and neutral; χ2 = 11.822, p < .05), but not in frequency (χ2 = 2.198, p>.05). Children pointed at 15 months longer during the positive trials than during the neutral trials (Z = -3.054, p < .05) and longer during the
negative than during the neutral trial (Z = -2.768, p < .05). However, between the positive and negative trials, there was not a significant difference in pointing duration, Z = -.038, p > .05. Summarizing, this means that the children pointed the shortest when seeing the neutral stimulus, but pointed equally long when seeing the positive and negative stimuli. They showed no differences in how often they pointed across the different trials.
No significant effect of the counterbalanced condition of stimuli presentation (either positive-negative-neutral or positive-neutral-negative) was spotted, neither at 12 nor at 15 months, neither for pointing frequency (U = 210.000, p > .05 at 12 months; U = 130.000, p > .05 at 15 months) nor for pointing duration (U = 218.000, p > .05 at 12 months; U = 164.000, p > .05 at 15 months). This indicates that the order in which the stimuli were presented did not influence how often, how long and how fast after the stimulus appeared the children pointed, at both ages.
1. 2. 3. 4. 5. 6. 7. 8. 9.
Dur Freq Dur Freq Dur Freq Dur Freq Dur Freq Dur Freq Dur Freq Dur Freq 12mo 15mo
1. DP: All trials 12 months
Duration -
Frequency ,935** -
2. DP: All trials 15 months
Duration ,450* ,491** -
Frequency ,308 .363* .905** -
3. DP: Positive trials 12 months
Duration .874** .788** .509** .362* -
Frequency .854** .856** .569** .427* .919** -
4. DP: Positive trials 15 months
Duration .384* .417* .819** .777** .391* .493** -
Frequencies .350 .427 .746** .850** .372* .474** .885** -
5. DP Negative trials 12 months
Duration .813** .840** .274 .258 .598** .658** .192 .217 -
Frequency .763** .847** .295 .281 .580** .633** .158 .258 .940** -
6. DP Negative trials 15 months
Duration .335 .466** .862** .810** .317 .435* .642** .589** .481** .489** -
Frequency .230 .318 .794** .860** .242 .334 .584** .610** .414* .386* .918** -
7. DP: Neutral trials 12 months
Duration .757** .770** .220 .096 .476** .561** .330 .272 .609** .534** .154 .059 -
Frequency .746** .785** .301 .175 .487** .588** .385* .323 .606** .543** .245 .148 .970** -
8. DP: Neutral trials 15 months
Duration .309 .278 .813** .774** .477** .405* .514** .509** .079 .142 .623** .585** .025 .119 - Frequency .223 .185 .762** .820** .378* .327 .466** .531** .077 .122 .615** .629** -.077 .020 .948** - 9. DC: Comprehension 12 months .206 .224 .123 .115 .187 .157 .021 .158 .089 .182 .005 -.026 .118 .062 .179 .171 - 15 months .052 .187 -.066 -.033 -.051 -.070 -.147 -.092 .182 .288 .107 .147 .124 .155 -.044 -.071 .210 - Table 5
Spearman rho correlations among experimentally observed pointing ariables
Note. DP = declarative production task. DC = declarative comprehension task. Dur = duration. Freq = frequencies. 12m = 12 months. 15m = 15 months.
Temperament and Pointing abilities
Extraverted children. There was no significant relation between the extraverted temperament of the children and their declarative comprehension, neither at 12 (mothers: rs =
.082; fathers: rs = .166, p > .05) nor at 15 months (mothers: rs= .170; fathers: rs = .055, p >
.05), indicating that the child’s extraversion temperament did not give any information about the child’s declarative pointing comprehension.
A negative correlation was found between the extraversion of children according to mother and the observed frequency and duration of pointing during all trials at 12 months (rs
= -.387, p < .05 and rs = -.324, p < .05 respectively). A linear regression analysis showed that
for this correlation the extraverted temperament defined by the mothers at 12 months also had a predictive value for the frequency of pointing during all the trials at 12 month, but not for the duration of pointing, R2 = .153, F(1,37) = 6.709, p <.05 and F(1,37) = 3.420, p > 0.05 respectively. The same results were not found when fathers reported the child’s temperament (frequency: rs =-.157, p > .05; duration: rs = -.101, p > .05), indicating that only when the
mother reported a higher score on extraversion for her child, the child showed less often and shorter pointing gestures during the observation for all the trials. In addition, the extraverted temperament as described by mothers partly predicted how often the child would point across all trials at 12 months.
When investigating children’s pointing abilities within the valence of the trials at 12 months, a negative correlation was found between the extraverted children as reported by mothers and the frequency of pointing during the positive trials (rs = -.361, p < .05), but not
for duration (rs = -.216 for duration, p > .05). In this specific case, the extraverted
temperament of the children also had a predictive value for the frequency of pointing during the positive trials, R2 = .166, F(1,37) = 7.370, p < .05. In other words, a higher score on extraversion as defined by mothers, predicted partly that the child would point less often during the positive trials at 12 months.
Equally, during the neutral trials at 12 months, the child would point less and shorter, when mothers defined their child as extraverted, rs = -.513, p < .05 and rs = -.497, p < .05.
Interestingly, opposite findings were found for the father’s perception of their extraverted children and their pointing frequency and duration towards the neutral stimulus at 15 months, rs = .431, p < .05 and rs = .395, p < .05. Please see Appendix G, Table G1 for more
information about the non-significant results. An extraverted child as reported by the mother at 12 months, pointed less and shorter during the neutral trials at 12 months, but when this was reported by the fathers at 15 months, the children pointed more and longer during the
neutral trials at 15 months. In this specific case, the measurement of extraversion by mother predicted partly the frequency and duration of pointing (frequency: R2 = .183, F(1,37) = 8.292, p < .05; duration: R2 = .220, F(1,37) = 10.465, p < .05) and the measurement of extraversion by fathers predicted the frequency, but not the duration of pointing during the neutral trials (frequency: R2 = .227, F(1,27) = 7.932, p < .05; duration: R2 = .105, F(1,25) = 3.167, p > .05). Finally, there were no relations found between the extraverted temperament of the children and their pointing abilities during the negative trials as observed in the lab, neither when temperament was described by mothers (range: rs = -.222 - .114, p >.05) nor
when it was described by fathers (range: rs = -.020 - .188, p >.05). At the same time, there
was no relation between the extraverted temperament described by both parents at both time points and the declarative pointing production based on the QPOINT questionnaires (mothers at 12 months: rs = -.003, p >.05; mothers at 15 months: rs = .087, p >.05; fathers at 12 months:
rs = .286, p>.05; fathers at 15 months: rs mothers = .068, p >.05).
Negative affective children. There was no significant relation between the negative affective temperament of the children and their declarative comprehension, neither at 12 (mothers: rs = -.263, p >.05; fathers: rs = -.009, p > .05) nor at 15 months (mothers: rs = -.102,
p >.05; fathers: rs = -.037, p > .05). Equally, there was no relation between the negative
affective children and their declarative pointing production, neither as observed in the lab (range: rs = -.139 - .100, p > .05) nor as measured with the questionnaires about the pointing
production (mothers at 12 months: rs = .146, p > .05; mothers at 15 months: rs = -.016, p > .05;
fathers at 12 months: rs = -.117, p > .05; fathers at 15 months: rs = .057, p > .05). This means
that the child’s negative affective temperament is not linked to the child’s declarative pointing comprehension.
Effortful control children. There was no significant relation between the effortful controlled temperament of the children and their declarative comprehension, neither at 12 (mothers: rs = .293, p >.05; fathers: rs = -.030, p > .05) nor at 15 months (mothers: rs = -.013, p
>.05; fathers: rs = .125, p > .05). In other words, the effortful controlled temperament of the
children is not linked to their abilities to understand the pointing gesture of someone else. Regarding the declarative production, when effortful controlled temperament was defined by mothers, no significant relation was found with the observed pointing frequency and duration at both time points (range rs = -.209 - .227, p > .05). Please see Appendix G,
Table G1 for more information about the non-significant results.
Instead, when effortful controlled temperament was defined by fathers, a positive correlation between effortful controlled temperament and the frequency and duration of
pointing production during the positive trials at 12 months was found (frequency: rs = .365, p
< .05; duration: rs = .428, p < .05). This correlation also had a predictive value for both the
frequency and duration during these positive trials at 12 months (frequency: R2 = .124, F(1,32) = 4.524, p < .05; duration: R2 = .280, F(1,32) = 8.409, p < .05). Please see Appendix G, Table G1 for more information about the non-significant results. These results mean that effortful controlled children, as defined by their fathers, showed more often and longer pointing during the positive trials and that within this relation the effortful controlled temperament also could partly predict the frequency and duration of pointing during these positive trials. When fathers described the temperament, there was no significant relationship at 15 months for the observed pointing abilities (range: rs = -.270 - .268, p > .05).
When looking at the pointing production measured by the questionnaires, the mothers reported a better declarative pointing production of their effortful controlled children at 12 and 15 months, r = .327, p < .05 and r = .355, p < .05 respectively. Within this relation, effortful controlled temperament, defined by mothers, was found to be predictive for the declarative pointing productions as described by mothers at both ages (12 months: R2 = .107, F(1,45) = 5.399, p < .05; 15 months: R2 = .089, F(1,40) = 3.895, p < .05). For fathers this relation was not found (12 months: r = .237, p > .05; 15 months: r = .145, p > .05). In other words, when the mother described her child as effortful controlled, she would also report that the child would point more, which was partly predictable based on the effortful controlled temperament.
Influence of Gender
The Mann-Whitney U tests showed no significant gender effect for the declarative pointing comprehension, neither at 12 months nor at 15 months (12 months: U = 233.000, p > .05 at 12 months; 15 months: U = 110.000, p > .05), neither for the frequency and duration of declarative production task at 12 months (frequency: U = 206.000, p > .05; duration: U = 207.000, p > .05) and nor at 15 months (frequency: U = 171.000, p > .05; duration: U = 146.000, p > .05). In contrast, girls did show more advanced declarative pointing production according to the mothers (based on the questionnaire) at 12 months (U = 136.500, p < .05), whereby the mean score for girls was M = 30.68 and for boys M = 19.06. This result was not found when fathers filled in the questionnaires (U = 162.000, p > .05), nor at 15 months for both parents (mothers: U = 196.000, p > .05; fathers: U = 135.000, p > .05).
Discussion
In the current research, the relation between the declarative pointing abilities of children at 12 and 15 months and their temperament characteristics, such as extraversion, negative affectivity and effortful control, was investigated. In addition, the role of gender and the pointing duration have been examined. Consistent with Camaioni and colleagues (2004), the declarative comprehension and production increased over time, showing a developmental improvement of pointing abilities from 12 to 15 months.
Declarative Pointing Abilities and Temperament
The observed declarative comprehension abilities did not relate to the temperament of the children. This is contradicting the expectation, but it is in line with research about the broader construct joint attention. In previous research of Vauhan and colleagues (2003), no relation was found between reacting to joint attention (e.g. reacting to the pointing gesture of someone) and positive and negative emotional reactivity, which are important behaviors defining the variability in temperament across children. When interpreting the pointing gesture of an adult, the child relies on the shared experience with that specific person (Tomasello et al., 2007) and on their general cognitive mechanisms (Gliga & Csibra, 2009; Moore, Call, & Tomasello, 2015). Consequently, the variance between children cannot be explained by the individual differences in emotional reactivity and self-regulation (i.e. temperament), but by the dyadic interaction with the experimenter and the cognitive abilities of the children. This research included tasks, prior to the pointing tasks, to promote joint attention between the child and experimenter, but this interaction might have been too short to have an impact on the child’s interpretation of the pointing gesture. The role of cognitive abilities was not measured at all. In conclusion, declarative comprehension is not related to the child’s temperament. The cognitive abilities and the shared experience with the person making the pointing gesture might explain the differences in pointing better.
Regarding the declarative production, extraverted children as defined by mothers pointed less often and shorter during all trials of the production task at 12 months. They pointed even less during the positive trials, though it was expected that they would point more during these trials due to their sociable character. It is important to note that these results at 12 months were only found when mothers reported about the child’s temperament, but not when fathers reported about the child’s temperament. At the same time, though, at 15 months, the extraverted children pointed more often and longer during the neutral trials, when fathers indicated their children as having an extraverted temperament, but this relation was not found
when mothers reported about the child’s temperament. Furthermore, extraversion
temperament explained only a small percentage of the variance between children, indicating that other factors might also influence the pointing abilities.
What is surprising about these results is the difference between what is reported by mothers and fathers, especially because correlation between parents was found for almost all subscales. The discrepancy may be due to the different reactions that parents elicit in their children and possible over- or underestimation of the parents (Clark, Durbin, Donnellan, & Neppl, 2017). Nonetheless, this does not diminish the advantages of including parents as informants, mainly for what concerns the high ecological validity (Clark et al., 2017).
These surprising results can be explained by the high activity level in extraverted children. Indeed, not only are extraverted children normally sociable, but they are also very active and constantly trying to explore their environment (Berdan, Keane, & Calkins, 2008). The tasks tested in the current research were part of a bigger battery of tasks, which all together lasted around 30 minutes. For all children, it is difficult to stay focused for such a long time (Bornstein, Arterberry, & Lamb, 2014), but this is even more difficult for children who are in general more active (i.e. extraverted children). This explorative, active behavior could hold the child back in maintaining the attention. Consequently, since the child would not be focused anymore, s/he would also point less often or less long. At a later age, the children’s attention span is longer (Bornstein et al., 2014), thus it would be easier for the 15-month-olds to stay focused on the task, giving a possible explanation why the children pointed more and longer at 15 months. In conclusion, there is a relation between extraversion and the pointing production of the children. Yet, the results seem to be dependent on the parent’s view on temperament and therefore not giving a fully established relation.
The expectation that negative affective children would show less declarative production is not confirmed. The negative affective temperament does not relate to the declarative pointing abilities of the children. This is in line with studies about joint attention stating that negative affective children show infrequent initiating joint attention (Todd & Dixon, 2010). Some research proceeds and even concludes that the initiation of joint attention is led only by positive affect, whereas negative affect shows the opposite effect and reduces initiating joint attention (Nichols et al., 2005; Todd & Dixon, 2010; Vaughan et al., 2003). The character of negative affective children is, indeed, more characterized by their negative affect, suggesting that they do initiate less, or at least less frequently, joint attention and, consequently, point less often or shorter. In conclusion, negative affective temperament does
not link to the declarative pointing production due to the negative affect that influences the joint attention behavior, among which the pointing production.
Children with an effortful controlled temperament, defined by fathers, pointed more often and longer towards the positive stimuli at 12 months. Similarly, the declarative
production measured via the questionnaire showed that effortful controlled children pointed more at 12 months. Thereby it can be concluded that children do not point less due to their reaction of distress/avoidance. The role of the amygdala, a temporal lobe in the human brain that responds to novel stimuli, could explain this unexpected positive relation (Blackford, Allen, Cowan, & Avery, 2012; Blackford, Avery, Cowan, Shelton, & Zald, 2010). The
amygdala is found to be particularly active in children with an inhibited temperament, such as the effortful controlled temperament (Blackford et al., 2010; 2012). This means that effortful controlled children respond more actively to novel stimuli, because they are more alert. A similar result has been found for joint attention (Nichols et al., 2005). Children at the age of 9 months showed a decreased inhibition of the amygdala when initiating joint attention. Nichols and colleagues (2005) highlighted, however, that the attention of the child says more about their information seeking or alertness than their willingness to share information.
In conclusion, effortful controlled children showed more and longer pointing during the positive trials. This can be explained by the hyperactive amygdala in effortful controlled children (Blackford et al., 2012), meaning that they respond more actively to novel stimuli. Yet, an increased attention that leads to more pointing does not necessarily mean that the child wants to share more experiences with the other person.
Latency and Gender
The declarative pointing abilities of children were not related to the child’s gender. This is in line with the research of Farkas and Valloton (2016) finding no gender differences for temperament, but conflicting with the expectations of this research and with the meta-analysis of Else-Quest and colleagues (2006), which did report gender differences in temperament. However, Else-Quest and colleagues (2006) point out that gender differences are larger when the children are older. In conclusion, the gender differences in temperament might not be well-established (yet) and therefore do not show differences at the beginning of the development of the pointing abilities.
Concerning the explorative hypothesis about latency, the results of this research about the latency duration have been integrated in the discussion above about the different
temperament factors. However, it is interesting to explicitly point out that the duration of pointing intermittently gave different information about the declarative production compared to the frequency of pointing. For example, children did point longer but not more often to the positive trials compared to the neutral trials at both ages. The discrepancy between duration and frequency suggests that the pointing duration might play a role in the pointing
development of the children. This is supported by the research of Cochet and Vauclair (2010), who found that the declarative pointing of children last longer than the imperative pointing. In conclusion, some significant results for latency have been found, making it worth to further examine the latency of the pointing gestures in new research. This could give us a more complete insight in the pointing abilities of children and how they develop across different situations.
Limitations and Strengths
The current research has a few limitations. First, the vast majority of the parents were highly educated. This is particularly significant for the current research because highly
educated parents define their own children as more extraverted (Farkas & Vallotton, 2016). In other words, the parents are biased in their answers and the sample is not representative and generalizable to the whole (Dutch) population. Secondly, results must be interpreted with caution because of the internal validity of the temperament questionnaires. Though still comparable to other research (Farkas & Vallotton, 2016), the internal validity for the IBQ-R/ECBQ were relatively low (Enlow, et al., 2016).
Next to the limitations, this research also has its strengths. First, it is unique in examining the relation between declarative pointing abilities and temperament. Even though previous literature showed enough relevance, no research had ever looked at this specific relation before. Secondly, the multi-method for the declarative production strengthened this research: indeed, the combination of lab observations and questionnaires allowed for this research to create a more complete view of the pointing abilities of children (Taskhakkori & Teddlie, 2003). Finally, the inclusion of the latency (i.e. duration and for reaction time, please see Appendix A) contributed as well to a more complete view. The latency of pointing has not (yet) been examined as extensively as future research could do, but as the results of the
current research shows, it does provide some innovative insights into the pointing abilities. This research gives several implications for further research. First of all, the role of cognitive abilities in the declarative comprehension could be examined more in detail.
