The Development of Crying in 3- to 7- year-old Children; Parental Socialization, Gender, and Cultural Differences.
De ontwikkeling van huilen en het ontstaan van gender -en culturele verschillen daarin gedurende de kindertijd.
Masterscriptie Orthopedagogiek, Pedagogische en Onderwijskundige Wetenschappen, Universiteit van Amsterdam B. Trommel Begeleiding: dr. F.C Jellesma & prof. dr. P.F de Jong Amsterdam, (januari, 2017)
The Development of Crying in 3- to 7- year-old Children; Parental Socialization, Gender, and Cultural Differences.
De ontwikkeling van huilen en het ontstaan van gender -en culturele verschillen daarin gedurende de kindertijd.
Masterscriptie Orthopedagogiek, Pedagogische en Onderwijskundige Wetenschappen, Universiteit van Amsterdam B. Trommel Begeleiding: dr. F.C Jellesma & prof. dr. P.F de Jong Amsterdam, (januari, 2017)
Index
Abstract 2
Introduction 3
• Gender difference in crying 3
• Cultural differences in crying 6
• Parental socialization of crying 9
• The present study 10
Method 11
• Participants and procedure 11
• Instruments 12
• Questionnaire 13
o Data analysis of the questionnaire 13
o Semi-structured interview 13
o Coding of the semi-structured interview 13 o Analysis of the semi-structured interview 14
Results 15
• Effects of age, culture and gender on crying frequency 15 • Effects of age, culture and gender on reason for crying 17 • Effects of age, culture and gender on parental socialization 19 • Effects of parental socialization on crying frequency
• Effects of parental socialization on reason for crying
• Parental beliefs about crying 22
Discussion 25
• Age 26
• Gender differences 27
• Cultural differences 28
• Parental socialization of crying 29
• Parental beliefs about crying 29
• Study limitations 30
• Conclusion 31
References 32
Abstract. The aim of this study is to document the development of age, gender and cultural differences in crying. Parents reported on their children's (3 to 7 years old) crying frequency, antecedents of crying and the socialization of crying. Gender, age and cultural differences in these concepts were assessed cross-sectionally. A total of N = 512 parents with three different cultural backgrounds (Dutch, Moroccan-Dutch and Xhosa-South African) filled out a questionnaire and a subsample n = 61 participated in a semi-structured interview on parental beliefs about crying. Results show that during development crying frequency decreases, the antecedents of crying change and gender differences in crying frequency develop when children are 7 years old. Also, the study showed Xhosa- South African children cry less frequently than Dutch and Moroccan-Dutch children.
Keywords: crying, crying frequency, antecedents of crying, culture, gender, children, socialization
Samenvatting. Deze studie had als doel de ontwikkeling van leeftijds, gender -en culturele verschillen in huilen te beschrijven. Daartoe rapporteerden ouders met kinderen in de leeftijd 3 tot 7 jaar, de huilfrequentie en reden voor huilen van hun kinderen. Er is cross sectioneel gekeken of deze variabelen gender of culturele verschillen vertoonden. In totaal vulden 512 ouders met drie verschillende culturele achtergronden (Nederlands, Marokkaans-Nederlands en Xhosa-Zuid Afrikaans) een vragenlijst in over een van hun kinderen. Een deel van deze ouders (n=61) nam ook deel aan een semi-gestructureerd interview gericht op hun ideeën over de ontwikkeling van huilen bij kinderen. De resultaten van de studie laten zien dat kinderen minder gaan huilen en om andere redenen wanneer ze opgroeien. Vanaf zevenjarige leeftijd huilen meisjes vaker dan jongens. Ook laat deze studie zien dat Xhosa-Zuid Afrikaanse kinderen minder huilen dan Nederlandse en Nederlands-Marokkaanse kinderen.
Introduction
Recent studies consistently show interindividual differences in how often people cry. Research has demonstrated that women cry more often than men (Vingerhoets & Scheirs, 2000) and in some countries people cry more often than in other countries (van Hemert, van de Vijver & Vingerhoets, 2011). These cultural -and gender differences in crying frequency seem to emerge in childhood.
Whereas at first, newborns cry often and mostly because they are in pain, hungry or in need of attention (Soltis, 2004; Zeifman, 2001), somewhere during development these antecedents of crying change. The reason for this is newborns to an important extent rely on crying to communicate their needs. They cry when they are sick, hungry or need a diaper change. Adults, on the other hand, are less dependent on others and have learned other means to communicate their needs. They seldom cry and when they do it is often because of more complex emotions like loss, despair or certain forms of happiness and shame, that newborns not yet experience (Provine, Krosnowski, & Brocato, 2009). This study is the first to document the development of crying frequency and antecedents of crying in early childhood across different cultural groups and gender. Crying frequency and the antecedents of crying will be measured by asking parents to report on their children’s crying. Also, the possible influence of parents on the development of crying in children will be explored. It's important to understand the development of crying because of the potential insight this gives into the origins of differences in the emotions underlying crying. An understanding of these differences can help mental health professionals better respond to gender and cultural differences. The remainder of this introduction will first focus on gender differences in crying and then discuss cultural differences in crying. In these sections, it will become apparent that cultural and gender differences are likely both related to social norms related to crying. It was thus chosen to discuss the influence of parents on cultural and gender difference in crying together in a combined section. This section will follow the sections on gender and cultural differences. The introduction is concluded by sketching the goal and expected findings of this study
Gender differences in crying
The gender differences in crying frequency observed in adult populations do not exist at birth. No differences in crying were found between male and female newborns (St James-Roberts & Halil, 1991). It remains uncertain when gender differences in crying frequency emerge as there is only limited research on the development of crying. Looking at the available research, gender differences in crying frequency probably arise before children turn eight (Jellesma &
Vingerhoets, 2012). Limited research is available on the developmental mechanisms causing differences in crying frequency. It was proposed that gender differences develop due to hormonal changes after menarche (Frey, 1985). Tilburg, Unterberg and Vingerhoets (2002), however, found gender differences are unaffected by menarche. Also, Jellesma and Vingerhoets (2012) found gender differences exist between children aged 8 when menarche hasn’t happened for most girls. As there are also no gender differences between newborns it is unlikely biological factors determine differences in crying frequency. This idea is further strengthened by the fact that the magnitude of gender differences varies between cultures (van Hemert, van de Vijver, & Vingerhoets, 2011). After all: biological factors would cross-culturally result in the equal gender differences in crying behavior.
Beside biological factors, other factors have been put forward to explain gender differences in crying frequency. Bekker and Vingerhoets (2001) suggest the following factors: exposure to emotional events, a more dramatic impact of events on psychological state variables, different appraisal of similar events and social learning processes. These factors will now be discussed separately.
The factor exposure to events suggests that women are more often exposed to emotional events that elicit crying. Bekker and Vingerhoets (2001) mention that women are possibly more often to exposed stressful life events like pregnancy and traumatic events and experience the loss of relatives more often because they have a higher life expectancy. It, however, seems unlikely these events can explain gender differences in crying between children for the obvious reason that children are not affected by pregnancy or experience the loss of relatives unequally. Other stressful life events to which boys and girls are unequally exposed also do not come to mind. As this study focuses on differences in crying between children, it was chosen not to consider the possible effects of a different exposure to events.
As a second factor to explain gender differences in crying frequency Bekker and Vingerhoets (2001) suggest a more dramatic impact of events on physical and psychological state variables that may moderate crying. In their preliminary model of crying they propose physical and psychological state variables constitute a crying threshold. Anecdotal evidence shows after certain events like pregnancy, stillbirth and a heart attack, people cry more often or much less indicating their crying threshold has changed. Bekker and Vingerhoets note the power of this factor to explain differences in crying is rather limited because gender differences are also found between groups that only exceptionally experience events that may change the crying threshold dramatically. As the focus of this study is children and the sample size is too small to pick up the effect of events that are experienced by a small percentage of children this factor is not further addressed
Bekker and Vingerhoets (2001) further suggest a different appraisal of similar events as a factor explaining gender differences in crying frequency. In their preliminary model of crying the appraisal of a situation creates a mental representation of that situation which causes an emotional state. In turn, this emotional state causes crying. Differences in the emotional states that women and men experience would thus indicate a different appraisal of events.
Although it's commonly assumed that women are more emotional than men, this claim is not supported by the available research (Barrett & Bliss-Moreau, 2009; Barrett, Lane, & Schwartz, 2000; Else-Quest, Higgins, Allison, & Morton, 2012). In studies using direct or momentary assessments of the emotion generally no clear gender difference is found in the emotions people experience (Barrett & Bliss-Moreau, 2009). When these direct or momentary assessments of emotion do reflect the emotional state caused by the appraisal of an event, it can be assumed women and men appraise events the same. Because there’s no robust evidence indicating gender differences in appraisal this effect is not examined in this study.
In contrast to studies using momentary assessments of emotion studies using less direct measures of emotion do find gender differences in the emotions men and women report. It is suggested that this difference in the emotions people experience and the emotions they report to have experienced, is caused by gender expectation norms (Barrett & Bliss-Moreau, 2009; Barrett, Robin, Pietromonaco, & Eyssell, 1998; Whittle, Yücel, Yap, & Allen, 2011). This means that men and women report different emotions than they experience to conform to social expectations. Bekker and Vingerhoets (2001) propose that a similar link exist between crying and social expectations. As a fourth factor accounting for gender differences in crying they put forward the influence of social learning processes. They suggest social learning influences the formation of gender stereotypes which affect crying. Gender stereotypes are the societal norms related to how a person should behave based on its sex. Gender stereotypes share common features with the concept display rules. Display rules are to be understood as the prescriptive social norms that dictate how, when and where emotions can be expressed in a particular culture (Brody, 2000). Research with predominately western middle-class Caucasian samples shows display rules related to gender generally include the notion that men are less emotional, show fewer emotions that indicate weakness, submission or vulnerability and should behave accordingly (see MacArthur & Shields, 2014 for a review). Because gender stereotypes are more favorable to female emotion expression, Fischer and LaFrance (2014) and Bekker and Vingerhoets suggest that gender stereotypes are related to gender differences in emotional expression, notably crying. It is thus expected that gender stereotypes cause women to cry more in general and to cry more of emotions indicating weakness like sadness. The formation of gender stereotypes begins early in development. Children aged 2 first show knowledge of rudimentary gender stereotypes that
during development grow more complex and sophisticated (Martin & Ruble, 2010). Children between 4 and 5 show gender stereotypes about physical aggression and relational aggression (ibid). Gender stereotypes about crying are of similar nature and complexity and could thus develop in the same time frame.
Bekker and Vingerhoets (2001) do not discuss emotion regulation as a possible factor to explain gender differences in crying. As will follow, however, emotion regulation possibly influences the development of gender differences in crying and is thus important to discuss. Emotion regulation refers to the activities that allow an individual to monitor, evaluate, and modify the nature and course of an emotional response, to reach his or her goals and to appropriately respond to environmental demands' (Nolen-Hoeksema, 2012). There are small differences between men and women in emotion regulation. Women use emotion regulation strategies like rumination, reappraisal, problem-solving, seeking social support, acceptance, and distraction more often (Nolen-Hoeksema, 2012; Tamres, Janicki, & Helgeson, 2002). It is possible that some of these strategies like rumination and seeking social support may lead to a higher crying frequency. Morrison and O’Connor (2005), for example, found that rumination predicts higher distress which could lead to more crying. Also, crying seems to function as a signal that a person needs social support (Soltis, 2004). It could be because women seek more social support as an emotion regulation strategy they also cry more often. Men compared to women might use more automatic non-conscious emotion regulation strategies which are more effective than conscious emotion regulation strategies (Barrett et al., 2000; Koole, Webb, & Sheeran, 2015; Williams, Bargh, Nocera, & Gray, 2009). This is in accordance with the observation that girls score higher on effortful control compared to boys (Nolen-Hoeksema, 2012). Men could cry less because they use more emotion regulation strategies that are more effective. Both emotion regulation and gender stereotypes to an important extent develop when children are 3 to 5 years old (compare Martin and Rumble, 2010 and Kiel and Kalomaris, 2015). It is likely that the development of crying reflects this development.
Cultural differences in crying frequency
Cultural differences in crying frequency exist in both newborns and adults, but different mechanism likely cause the cultural differences found in these two periods of life. The differences found between newborns seem to be related to soothing and psychical contact (Abdulrazzaq, Al Kendi, & Nagelkerke, 2009; Bleah & Ellett, 2010; St James-Roberts et al., 2006). In their review of the available studies, Bleah and Ellett (2010) conclude that in countries where babies are being carried more often, parents are closer by and soothe children sooner, children cry shorter and less often. Both soothing and psychical contact are forms of external
regulation. It is likely the effects of external regulation on crying frequency fade when children develop forms of internal regulation of their emotions.
In a study across 37 countries, van Hemert, van de Vijver and Vingerhoets (2011) showed that adults in more democratic, affluent, individualistic, extravert countries cry more often. They suggest these characteristics reflect differences in display rules and tolerance of emotion. Further evidence on the influence of display rules is found in a recent review by Kiel and Kalomiris (2015) on emotion regulation who suggest the development of emotion regulation is moderated by social norms like display rules. Following these studies, it is expected that display rules play an important role in the development of cultural differences in crying frequency.
The development of display rules is often tested by using a disappointing gift paradigm (Tobin & Graziano, 2011). The disappointing gift paradigm tests whether children hide their disappointment when they get an undesirable gift. Studies employing this paradigm show children up from age 4 use display rules and their ability to do so progresses with age (see Kromm, Färber & Holodynski, 2014 for a review). Using a disappointing gift paradigm Garrett-Peters and Fox (2007) found European American children showed more positive expressive behaviors than Chinese American children aged 4 to 7. Other studies not using the disappointing gift paradigm also find cultural differences related to the expression of emotion. Cole, Bruschi and Tamang (2002) found Brahman children compared to Tamang and U.S. children were less likely to communicate negative emotions. Lewis, Takai-Kawakami, Kawakami and Sullivan (2010) found Japanese children compared to American children show less shame, pride and sadness and more embarrassment. Novin, Rieffe, Banerjee, Miers, and Cheung (2011) compared Chinese and Dutch children differ in how they react to aggression. Chinese react more tolerant and Dutch children are more likely to confront their aggressor trying to reinstate their personal goals. Safdar et al. (2009) compared display rules of college students from Japan, Canada, and the U.S. They found Japanese students compared to Canadian and U.S. students have more diverse display rules for different interaction partners, permit the expression of powerful emotions and discourage the expression of positive emotions. In sum, the limited research available suggests different aspects of the role of display rules in the expression of emotion. Between cultures, display rules differ for individual emotions, differ depending on the social context and children differ both in the strategies and goals they employ in situations where emotions are expressed.
Let us have a closer look at why these differences exist. Research shows display rules and actual emotional expressions are related to cultural value orientations (Koopmann-Holm & Matsumoto, 2010; Matsumoto, Seung Hee Yoo & Fontaine, 2008; Novin, Banerjee, Dadkhah, & Rieffe, 2009; Wong, Bond, & Rodriguez Mosquera, 2008). Often used contrasts in value
orientations to explain cultural differences are interdependence/independence and a horizontal/vertical distribution of power, both put forward by Hofstede (2001, see Halberstadt & Lozada, 2011 for a review). The contrast interdependence/independence is based on the idea that cultures put more value on either the individual (individualism) or social relations (communalism). Questions like: “should children inhibit their reactions to uphold social harmony?” revolve around this contrast. The contrast horizontal/vertical is based on the idea that cultures differ in how power is distributed between people in a culture and revolve around questions like: "should children listen to their parents?" Although these contrasts are often used, they are not without controversy. Most recent studies using these models mention cultural differences are more complex and dynamic than these models are capable of capturing and research shows interdependence and independence are not mutually exclusive (Raeff, 2010; Wilson, Raval, Salvina, Raval, & Panchal, 2012).
In this study, three different cultural groups have been included: Dutch, Moroccan-Dutch and Xhosa-South African. These groups were included because they were accessible to the researcher and likely differ in how much they value independence/interdependence. Xhosa is one of the eleven official languages of South Africa. About 18% of the population uses it as its primary language (Statistics South Africa, 2011). Different clans who share a common culture use Xhosa as their primary language (Afoyalan, 2004). As value orientations are dynamic and subject to change, it is important to consult recent research to gain information on the value orientations of a particular group. Recent research documenting the value orientations of Xhosa-South African people, however, was very scarce. Only research mapping the traits different South African cultural groups use to describe persons or themselves was found. This research shows compared to white and colored South Africans, Xhosa-South African use the same traits to describe themselves but use more social relational terms (Adams, Van de Vijver, & De Bruin, 2012; Valchev, van de Vijver, Alewyn Nel, Rothmann, Meiring, & de Bruin, 2011). This research seems to suggest compared to South African whites and coloreds Xhosa-South African put more value on interdependence. No research was available on how power is distributed between Xhosa-South Africans. Recent studies mapping the value orientations of Dutch, Moroccan-Dutch were also scarce. The available research suggests the Dutch cultural group values independence highly and has a horizontal power distribution (Harkness, Super, & van Tijen, 2000; Huijbregts, Tavecchio, Leseman, & Hoffenaar, 2009). The Moroccan-Dutch group consists of descendants of guest workers and recent immigrants from Morocco. The available research suggests Moroccan-Dutch value vertical power structures and compared to the Dutch value interdependence more (Huijbregts et al., 2009). In sum the Dutch seem to value independence the most, followed by Xhosa-South African. Moroccan-Dutch value independence
the least. As mentioned before adults in more democratic, affluent, individualistic, extravert countries cry more often. (van Hemert, van de Vijver, & Vingerhoets, 2011). Countries that are more democratic, affluent and individualistic tend to be best characterized as having a value orientation centered on independence. It can thus be expected that children in countries that value independence cry more.
Parental socialization of crying
This introduction thus far has only briefly discussed the possible role of parents in gender and cultural differences. However, it is also a goal of this study to uncover whether and how parents influence the development of crying. It is important to look at the role of parents because it can give information on how socialization influences the expression of emotions.
No research was found on the parental socialization of crying in early and middle childhood. Though research on characteristics related to crying is well available and shows parents play an important role in the socialization of emotion (see Morris, Silk, Steinberg, Myers, & Robinson, 2007 for a review). When children cry, this attracts attention. In early childhood often it will be parents who attend to their child when it cries. When parents react, this opens the possibility for them to coach their child, model behavior and talk about emotions, all of which are important means for socializing emotions (Rogers, Halberstadt, Castro, Mac-Cormack & Garret-Peters, 2016). And there is robust evidence there are cultural and gender differences in these behaviors. Raval, Raval, Salvina, Wilson and Writer (2013) for example found mothers from India respond to children’s emotions more by explaining what children are experiencing when compared to US mothers who respond more often by offering solutions. With regard to gender differences van der Pol et al. (2015) found parents from the Netherlands associate anger with boys and sadness and happiness with girls when they talk about emotions with their children. No studies were found that reported an effect in children of gender or cultural differences in parental emotion socialization. Given the number of studies that report these differences, this is surprising. Brody (2000) suggests cultural and gender differences are to a large extent caused by indirect socialization and this might explain the lack of studies reporting an effect of gender and cultural differences in parental socialization. Because of the possible role of indirect socialization this study makes use of a two-track approach studying both the direct and indirect socialization of crying.
The effects of direct socialization will be measured by asking parents how they respond to crying and why they respond in a particular way. Although there is no evidence supporting the effect of parental emotion socialization on gender and cultural differences this study still expects to find an effect of parental response and the goal of this respond. This study expects to find a relation
because crying compared to other behaviors more likely elicits a response by parents. It is thus expected it is also more likely parents influence their children through their behavior when their children cry. It is thus expected the parental response to crying and the goal of this response cause gender and cultural differences in how often and why children cry. Studies further show parents react differently to different emotions expressed by their child (O'Neal, Magai, 2005). When children cry, this can signify different emotions (Soltis, 2004). It is thus expected that parental reactions are related to the emotion that causes crying. The way parents react to different emotional displays by children is further related to children is age, gender and parents’ cultural background (age: Van der Pol et al., 2015; gender: Kennedy Root & Denham, 2010; culture: Keller & Otto, 2009). It is thus expected parental reactions to crying are also related to their child’s age, gender and parents’ cultural background.
The second track in this study focuses on parental beliefs. Parental beliefs reflect indirect socialization because both reflect wider held cultural held beliefs (Halberstadt & Lozada, 2011). Parental beliefs also help guide parental behavior and influence subsequent child outcomes (Eisenberg, Cumberland & Spinrad, 1998). Research further shows there are cultural differences in these parental beliefs (Halberstadt & Lozada, 2011). Documenting the beliefs of parents with regard to crying can help understand differences in the direct socialization of crying and more important indirect socialization of children. It can also help generate hypothesis for future research. The semi-structured interview shall focus on parental theories regarding the meaning of crying, normal development of crying, socialization of crying and gender differences in crying because these topics are expected to influence parental socialization and reflect cultural theories about crying.
The present study
This study is the first to describe age-, cultural- and gender differences in crying frequency and the antecedents of crying. To this aim, the longitudinal development of crying frequency and the antecedents of crying will be assessed cross-sectionally in children aged 3 to 7 from three different cultural backgrounds (Dutch, Moroccan-Dutch, and Xhosa). This will be done by asking parents to report on their children’s crying and their own behavior. Age, gender, and cultural differences are likely related to emotion regulation and social norms (display rules, value orientations and gender stereotypes). Children develop emotion regulation and learn social norms for an important part between 3 and 5 years old and continue to develop after that. It is thus expected crying frequencies gradually decline and cultural and gender differences become visible between children aged 5. Because differences in crying seem to be related to social norms, it is expected differences in crying will develop in a pattern consistent with these norms. Older
children are thus expected to cry less because of emotions which are deemed childish like fear or frustration. Boys are expected to cry less than girls and less because of antecedents that indicate weakness or openness to other people's emotions. Dutch children are expected to cry the most, followed by Xhosa-South African and last Moroccan-Dutch children.
Besides assessing the development of the antecedents and the prevalence of crying the exploration of the possible effects of parents on the development of differences in crying is a goal of this study. To this end, parents will report their response and the goal of this response the last time their child cried. Parental reports on their reaction and goals will be coded into categories which will be based on these reports. It is expected parental responses and the goals of these responses are related to crying frequency, the antecedents of crying, child’s age and the gender of the child and parents’ cultural background. As research shows, parental behavior related to emotion expression, in general, is related to these characteristics.
Finally, a semi-structured interview is with a sub-sample of parents is part of this study. The goal of this interview is to describe parental beliefs about crying. This can help generate hypotheses for future research. In the semi-structured interview the meaning, socialization, development and gender differences in crying will be targeted.
Method
Participants and procedure
Using a questionnaire and a semi-structured interview, data were collected from parents with a Dutch, Moroccan- Dutch and Xhosa- South African cultural background. Participants (n=512) were approached in public spaces to participate in the study. Parents willing to participate filled out a questionnaire after which each participant was asked to be interviewed until the sample for the semi-structured interview (n=61) included 20 participants from each cultural group.
The Dutch group (n=178, Mage= 33.76 SD = 3.78 / interview: n=21, Mage= 33.38 SD = 2.711) consisted of parents living in the Netherlands born to parents with a Dutch nationality. Included in the Moroccan-Dutch group (n=164, Mage=32.20 SD=3.50 / interview: n=21, Mage= 32.05 SD=4.13) were parents living in the Netherlands who have at least one parent born in Morocco. Both groups of parents from the Netherlands were approached in shopping malls in Utrecht and outside a zoo in Rotterdam. Both Utrecht and Rotterdam are relatively large urban cities. The Xhosa-group (n=170, Mage=27,20 SD = 3.83 / interview: n=19, Mage= 28.74 SD = 3.081) consisted of parents whose primary language is “isiXhosa". The Xhosa sample was the only subsample not equally distributed and consisted of more mothers (104 mothers and 66
fathers compared to 98 and 80, and 85 and 79 for the Dutch and Moroccan-Dutch sample) reflecting the high number of mothers raising their children alone (Hosegood & Madhavan, 2012). The Xhosa sample was recruited in shopping malls, a young mothers only church gathering and at train stations. Recruitment took part in the most affluent neighborhoods of Cape Town.
The questionnaire/interviews were conducted by the author of this dissertation, a Caucasian male, and in the introduction to the questionnaire, it was mentioned that the study was part of the interviewer's university education. About 25% of the people who were approached were willing to participate in the study in all three cultural groups. People mentioned a lack of time for not willing to participate. All participants who completed the questionnaire were asked to participate in the semi-structured interview until the sample size for each cultural group was 20. About 50% of the participants agreed to also be interviewed. Again, participants were unwilling to participate mostly because of a lack of time. An English version of the questionnaire was used for the Xhosa group, and a Dutch translation of the questionnaire was used for the Dutch or Moroccan-Dutch group. In case language comprehension of one of these two languages was insufficient, the participant was removed from the sample. This was done 11 times for a Xhosa-South African participant and twice for a Moroccan-Dutch participant.
Instruments
Questionnaire
Crying frequency was operationalized by asking parents the estimated number of the times his or her child cried in the last 4 weeks per week.
The cause for crying was operationalized by the following question: why did your child cry the last time you saw him/her crying? The answers to this open question were categorized afterward, which resulted in the following six codes: 1) physical pain, 2) frustration, 3) anger, 4) sadness, 5) anxiety and 6) shame. These categories were chosen to reflect best the causes for crying as stated by parents.
The response of parents to crying was operationalized by asking how they reacted the last time they saw their child crying and why they responded this way. The answers to these questions were categorized afterward and resulted in 2 different categories of codes: one about parents’ response which was coded as positive (soothing, problem-solving, rewarding behavior), negative (punitive, correcting, limiting behavior) or neutral/mixed (…). The other set on the goal of parents’ response which contained the following codes: 1) focused on parental interest/well-being, 2) decrease of distress child, 3) learning child to cope with the emotion that
caused crying, 4) taking away the situation that caused crying, 5) learning child how to deal with the situation that caused crying and 6) reduce the effects of crying for the social environment.
The answers to the three open questions in the questionnaire were coded into categories. The categories were chosen primarily to reflect answers as closely as possible. Secondarily other research was taken into account in choosing the categories. The categories chosen in the coding of the goal of parental response reflects the factors used in the Coping with Children's Negative Emotions scale by Fabes, Eisenberg and Bernzweig (1990).
Semi-structured interview
The semi-structured interview was used to explore parental theories and beliefs about the meaning of crying, development of crying, socialization of crying and gender differences in crying. The following topics were chosen to explore these themes: 1) differences between baby's and adults crying and that of children, 2) potential parental worries related to their child’s crying, 3) parental influence on the development of crying 4) gender differences in crying. This topic list was used to formulate open-ended questions framed at capturing parental values and beliefs with regard to the scope of the semi-structured interview. Different questions and probes were used depending on the course of the interview trying to cover each topic sufficiently. Often used questions, their answers, and corresponding categories can be found in Appendix 1. In all cultural groups at least the last six interviews did not generate new codes suggesting data saturation had occurred.
Coding of the semi-structured interview.
After consent, the interviews were recorded and also notes were taken. The audio file was used to transcribe thematic units. When the audio recording was unclear, the notes taken were used to reconstruct the thematic units. The thematic units were then coded in close resemblance to their content and categorized in broad categories on the basis of their theme. The coded thematic units within a category were then divided into subcategories by their content. The categories and subcategories can be found below. After choosing the categories and subcategories all thematic units were coded again using the new categories. This process resulted in the following categories and codes:
1) Parental attitudes about crying a. Negative.
b. Positive
a. Parental comfort: parents aim to prevent/end crying because of the nuisance it causes them.
b. Develop coping: parents aim to develop children’s coping, so they cry less. c. Limit distress: parents aim to prevent or shorten the suffering, or abnormal state
crying indicates in the child
d. Develop socially functional behavior: parents aim to teach their children behavior that meets social expectations and promotes relationships.
3) Parental theories regarding the development and socialization of crying.
a. Prevention: children cry less frequently because their own or parents’ behavior prevents possible antecedents of crying.
b. Emotional reactivity: development changes in, and parental influence on, emotional reactivity reduces crying.
c. Better knowledge/ learning rules: children cry less because they have learned different ways to communicate and when it is appropriate to cry.
d. Aging: children cry less because it’s no longer age appropriate behavior. 4) Beliefs regarding the meaning of crying
a. Supernatural influence: parents state their children’s crying is caused by influences outside the child that are for example demonic or spiritual.
b. Communicative behavior: parents understand their children’s crying as involuntary signal
c. Functional behavior: parents understand crying as a behavior that children consciously engage in to stimulate their parents to do something.
d. Emotional response: parents understand crying as a reflection of a child’s (overwhelming) emotions.
Examples of the how the interview was coded can be found in appendix 1.
Analysis of the structured interview
The number of participants was counted that used a code within their interview. This was done irrespective of how often a participant used the code in an interview. Because whether a participant used a code was counted it was possible participants were counted using multiple codes from the same category. It was expected the number of participants using a code reflects its relative importance. Because participants could name multiple codes within a category it was not possible to compare frequencies using chi-square analyses as this test assumes independence of observations. The results of the semi-structured interview were thus only described quantatively but not statistically analyzed.
Results
Effects of Age, Culture and Gender on Crying Frequency
Differences in crying frequency were examined using multivariate analyses of variance. This type of analysis assumes crying frequency has a normal distribution and the variance of crying frequency is equal across groups. To meet these assumptions two steps were taken. First crying frequency was transformed to its common logarithm to limit the skewness and kurtosis of crying frequency. Before the transformation the z-score for kurtosis was 4.07, and the z-score for skewness was 10.61, after the transformation the z-score for kurtosis was -1.46 and for skewness was -0.44. These scores indicate before transformation the distribution had significant kurtosis and skewness (p < .001) and after transformation the scores for kurtosis and skewness were no longer significant (p > .05). This means the transformed distribution of crying frequency meets the assumption of normality. Second, to meet the assumption of equal variances across groups two separate analyses of variance were performed instead of one analysis that included all the effects. One analysis that included the effects of culture and age on crying frequency and one that included the effects of gender and age on crying frequency.
The development of crying frequency by age and culture can be found in Figure 1. The effects of age and culture on crying frequency was tested using a two-way analysis of variance This test showed a main effect for age, (F(4, 497) =31.11, p < .001). A Pearson product-moment correlation coefficient was computed to asses this relation. As expected, older children cry less frequent than younger children as shown by a moderate negative correlation between crying frequency and age (r =-.43 p = < .001; see also Figure 1). There also was a main effect for culture, (F(2, 497) = 32.11, p < .001) which indicates there are differences in the crying frequency between the cultural groups. Post hoc tests with the Bonferroni correction showed Xhosa children cry less than Dutch and Moroccan-Dutch children (p < .001), no difference was found between Dutch and Moroccan-Dutch children (p = .62). Unexpectedly, there was no Culture × Age interaction effect on crying frequency, (F(8, 497) = 0.70, p = .70).
Figure 1. Mean crying frequency at age 3 to 7 year for Dutch, Moroccan-Dutch and Xhosa children
The crying frequency by gender and age can be found in Table 1. It was hypothesized that gender differences in crying frequency develop from age 5. With a two-way analysis of variance the main effects of gender and age as well as the Gender × Age interaction on crying frequency were examined. The above described main effect of age was logically confirmed, (F(4, 502) = 31.78, p < .001). There was a main effect of gender, (F(1, 502) = 5.27, p = .02) There also was a Gender × Age effect on crying frequency, (F(4, 502) = 2.51, p = .04). Simple main effects analysis showed gender differences in crying frequency are only present between children aged 7 (F(1, 502) = 12.83 p < .001), partially confirming the hypothesis gender differences develop from age 5. As can be seen in Table 1, at age 3 to 6 boys and girls cry with approximately the same frequency, whereas at the age of 7 girls cry significantly more frequent than boys.
1 2 3 4 3 4 5 6 7 M ean cr y in g f req u en cy Age in years Dutch Dutch-Moroccan Xhosa
Table 1
Boys and girls crying frequency at age 3 to 7 year
Boys Girls Total
age M n SD M n SD M n SD 3 3.57 49 2.20 3.25 53 2.06 3.40 102 2.12 4 2.38 47 1.70 2.45 53 1.69 2.42 100 1.69 5 2.20 56 1.52 2.46 54 1.57 2.33 110 1.55 6 1.57 51 1.22 1.72 53 1.12 1.64 104 1.17 7 .78* 49 .80 1.47* 47 1.00 1.11 96 .96 total 2.10 252 1.80 2.29 260 1.66 2.19 512 1.73
Note. This table reports the original data before transformation.* p = > .001
Effects of Age, Gender and Culture on Reason for Crying
To test for effects of age, culture, and gender on reason for crying, log-linear analyses were carried out. Log-linear analyses test whether the deletion of an effect from a model affects the fit of that model. Two kinds of statistics are reported for this analysis in this study. The likelihood ratio chi-square statistic of the model, which reflects whether the frequencies predicted by the model are different from the observed frequencies. When the model predicts the data correctly this statistic is insignificant. And the chi-square statistics for individual effects, which reflects whether the distribution of an effect is different from the expected distribution. When the distribution is significantly different the effect cannot be deleted from the model without affecting the fit of the model and is retained.
Log-linear analyses have the assumption that the expected frequency for each category is at least 1 and that less than 20% of categories have an expected frequency below 5. To meet this assumption the number of categories compared in the analyses had to be limited. This was done in two ways. First the variable children’s age was categorized into two groups. One group including children aged 3 and 4 and one group including children aged 5, 6 and 7 years old. This particular division was chosen because gender and cultural differences in reason for crying are expected to develop when children turn 5 years old. Second, instead of analyzing Age × Culture × Gender × Reason for Crying in one analysis, two separate analyses were carried out: one for Gender × Age × Reason for Crying and one for Culture × Age × Reason for Crying.
In the first analysis of Gender × Age × Reason for Crying the final model did not retain the highest order three-way interaction effect of Gender × Age × Reason for Crying (χ2 (5) = 5,83 p = .32). The final model with a likelihood of ratio χ2 (6) = 6.59, p = .36 did retain the Age × Reason for Crying interaction (χ2 (5) = 22.90, p < .001) and the Gender × Crying interaction (χ2 (5) = 98.25, p < .001). This indicates there are gender and age differences in reason for crying.
Odds ratio and standardized residuals were calculated to break down this result. Boys had 2,3 times the odds of girls to cry because of anger and pain and girls had 1,8 times the odds of boys to cry because of anxiety and sadness. Based on the odds ratio children aged 3 and 4 compared to children aged 5 to 7 year have 4,6 times the odds likely to cry because of frustration and anxiety. Children age 5 to 7 compared to children age 3 and 4 year on the other hand have 16,6 times the odds to cry because of shame. Percentages of reason for crying separated for gender and age can be found in Table 2. In this table the categories with a standardized residual higher than 1.96 are marked, this indicates the observed frequency was different from the expected frequency at a significance level of p < .05.
Table 2
Reason for crying by age and gender
Boys Girls 3-4 years olds 5-7 years old
Pain 31.0% 20.4% 22.8% 27.4% Frustration 12.7% 15.0% 23.3%* 7.7%* Anger 17.9%* 8.8%* 12.9% 13.5% Shame 12.7% 19.2% 2.0%* 25.2%* Sadness 11.5% 18.1% 12.4% 16.5% Anxiety 14.3% 18.5% 26.7%* 9.7%* * standardized residual > 1.96, p < .05
A second three-way log-linear analysis was carried out to test Culture × Age × Reason for Crying. This analysis produced a model that retained all effects. The likelihood ratio of this model was χ2(0) = 0.00 p = 1. This indicates the Age × Culture × Reason for Crying interaction was significant (χ2(10) = 26.53, p = .003). To break down this effect separate chi-square tests were performed on culture and reason for crying, for children aged 3 and 4 and for children aged 5 to 7. These tests showed there are no cultural differences in reason for crying between children aged 3 and 4 (χ2(10) = 14.93, p = .13), but there are cultural differences in reason for crying between children aged 5 to 7 (χ2(10) = 19.02, p = .04). This confirms the hypothesis
cultural differences develop from age 5. Reason for crying by age and culture can be found in Table 3. In this table the differences between Dutch and Moroccan-Dutch children on frustration and Dutch and Xhosa children on anger stand out. Moroccan-Dutch children aged 5 to 7 have 5.8 times the odds of Dutch children aged 5 to 7 to cry because frustration. Dutch children aged 5 to 7 have 2.9 times the odds of Xhosa children aged 5 to 7 to cry because of anger. Standardized residual were calculated for the percentages in table 3 but none were higher than 1.96.
Table 3
Reason for crying separated by age and culture
3-4 year olds 5-7 year olds
Dutch Moroccan-Dutch Xhosa Dutch Moroccan-Dutch Xhosa
Pain 15.9% 23.1% 29.4% 30.3% 23.2% 28.4% Frustration 31.9% 20.0% 17.6% 2.8% 14.1% 6.9% Anger 8.7% 18.5% 11.8% 19.3% 13.1% 7.8% Shame 0.0% 3.1% 2.9% 26.6% 25.3% 23.5% Sadness 8.7% 15.4% 13.2% 14.7% 15.2% 19.6% Anxiety 34.8% 20.0% 25.0% 6.4% 9.1% 13.7%
Effects of Age, Gender and Culture on Parental Socialization.
Parental socialization was measured as parental response to crying (response) and the goal of this response (goal). To test whether there are age, gender and cultural differences in parental socialization two three-way log-linear analysis were carried out: one testing Gender × Response × Goal and one testing Culture × Response × Goal. The effect of age on response and goal was tested using two chi-square analysis because Age× Response × Goal, did not meet the assumptions for a log-linear analysis.
The three-way log-linear analysis testing Gender × Response × Goal, did not retain Gender × Response × Goal, as exclusion of this three way interaction effect did not significantly worsen the model fit (χ2 (10) = 3.50, p = .97). Similarly, the two way interactions of Gender x Response (χ2 (2) = 2.28, p = .32) and Gender x Goal (χ2 (5) = 5.97, p = .31) were removed from the model. This indicates that there are no gender differences in parental socialization. The final model with a likelihood ratio of χ2 (18) = 11.85, p = .86 did retain Response x Goal (χ2 (5) = 69.28, p < .001). This indicates parental response is related to parental goal.
The three-way log-linear analysis testing Culture × Response × Goal, did not retain Culture × Response × Goal (χ2 (20) = 25.85, p = .17). The final model with a likelihood of (χ2
(20) = 25.85, p = .17) did retain Culture x Goal (χ2 (10) = 31.66, p < .001), Culture x Response (χ2 (4) = 23.35, p < .001) and Response x Goal (χ2 (10) = 67.25, p < .001). This indicates that there are cultural differences in parental response and parental goal. The cultural differences in parental response seem to be caused by the fact Xhosa parents more often react negative. The cultural difference in parental goal seems to be caused by the fact Dutch parents more often have parental interest as a goal and less often have limit distress as a goal. Parental goals and parental response for each cultural group can be found in table 4. Categories with a standard residual higher than 1.96 were marked in this table.
Table 4
Parental response and parental goal for culture
Dutch Moroccan-Dutch Xhosa Parental response Positive 43.8% 54.3% 40.6% Negative 39.9% 32.9% 54.7%* Neutral/mixed 16.3%* 12.8% 4.7%* Parental goal Parental interest 14.0%* 9.1% 4.1%* Limit distress 19.7%* 33.5% 35.3% Develop coping 28.7% 22.0% 25.3% Solve problem 6.7% 3.7% 5.9%
Develop problem solving 24.7% 22.6% 14.7%
Social environment 6.2% 9.1% 14.7%*
* standardized residual > 1.96 p < .05
The relation between children’s age and their parents’ response and goal was tested using two chi-square analyses. It was found there are no differences in parents their goal (χ2 (20) = 25.72, p = .18) and response (χ2 (8) = 4.22, p = .84) when children of different ages cry.
Effects of Parental Socialization on Crying Frequency
It was expected that parental socialization (i.e. parental goal and parental response) explains cultural and gender differences in crying frequency. This hypothesis was tested separately for gender, culture and age. The effects of parental socialization on differences in crying frequency were evaluated using 3 three-way independent analysis of variance, one testing the effects of Culture x Response x Goal, one testing Gender x Response x Goal and one testing
the effects of Age x Response x Goal . In the analyses the natural log of crying frequency was used to meet the assumption of normality.
The first three-way analysis of variances that was carried out tested the effect of culture, response and goal on crying frequency. This analysis found the main effect of goal (F(5, 461) = 1.90, p = .09), the main effect of response (F(2, 461) = 0.37, p = .69) the interaction effect of Culture x Response x Goal (F(17, 461) = 1.34, p = .16) and the interaction effect of Culture x Response (F(10, 461) = 1.00 p = .44) were all insignificant. A significant interaction effect of Culture x Goal on crying frequency was found (F(10, 461) = 2.00 p = .03). Simple main effects analysis showed the parental goals: social environment (F(2, 461) = 3.88, p = .02) and develop problem solving (F(2, 461) = 5.42, p = .005), are related to cultural differences in crying frequency. Pairwise comparisons showed Dutch children whose parents had develop problem solving as a goal cried more often than Xhosa (p = .002) and Moroccan-Dutch children (p = .03). Pairwise comparisons further showed Xhosa children whose parents had social environment as a goal cried less than Dutch (p = .015) and Moroccan-Dutch children p= .010). This suggest the parental goal develop problem solving is related to more crying in Dutch children and the parental goal social environment is related to less crying in Xhosa children.
The second three-way multivariate analysis of variances tested the effects of gender, goal and response on crying frequency. The insignificant main effects of goal (F(5, 477) = 2.10, p = .06) and response (F(2, 477) = 0.62, p= .54) confirmed the results of the analysis of variances including culture. There also was no main effect of gender, (F(1, 477) = 2.07, p = .15). Further, the interactions Gender x Response x Goal (F(9, 477) = 1.21, p = .28), Gender x Goal (F(5, 477) = 0.89, p = .47) and Gender x Response (F(2, 477) = 2.52, p = .08) were all insignificant. In sum these results suggests gender differences in crying are not related to parental socialization and there is no effect of parental socialization on crying frequency. This was not expected.
The third three-way multivariate analysis of variances tested the effects of age, goal and response on crying frequency. The main effect of age (F(4, 432) = 12.99, p < .001) was significant, confirming the results of earlier analyses. No combined effect of age and parental socialization on crying frequency was found. The interaction effects: Age x Goal (F(20, 432) = 1.32, p = .16), Age x Response (F(8, 432) = 0.16 p = .995) and Age x Response x Goal (F(30, 432) = 0.79 p = .78) were all insignificant. The other effects in the analysis confirmed the results of the other two analyses in this section.
Effects of Parental Socialization on Reason for Crying
It was expected that parental socialization (i.e. parental goal and parental response) explains age, cultural and gender differences in reason for crying.
It was not possible to test the combined effect of parental response and age, culture or gender on reason for crying because for each corresponding log-linear analysis more than 20% of the categories had an expected frequency below 5. This assumption not being met causes a significant loss of power rendering the analysis useless. Instead a chi-square analysis was performed on parental goal and reason for crying. This analysis found parents their goals are different when children cry for different reasons, (χ2 (25) = 64.68, p = p < .001). This seems to reflect parents have different goals when children cry because of frustration, anger, shame and sadness, see Table 5.
The effect of parental response on age, cultural and gender differences in reason for crying was tested using three three-way log-linear analysis. The first analysis did not retain the highest order effect of Gender x Response x Reason for Crying (χ2 (10) = 11.00, p = .36) in its final model. The second analysis also did not retain the highest order interaction Culture x Response x Reason for Crying (χ2 (20) = 17.82, p = .60). And the third analysis also did not retain its highest order Age x Response x Reason for Crying interaction (χ2 (20) = 17.82, p = .60). Combined the results of these three analyses suggest parental response is not related to cultural, gender and age differences in reason for crying. All three analyses did retain Response x Reason for Crying (χ2 (10) = 37.88, p = p < .001) in their final model. This indicates parents respond different when their children cry for a different reason. This seems to reflect parents respond negative more often to anger, see Table 5.
Table 5
Parental response and goal for reason for crying
Pain Frustration Anger Shame Sadness Anxiety Parental response Positive 51.9% 40.8% 20.6%* 50.0% 47.4% 57.1% Negative 43.5% 39.4% 61.8%* 35.4% 43.4% 34.5% Neutral/mixed 4.6%* 19.7%* 17.6% 14.6% 9.2% 8.3% Parental goal Parental interest 9.2% 16.9%* 8.8% 11.0% 6.6% 3.6% Limit distress 33.6% 8.5%* 22.1% 25.6% 44.7%* 35.7% Develop coping 21.4% 16.9% 38.2%* 20.7% 27.6% 31.0% Solve problem 6.1% 9.9% 7.4% 2.4% 1.3% 6.0%
Develop problem solving 21.4% 32.4%* 14.7% 31.7%* 9.2%* 14.3%
Social environment 8.4% 15.5% 8.8% 8.5% 10.5% 9.5%
Parental beliefs about crying
Semi-structured interviews were conducted to explore the parental beliefs related to crying. A description of the categories and codes used in the coding of the interview can be found in the methods sections. An overview of the number of interviews each code occurred in can be found in table 6. Multiple codes from a single category could be used in the same interview. Percentages thus reflect the relative importance of a code within a cultural group. In this section, notable trends in the interviews and results not captured by the codes will be highlighted.
In the semi-structured interviews, it became clear parents irrespective of their cultural background do not want their child to cry. This was captured by the code negative attitude. Often parents would also explain their negative attitude. They, for example, told: "I don't like it when my child cries because it should have nice life." This explanation was categorized as a parental goal because the explanation overlapped with the goals parents had. For example "I want my child to have friends" and "I don't want my child to cry because it will be harder to make friends" were both coded as develop socially appropriate behavior although in the second quote parents not specifically mention a goal.
In the category parental goals, Xhosa parents were the only ones to mention they should prevent their child from crying by protecting it from harm. One Xhosa parent, for example, told: as a parent, it is your obligation to make sure your child has nothing to cry about. The idea that it is parents’ duty to prevent and protect their children from harm was often mentioned by Xhosa parents but not by Dutch and Moroccan-Dutch parents. This is not visible in the codes because this notion of protecting your child was coded as limit distress, a code that also captured the idea children shouldn't experience distress without stating parents’ role in protecting a child from harm. A Dutch parent, for example, said: "children cry when they are in pain. And they should not be in pain, right?” Dutch parents, compared to the other parents mentioned the goal: parental comfort often. One Dutch father explained: "it is annoying when X cries because daddy needs to work."
In interviews, parents mentioned both beliefs about the developmental changes in children that affect crying and beliefs about how parents think they can influence the development of crying. In the coding process, similar codes came forward in both these groups of beliefs. Therefore, both were coded into the category: Parental theories regarding the development and socialization of crying. Compared to Dutch parents some of the Moroccan-Dutch and Xhosa parents had particular ideas about the effect age, stating the effect of age is not gradual but involves milestones after which different things can be expected from children. They, for example, mentioned: "When Y becomes a boy he can’t longer cry about everything”. Dutch parents (n=2) only mentioned going to school or daycare as a milestone after which children
should behave differently. Moroccan-Dutch (n=6) and Xhosa parents (n=5) mentioned a transition between early childhood and childhood, usually around age 5 after which different things could be expected of children. Both a gradual change and transitional change were coded as aging
Parents from all three cultural groups understand crying as an emotional response. In the interviews, they, for example, mentioned crying to reflect pain, anger, sadness. Other aspects of crying were also named by parents. Xhosa parents often highlighted the communicative aspects of crying as a call to action (communicative behavior). Also some Xhosa parents mentioned crying as a result of supernatural influences on children. One Xhosa parent, for example, told she prayed when her child cried to free it from demonic influences. Especially young children were believed to open to supernatural influences like demons and bad spirits by Xhosa parents. Dutch and Moroccan-Dutch compared to Xhosa less often mentioned the communicative aspect of crying and did not mention supernatural influences. Instead, Dutch and Moroccan-Dutch seem to understand crying as deliberate behavior this categorized as functional behavior. One mother for example mentioned “she cries, so her daddy lets her stay up late” and “he cries to get help”. Possible cultural models underlying these results shall be highlighted in the discussion.
Table 6
Parental beliefs on socialization of crying percentage of parents separated for cultural group
Dutch Moroccan-Dutch Xhosa
Category Code N=21 N=21 n=18 Parental attitude Negative 95% 90% 100% Positive 20% 19% 11% Parental goals Parental comfort 33% 5% 6% Develop coping 24% 19% 44% Resolve 43% 52% 78% Social competence 43% 57% 44%
Theories on changes Prevention 19% 43% 44% Emotional reactivity 33% 43% 33% Advanced knowledge 48% 29% 33% Aging 19% 52% 50% Meaning of crying Supernatural influence - - 16% Communicative behavior 19% 33,% 67% Functional behavior 62% 48% 17% Emotional response 86% 90% 89% Discussion
The aim of this study was to document the development of crying in children aged 3 to 7 year. Special attention was paid to the development of gender and cultural differences in crying and the role of parents in the development of crying. This study shows crying declines from on average 3.4 times per week when children are 3 years old to 1.1 times a week when children are 7 years old. Parents reported children cry for the following reasons: psychical pain, frustration, anger, sadness, anxiety, and shame. Children aged 3 and 4 compared to children aged 5 to 7 cry more often because of frustration and fear and less because of shame.
Gender and cultural differences in crying were also visible. Boys cried more often than girls because of pain and anger and less because of shame and anxiety. Gender differences in crying frequency were only found between boys and girls aged 7. Xhosa children cried less than Dutch and Moroccan-Dutch children. Children’s age did not affect these cultural differences. Children’s age did affect cultural differences in reason for crying. Cultural differences in reason for crying were only found between children aged 5 to 7 years old and not between children aged 3 and 4. The observed gender and cultural differences are similar to what could be expected on the basis of social norms and seem to confirm their role.
Parental socialization was operationalized as parents’ self-reported response to crying and the goal of this response. Parental response to crying was classified as positive, negative or neutral/mixed. The goal of parents their response to crying was classified as: focused on parental interest/well-being, decrease of distress child, learning child to cope with emotion that caused crying, taking away situation that caused crying, learning child how the deal with the situation
that caused crying or reduce the effects of crying to the social environment. Cultural differences were found in parental response and parental goal. For parental response, Xhosa parents their response was more often classified as negative. For parental goal, Dutch parents more often mentioned parental interest and less mentioned often limit distress. Also Xhosa parents more often mentioned social environment and less often mentioned parental interest. Cultural differences in parental goal were related to crying frequency. Dutch children whose parents had develop problem solving as a goal cried more often and Xhosa children whose parents had social environment as a goal cried less often. No relations were found between parental goal or parental response and gender differences in crying frequency, gender differences in reason for crying or cultural differences in reason for crying.
A semi-structured interview was part of the study to describe parental beliefs about crying. Parents thought of crying primarily as an emotional response and have multiple not mutually exclusive goals related to crying: They want to limit the distress crying indicates, want their child only to cry in socially appropriate situations, want their child to develop emotional resilience and want to limit the nuisance crying causes them. These goals reflect the goals found by the questionnaire. When asked why crying changes, parents named the development of coping, better knowledge of how to behave, the development of self-reliance and changes in social expectations related to age as factors that affect differences in crying.
Age
As expected children cry less when they grow older and their reasons for crying change. This was expected because between 3 and 7 year children develop emotion regulation and learn social norms. It was argued that when children develop emotion regulation they can better control crying and thus cry less. It was further argued in which situations children control crying is dependent on social norms. When children learn social norms and have developed a sufficient level of emotion regulation they will thus cry less when social norms stipulate crying is not appropriate. And thus their reason for crying changes when there are differences in social norms. The data confirms reason for crying changes when children grow older, but suggests the influence of social norms is more complex than expected.
In the semi-structured interview, multiple parents mentioned children temporarily cry more often because of anxiety when they first go to school or kindergarten. It was also observed children aged 3 and 4 cry more often because of anxiety. It is possible, children aged 3 and 4 year cry more often because of anxiety because they first go to kindergarten or school. It was further observed children aged 5 to 7 cry more often because of shame. Eccles (1999) argues middle childhood is an important period for the development of self-esteem. It possible, this is
why, this study found children cry more often because of shame in this lifespan. In sum, it seem possible reason for crying also reflects the particular developmental challenges children face at particular ages. Beforehand this was not expected. In part this was caused by this study's focus on cultural and gender differences in crying. These differences seem to be more likely explained by social norms and thus relatively little attention was paid to the broader development of children. The age related differences in crying are however more likely caused by children’s broader development. The data suggest at different ages children struggle with different situations in which they learn new skills, situations like being separated from their parents, making friends and going to school. Being in these situations increases the change they cry for a particular reason. It seems possible the course of these situations can explain age related changes in reason for crying. More research is needed to confirm this pattern. Such research could show whether the development of control over crying is specific for different emotions and contexts, or acts independent of such factors.
Gender differences
It was expected gender differences in reason for crying and crying frequency develop from age 5. This hypothesis was partially confirmed. Gender differences in both crying frequency and reason for crying frequency were observed. However, gender differences in crying were only observed between 7-year-old's and age did not influence gender differences in reason for crying.
It is possible no gender differences in crying frequency were found between 5 and 6-year-olds because the development of emotion regulation is slower for boys. On average more boys than girls their development of emotion regulation lags behind (Montroy, Bowles, Skibbe, McClelland & Morrison, 2011). It is likely the boys whose development lags behind cry more often when they are 5 and 6 and offset the difference made by social norms.
This does only partly explains why no combined effect of gender and age on reason for crying was observed. It is possible the effect of social norms on reason for crying was not visible because of the effect of children’s broader development on crying. As was argued before, children are more likely to cry because of certain reasons depending on the developmental challenges they face. There for example seems to be a higher chance children cry because of anxiety when they first go to school. Both boys and girls first go to school at the same age and thus have a higher chance of crying for the same reason at the same time. Other developmental challenges might also cause boys and girls to cry for the same reason at the same age. In comparisons with the effect of these common age related changes in reason for crying the combined effect of age and gender on reason for crying is small. It is possible the small combined effect of gender and age was not visible due to the big effect of age on reason for
crying. It is possible; a study with a higher power would have found a combined effect of age and gender on reason for crying.
Cultural differences
Cultural differences in crying frequency were expected to develop from age 5. However, age did not influence the size of the cultural differences in crying frequency. It was found Xhosa children cry less than Dutch and Moroccan-Dutch children. No difference was found between Dutch and Moroccan-Dutch children. This was also not expected. The hypothesis cultural differences in reason for crying develop from age 5 was confirmed by the data.
Difference between Dutch and Moroccan-Dutch children’s crying frequency were expected because research shows adults cry more often in countries that value independence (van Hemert, van de Vijver & Vingerhoets, 2011). As Dutch parents value independence more than Moroccan-Dutch parents, it was expected Dutch children cry more often. It is likely no difference in crying frequency was found because the differences in value orientation between Dutch and Moroccan-Dutch are small (Huijbregts, Tavecchio, Leseman, & Hoffenaar, 2009). Also, it is possible Moroccan-Dutch parents with a more traditional cultural background were less willing to participate in the study making the difference even smaller. In combination with the low power of this study, this could explain why no difference was found between Moroccan-Dutch and Moroccan-Dutch children.
It was expected cultural differences in crying frequency develop from age five because children then learn social norms. No effect of age on the size of cultural differences was however found. It is possible no effect of age on the size of cultural differences in crying frequency was found because parents influence crying by regulating younger children’s crying. It is likely this regulation mimics the effect of social rules that affect the crying of older children because parents and children share the same social norms. Put simpler, parents help children to regulate crying when children are young. When they help their children they employ social norms and their children learn these social norms. When children grow older, they start to regulate their own crying and will apply the social norms they have learned when their parents helped them regulate crying. Because parents and children apply the same social rules when they regulate crying, there is no effect of children learning social norms. This reasoning is far-fetched but could explain why cultural differences in crying frequency are found both before and after children learn social norms. Children's reason for crying is however affected by age. This makes this argument less convincing. Also, it is unclear to what extent parents help their children not to cry and it is likely socials norms are different for children of different ages. In the semi-structured interview, parents from each cultural group told they have different expectations with regard to crying
