Keywords

Introduction

In studies of children’s language, gender is a factor which is often taken into account in individual variations. However, gender difference in child language is still a debatable issue. Some findings highlight the superiority of one gender over the other, while other studies show that gender difference has no effect on language use. For example, in language assessment, girls outperformed boys (Bornstein et al., 2004), as well as with narrative quality (Fey et al., 2004). Lange (2016) performed several tests on German children in order to assess their language competence (vocabulary, grammar, speech comprehension, and other linguistic skills.). The study found that girls were significantly more advanced in all domains, whereas boys varied considerably in their language competence. This finding was similar to that of Razmjoee et al.’s (2015) study on Iranian children. On the other hand, Roberts (1994) found that when it comes to the use of the standard variant of a language (in this case, English), boys are more advanced than girls. Barbu et al.’s study (2015) focused on the relationship between gender, socioeconomic status (SES), and language development in a sample of French children (ages 2–6). The results showed a significant difference in the language competence of the low-SES boys and girls, whereas no significant difference was found among the high-SES boys and girls. In addition, the performances of the low-SES boys were poor, while the performances of the low-SES girls were intermediate. Conversely, the performances of the high-SES boys and girls were higher than their low-SES counterparts.

Regarding narrative skills, the issue of gender superiority is also still debatable. For example, in language assessment, girls outperformed boys (Bornstein et al., 2004), as well as with narrative quality (Fey et al., 2004). According to Barra and McCabe’s (2013) study on Chilean children (aged 4–6), girls are more fluent in telling stories than boys. In a study of African American preschoolers’ oral narrative skills, Gardner-Neblett and Sideris (2018) found that boys had lower scores than girls. Yet, studies by Sperry and Sperry (1996) found that Afro-American boys talk more, producing more stories than do Afro-American girls. At the same time, Champion and Maines (as cited in Horton-Ikard, 2009) found that children who use African American English do not violate gender restrictions in terms of pronoun usage in narratives.

Studies on children’s narratives focusing on gender differences, as well as on competence in using cohesive devices, are quite plentiful. However, studies focusing on both issues—gender differences and cohesion competence—are still rare, let alone with Indonesian data. The present chapter discusses Indonesian-speaking young children’s narrative production focusing on the gender difference in grammatical cohesive devices. This study departs from a question: do girls and boys have differences in using cohesive devices in narratives? Our participants were children aged 4–6 years old enrolled in a national curriculum-based preschool located in Bogor, West Java. This chapter is part of the first author’s (Ferhadija, 2017) thesis, under the supervision of the second author. The remainder of this chapter includes a brief theoretical review concerning factors dealing with language development, followed by a discussion on cohesive devices and a description of the research method and findings. The final section presents the conclusions as well as the recommendations for future studies.

Narrative Ability and Cohesive Devices

As already mentioned previously, the focus of this chapter is the use of grammatical cohesive devices. This section presents a brief discussion on cohesion, preceded by a brief discussion on the influencing factors in language development.

Language Development and the Influencing Factors

In language development, several factors should be considered, namely culture, socioeconomic status, and gender. These factors deal with the relation between children, their surroundings, and linguistic input. Interactions between children and their caregivers may reflect cultural differences. There are mothers who are more information-oriented and those who are more emotion-oriented. Research by Toda et al. (as cited in Owens, 2012, p. 135) shows that American mothers use more questions and correct grammar in interaction with their children, whereas Japanese mothers use more environmental sounds and baby talk (see also Papalia & Martorell, 2015). Socioeconomic status may also influence language development. According to Owens (2012), middle-class mothers tend to ask more questions and are more verbal, so that their children’s language is stimulated. Meanwhile, lower-class mothers tend to use more imperatives or directives and are less verbal.

Gender difference is also regarded as a factor affecting language development. Fenson et al. (as cited in Rowland, 2014) suggested that in the early years, girls tend to be more advanced in language than boys since girls tend to mature earlier than boys. Furthermore, the gender difference is reflected in the parenting and socialization process. Owens (2012, pp. 135–137) suggested that generally, mothers tend to be closer to their daughters than to their sons. It is also suggested that mothers tend to speak in longer sentences to daughters than to sons. Furthermore, parents tend to talk about different things to boys than they do to girls (Rowland, 2014, p. 211). Nevertheless, Owen (2012) also affirms that the difference is not always related to the child’s linguistic behavior. Findings in the aforementioned studies have addressed this issue.

Several studies show that gender difference can be observed in narrative styles and preferences in children’s language (e.g., Nicolopoulou, 2008; Richner & Nicolopoulou, 2001). Girls tend to tell home-centered stories and establish the characters in their network relationships, whereas boys’ stories are usually marked by conflict, movement, and disruptions and portray disconnected characters (Nicolopoulou, 2008, pp. 310–311).

Development of Narrative Ability

In children’s narrative studies, Hickmann (2003, 2004) suggests that linguistic and cognitive abilities are needed in narrative reference, which includes distinction between deixis and anaphora, and between given and new information. Other requirements are shared information between narrator and interlocutor, and a good command of the theory of the mind (as suggested by Berman & Slobin, 2013; see also Berman, 2009a, 2009b, p. 359).

A child’s ability to narrate a story is usually apparent at preschool age (4–6 years) since language skills rapidly develop during this time period. More specifically, a child at this age is capable of producing simple sentences as well as some more complex sentences with conjunctions (Papalia & Martorell, 2015, p. 221). From the “frog story,” (English-speaking) children’s narrative samples, Berman and Slobin (2013) found three developmental linguistic categories: prepositions, suffix -ing, and the connective and. The findings show that through narrative, children show their abilities with perception of space, time, and order.

As in other aspects of linguistics, the use of cohesive devices becomes interesting when it is associated with a child’s language skills. According to Colozzo and Whitely (2014), grammatical cohesive devices begin to appear when a child is between 5 and 6 years of age. Meanwhile, Hickmann (1995) stated that cohesive devices, in the form of references, also appear when a child reaches the age of 5 or 6. However, a narrative produced by a child at that age is still egocentric in nature. In addition, the use of pronominal references or temporal markers in sentences indicates that he/she still finds it difficult to place him/herself in the audience’s perspective (Hickmann, 1995, p. 106). Thus, in some cases, the use of references does not clearly refer to certain objects. Peterson et al. (1999) showed that at the age of 6, children are quite capable of providing a complete structure of narrative with a clear ending. Meanwhile, five-year-old children can tell an almost complete story, albeit without a clear ending, whereas the four-year-olds tend to jump from one event or point in a story to another, which is a result of their inability to concentrate on a specific reference. Other studies claim that children are actually quite capable of reconstructing an event in a chronological manner and reiterating it in a coherent story (e.g., Thompson & Myers, as cited in Berk, 2006).

Cohesive Devices in a Narrative

According to Halliday and Hasan (1976, p. 4), cohesion is a semantic concept that refers to the relationship between meanings in a discourse. Thus, cohesion can be used to determine whether a written text is actually a cohesive and coherent discourse. Halliday and Hasan (1976) also distinguished grammatical and lexical cohesions. As for grammatical cohesions, they are categorized into five grammatical cohesive devices: references, substitutions, ellipses, and conjunctions. Lexical cohesions refer to the semantic relationship between elements that form a discourse by utilizing lexical elements or words (Yuwono, 2005, p. 98). Lexical cohesive devices are divided into reiterations (consisting of repetitions, synonymies, superordinates, metonymies, and antonymies) and collocations. Since the focus of the present study is on the use of grammatical cohesions, the following provides a more detailed description of such cohesive devices with some examples from Indonesian adult language.

Reference is cohesive device connecting a marker with the object to which it refers. Halliday and Hasan (1976) categorized a reference into two types: exophoric and endophoric. An exophoric reference refers to something outside the text, whereas an endophoric reference refers to something within the text. Based on the position of its reference, the endophora is divided into anaphoric and cataphoric references. An anaphoric reference is located before the referent’s marker, whereas a cataphoric reference is placed after the referent’s marker. Moreover, a reference can be classified according to its marker, i.e., personal reference, demonstrative reference, and comparative reference. In Bahasa Indonesia, personal references are pronouns, namely aku, saya “I,” kamu, Anda “you,” dia “s/he,” kami “we/us,” kalian “2nd. PL,” and mereka “they/them” [see, for example, (1)]. Demonstrative references are demonstratives, such as sini “here,” sana “there,” ini “this,” and itu “that” [see (2)]. Meanwhile, comparative references can be found in sama dengan “same with,” berbeda dengan “different from” [see (3)].

(1)

Personal reference

Kemarin Nana pergi ke pasar. Dia membeli sayur-sayuran.

“Yesterday Nana went to the market. She bought vegetables.”

(2)

Demonstrative reference

Akui baru saja pergi ke perpustakaan dan meminjam buku-buku di sana .

“I just visited the library and borrowed some books there.”

(3)

Comparative reference

Fina berumur dua belas tahun . Umur Zakiyah sama dengan umur Fina.

“Fina is twelve-year-old. Zakiyah’s age is the same with Fina.”

Substitution is a cohesive device replacing one language element with another that includes the same meaning. A substitution is also construed as the relationship between a word and another word it replaces to avoid repetition. Moreover, a substitution is classified according to its form, i.e., nominal substitution (4), verbal substitution (5), and clausal substitution (6). The examples are presented in Indonesian.

(4)

Nominal substitution

Didi sering mengunjungi orang tuanya di Medan. Anak yang baik itu memang selalu berusaha menyenangkan hati mereka.

“Didi often visits his parents in Medan. That kind-hearted boy is always trying to please them.”

(5)

Verbal substitution

Mereka sudah berusaha. Kami juga begitu.

“they had already tried. So did we.”

(6)

Clausal substitution

Keluarga saya sering mengadakan upacara adat. Keluarga suami saya juga demikian.

“My family often holds traditional ceremonies. My husband’s family is the same.”

Ellipsis is a deletion/removal of a language element that can be traced back by referring to the previous element. In some cases, an ellipsis is classified as a zero substitution, as what the speaker is referring to is not restated in the following sentence. Moreover, an ellipsis is classified into three types, i.e., nominal ellipsis (7), verbal ellipsis (8), and clausal ellipsis (9). The examples are presented in Indonesian.

(7)

Nominal ellipsis

Aduk adonan. Diamkan Ø sebentar.

“stir the batter. Let Ø sit”

(8)

Verbal ellipsis

Mereka berteriak. Anak itu Ø juga.

“they shout. The child Ø also”

(9)

Clausal ellipsis

Saya sudah mengerjakan tugas. Dia juga Ø.

“I have done my assignment. He also Ø”

Conjunction is a cohesive device whose function is to link one sentence to another. Unlike other grammatical cohesive devices, a conjunction does not refer to an element that is already in the text. The purpose of a conjunction is to mark the relationship between parts of a text, thereby allowing the text to be understood. Halliday and Hasan (1976) categorized a conjunction into four types, i.e., additive (10), adversative (11), causal (12), and temporal (13). The examples are presented in Indonesian.

(10)

Additive conjunction

Kami tidak bisa membeli roti karena toko-toko sudah tutup. Lagipula, sekarang hampir tengah malam.

“we cannot buy bread as the shops have already closed. Moreover, it is almost midnight”

(11)

Adversative conjunction

Kakinya berdarah. Namun ia berusaha untuk berjalan.

“his foot is bleeding. But he tries to walk”

(12)

Causal conjunction

Keadaannya sudah membaik. Karena itu, dia tidak perlu minum obat.

“his condition is getting better. Therefore, he does not need to take the medicine”

(13)

Temporal conjunction

Ia duduk di sebuah sudut. Lalu ia memesan makanan.

“she sits in the corner. And then she orders some food”

Method

At preschool age, children still find it difficult to grasp abstract concepts (Papalia & Martorell, 2015, p. 269). Thus, the present study employs the images of animals as the research instrument, as children are generally familiar with the concept of animals. This study applied semi-structured elicitation (Eisenbeiss, 2010, p. 21). We prepared some short questions to encourage participants to describe displays of events or objects.

Before collecting the data, the researchers conducted a pilot test with a group of 4-year-olds. Using a wordless picture book containing a story of animals, the pilot test revealed that the children found it difficult to focus on the researchers’ questions, as there were many characters being referenced. It was also found that there were several objects which were unfamiliar to the children. Mäkinen et al. (2014) suggested that, to assess narrative using a storybook, familiarity is an important thing. Therefore, we tried another wordless picture book, with only several characters, and found that they were more easily recognized by the children. We also found that to elicit their stories, children needed to be guided. For example, when they turned the page, they did not say anything, unless we asked them. Besides, any familiarity with the reviewer could potentially affect the children’s responses. As a result, we prepared a scenario, which helped the children be more attentive and better able to produce the narratives. We arranged a short question for each page of the book, such as “Ada apa ini?” “What is happening?” or “Ini sedang apa?” “What are they doing?”.

For the purpose of this study, the children were asked to narrate an eight-page, wordless picture book. Besides obtaining the children’s utterances through the picture-series elicitation, the analysis was supported by questionnaires about the children and their parents. The results of the comparison regarding the use of cohesive devices between both genders were quantitatively analyzed.

The main instrument in this study is an eight-page, wordless picture book by Watik Ideo (2015), titled Makan Rame-Rame (“Let’s Eat Together”). This book, which depicts a story about a flock of birds preparing to have a feast, is used to elicit the children’s utterances (see also Herningtias & Kushartanti, in this volume, who also used this instrument in their study).

We also administered questionnaires to the parents in order to obtain factual data pertaining to their children, the parents themselves, and the language that they used in various contexts. The information, in this case, included the children’s date of birth, the parents’ employment and ethnicity, and the language used by the parents on a daily basis. The questionnaires were adapted from Kushartanti’s (2014) study on the acquisition of language among Indonesian-speaking children.

Participants

The participants of this study consisted of 28 preschoolers (14 boys and 14 girls; 4–6 years of age) attending a national based-curriculum preschool in Bogor, West Java. The preschool was monolingual; the only language of instruction was Indonesian. We selected children using the following criteria: 1) aged between 4 and 6 years; 2) their primary language was Indonesian; and 3) possessed good communication skills with clear articulation. The school in which the research was conducted applied the metode sentra or the “center method,” which related to the Beyond Center and Circle Time (BCCT) curriculum.Footnote 1

Personal information about the participants was obtained from two sources: the school and the parental questionnaires. From the school, the researchers obtained personal information, such as the place and date of birth, birth order, and the language used by the participants in their daily activities. From the parental questionnaires, we obtained a demographic picture of the participants, such as the parents’ occupation and ethnicity, and the language used by the parents in their daily activities. Included in the parental questionnaire was a consent form. From the 28 questionnaires distributed, 21 were returned. Nevertheless, all the parents gave their permission to let their children be involved in this study, as confirmed by the teachers.

From the parental questionnaires, we found that the majority of the children were born in Bogor, which is in line with the location of this study. All of the children spoke Indonesian in their daily activities at school, at home, and between their friends and families. Only a few children spoke a foreign or local language outside of these situations.

The parental questionnaires revealed that the majority of the participants’ fathers were either employed by private companies or working as entrepreneurs (i.e., among the 21 responses, 9 were entrepreneurs and 9 were employed by private companies). The information about the fathers’ occupations implies that the participants of this study come from middle-class families. Most of the mothers were housewives (12 of the 21 responses). Moreover, the majority of the parents were either Javanese or Sundanese (15 of the 21 responses).

Procedure

The school allowed the researchers to collect the data when the children were participating in their classroom activities. On the day of the data collection, individual interviews with the children were conducted. As an “ice-breaker,” each child played a game as the researchers engaged them in light conversation. After establishing a good rapport, the child was presented with the picture book and asked several questions (prepared beforehand based on the scenario) that were related to the contents of the book. The purpose was to familiarize the child with the contents of the book, thus allowing him/her to produce a more coherent narrative. Afterward, the child was asked to retell their story. Initially, we assumed that children in our study were going to be too shy to speak; therefore, we prepared the scenario. Nevertheless, all the children in this study were actively engaged and capable to retelling their own stories without intervention. All we did was to give prompts, such as mmm, uh huh, or iya “yes.” The children’s re-tellings were recorded. At the end of the session, the researchers gave each child a token of appreciation. It is important to note that all the children in this study were able to provide a narration to accompany the pictures until the end of the book.

The recordings of the children’s narratives were transcribed into orthographic transcriptions, from which we examined the number of words and clauses. Afterward, we identified the grammatical cohesive devices and calculated the occurrences, from which we compared the boys’ and girls’ speech. The results are presented in tables in the following section.

Cohesive Devices in Children’s Narrative Productions

Totally, we have gathered 1051 words from 239 clauses as uttered by the children in our study. From the data, we found four kinds of cohesive device: reference, substitution, ellipsis, and conjunction. In this section, we will present some examples found in the children’s narratives. We will then move further to the comparison between the two genders, based on the quantitative analyses.

The Use of Cohesive Devices

In this study, we found that children’s foci are primarily on the characters of their story and then the events. It should be noted that, as we used a wordless picture book, the children pointed to the objects they saw. We found that the most frequently used device was personal references, or pronominal references, (the others were anaphoric marking references and demonstrative references) to refer to the characters. The finding was in line with Colozzo and Whitely (2014), who studied Canadian children using wordless picture books. In this section, we present an example.

(14)

Ada se-ekor burung yang mau turun ke tanah.

exist ONE-tail bird REL want go.down to ground

 

Lalu         dia men-cari makan.

And.then 3SG ACT.TR-look.for eat

 

“There is a bird that would like to go down to the ground. Then, she looks for some food.”

The example shows that dia “3SG” refers to seekor burung “a bird” as the child saw in the picture.

Example (14) also shows how the child connected the events chronologically by using the conjunction lalu “and then.” The term lalu means the same as the term terus “and then.” While the former is usually used in a formal situation, the latter being is used in a counterpart situation, as is also found in another piece of data, presented in Example (15).

(15)

Ada burung.

exist bird

 

Terus     burung-nya cari           makan.

and.then bird-DEF    search.for eat

“There is a bird. And then the bird is looking for food.”

Conjunctions were also frequently used by the children. The use of conjunctions—especially temporal conjunctions—indicates that these children had learned to compose stories in an orderly and chronological way. The findings are in line with McGregor (as cited in Horton-Ikard, 2009) that preschool-aged children, as well as the older ones (Horton-Ikard, 2009), predominantly use temporal conjunctions in storytelling.

We also found that children were able to use ellipses in their narrative productions. Mostly, the omitted components were the character of the story, as shown in Example (16).

(16)

Terus       burung-nya ada lagi.

and.then bird-DEF      exist again

 

Terus     Ø     dapet makan-an.

and.then     receive eat-NOUN

‘And then, comes another bird. And then (they) got food’

The use of ellipses indicates that the children were able to maintain the reference. As Berman (2009a, 2009b, p. 371) mentioned, same subject elision “[…] plays a role in narrative connectivity”; as it is an important means of reference maintenance.

It is found that the use of substitutions between the boys and the girls was rather miniscule. In fact, their vocabulary was still limited. Therefore, the ability to replace a certain word with another that includes the same meaning was still restricted. However, we found that both genders tended to substitute the “name” of the characters with kinship terms, as exemplified in (17).

(17)

Ibu-nya        me-marah    karena anak-nya    terbang.

Mother-DEF ACT-angry because child-POSS fly

 

Lalu         anak-nya    di-gendong sama mamah-nya.

and.then child-DEF PASS-carry with mother-POSS

“The mother is angry because her child flies away. Then, the child is carried by her mom.”

Example (17) shows that the child uses different names to refer to the same object. Here, we witness how they use other strategies to maintain the reference.

Comparisons Between Both Genders

This section discusses the analyses on the use of cohesive devices, based on the quantitative analyses. Before we start with our main focus, the production of words and sentences was analyzed. Table 1 presents a comparison between the girls and the boys in word production.

Table 1 Production of words in both gender groups (mean scores and standard deviations)
Full size table

Table 1 shows that word production among the girls (M = 37.43; SD = 14.88) was almost similar to that of the boys (M = 37.64; SD = 18.01). The table also shows that the standard deviation for each group was smaller than the mean for each group. Thus, it can be concluded that the ability to produce words was evenly distributed among both genders. The result from an independent sample t-test indicated that there was no significant difference between the two groups (p = 0.875).

Meanwhile, the comparison between the boys’ and the girls’ sentence production is presented in Table 2 and shows similar results to previous analyses, in terms of indifference.

Table 2 Production of clauses in both gender groups (mean scores and standard deviations)
Full size table

Table 2 shows that the girls’ clause production (M = 8.57; SD = 1.28) was almost similar to that of the boys (M = 8.50; SD = 1.09). The table also shows that the standard deviation for each group was smaller than the mean for each group (SD = 1.28 for the girls and SD = 1.09 for the boys). Thus, it can be concluded that, as with word production, the ability to produce clauses was evenly distributed among both genders. The result from an independent sample t-test also indicated that there was no significant difference between the two groups (p = 0.973).

Further, we examined the overall use of grammatical devices. The results from both genders are presented in Table 3.

Table 3 Use of grammatical cohesive devices split by gender
Full size table

Table 3 shows that among the girls, the use of grammatical cohesive devices ranged from 3 to 24, whereas such use among the boys ranged from 5 to 36. Despite the difference in usage, the mean values for the use of cohesive devices were the same for both groups; that is, both genders tend to use cohesive devices in nearly equal numbers. Moreover, the independent sample t-test confirmed that there was no significant difference between the two groups (p = 1.00).

A further analysis was conducted to examine the children’s use of each grammatical cohesive device. Figure 1 shows the comparison of the means of individual ratio of the use of grammatical cohesive devices between both genders.

Fig. 1
A double bar graph plots the use of devices by both genders. One set of bars on the graph demonstrates relatively balanced usage, while in another set, device usage among women is notably higher. The remaining two bars emphasize significantly elevated device usage among men.

Comparison of the mean scores (of individual ratio) of the overall use of grammatical cohesive devices between both genders

Full size image

As previously discussed, it is indicated that references were the most frequently used by both genders, whereas substitutions were the least used among both groups. Based on this finding, it is possible to surmise that many of the children in this study have not mastered the use of substitutions in narrative productions.

We also conducted quantitative analyses to examine the difference between both genders, in terms of the use of each cohesive device. In Table 4, we present the range scores (minimum use and maximum use), mean scores, and standard deviation. Further discussion on these results is based on Fig. 1 and Table 4.

Table 4 Comparison of the overall use of grammatical cohesive devices between both genders
Full size table

Both Fig. 1 and Table 4 suggest that there were no significant differences between both genders. Nevertheless, to confirm the finding, we conducted an independent sample t-test for each cohesive device. Based on Fig. 1 and Table 4, we found that both genders have equal ability in the use of references. From the range scores (0.0–0.78), we found that there were girls who did not use any references. Nevertheless, the result from an independent samples t-test confirms that there is no significant difference between the two genders (p = 0.835).

Another finding showed that the use of substitution was very limited in both genders. We were able to confirm that there is no significant difference between them (p = 0.927). The boys’ range score in the use of ellipses was larger (0.0–0.75) than girls’ (0.0–0.67), but the mean score in the counterpart group was higher. However, we can confirm that there is no significant difference on the use of ellipses (p = 0.130).

Based on the range score, there are girls who did not use conjunctions, and the score is slightly larger (0.0–0.69) than the boys’ score (0.13–0.71). While the girls’ mean score was lower than boys’, we found that there was still no significant difference in the use of conjunctions (p = 0.133).

All the comparisons confirmed that there was no difference between the girls and the boys in narrative production, especially in terms of the use of grammatical cohesive devices. The finding did not align with other studies on gender differences in children’s narrative production, especially on the superiority of a certain gender (Bornstein et al., 2004; Barra & McCabe, 2013; Fey et al., 2004; Gardner-Neblett & Sideris, 2018). Note, however, that the foci in these studies is on different aspects, such as language assessment (Bornstein et al., 2004; Fey et al., 2004), performance (Barra & McCabe, 2013; Gardner-Neblett & Sideris, 2018; Sperry & Sperry, 1996). Thus, an explanation of the findings rests mainly with the variables in this study, namely the cohesive devices in Indonesian children’s narrative production.

Conclusion and Suggestions for Future Research

The purpose of the present study was to determine whether gender difference has any effect on the use of grammatical cohesive devices in the production of narrative in Indonesian-speaking preschoolers. The findings show that both genders (on average) used grammatical cohesive devices (i.e., references, substitutions, ellipses, and conjunctions) somewhat equally. Among the devices, references (especially personal references, represented by the word dia “he/she”) and conjunctions (especially temporal conjunctions, represented by the words terus “then” and lalu “then”) were used the most frequently. Moreover, the boys tended to use references and conjunctions more than the girls, whereas the girls used ellipses more than the boys. As for substitutions, both groups showed a similar tendency.

Given that gender effect is still found in other studies, the explanation may have something to do with the observed linguistic variables—the grammatical cohesive devices—and the nature of the observed language, Indonesian. Besides, the number of participants in this study was very limited, Therefore, additional research should be conducted in order to highlight any potential general tendencies associated with gender. In addition, future research should utilize a larger sample than the one in this study in order to generalize the results. Finally, it is important to note that the use of grammatical cohesive devices by Indonesian children in their production of narrative is still relatively unexplored, especially in terms of gender. Therefore, this study can be viewed as a preliminary study with regard to the difference in language skills between Indonesian-speaking boys and girls.