CC BY-NC-ND
33
Psychology. Journal of the Higher School of Economics. 2016. Vol. 13. N 1. P. 192-213.
COHESION, SIMILARITY AND VALUE IN PARENT-CHILD REPRESENTATIONS OF ALBANIAN AND SERBIAN IMMIGRANT AND ITALIAN NATIVE CHILDREN
R. DIMITROVAa
a Stockholm University, 14 Frescati Hagv, Stockholm, SE-106 91, Sweden
Abstract
The study of parent-child representations across cultures is important in order to obtain a proper understanding of the attributes, size and positioning of such figures as indicators of different interaction patterns across cultures. A thorough base of research evidence for the interpretation of children's drawings may facilitate work in multicultural educational settings and enhance our understanding of cultural diversity in schools. Italy provides an ideal context for the study of parent-child representations, as the country has witnessed increasing cultural diversity in recent years with the immigration of various ethnic groups. This study examined the extent to which this context influences children's representations in domains of Cohesion (interpersonal bonding), Similarity (affinity) and Value (spatial relevance) among parent-child figures because these domains inform important representational processes of interpersonal bonding with parents across specific cultures. The Pictorial Assessment of Interpersonal Relationships (PAIR) was used to codify drawings of 326 children with Albanian (n = 59), Serbian (n = 85) and Italian (n = 182) backgrounds. The results showed that in drawings made by Albanian and Serbian children parental figures were drawn similar to and close to the child figure representing their less independent reciprocal stance. The parental figures drawn by Italian children appear bigger and farther apart. Important implications may be derived from the results in facilitating work in multicultural educational settings, by enhancing knowledge regarding cultural diversity in schools.
Keywords: drawings; parent-child representations; Albanian and Serbian immigrant children, Italian children, PAIR.
This study explores parent-child representations in the drawings of Albanian and Serbian immigrant and Italian native children in Italy. Although research shows the existence of context-dependent influences in parent-child representations in terms of interpersonal bonding, and the distance between parents and children in Swedish, British or Arab cultures (Andersson & Andersson, 2009; An-
dersson, 1995; Golomb, 2004; Lev-Wie-sel & Al-Krenawi, 2000), this research is still emerging especially with regard to the comparison of drawings of children in Italy. The study of parent-child representations across cultures is important for three reasons. Firstly, the available research consistently shows that the size, positioning and the attributes of depicted objects in children's drawings can be reliably interpreted as
an index of children's perception of the figures and objects being depicted (Burkitt, Barett, & Davis, 2003; Koppitz, 1969; Thomas & Jolley, 1998). Secondly, it is important to obtain a proper understanding of the attributes, size and positioning of such figures as indicators of different interaction patterns across cultures. Lastly, understanding the cultural variations in children's drawings has potential applications. A thorough base of research evidence for the interpretation of children's drawings may facilitate work in multicultural educational settings, as this will enhance our understanding of cultural diversity in schools. Italy provides an ideal context for the study of parent-child representations as the country has witnessed increasing cultural diversity in recent years with the immigration of various ethnic groups, most notably Albanian and Serbian. This study examined the extent to which this context influences children's representations in the domains of Cohesion (interpersonal bonding), Similarity (affinity) and Value (spatial relevance) among parent-child figures.
Children's drawings across cultures
The available literature provides evidence for the notion that differences in children's representations with significant others assume clearly distinguishable meaning, according to their cultural context. For example, Pinto, Bombi and Cordioli (1997) compared children's drawings of themselves together with a friend in Italy, Bolivia and Lebanon. The results indicated that children from Bolivia drew more similarities in the figures of themselves and their friend compared with those in
Italy and Lebanon. Similarly, Vasquez (1988) compared the drawings of Chinese and American children in elementary school. American children included their parents and other family members significantly less often in their drawings than did Chinese children (Nuttal, Chieh, & Nuttal, 1988). Andersson (1995) also found cultural variations in drawings of children from an urban school in Tanzania, who drew family figures as aligned groups, with reduced social distance among figures. In contrast, Swedish children drew greater distance between children and parents, and included more decorative objects than did African children. Steward, Furuya, Steward, and Ikeda (1982) report cultural variations, indicating that American children were more likely than those in Japan to complete their drawing of a human figure by using more color, more facial features and clothing details. Japanese children, as compared to Americans, were also found to draw significantly fewer smiles, larger and taller figures and a larger number of details (La Voy et al., 2001). Interesting cultural differences are also reported in rural Cameroon compared to urban German samples. The figure size and the head size in drawings of children in rural Cameroon were found to be substantially smaller than those of urban German children (Rubeling et al., 2010).
Furthermore, the study conducted by Pinto and Arcienega (2001) supports the importance of cultural influence in the representations of children from two contexts — a rural collectivis-tic community of Guaranн Indians and an individualistic urban community in Bolivia. Drawings provided by children in the collectivistic community were
characterized by more details of closeness and proximity to their parents compared to those of the individualistic setting. Finally, results reported by Arace (2006) add to the relatively few data on the way in which children from different ethnic groups in Italy reflect their perception of parent-child relationships in their drawings. The study was conducted with a sample of Moroccan and Italian elementary schoolchildren. The results showed that drawings produced by the Moroccan children compared to those by Italian children were characterized by more indicators of interaction and socio-emotional closeness between the self and parental figures.
Drawing on the above-presented work, this study examines parent-child representations among Albanian, Serbian immigrant and Italian native groups that are particularly interesting in the context of these ethnic communities and presumably affect their representations in a unique way. These groups can also be viewed in terms of their representativeness for individualistic and collectivistic contexts (Hofstede, 2001), as discussed in the next section.
Individualistic and collectivistic cultures
In the literature, a distinction has been made between collectivistic and individualistic cultures in order to define and attempt to explain differences in behavior, attitudes, and emotional concerns of people in different cultures. It has been suggested that people from individualistic and collec-tivistic cultures differ in behaviour, relationships and the way they define
themselves (Shiraev & Levy, 2010). An essential attribute of collectivistic cultures is that individuals subordinate their personal goals to the goals of a stable collective in-group, for example their family (Hofstede, 2001). Major emphasis regards the concern for others, care for tradition and values as well as strong ties among family members. Goals of the group are seen as important; therefore an important concern regards the integrity among group members and emotional attachment to the group. Social relationships with unequal power are more common in collectivistic cultures and interdependence between parent and child is maximized. In contrast, individualistic cultures define the self as an entity that is distinct and separate from the group. There is an emphasis on personal goals accompanied by less emotional attachment to the group. Parents in individualistic countries tend to be more willing to increase the independence and autonomy for their children (Triandis, Bontempo, Villareal, Asai, & Lucca, 1988).
Closely related to individualism and collectivism is the dichotomy of the independent self and the interdependent self. In fact, considerable research evidence supports the notion that people differ with respect to their conceptions of self in terms of an independent or interdependent self-construal (Markus & Kitayama, 1991). The independent self concerns individual characteristics and personal qualities related to self-enhancement, self-reliance and self-expression. Prototypical examples of the independent self are found to characterize Western urban contexts and are represented by autonomy (Kagitciba^i, 2007; Keller, 2007;
Markus & Kitayama, 1991). It is important to note that the notion of autonomy is conceived here in terms of independence in contrast to the one used in self-determination theory (SDT; Deci & Ryan, 2000) emphasizing the choice and the ability to take initiative. The interdependent self focuses on strong reliance on social systems, mainly the family, and is characterized by the need for strong relationships, cooperation and the value of hierarchy and conformity. The core component of the interdependent self is relatedness with others, the family unit or members of close social networks (Keller, Kartner, Borke, Yovsi, & Kleis, 2005). A common feature of individualism/collectivism and independence/ interdependence models is their unidimensional nature in that they are mutually dependent constructs on the same continuum. However, recent the-orization of autonomy/relatedness suggests different conceptions of their interrelations (Keller, 2012). In this line of reasoning, autonomy and relat-edness may coexist because they are both universal human needs and an essential part of contextual demands for individuals (Kagitciba^i, 2007).
With regards to the above conceptualizations, Albanian, Serbian and Italian children were chosen as representative of individualistic and collec-tivistic orientations. Specifically, collectivism has been suggested to be higher in Albania and Serbia in contrast to Italy where more individualistic oriented attitudes are a common norm (Gyula, Takacs, Karacsonyi, & Imrek, 2002; Hofstede, 2001). On the contrary, Italy has been shown to present the prototypical characteristics of an individualistic culture (Hofstede,
2001). Italy is classified among the ten most individualistic countries in Europe and has been included in the list of cultures belonging to the ten highest rank countries on individualism (Yates, 2013).
Based on the above presented evidence, it is plausible to assume that children's representations of parent-child relationships may differ, and that these differences are expressed in children's drawings. This study was set up to investigate this assertion, by comparing drawings of children belonging to three cultures regarded as representative for collectivism and individualism dimensions. The focus of the present investigation is on differences in Albanian, Serbian and Italian children's drawings in relation to family representations which are related to more collectivistic Albanian and Serbian cultures characterized by higher family cohesion and ties compared to the more autonomous oriented Italian culture. In the following paragraphs, a description of parent-child relationships of the target groups and the study context are discussed, before presenting the hypotheses.
Parent-child relationships in Albanian, Serbian and Italian cultures
Parent-child relationships are shaped by cultural contexts embedded within cultural conceptions of independence and interdependence (Bornstein & Cote, 2006; Lamm & Keller, 2007). In broad terms, parents from cultures with a dominantly independent orientation value self-reliance, indi-viduation and autonomy, whereas those from interdependent cultures view
obedience to authority as an important socialization value for their children. The present study investigated parent-child representations in three samples of Albanian, Serbian and Italian children living in Northern Italy. Parenting norms are quite different in these three cultures, making them interesting settings in which to investigate the extent to which associations of cultural background and representations of parent-child relations are similar or different across groups.
The traditional Albanian family is characterized by social norms of patriarchal values and obedience to authority. The family is the basic unit of the country's social structure, where children are brought up to respect their elders and, above all, their father, whose word is law within the family (Doja, 2010). As a consequence, an important feature of the Albanian family is the role of the father with his highly respected position within the family and society. An interesting feature of the traditional Albanian family is its large extended structure, which has its beginning in the pro-natal policy pursued during the communist regime. The extended family, headed by an elderly male, usually comprises all of his sons and their wives and children, all living under the same roof. In some cases, an extended family may inhabit several adjoining houses, where household and labor tasks are performed jointly (Gruber & Pichler, 2002). However, recent reports state that the nuclear family is on the rise in almost all parts of the country although larger families are more prevalent in rural areas and the majority of these belong to the group with the lowest level of wealth. The average number of chil-
dren in Albanian families is three or more with a clear peak for rural areas where many families also have more than three children, supporting the presence of more extended families in rural areas (INSTAT, 2011).
Similar to the Albanian family structure, ample evidence documents strong family features of the Serbian community. Studies report on strong family ties, intergenerational connections and solidarity within and between families as well as the central role of the paternal figure (Brannen, Lewis, Nilsen, & Smithson, 2002; Wallace & Kovatcheva, 1998). There is a strong moral obligation for parents to support their children financially throughout their education, in starting an independent household and in child-care. Related to that, parents rely heavily on their social ties as a support in parenting (Tomanovic, 2005). Interfamilial ties are based on a strong sense of reciprocity; this was recognized by the policies of the socialist system, which oriented the majority of rights and privileges towards the family through their housing policy. Serbia is characterized by a low proportion of children within the family as the average household size is 3 persons. Recent estimates indicate negative population growth during the past decade with 38 percent of households having only one child under18 (UNICEF, 2007).
Although classified as an individualistic culture, Italy is characterized by close familial relationships typical of Mediterranean cultures, where the importance of family bonds and values are still the focus of their society (Georgas, Berry, van de Vijver, Kagitci-basi, & Poortinga, 2006). Currently the modal number of children in Italian
families is one (Ciairano, Kliewer, & Rabaglietti, 2009). It is also of interest that children typically live with their families until early adulthood (Bonino, Cattelino, & Ciairano, 2006). In fact, youths and young adults leave their family to live independently very late — usually when they marry at around 30 years of age. This may be due to the central role of parental control and support in Italian families (Ciaira-no, Kliewer, Bonino, & Bosma, 2008). A distinctive feature of the Italian family model is the role of the mother, who holds a highly respected position in Italian society. She is the dominant figure in the Italian family, typically running every aspect of domestic life, and helping her family in all spheres. Italians consider the mother to be the primary figure who keeps the family together, who creates and maintains the familial links, and who is the bond that unifies all family members (Ma-netti & Schneider, 1996).
With regard to gender, there are some important differences among Albanian, Serbian and Italian cultures. In the past, Italy had a more traditional gender-role division, with men working and women taking care of children and the household. However, during the last twenty years this pattern has changed, and now the majority of women are employed full-time (Buzzi, Cavalli, & de Lillo, 1997). Relative to Albanian and Serbian cultures, there seem to be fewer traditional gender roles in the Italian family, due to new societal trends. Related to that, gender differences might be evident in comparing Italy with cultures characterized by more traditional sex roles such as Albania and Serbia. Additionally, across all three cultures, boys are
socialized to be more independent and to individuate from the family more quickly than girls; therefore, parent-child representations may have different interpretations for boys and girls.
In summary, despite commonalities in family bonds and values, Albanian, Serbian and Italian families differ in their household composition, role models and gender trends. Whereas Albanians and Serbians put more emphasis on the paternal role, the mother is the dominant figure for Italians. Italian compared to Albanian and Serbian families have less traditional gender roles and lower numbers of children. It is also worth noting that both Albania and Serbia are former communist states that spent several decades under the influence of the Soviet Union, a socialist society, whereas Italy did not. Arguably, such relevant historical characteristic coupled with differences in dominant religion in each culture (e.g., Roman Catholicism; Eastern Orthodoxy; atheism) and the individualism/ collectivism dimension, might account for important cultural differences in these groups. Therefore, differences among Albanian, Serbian and Italian cultures create particularly interesting settings to compare children's representations, and in addition, to investigate whether relations between parents and children are similar or different in these three cultures in relation to gender effects.
Study rontext
The present study was conducted in Italy, which, like other European regions, has been an experiencing marked immigrant inflow since the early 1990s. Italy is considered to be a desirable
European country to migrate to and settle in because of the need for immigrant labor and frequent regularization practices (Ceccagno, 2003). An important immigration characteristic is the growing number of children, considering that the birth rate of immigrants is twice that of native Italians, thus contributing to the national population growth rate (Farina, 2002). According to official statistics, the current number of immigrants in the country is close to four million (i.e. between 3.8 and 4.0 million), constituting 7 per cent of the total population (Caritas e Migrantes, 2010). Major migratory groups are mainly from Eastern Europe, with the largest ones from Albania, Romania and Serbia. In particular, the Albanian and Serbian populations are distributed mainly in Northern Italy, which presents specific demographic characteristics with respect to the distribution of immigrant populations throughout the rest of the country. In fact, there has been a great increase in the immigrant population in recent years, and this phenomenon has established the area at fifth place nationally in terms of the ratio of immigrants to native residents. Moreover, the factors sustaining migratory movement and the stable settlement of the immigrant population are the high number of family reunions, the acquisition of Italian nationality, and mixed marriages (Marra, 2002).
Some important characteristics of the Albanian and Serbian groups in the area need to be outlined. These communities are the largest immigrant groups, and they are well represented in the local social and economic context. They have settled in the past two decades with their families and chil-
dren, creating a well-structured community with the tendency to re-create family groups. Moreover, the percentage of Albanians and Serbians with children is particularly high, as are nationality acquisitions and the number of naturalizations (Piperino, 2002). In less than ten years, Albanian immigration has developed rapidly with gradual stabilization of families and a large increase of the gender structure of their community. Albanian immigrant women and children have much higher growth rates than men, and now represent over one-third of the community (Bonifazi & Sabatino, 2003). Similarly, annual increase of the Serbian community has been registered with a significant female component representing 45 percent of the total Serbian population in Italy (Fincati, 2007). Another salient characteristic of these groups regards their difficult integration and discrimination by the host population. In fact, the integration of these immigrants is very difficult and related to social exclusion and negative stereotyping which in turn reinforce poor labor inclusion and socio-cultural integration (Piperino, 2002).
This study included immigrant groups of children who had emigrated from Serbia and Albania. Despite the great numerical presence of these minority groups, a main reason for their inclusion in the study was their representativeness as prototypical cultural groups in the territory considered. Albanian and Serbian ethnic minorities in this area were formed by involuntary immigrant movement from their homeland because of inter-ethnic and political conflicts. This situation is further amplified by the fact that their migration status is more frequently related to
discrimination and difficult integration into the Italian context. Arguably, such differences in migration status might also be a relevant factor determining similarities and differences among these groups.
The present study
The purpose of the present study is to investigate similarities and differences in parent-child representations as reflected in drawings of Albanian, Serbian immigrant and Italian children in Italy. The Cohesion, Similarity and Value of the drawn figures of the children and their parents at the dyadic level were chosen for the study because they may inform us regarding important representational processes of interpersonal bonding with parents across three specific cultures. Prior evidence suggests that drawings of children from col-lectivistic communities show higher interpersonal bonding between parents and children, while those from individualistic contexts, focus on autonomy and independence among family members (Andersson, 1995). Accordingly, it was hypothesized that children's representations in terms of physical proximity, similarity and value between self and parental figures vary across groups. In particular, the study was expected to observe differences in children's drawings, which would be characterized by less emphasized independence and autonomy (lower scores on Similarity and higher on Value and Cohesion) in the Albanian and Serbian groups compared with the Italian group (Hypothesis 1). In addition, gender differences among Albanian, Serbian and Italian children's parent-child representations were investigated. Earlier research de-
monstrated that the height of figures was significantly bigger in drawings of girls (Andersson, 1995; La Voy et al., 2001) and girls compared to boys draw significantly more essential details, female or mother figures taller than males (Cherney, Seiwart, Dickey, & Flichtbeil, 2006). Building on this prior evidence, this study aimed at exploring gender effects and interactions in the parent-child representations as reflected in the drawings of Albanian, Serbian immigrant and Italian native children. Because of less traditional gender roles in the Italian family due to new societal trends (Buzzi et al., 1997), Italian girls were expected to represent greater independence from their parents (lower scores on Similarity and Cohesion) in their representations relative to the Albanian and Serbian girls (Hypothesis 2). Finally, ethnic group differences between mother-child and father-child dyads were explored. Prior work has demonstrated the central role of the maternal role model as a dominant figure in Italian families (Ciairano et al., 2008; Manetti & Schneider, 1996) and this study sought to explore this role in the parent-child representations. Given the strong role model of the mother in Italian culture, Italian children were expected to represent higher affinity and spatial relevance with their mother (higher scores on Similarity and Value) than their father compared to both Albanian and Serbian children (Hypothesis 3).
Method
Participants
The participants were 326 children from five elementary schools, aged 6 to
13 years, in middle-size urban areas in Northern Italy. The mean age of the sample was 8.69 years (SD = 1.39). There were 59 Albanian, 85 Serbian and 182 Italian children. 52% of all children were female, and 48% were male (for overview see Table 1). With the help of school teachers first-generation (born abroad) Albanian and Serbian children were identified (i.e. those whose parents were both from the same country of origin, i.e., Albania or Serbia). Prior to data collection, teachers were asked about the average length of stay of the immigrant children. All Albanian and Serbian children had been born abroad and living in Italy from minimum one to maximum ten years. Albanian and Serbian groups did not differ with respect to the number of years since their immigration, ^(1, N = 46) = 3.07, p = .08. Additional analyses revealed no significant group differences with respect to the age of participants, F(2, 325) = = 1.41, p = .27. Groups also did not differ with respect to gender, x2 (2, 326) = = 0.07, p = .96. School registers containing the occupation status of both parents (unskilled, semiprofessional and professional) were used to obtain information about the participants' socio-economic status (SES), resulting in three levels of SES — low, middle, and high. Ethnic groups differed with regard to family SES, with Italian children (not surprisingly) having a higher SES, x2(4, 326) = 64.35, p < .001.
Measures
The Draw-a-Person-Test (DPT)
in the Italian version of the original Goodenough-Harris test (Goodeno-ugh, 1926; Harris, 1963; Polacek &
Carli, 1977) was applied. This is a widely used measure within different cultural groups, because the subject to be drawn is universal, the materials needed are simple and the instructions are easily comprehended. Children are asked to draw a picture of a complete human figure (male or female), to the best of their ability. The final score is ascertained from total scores of 73 and 71 points, derived from differential aspects of the child's drawing of a human figure. Items consider the presence of specific graphic elements, and their gradual complexity of representation (e.g. eyes, legs, hands, fingers, the correct number of fingers, the proportion of all elements, clothing etc.). The manual provides a reliable and valid scoring system for children aged 3 to 15 years, and has been validated within the Italian population by Polacek and Carli (1977).
The Pictorial Assessment of Interpersonal Relationship (PAIR) developed by Bombi, Pinto and Cannoni (2007) was applied in order to investigate parent-child relationships within the children's graphic representations. PAIR uses six subscales of Cohesion, Distance, Similarity, Value, Emotions and Conflict. Because Cohesion and Distance scales are conceptually related (Bombi et al., 2007), as also indicated by significant correlation between the scales for the present sample (mother-child (r(319) = -.39, p < .001; father-child comparisons (r(311) = -.38, p < .05), the analyses were conducted only with the Cohesion Scale. For the present study only Cohesion, Similarity and Value scales were considered, as these indicators were of particular interest in investigating children's representations of familial relationships, as
Table 1
Sample Characteristics and Means by Ethnic Group
Albanian (n = 59) Serbian (n = 85) Italian (n = 182) Group differences
Range 7-12 7-13 6-12
Mean (SD) 8.92 (1.43) 8.76 (1.45) 8.85 (1.35) n.s
Gender
Boys 29 41 86 n.s
Girls 30 44 96
SES, n
Low 43 52 43 X2(4, 326) = 17.55***
Middle 15 33 124
High 1 0 15
Length of residence, n
1-5 years 13 19 - n.s
5-10 years 2 12 -
Drawings
Mother — Child, M (SD)
Cohesion 1.52 (1.12) 1.61 (1.24) 1.43 (1.27) n.s
Similarity 6.00 (1.18) 5.59 (1.57) 5.17 (1.71) F(2, 308) = 7.43***
Value 2.87 (1.49) 2.63 (1.74) 2.41 (1.64) n.s
Father — Child, M (SD)
Cohesion 1.24 (1.09) .99 (1.08) 1.44 (1.43) n.s
Similarity 5.63 (1.24) 5.42 (1.48) 4.90 (1.76) F(2, 308) = 6.11***
Value 3.15 (1.51) 2.75 (1.78) 2.72 (1.76) n.s
*** p < 0.001, n.s. = non-significant.
they could be reliably identified through drawings in different cultures (Bombi & Pinto, 2000; Pinto & Arcienega, 2001).
The Cohesion Scale measures interpersonal bonding between figures using six subscales referring to: C1 — Glance (if one figure perceives the head and/or major part of the body of the other figure); C2 — Approaching (the position
of the figure should suggest a reduction of the space between him/herself and the second figure); C3 — Coordinate Activity (whether or not each of the figures is involved in an action, and whether these actions are coordinated among the figures and/or are independent); C4 — Nearness (whether the two figures are "at hand"); C5 — Common Area (when figures are together in a def-
inite place, or when the space that surrounds the figures is delimited by some elements that mark the borders); C6 — Union (two figures are considered a unit when they are directly or indirectly linked). Dichotomous scores (zero-one) are assigned to each subscale, and these are summed to obtain the overall score for Cohesion. Higher scores indicate a greater degree of Cohesion.
The Similarity Scale measures affinity among figures. Similarity represents the degree to which figures are perceived to be similar in a child's drawing. More precisely, it refers to physical similarity in terms of height, body, position, and attributes between figures as an expressive way of representing the psychological similarity among figures. The scale is composed by the following subscales: S1 — Dimensions (figures' height and width); S2 — Position (e.g. standing, sitting, frontal, in profile, etc.); S3 — Body (shape of trunk, facial features, hair color, etc.); S4 — Attributes (type, shape, and color of clothes and accessories). Scores in each sub-scale range from zero (little or no similarity) to two (great similarity), with the sum of these subscales representing the overall Similarity score.
The Value Scale includes five sub-scales related to spatial relevance for each figure: V1 — Space Occupied (whether figures occupy equal amounts of space); V2 — Dominant Position (relative equal or dominant location of the figures on the page); V3 — Body Detail (an equal number of body parts, or whether one figure has more parts); V4 — Number of Attributes (whether figures have the same number of attributes, or whether one has more); and V5 — Number of Colors (whether figures have the same number of colors or
whether one has more). Scores in each subscale range from zero (identical or very similar value) to two (great disparity in value).
Individual scores for each scale are calculated based on the mean of the independent subscales with higher scores indicating higher Cohesion (interpersonal bonding), Similarity (affinity) and Value (spatial relevance) among parent-child figures. PAIR has been found to be a reliable tool for the investigation of children's representations of their social world, based on research from different cultural contexts. The system of encoding has also been proven to present satisfactory psychometric properties in terms of discriminant, concurrent and construct validity (Bombi et al., 2007).
Procedure
Participants were recruited from five public elementary schools. Prior to data collection, the principals of all the schools were contacted in order to explain the nature of the study to ensure their participation. Signed informed consent was obtained from each principal and parent for their schools and children to participate. Children were selected based on the written consents from parents and schools. Drawings were obtained during regular classroom teaching time by three research assistants. Albanian and Serbian children were able to speak Italian fluently enough to understand the instructions in that language. Additionally, supervision and support in the school setting was provided by two bilingual Albanian and Serbian research assistants. Children were asked to produce two drawings, one of
a whole human figure and one of themselves with their mother and father. They were also provided with two sheets of paper and six colored pencils (Bombi et al., 2007). The paper size for both drawings was A3. If children requested to add other relatives or specific background features (such as pets or environmental characteristics), they were given free choice to do so. After completing the drawing of themselves with their parents, each child was asked to identify the figures for accurate identification through the coding stage. Children were given a maximum time of 45 minute to complete the drawings. The average time for completion of the two drawings ranged between 15 to 45 minutes and was recorded for each class. For the purposes of this study, analyses of each drawing considered mother-child and father-child dyads, applying the PAIR scoring system. For each drawing, two independent assessments were performed: one comparing the dyadic interaction between mother and child, and the other that between father and child. Two research assistants independently coded each of the drawings. A third coder intervened in cases of disagreement; the inter-rater agreement ranged between 86% and 96%. The three coders had all received training with this coding system and undertook targeted practice before coding the children's drawings.
Analytic plan
Preliminary analyses were conducted to examine ethnic group differences in the Draw-a-Person Test (DPT), by running ANCOVA with an independent variable of group (Albanian,
Serbian and Italian), and a dependent variable of DPT total score, with age and gender as covariates. The results did not reveal significant differences. There were no ethnic group differences for Albanian (M = 43.63, SD = 10.97), Serbian (M = 41.55, SD = 10.24) and Italian (M = 40.93, SD = 11.46) children, F(2, 323) = 1.32, p = .26. However, to account for the possibility that high scores in similarity scales may relate to a lack of differentiation between persons (rather than an intended similar depiction of family members), the DPT scores were correlated with the Similarity scale for both mother-child and father-child comparisons. The DPT score was significantly correlated with Similarity for mother-child comparisons (r(319) = -.17, p < .001); therefore, DPT was controlled for in further analyses. Next, analyses were conducted in two steps following the main predictions of the study. First, ethnic group and gender differences of parent-child representations were investigated in Albanian, Serbian and Italian children, using multivariate analyses of variance. Second, within-group differences were examined by applying a paired sample t-test conducted separately for each cultural group.
Results
In order to test Hypotheses 1 and 2 (dealing with group and gender differences in parent-child representations), a MANCOVA was carried out with group (3 levels) and gender (2 levels) as independent variables; Cohesion, Similarity and Value scores for mother-child and father-child dyads (6 levels) as dependent variables; and age and
Table 2
Cohesion, Similarity and Value in Parent-Child Relationships According to Gender
and Ethnic Group
Albanian Serbian Italian Gender Ethnicity Ethnicity X Gender
Mean Boys Girls Boys Girls Boys Girls F v2 F v2 F v2
Mother — Child Dyad
Cohesion 1.42 1.61 1.65 1.58 1.46 1.40 0.63 .00 0.00 .00 0.24 .00
Similarity 5.88 6.11 5.20 5.95 5.10 5.24 7 43*** .04 5.97** .01 1.33 .00
Value 3.15 2.61 2.85 2.42 2.43 2.40 1.29 .00 4.44** .01 0.87 .00
Father — Child Dyad
Cohesion 1.50 1.00 0.93 1.05 1.34 1.53 3.38** .02 0.26 .00 1.54 .01
Similarity 5.92 5.36 5.48 5.37 5.00 4.80 611*** .03 1.30 .00 0.37 .00
Value 3.35 2.96 2.83 2.67 2.80 2.63 0.98 .00 2.06 .00 0.10 .00
**p < 0.01, ***p < 0.001.
DPT as covariates. The multivariate tests of ethnic (Wilks' lambda = .88, F(2, 308) = 3.16, p < .001) and gender differences (Wilks' lambda = .94, F(1, 308) = 2.92, p < .001) were significant. With regard to group differences, the analyses revealed Italian children showing lower Similarity with their mother (F(2, 308) = 7.43, p < .001, r]2 = .04) and father (F(2, 308) = 6.11, p < .001, r 2 = .03) than Albanian and Serbian children (Table 2)1. These results are in line with the expected ethnic group differences regarding the higher scores of Similarity scale for Italian in contrast to Albanian and Serbian children. Similarly, with regard to gender differences, there were significant results for
the Similarity scale with the mother (F(1, 308) = 5.97, p < .01, r2 = .01), where girls scored higher than boys. Gender differences were also found for the Value scale in the mother-child dyad, with boys obtaining higher scores than girls, F(1, 308) = 4.43, p < .05, r] = .01). There were no significant gender and ethnic group interaction effects.
In a further step, the third prediction regarding differences between mother-child and father-child dyads within each ethnic group was explored. Mean differences in Cohesion, Similarity and Value for mother-child and father-child comparisons were explored within each group by running a series of paired sample t-tests. The results for
1 Because immigrant status was not included in the analyses, additional analyses look at differences in drawings according to the immigration status of immigrant children only. A MANCOVA was carried out with group (2 levels), gender (2 levels), immigrant status (2 levels) as independent variables, Cohesion, Similarity and Value (6 levels) as dependent variables, and age and DPT as covariates.
Results did not show significant effects for immigrant status as well as other independent variables on Cohesion, Similarity and Value scores for mother-child and father-child dyads.
the Serbian group revealed that Cohesion with the mother was more salient than Cohesion with the father, t(82) = = 4.89, p < .001. In addition, Italian children obtained higher scores for the Value scale with their father than with their mother, t(171) = 3.52, p < .001. Italian children were also found to attain higher scores on the Similarity scale with their mother rather than their father (t(171) = 2.39, p < .01), which is in line with the expectation regarding the strong role model of the mother in Italian culture (Hypothesis 3).
In summary, parent-child relationships, and specifically Cohesion and Similarity components in drawings, showed different pathways for Albanian, Serbian and Italian children. This pattern was particularly pronounced in the Albanian and Serbian children's representations, as documented by higher levels of Similarity indicators to both parents (Figure 1). Gender differences also showed greater affinity between daughters and mothers, as well
as a higher value of the mother-child dyad for boys. As expected, within-group differences revealed greater identification and affinity with the mother than the father in Italian (rather than Albanian and Serbian) children's representations.
Discussion
The goal of this study was to investigate parent-child relationships encompassing Cohesion, Similarity and Value as depicted in drawings of children from different background in Italy. This is the first study to address these topics in Albanian and Serbian immigrant and Italian native groups, by exploring how children of these groups perceive parent-child relationships through their drawings. In addition, gender effects in parent-child representations were considered, as well as differences between mother-child and father-child dyads within each ethnic group. The findings are discussed in
Figure 1
Group Differences in Parent-Child Representations
A. Drawing of Albanian Girl Aged 12 (self figure on the left) showing High Similarity between the Self and the Parent.
B. Drawing of an Italian Girl Aged 12 (self figure in the middle) showing Low Similarity between the Self and the Parent.
terms of similarities and differences in parent-child representations, and their relation to specifics of the three groups considered.
The first hypothesis of this study dealt with group differences in that Albanian and Serbian children's drawings would be characterized by less emphasized independence and autonomy (lower scores on Similarity and higher on Value and Cohesion) compared to their Italian peers. With regard to the expected differences in parent-child representations, significant group differences emerged in terms of Similarity. The results revealed that Italian children show lower Similarity with their parents compared with their Albanian and Serbian immigrant peers. In both mother-child and father-child comparisons, the figures of the parents compared to those of children drawn by Italian children were characterized by higher autonomy, independent activities and separation. Albanian and Serbian children drew themselves as being more similar to their parents, which was denoted by lower spatial and perceptual distance between the self and parental figures. This result is in line with expectations regarding differences between independent and interdependent self-construal (Kagitci-ba§i, 2007; Keller, 2007; Markus & Ki-tayama, 1991). It also suggests that when society places particular emphasis on the value of collectivity and social cohesion, it is possible to perceive oneself as more similar to one's parents. Albanian and Serbian children indicate more interdependent types of parent-child representations, where having someone who resembles and is similar to the self carries more weight
than the representations of importance, expressed through Cohesion and Value dimensions. This type of relationship may be particularly satisfying for Albanian and Serban children in Italy supporting the idea that a certain overall degree of affinity is necessary for expressing parent-child representations in these cultures. Results are also consistent with the traditional Albanian and Serbian family structure, being characterized by strong family ties, intergenerational connections and solidarity (Brannen et al., 2002; Doja, 2010; Wallace & Kovatcheva, 1998). It seems that Italian children opt for a more independent and autonomous type of relationship with their parents. Group differences in spatial, size and proximity relations in the drawings of mother and father reflect a similar pattern documented in other cross-cultural studies (Andersson, 1995; Arace, 2006; Pinto & Arcienega, 2001; Pinto et al., 1997). The findings also support the notion that children's representations in terms of proximity and similarity between self and parental figures may be differently expressed in independent cultural settings compared to interdependent ones, providing support for the important differences in children's pictorial representations of social interactions and close interpersonal relationships (Keller, 2007; Keller et al., 2005).
The second hypothesis of this study predicted greater independence from their parents (lower scores on Similarity and Cohesion) in Italian rather than Albanian and Serbian girls. With regard to the expected gender differences for Italian girls, representing more independence from parents in their representations compared to
those of Albanian and Serbian girls, no significant results were found. The data presented here indicates that there are more affinities than differences among groups with regard to gender. Arguably, there might be more similarities in parent-child representations for girls belonging to interdependent (Albanian and Serbian) and independent (Italian) cultures. However, overall gender influences on children's pictorial representations were found. Girls showed more pronounced Similarity with their mother, whereas boys attributed more Value to their father. Gender results in similarity with the mother can be explained in terms of differential socialization of boys and girls, or, in conceptual terms, a tendency for girls to identify with their mother (Russell & Saebel, 1997). It may also be more important for girls to denote their personal identity through a female appearance that depends on their maternal influence in relation to social demands concerning the female role. The greater similarity of girls to their mothers lends support to prior work, showing the influence of gender in children's graphic representations. Gender differences are also reported in terms of differences in clothing, size of figures and stereotypical details in which girls express specific gender-related details. For example, it has been found that girls include more body parts and clothing in their drawings than do boys (Koppitz, 1983). Girls tend to draw more stereotyped details (such as fingernails, hairstyles) than boys do in their drawings (Skybo, Ryan-Wenger, & Su, 2007), suggesting that girls develop a scheme of what it is to be 'female' by drawing objects and figures as they should look according to speci-
fic gender schemes (Cherney et al., 2006). These and other results (Dag-lioglu, Alemdar, Calisandemir, & Bencik Kangal, 2010) suggest that children's drawings represent different perceptions of boys and girls within social relationships, which in turn might be of importance in representations of their parent-child dyads. Additionally, the finding that in parent-child representations boys attributed greater Value to their father than their mother might be related to strong social gender roles in the groups considered. As specified in the introduction, boys are socialized to be more independent from the family than girls, and this is particularly true for the expected male role model in the three societies (Brannen et al., 2002; Ciairano et al., 2008; Doja, 2010; Wallace & Kovatcheva, 1998). Arguably, boys may view their father as the more important figure with whom to identify, which is also reflected in their representations.
The third hypothesis of this study predicted Italian children to represent higher scores on Similarity and Value with their mother than their father, given the strong role model of the mother in Italian culture. In line with expectations, group differences were found between mother-child and father-child dyads. Italian children were found to attain higher scores on Similarity with their mother rather than their father. Italian children represented higher identification and affinity with their mother than their father compared to both Albanian and Serbian children, which confirms the expectation regarding the strong role model of the mother in the Italian culture (Ciairano et al., 2008). The results also revealed that Serbian immigrant
children represent more Cohesion with their mother, which may be due to stronger bonds with the mother rather than the father figure. Such distinctive-ness in relationships with the mother may be due to the importance of mothers to Serbian children's development in Italy, which is a speculation in need of further investigation.
Limitations
Despite the importance of the study as the first documentation of parent-child representations in drawings of children from different ethnic communities in Italy, some limitation should be acknowledged. It is necessary to be cautious with generalizations, due to the necessity of applying additional quantitative indicators of family relationships (Di Leo, 1973; Hammer, 1997). The analyses were based on a quantitative interpretation of family drawings, which were not complemented with additional measures under test and questionnaire conditions. Although the results suggest important representational aspects reflected in children's drawings in different groups, future studies should include questionnaire-based data. It may be also beneficial to apply qualitative interview techniques, which would be useful in gaining a better and more in-depth understanding about family structures and relationships across cultures. Future analyses should incorporate both qualitative and quantitative methods concerning interpersonal relationships as perceived by parents and children in different cultural settings. Additional consideration should be paid to the possibility to examine drawings of different family structures (e.g., single-
parents, extended family), which may also influence representations of parent-child relationships (Fan, 2012). Further research investigating emotional (Koppitz, 1969, 1983) and cognitive indicators may add valuable insights to the drawings literature. On a methodological note, family SES, unbalanced sample size and lack of control of immigrant status may have an influence on family representation. Finally, it is essential to conduct further studies in order to ascertain whether the effects of different parent-child representations as expressed in the drawings of Albanian, Serbian and Italian children also occur in children from other contexts. Further research should include Albanian and Serbian children in Albania and Serbia to have a better understanding of cultural differences in parent-child representations. The role of immigration experience and acculturation in drawings of children with different ethnic background needs to be further assessed and may play a role in the production of the drawings.
Implications and conclusions
The findings of this study, which support the existence of cultural differences in representation of parent-child relationships in drawings of children across three cultural groups, have important implications in practitioner-oriented and educational fields. This study has potential implications for intercultural relations, where it is important to understand that representations of parent-child relationships may be expressed in different ways. Teachers and school counselors may use this knowledge in teaching students to use drawing as a strategy for gaining a
better knowledge of representational processes across different cultures. Related to that, drawing should be used in classroom to improve learning and intercultural interactions among students in increasingly diverse schools in European countries. The emphasis on different aspects, meanings and perceptions of close relationships may lead both teachers and students to communicate better in a multicultural setting. Understanding how a child sees him- or herself within his or her culture can expand our knowledge of the development of identity, and the interrelationships of its many components, which may be a critical step towards productive relationships. There is much yet to be implemented in that direction, but drawings produced by Albanian and Serbian immigrant and Italian native children may facilitate the understanding in teachers and students and their appreciation of cultural diversity.
In conclusion, this study adds to existing research by providing unique evidence concerning parent-child representations of Albanian and Serbian immigrant and Italian native children in Italy that have only scarcely been examined in prior research. Furthermore, the study shows the usefulness of drawing for accessing children's parent-child representations in different cultures. The drawing of the self together with both parents appears to be sensitive with regard to revealing children's representations, by showing the distinctiveness of specific appearances, linked to greater or lesser independent characteristics of their culture. The most salient result is that cultural dynamics are expressed in children's drawings, suggesting important differences in mother-child and father-child representations in terms of perceptual affinity and identification between parents and children.
References
Andersson, I., & Andersson, S. B. (2009). Aesthetic representations among Himba people in Namibia. International Art in Early Childhood Research Journal, 1, 1-14.
Andersson, S. (1995). Social scaling in children's family drawings: A comparative study in three cultures. Child Study Journal, 25, 97-122.
Arace, A. (2006). Attaccamenti, separazioni, perdite. Eventi critici nello sviluppo del Se e dei legami familiari [Attachment, separation and loss. Critical events in the development of self and familial bonds]. Milan: Edizioni Unicopli. (in Italian).
Bombi, A. S., & Pinto, G. (2000). Le relazioni interpersonali del bambino [Child interpersonal relationships]. Roma: Carocci. (in Italian).
Bombi, A. S., Pinto, G., & Cannoni, E. (2007). Pictorial Assessment of Interpersonal Relationship (PAIR). An analytic system for understanding children's drawings. Firenze: Firenze University Press.
Bonifazi, C., & Sabatino, S. (2003). Albanian migration to Italy: What official data and survey results can reveal. Journal of Ethnic and Migration Studies, 29, 967-995. doi:10.1080/ 1369183032000171320
Bonino, S., Cattelino, E., & Ciairano, S. (2006). Italy. In J. J. Arnett (Ed.), International encyclopedia of adolescence (pp. 510-523). London/New York: Routledge.
Bornstein, M., & Cote, L. (2006). Acculturation and parent-child relationships: Measurement and development. Mahwah, NJ: Lawrence Erlbaum.
Brannen, J., Lewis, S., Nilsen, A., &. Smithson, J. (2002). Young Europeans, work and family: Futures in transition. London: Routledge.
Burkitt, E., Barrett, M., & Davis, A. (2003). The effect of affective characterizations on the size of children's drawings. British Journal of Developmental Psychology, 21, 565-584. doi:org/10.1348/ 026151003322535228
Buzzi, C., Cavalli, A., & de Lillo, A. (Eds.). (1997). Giovani verso il Duemila. Quarto Rapporto IARD sulla condizione giovanile in Italia [Young people towards the 21st century. Fourth report on the status of youth in Italy]. Bologna, IT: Il Mulino. (in Italian).
Caritas e Migrantes. (2010). Immigrazione: Dossier statistico 2010 [Immigration: Statistic Dossier, 2010]. Rome: Anterem. (in Italian).
Ceccagno, A. (2003). New Chinese migrants in Italy. International Migration, 41, 187-213. doi: org/10.1111/1468-2435.00246
Cherney, I. D., Seiwert, C. S., Dickey, T. M., & Flichtbeil, J. D. (2006). Children's drawings: A mirror to their minds. Educational Psychology, 26, 127-142. doi:org/10.1080/01443410500344167
Ciairano, S., Kliewer, W., Bonino, S., & Bosma, H. (2008). Parenting and adolescent well-being in two European countries. Adolescence, 43, 99-117.
Ciairano, S., Kliewer, W., & Rabaglietti, E. (2009). Adolescent risk behavior in Italy and the Netherlands. European Psychologist, 14, 180-192. doi:10.1027/1016-9040.14.3.180
Daglioglu, H. E., Alemdar, M., Calisandemir, F., & Bencik Kangal, S. (2010). Examination of human figure drawings by gifted and normally developed children at preschool period. Elementary Education Online, 9, 31-43.
Deci, E. L., & Ryan, R. M. (2000). The "what" and "why" of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11, 227-268. doi:10.1207/ S15327965PLI1104_01
Di Leo, J. H. (1973). Children's drawings as diagnostic aids. New York: Brunner Mazel.
Doja, A. (2010). Fertility trends, marriage patterns, and savant typologies in Albanian context. Journal of Family History, 35, 346-367. doi:10.1177/0363199010381045
Fan, R. J. (2012). A study on the Kinetic Family Drawings by children with different family structures. The InternationalJournal of Arts Education, 173-204.
Farina, P. (2002). La presenza straniera in Europa [Foreigners in Europe]. In ISMU (Eds.), Settimo rapporto sulle migrazioni 2001 [Seventh report on migration 2001] (pp. 27-35). Milano: Franco Angeli. (in Italian).
Fincati, V. (2007). Gli immigrati serbi e montenegrini in Italia e Veneto [Serbian and Montenegrine immigrants in Italy and Veneto]. Regine del Veneto, Osservatorio e Ricerca. (in Italian).
Georgas, J., Berry, J. W., van de Vijver, F. J. R., Kagitcibasi, C., & Poortinga, Y. H. (2006). Families across cultures. A 30-nation psychological study. New York: Cambridge University Press.
Golomb, C. (2004). The child's creation of a pictorial world. Mahwah, NJ: Lawrence Erlbaum Associates.
Goodenough, F. L. (1926). Measurement of intelligence by drawings. Yonkers-on-Hudson, NY: World Book.
Gruber, S., & Pichler, R. (2002). Household structures in Albania in the early 20th century. The History of the Family, 7, 351 374. doi:org/10.1016/S1081-602X(02)00106-9
Gyula, B., Takacs, S., Karacsonyi, A., & Imrek, V. (2002). Eastern European cluster: Traditions and transition. Journal of World Business, 37, 69-80.
Hammer, E. F. (1997). Advances in projective drawing interpretation. Springfield, IL: Thomas.
Harris, D. B. (1963). Children's drawings as measures of intellectual maturity: A revision and extension of the Goodenough Draw-a-Man Test. New York: Harcourt, Brace and World, Inc.
Hofstede, G. (2001). Culture's consequences: Comparing values, behaviors, institutions, and organizations across nations. Thousand Oaks, CA: SAGE Publications.
INSTAT. (2011). Socio demographic changes of the family and child development in Albania. Tirana, Albania: National Institute of Statistics.
Kagitciba^i, C. (2007). Family, self, and human development across countries. Theory and applications. Mahwah, NJ: Lawrence Erlbaum.
Keller, H. (2007). Cultures of infancy. Mahwah, NJ: Lawrence Erlbaum.
Keller, H. (2012). Autonomy and relatedness revisited: Cultural manifestations of universal human needs. Child Development Perspectives, 6, 12-18. doi:10.1111/j.1750-8606.2011.00208.x
Keller, H., Kartner, J., Borke, J., Yovsi, R. D., & Kleis, A. (2005). Parenting styles and the development of the categorical self. A longitudinal study on mirror self recognition in Cameroonian Nso farming and German families. International Journal of Behavioral Development, 29, 496-504. doi: 10.1177/01650250500147485
Koppitz, E. (1969). Emotional indicators on human figure drawings of boys and girls from lower and middle-class backgrounds. Journal of Clinical Psychology, 25, 432-434. doi:10.1002/1097-4679(196910)25:4<432::AID-JCLP2270250428>3.0.C0;2-Q
Koppitz, E. (1983). Projective drawings with children and adolescents. School Psychology Review, 12, 421-427.
La Voy, S. K., Pederson, W. C., Reitz, J. M., Brauch, A. A., Luxenberg, T. M., & Nofisnger, C. C. (2001). Children's drawings: A cross-cultural analysis from Japan and the United States. School Psychology International, 22, 53-63. doi:org/10.1177/0143034301221005
Lamm, B., & Keller, H. (2007). Understanding cultural models of parenting. The role of intracultural variation and response style. Journal of Cross-Cultural Psychology, 38, 50-57. doi:10.1177/ 0022022106295441
Lev-Wiesel, R., & Al-Krenawi, A. (2000). Perception of family among Bedouin-Arab children of polygamous families as reflected in their drawings. American Journal of Art Therapy, 38, 98-106.
Manetti, M., & Schneider, B. H. (1996). Stability and change in patterns of parental social support and their relation to children's school adjustment. Journal of Applied Developmental Psychology, 17, 101-115. doi:10.1016/S0193-3973(96)90008-8
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition, emotion, and motivation. Psychological Review, 98, 224-253. doi:10.1037//0033-295X.98.2.224
Marra, C. (2002). Il monitoraggio dei fenomeni migratori nel Friuli-Venezia Giulia. Una rassegna bibliografica [Monitoring migration phenomena in Friuli-Venezia Giulia. A review]. Migration Studies, 147, 702-711.
Nuttall, E., Chieh, L., & Nuttall, R. (1988). Views of the family by Chinese and U.S. children: A comparative study of Kinetic Family Drawings. Journal of School Psychology, 26, 191-194.
Pinto, G., & Arcienega, R. C. (2001). Cohesion and distance between parents and children: A cross-cultural study. Rassegna di Psicologia, 18, 51-72.
Pinto, G., Bombi, A. S., & Cordioli, A. (1997). Similarity of friends in three countries: A study of children's drawings. InternationalJournal of Behavioral Development, 20, 453-469. doi:10.1080/016502597385225
