Perceptions of decisional uutonomy of Turkish adolescents and their parents

A percepção de adolescentes e de pais sobre decisão e autonomia

EMPIRICAL RESEARCH

Perceptions of decisional uutonomy of Turkish adolescents and their parents

A percepção de adolescentes e de pais sobre decisão e autonomia

Nermin Celen^I,1; Figen Cok^II; Harke A. Bosma^III; H. Zijsling Djurre^III

^IMaltepe University , Istanbul, Turkey

^IIUniversity of Ankara, Ankara, Turkey

^IIIUniversity of Groningen, The Netherlands

ABSTRACT

This study attempted to investigate decisional autonomy in Turkish adolescents from 12 to 18 years. The Perspectives on Adolescent Decision Making (PADM) questionnaire was administered to 372 middle class adolescents who attend middle and high schools and to their parents. The PADM assess if adolescents decide for themselves, or parents impose restrictions or adolescents and parents have arguments about the topic. MANOVA analyzes were used. Results showed that affirmative answers increased with age. From adolescent and parents' perspectives adolescent decisional autonomy grows with age, parental control decreases, conflicts between them tended to decrease, on the perspective of parents. There was minor gender differences: girls have higher level of decisional autonomy; boys experience more conflict. Adolescents' decisional autonomy expectations tended to be higher than those of parents. Fathers' and mothers' perspectives on decisional autonomy were very similar. The results support the new family model proposed by Kaðýtçýbaþý.

Keywords:Decisional Autonomy, Turkish Adolescents, Turkish Parents.

RESUMO

Estudo objetivou investigar autonomia de decisão de adolescentes turcos de 12-18 anos de idade. Questionário PADM (Perspectivas na Tomada de Decisões por Adolescentes) foi administrado a 372 adolescentes de classe média, alunos de ensino fundamental e médio e a seus pais. O PADM verifica se adolescentes decidem por si mesmos ou se os pais impõem restrições, também se discutem o assunto. Foram usadas análises MANOVA. Os resultados mostraram que respostas afirmativas aumentam com a idade. Na perspectiva de pais e adolescentes a autonomia para tomada de decisões aumenta com a idade, o controle parental diminui, conflitos tendem a diminuir. Houve pequena diferença de gênero: meninas têm maior nível de autonomia decisional, meninos experimentam mais conflitos. As expectativas de autonomia decisional dos adolescentes tendem a ser maiores que as dos pais. Perspectivas de pais e mães sobre autonomia foram bastante similares; os resultados suportam o modelo de família proposto por Kaðýtçýbaþý.

Palavras-chave: Autonomia Decisional, Adolescentes Turcos, Pais Turcos.

Introduction

Conceptions of individual autonomy in the modern sense began to arise in the Renaissance (Hill & Holmbeck, 1986) and gained further acceptance during the protestant reformation. Industrialization and the increases in the division of labor have made individual autonomy respectable. Durkheim (1933, cited in Durkin, 1995) claimed that when the division of labor in society is small people are bound together by their similarities. Industrialization, however, brought a large division of labor that led to specialization and individual autonomy. For a long time already, autonomy and self-regulation have been a central theoretical and research topic in adolescent psychology. Autonomy has been regarded as a sense of separateness, emotional independence, and as a striving to gain freedom from parents and other social influences (Erikson, 1950). The concept of autonomy emphasizes 'independence' and is associated with concepts like 'conflict' and 'self-regulation'. Crittenden (1990) defined autonomy in terms of the capacities to take responsibility for one's behavior, to make decisions regarding one's life, and to maintain supportive relationships. Hill and Holmbeck (1986) have suggested that autonomy does not refer to freedom from parents but to the adolescents' freedom to carry out actions while maintaining appropriate connections to significant others. Maccoby (1984) argued that the gradual transition from parental regulation to adolescent autonomy and self-regulation entails regulation strategies, including shared decision making and parental monitoring of autonomous action (cited in Feldman & Quatman, 1988).

During adolescence internal (intrapsychic, cognitive, and biological) and external (societal expectations, interpersonal relationships, role responsibilities, family composition) changes dictate the developmental tasks of this period. It is a period of realignment of family relations during which there can be temporary perturbations in the parent-adolescent relationship (Collins 1990; Sessa & Steinberg, 1991; Steinberg, 1990). This is different from the storm-and-stress view of adolescence as suggested by Hall and Freud. If the family creates a warm and democratic atmosphere, accepts the adolescent emotionally, and gives him or her room to make his or her own decisions, a healthy, autonomous development is promoted (Hill & Holmbeck,1986; Steinberg, 1990).

Autonomy is not a unidimensional achievement. It involves progress in different domains, while the pace of development may not be consistent across domains. When individuals relinquish their childish dependence from the parents, they achieve autonomy in the family domain, but not automatically in the friendship domain. They may still be effected by peer pressure. Peer norms in the domains of hobbies, reading material, music and clothes might not be in line with parental norms and values. These differences could lead to different perspectives across generations. In general parents' roles are essentially conservative; parents tend to invest in conserving, protecting, maintaining, and promoting the safe growth of their offspring. Youngsters, on the other hand, are inclined to experiment, to seek new, expansive roles, have different expectations with regard to their life. Both parents and adolescents have to adapt to changes. Smetana (1993) suggests that parent-adolescent conflicts reflect the parties' different roles in the family. Family members attempt to negotiate their conflicts by coordinating their social cognitive perspectives. When the adolescents feel free of excessive dependency upon others (parents and peers), they can take initiative and have a feeling of control over their life.

From a social cognitive point of view, autonomy is a multiple perspective-taking and inferential social reasoning process. In the individualistic stage of ego development what is considered right or wrong depends upon one's personal evaluation of circumstances. Adolescents in this stage begin to evaluate rules, standards, long-term goals and ideals in terms of their own ethical principles. Self-criticism and self-respect increase, and the awareness of inner conflicts decreases. The autonomous stage of ego development is the stage in which the individuals have the ability to cope with conflicting needs and responsibilities (Muuss, 1996). The adolescent with the increased cognitive capacities of formal-operational thinking is able to unite and integrate ideas that appear as incompatible alternatives to younger adolescents. Not only do they acknowledge interpersonal relatedness but they also respect each other's autonomy. (Cognitive autonomy is not only the resolution of internal conflicts but also the recognition of the other's need for autonomy). This type of autonomy develops in the social context of the family, and especially in the context of peergroup interactions (Youniss & Smollar, 1990).

Achieving autonomy is one of the major developmental tasks of adolescence. Smetana (1988) has shown that parents' and adolescents' conceptions of authority versus adolescent autonomy change between the ages of 11 and 16. These changes concerned certain domains and not others. Thus, adolescents may consider parental authority legitimate with regard to the moral and normative part of their lives but not so in their daily activities like style of dress, personal appearance and choice of friends. These latter domains tend to create conflicts.

The developmental changes seem to reflect age-graded normative expectations with regard to parental authority and adolescent autonomy. Adolescents have the freedom to make decisions about certain areas within their lives while their parents continue to have authority over the other areas, but in the course of development in adolescence, the adolescent takes on responsibilities which were previously held by the parents.

On the basis of the work of Goodnow and Collins (1990) it is assumed that parents' ideas and expectations about adolescence and adolescent development will guide the way in which authority is used and decisional autonomy is given to the adolescent. Furthermore, adolescents' view of adolescence and what is appropriate for adolescents to do or decide for themselves probably influences their responses to the actions of the parents. These processes depend on the nature of the topic involved (Bosma, et al., 1996; Smetana, 1988) and are influenced by cultural factors (Zani, Bosma, Zijsling & Honess, 2001). So it thus needs to be investigated in different cultures.

Cultural values associated with ethnic or national origin could be potent in shaping ideas, expectations and social cognition. Understanding parents' ideas may be enriched when studied in a cross-cultural perspective (Osterweil & Nogano 1991). Goodnow (1983) has found that parents from different cultural groups differ in the views held about 'desirable' behavior and 'reasonable' ages. These views were transmitted by formal and informal tasks, teaching practices and assessment procedures of a culture. Cultures that are collectivistic at the axis of individualism-collectivism emphasize conformity, obedience or an interdependence orientation toward the collective good and in-group members. Individualistic cultures are concerned with traits such as independence (Steinberg, 1990) and autonomy. Differences between cultures along the individualistic- collectivistic dimension have recently been shown to influence the beliefs about autonomy. The expectation with regard to age of autonomy is heavily influenced by cultural context (Feldman & Quatman, 1988). Steinberg (1990) suggested that in societies in which the rate of cultural change is slow and expectations for behavior change little from generation to generation 'obedience' is far more adaptive.

The Turkish society had been known as a collectivist culture and parental values as obedience and conformity are emphasized. From the 1950's on, there has been a shift from an agricultural economy to a large scale market economy. There is a rapid change in cultural values especially in the big cities because of migration (from rural to urban areas) and an increase in communication facilities. Changes in family structure (from an extended to a nuclear family) and family interaction patterns led to changes in child rearing beliefs. Parents felt encouraged to foster independency while at the same time maintaining closely-knit interaction patterns. A new family model, named 'emotional interdependence' combined the two basic human needs - autonomy and relatedness - with each other. Lin and Cha found the similar family models among Korean and Chinese parents (Kağıtçıbaşı, 1996). This model is different from the traditional psychological conceptualizations. There is a decrease in material dependencies and an increase in the demands of urban life but emotional interdependencies continue. Urban adolescents in Turkey could be seen as oriental, collectivist from the perspective of classical conceptualizations. Recent research, however, showed that family life in Turkey has been changing dramatically in urban regions, it now tends to have autonomy as well as relatedness characteristics. Kağıtçıbaşı (1996, b) suggested that relatedness and autonomy typically coexist in these families. Empirical support for this family type comes from studies in different countries including Turkey. Child-rearing emphasizes relatedness in these families, but in response to changing life styles children also get room for the development of autonomy. Kağıtçıbaşı (2000) maintained that the 'autonomous-related' self concerns an integrative synthesis derived from both autonomy and relatedness.

In Turkey research on parent and adolescent relationships from a developmental perspective is very limitied. Hortacsu (1989), Imamoglu (1997), LeCompte and LeCompte (1973), and Taylor and Oskay (1995) have done research on different aspects of adolescent-parent relationships in Turkey. But there seems no research on the autonomy of Turkish adolescents. Although the autonomy of a growing person has gained much importance in the Turkish culture, the development of autonomy still has to be researched.

Gender studies provided considerable data on gender differences in many psychosocial dimensions: communication, affiliation, empathy, nurturance, altruism, morality, aggression, assertiveness, dominance, intimate relationships and achievement (Basow, 1992). These dimensions may all contribute to the autonomy of the individual. However, the findings with regard to the gender differences in the autonomy of adolescents are not consistent. Steinberg (2002) pointed out that, contrary to the popular beliefs that boys expected more autonomy than girls, there were no sex differences in this respect. There were gender differences in the extent to which parents grant autonomy, though. Research also showed some evidence about the differences in certain developmental issues of adolescent boy and girls in Turkey (Kazgan, 2002). In a comprehensive research, Kağıtçıbaşı reported that the value of children from the parental perspective could be different for their sons and daughters. Sons could be valued as future caregivers and sources of economic support for their parents in their old age. Parents traditionally tend to live with their sons after his marriage in a typical extended Turkish family. In such a pattern sons are definitely valued higher than daughters. On the other hand, in the new generation parents, especially mothers tend to value their daughters as emotional support figures. The number of mothers with a preference for a daugther tends to increase. Furthermore, Turkish women in their mother role were found to encourage autonomy of their children if a training is provided (Kağıtçıbaşı, 1996,a). Family and gender expectations in the Turkish culture, thus, seem to be changing. So, in order to gain more insight into the issue of adolescent autonomy in the Turkish culture, it would be worthwhile to investigate gender differences in the adolescents' and parents' perspectives of adolescent autonomy.

This study was planned to investigate the perspectives of Turkish adolescents and their parents of decisional autonomy in adolescence. It will be studied in such a way that age and gender differences and differences between issues are taken into account.

It is hypothesized that (1) from the adolescents' and parents' perspectives adolescent decisional autonomy tends to increase with age, (2) that parental limitations with regard to the adolescent's decisional autonomy and parent-adolescent conflict about decisional issues will decrease with age in both adolescent and parental perspectives. Finally, we hypothesized that (3) boys will be viewed as more autonomous than girls regardless of their age in both the adolescents' and parents' perspectives.

Method

Adolescents in the sample were recruited from middle and high schools in Ankara and Bursa, Turkey. The mothers and fathers received the instruments via their adolescent daughters and sons. The school administrators gave permission to recruit the adolescents and their parents. The neighborhoods of the schools in the two cities were considered to represent a middle Socio Economic Status (SES). The adolescents volunteered to be tested. The instrument was administered in the classroom. While the students completed the tests, copies to take home to their parents were also handed out. A week later the questionnaires returned by the parents were collected at school. In the end 372 questionnaires from adolescents, 333 from mothers and 324 from fathers were obtained. Not every parent returned the questionnaire. In some cases it took the students more than a week to return the completed forms. They were reminded to bring these in the following week. Table 1. shows the final distribution of the number of boys and girls, and the number of mothers and fathers in the sample in the three adolescent age groups.

The Parent Adolescent Decision Making Questionnaire (PADM) has been used to collect the data. This instrument was developed by Bosma et al. (1996). Items of PADM were partly derived from the literature on adolescent development and conflict within the family and partly from pilot studies carried out in the Netherlands. The instrument comprises four questions about issues which involve adolescent and parents' decision making and which can be sources of conflict between parents and adolescent. These issues were: chores, bedtime, manners, language, visits, privacy, smoke, alcohol, sweets, body care, clothes, look, money, sports, hobbies, go out, time in, friends, sex, church (not adopted in this study) and homework.

Each of the four standard questions in the PADM reflects a key aspect of the adolescent and parental construction of adolescence:1. Whether the adolescent decides for him/herself on ... (issue), (here labeled as 'adolescent choice' , q1) 2. Whether the parents feel the adolescent should or should not... (issue), ('parental feeling', q2) 3. Whether there are often arguments between parents and adolescents about ...(issue), ('arguments', q3),4. Whether it's normal for someone of the adolescent's age to decide for him/herself on ...(issue), ('normality', q4)

The format of the quartet of questions for the adolescents was as follows (with 'smoke' as an example): I decide myself whether I smoke or not yes? no - My parents feel I shouldn't smoke yes? no - I often have arguments with my parents about smoking yes? No I think it's normal for someone of my age to decide for myself about smoking yes ? no.

The questions for the parents are exactly the same, but phrased from the perspective of the parents, 'My son/daughter decides for him/herself whether he/she smokes or not', 'I feel he/she shouldn't smoke', 'I often have arguments with my son/daughter about smoking', 'I think it's normal for someone of the age of my son/daughter to decide for him/herself about smoking'. The same answering format is used. The '?' means 'don't know'.

Before application in Turkish, the PADM questionnaire was translated into Turkish and back-translation to English was provided by two independent translators. Colleagues specialized in the psychology of human development reviewed the questionnaire in terms of its convenience with regard to language and its use by Turkish adolescents. Just minor changes were suggested. The final instrument consisted of 80 questions for 20 domains.

Data analysis

The PADM data were analysed in different ways. The first involved computation of each participant's sums of Yes answer for each of the four different questions across the 20 issues. This yields four interval scores for each subject which can range from 0 to 20. These scores (of the adolescents, the mothers and the fathers) will be used to test the hypotheses about developmental changes in adolescent behavioral autonomy. MANOVAs will be done with the adolescent, mother and father samples respectively, with these four sums as dependent and age (three age-levels: level 1 13-14, level 2 15-16, and level 3 17-18 years of age) and gender of the adolescents as independent variables.

Since there often were clear differences between the samples, the number of yes-answers of the adolescents, the mothers and fathers (as independent samples) will also be compared with oneway ANOVAs and post hoc tests. To see which issues mostly contribute to these differences Chi-square analyses will be used. According to the Bonferroni correction the critical value should be p<.0024 (p<.05 divided by 21).

Another approach to the data focused on response patterns across questions within each of the issues. This involved examining each subject's constellation of answers to the four questions concerning each separate issue. Only constellations of clear Yes and No answers were used. Omission of ? responses yields 16 possible constellations of YES and NO scores per issue (details of these analyses are shown in Bosma et al., 1996).

Four of the possible 16 constellations were used most frequently and were also regarded as being psychologically meaningful (Bosma et al. ,1996). These are the following:

YYYY Norm supported conflict

Adolescents decide for themselves,

parents have negative feelings,

arguments occur, and adolescents think it is normal to decide for themselves.

YYNY Norm supported compromise

Adolescents decide for themselves,

parents have negative feelings,

no arguments occur, and adolescents think it is normal to decide for themselves.

YNNY Norm supported autonomy

Adolescents decide for themselves,

parents have no negative feelings,

no arguments occur, and adolescents think it is normal to decide for themselves.

NYNN Accepted parental authority

Adolescents do not decide for themselves, parents have negative feelings,

no arguments occur, and adolescents do not think it is normal to decide for themselves².

In this study it is assumed that these constellations give an indication of how the participant perceived the family decision-making situation with regard to a particular issue. In this perception four aspects are taken together: Whether or not adolescents decide for themselves and whether that is normal (relative to what their peers do), whether or not parents impose restrictions, and whether or not adolescents often have arguments with parents about the issue in question. In this study these constellations were used to explore changes in the family decision-making situation. We expected to find changes into the direction of more symmetrical forms when the adolescent grows older.

Results

The number of "Yes" responses across issues for each of the four questions was examined for adolescents, mothers and fathers separately. These answers mean that adolescents see themselves, and parents their offspring, as making their own decisions about the issue (Yes 1), see that parents should impose limitations with regard to the behavior (Yes 2), report that there are often arguments (Yes 3), and see it as normal that the adolescents decide for themselves (Yes 4).

Adolescent choice and normality:An overview of the results for the yes-answers to the first question across issues (Yes 1) is given in Table 2. According to the analyses the number of issues for which the adolescents answer 'yes' to the first question increased significantly with age, from about 13 to almost 18 of the 20 issues ( F= 67.05; df= 1, 366; p<.000). The Yes 1 responses of the mothers and fathers similarly indicated a significant increase of adolescent choice with their adolescent child's age (mothers F= 43.78; df=2, 327; p<.000; fathers F= 38.57; df= 2, 318; p<.000). For each sample the differences between the three age categories were significant (bottom row of Table 2). Neither of the three samples reported significant differences according to the adolescents' gender, nor were there significant interactions of age by gender. The tendency that the scores were higher for the daughters almost reached significance in the mothers' data (F= 3.19; df= 1, 327; p=.075).

The perceived room for adolescent decision making, however, differed when the perspective of the adolescents was compared to the perspective of the parents. The mean score of the adolescent Yes 1 responses was about two issues higher than the mean scores of the parent samples (F=33.76; df=2,1026; p<.000).

By comparing the frequencies of the yes-answers per issue the differences between the parent and adolescent perspectives can be analyzed in more detail. This gives an indication of which issues contributed to the differences between adolescents and parents. In certain domains the frequencies of fathers' and mothers' Yes 1 responses were quite similar while the adolescents' Yes 1 responses were much more frequent. Table 3 gives an overview of all the domains (language, smoke, alcohol, go out, friends, sex and crime) where the differences were significant (Chi-square analyses, p<.003).

In general, there was a strong correspondence between the number of "Yes" answers to the first (adolescent choice) and the fourth question (normality) (Bosma et al., 1996). The nature of these two questions seem to be complementary. The logic behind this correspondence could be that if it is normal for somebody of my age (or my offspring's age) to decide.... , then I also decide for myself (or he/she decides for him/herself) whether...(of course, the reasoning could also be the other way around).

An overview of the results for the yes-answers to the fourth question across issues (Yes 4) is given in Table 4. The results were highly similar to the results of the first question. According to the analyses the number of issues for which the adolescents answered 'yes' to the normality question increased with age, from about 14 to about 18 of the 20 issues (F=43.91; df=1, 366, p<.000). The Yes 4 responses of the mothers and fathers similarly indicated an increase of adolescent choice with their adolescent child's age (mothers F= 37.00; df= 2, 327; p<.000; fathers F= 26.75; df= 2, 318; p<.000). For each sample the differences between the three age categories were significant (bottom row of Table 4). In contrast to the results for the first question the adolescents showed a gender difference: the normality scores for the female adolescents were higher (F=4.91; df=1, 366; p<.027). For the parent samples a similar but not significant tendency was found. There were no significant age by gender interactions.

The perception of what is normal differed for the parent samples when compared to the adolescent responses. The mean score of the adolescent Yes 4 responses was about two issues higher than the mean scores of the parent samples (F=32.15; df=2,1026; p<.000).

The issues in which adolescents and parents seem to be different in Yes 4 answers were almost the same as the Yes 1 responses. There were minor differences: Percentages of yes answers for the issues of language and crime of adolescents and parents were more similar. However, more adolescents than parents stated that it is normal for an adolescent to decide for him/herself on clothes.

In general, Yes 1 and Yes 4 responses were very similar, both showed a clear increase with the adolescent's age both in the adolescent and the parent groups. The issues in which adolescents and parents were clearly different in percentages were similar in Yes 1 and Yes 4 answers.

Only one significant sex difference was found in the "Yes 1" and "Yes 4" responses, in the adolescent data. Girls more often tended to expect that it is normal for an adolescent to decide for herself. This sex difference can be seen in most of the issues except go out, crime, time in and sex. In these latter issues more boys reported that it is normal to decide. These issues (go out, time in and sex) seem to fit the traditional understanding of being a young man who leaves home and arrives home whenever he wants, who does outside whatever he wants, who has the freedom and even is encouraged to be sexually active. This has also been found in similar data from Italy and the Netherlands (Zani, et al., 2001).

Parental feelings: Yes 2 responses reflect the parental feelings with regard to issues in which adolescent decision making is involved. The question is phrased in the form of whether the parents feel the adolescent should or should not …

MANOVA and ad hoc tests showed that there were significant differences between the oldest (17-18 years) and the two younger age groups (13-14 and 15-16 years) in the mothers' (F= 4.48; df= 2, 327; p<.012) and the fathers' (F= 5.82; df= 2, 318; p<.003) and also in the adolescents' (F=9.55; df=1, 366; p<.000) responses (see Table 6).

Parents expected the eldest group to be more autonomous, as compared to the younger two age groups. There was a parallel between the answers of the adolescents and parents in terms of the age at which the most substantial drop in parental limitations occurred (between middle and late adolescence), but the adolescents differed strongly from the parents with regard to the mean number of issues in which the parents impose limitations (F=95.24; df=2, 1026; p<.000). Fathers, in general, seemed to be even stricter than the mothers, a difference which was almost significant (p=.055). Regarding the adolescents' gender no differences were found in the adolescents', the mothers' and the fathers' responses to the second question, nor were there any interactions of gender by age.

In Table 7 the issues are listed (chores, bedtime, private, sweets, clothes, money, and homework) that showed significant differences in frequencies (Chi-square tests).

Yes 3 answers refer to the frequent presence of arguments between adolescents and parents in issues. In general the mean scores were quite low: frequent arguments were reported for about two to three issues. According to the MANOVA results (Table 8), there were significant differences among the parents of the different adolescent age groups (Mothers: F= 7.33; df= 2, 327; p<.001; Fathers: F= 8.31; df= 2, 318; p < .000) while no differences were found among the three age groups of adolescents themselves (F=1.42; df=1, 366; p=.24). Both mothers and fathers reported more conflict with the youngest age group (13-14 years) and less conflict with the eldest age group. Their reports with regard to the 15-16 year old adolescents were different. According to the mothers a significant drop in arguments with their adolescent children takes place at a later age. The adolescent age groups did not perceive a significant decline in arguments. Their mean score at the age of 17-18 years was comparable to the mean scores of the parent groups. The between groups variance of the total mean scores of the parents and adolescents was significant (MANOVA, F=4.04; df= 2, 1026; p= .018), but only the difference between the adolescents and the mothers was significant (ONEWAY, mothers: p=.005; fathers: p=.152). Significant gender differences were obtained in the parental perspectives of arguments. Both mothers (F= 10.31; df= 1, 327; p=.001) and fathers (F= 5.59; df= 1, 318; p=.019) reported more conflicts with their sons. No significant gender difference was found in the adolescent' reports (F=1.18; df=1, 366; p=.278). None of the three groups of respondents showed an interaction effect of gender by age in the number of arguments reported.

When we look at percentages of Yes 3 responses per issue, there appeared to be three (chores, bedtime and language) in which adolescents and parents were significantly different in percentages (chi-square analyses). Mothers reported slightly more conflict than fathers and adolescents (Table 9).

Although the frequencies of conflicts in these issues reported by the mothers were not very much higher than what the fathers reported, the differences could be a consequence of the fact that the mothers are more directly involved in their children's lives and as a result experience more conflicts.

Since the parents reported significant gender differences we also looked at the frequency distribution of conflict responses by gender. In the issues such as bedtime, language, alcohol, crime and homework, both mothers and fathers reported significantly more conflicts with their adolescent boys (Table 10). Especially for homework, this difference was very substantial.

The results so far seem to support the first two hypotheses of this study: Adolescent decisional autonomy increases with age of the adolescent. This is evident from the increases of Yes 1 and Yes 4 answers and decreases of Yes 2 and Yes 3 answers. Both adolescents and their parents reported these changes. The results don't seem to support the third hypothesis which suggested that boys are higher in the decisional autonomy as compared to girls. Minor gender differences have been obtained which, instead, seem to suggest that girls get more room to make their own decisions.

Constellations of PADM responses within issues: Examination of individual score constellations - how does an individual respond to the constellation of the four questions per issue - provides information of a different kind. It is assumed that these constellations give an indication of the family decision making situation. While the number of yes answers reflect individual perspectives, the constellations probably reflect more of a family decision style. Four different constellations have been distinguished on conceptual and empirical grounds (Bosma et al., 1996; see method section).

Constellations can be studied on an issue by issue basis, and, similar to the counting of yes answers across issues, one can also count how often a particular constellation is used across the 20 issues. In Table 11 the mean scores of the use of the four different constellations of the adolescent and parent samples are presented. Also included are the results of the MANOVA and ONEWAY analyses of the differences on the basis of age and gender of the adolescents.

Norm-supported conflict (YYYY): MANOVA analyses showed that there was no significant age-related difference in norm-supported conflict from the perspectives of the adolescents (F=.22; df=2, 366; p=.805) and the fathers (F= .70; df= 2, 318; p=.496), while a significant difference between the adolescent age groups was observed in the mothers' responses (F=3.28; df= 2, 327; p=.039). The mothers of the 15-16 year olds reported more norm-supported conflict, from their perspective mid-adolescence is most conflictive period. The mean number of issues (about 1) in which this decision-making style occurs, however, is low. Regarding the sex of the adolescent, the adolescents and the fathers did not perceive any substantial difference. In contrast, the mothers of boys reported significantly more norm-supported conflicts than the mothers of girls (F= 4.36; df= 1, 327; p=.038). The between groups variance of the three samples was also significant (F=6.49; df=2, 1026; p=.002). The adolescents reported significantly more norm-supported conflict than the fathers. The other group differences were not significant. There also were no significant interactions of gender by age within each group of respondents.

Norm-supported compromise (YYNY): MANOVA analyses showed that norm-supported compromise significantly differed across adolescent age groups, from the perspectives of the adolescents (F=11.45; df=2, 366; p<. 000), the mothers (F= 20.98; df= 2, 327; p<.000), as well as the fathers (F=12.93; df= 2, 318; p<.000). Of the four constellations norm-supported compromise was used most frequently. Its use increased from early to mid adolescence to a stable level of about 7 to 9 issues. This suggests that by mid adolescence in about a third to almost half of the issues decision making is characterized by compromising between adolescents and parents. Only in the fathers' perspective a significant gender difference has been observed indicating more use of this constellation among girls than boys (F= 4.03; df=1, 318; p=.046). A similar tendency, almost significant, can be seen in the perspectives of the mothers and adolescents (F=3.74; df=1,327; p=.054). There were no significant gender by age differences within each group. The three samples also showed significant overall differences (F=8.52; df=2, 1026; p<.000). The adolescents less often used norm-supported compromise than the mothers (p<.011) and fathers (p<.000). The parents did not differ in this respect.

Norm-supported autonomy (YNNY):The MANOVA analyses showed that norm-supported autonomy significantly increased across the adolescent age groups, from the adolescents' (F=16,30; df=2, 366; p<.000), the mothers' (F=12.78; df=2, 327; p<.000), as well as the fathers' (F=13.14; df=2, 318; p<.000) perspectives. In the perspectives of the adolescents and fathers each of three age groups showed a significant increase in the use of this constellation. From the perspective of the mothers the significant increase in the use of this constellation occurred after mid adolescence. Regarding the sex of the adolescent only the mothers showed a significant difference in their use of the norm-supported autonomy constellation (F=5.45; df=1, 327; p=.020). They used this constellation more often with their daughters. There were no significant gender by age interaction effects. The differences between the groups were very significant (F=107.28; df=2, 1026; p<.000). The mean score of the adolescents was much higher than the mean scores of the mothers (p<.000) as well as the fathers (p<.000). The parent groups did not differ.

Accepted parental authority (NYYN): The analyses showed that accepted parental authority showed significant decreases with adolescent age, from the adolescents' (F=41.11; df=2, 366; p<.000), the mothers' (F=23.41; df=2, 327 ; p<.000), as well as the fathers' (F=15.86; df=2, 318; p<.000) perspectives. The mothers and adolescents reported a significant drop after early adolescence, while the decline after mid adolescence, in their opinion, was not significant anymore. The fathers, in contrast, perceived a significant decline after early as well as mid adolescence. In none of the three samples significant sex differences were found in the use of accepted parental authority. There also were no gender by age interactions. The samples showed a significant between groups variance (F= 17.16; df=2, 1026; p<.000): The adolescents used this constellation less often than both the mothers (p<.000) and the fathers (p<.000). The parent groups did not differ in this respect.

Discussion

In conclusion, all constellations showed age related changes in adolescence in that norm-supported compromising and norm-supported autonomous decision making tend to increase and accepted parental authority tends to decrease with age. Apart from some differences in timing, these tendencies were observed in the perspectives of the adolescents, as well as in the perspectives of the parents. No such general tendency has been observed for norm-supported conflict. In general the family members were in agreement about the age related changes towards increased adolescent autonomy in decision making. The differences between the adolescents and parents in general mean scores (adolescents tend to see more norm-supported conflict and score higher on norm-supported autonomy, while the parents reported higher scores on norm-supported compromise and accepted parental authority) probably reflect their different positions in the ongoing process of negotiating adolescent autonomy and parental authority.

The results with regard to the family decision making patterns also support the first two hypotheses with increasing use of norm-supported compromise and norm-supported autonomy and decreasing use of accepted parental authority. On the other hand, the results do not support the third hypothesis that boys would score higher on decisional autonomy. Dramatic gender differences were not prevalent, there were even some significant findings favoring the opposite hypothesis. Boys seemed to use norm-supported conflict more often than girls (from the mothers' perspective) and girls used more norm-supported compromise (from the fathers' perspective) and norm-supported autonomy (from mothers' perspective). No gender differences were found in the adolescent data.

Achieving autonomy is one of the major developmental tasks of adolescence. The achievement of autonomy in adolescence also has a cultural base which means that the culture in which the adolescent lives contributes to the achievement of autonomy.

Turkey is a country which is basically traditional. On the other hand, values have rapidly been changing especially since 1950's. The ratio of the young population is very high compared to other European Countries. It, therefore, seems important to do research on autonomy of this young population in Turkey, which probably is different from the young generations in western countries as well as more traditional countries.

In general, the results showed that Turkish adolescents tend to gradually achieve decisional autonomy during the period of adolescence. One of the interesting findings is that the affirmative answers to "I decide myself " and "My son/daughter decides" (Yes 1) and "It is normal" (Yes 4) increased by age during the adolescent period in the perspectives of not only the adolescents, but also in the perspectives of the mothers and fathers. The autonomy scores (Yes 1 and Yes 4) of the adolescents and parents were not equally high, though. The adolescents gave themselves higher scores than their parents gave to them. This was especially apparent for the following issues: language, smoke, alcohol, go out, friends, sex, and crime. For all these issues many more adolescents claimed room to make their own decision than parents who said that their sons or daughters made their own decision. This could be seen in both the Yes 1 and Yes 4 responses. It is interesting to see the strong similarity between the fathers' and the mothers' responses. Traditionally, fathers are believed to be more authoritarian and not so close to their children as the mothers (Ekşi, 1990) so it could have been speculated that the autonomy views of the fathers would be more stringent than those of the mothers. This was not found.

In traditional (collectivist) societies age is considered to be the only criterion with regard to expectations about dependency and parental authority. According to child-rearing beliefs, parents should give permission for autonomy in late adolescence while the young are then expected to assert some independence when they reach late adolescence. In most of the cases family control mechanisms are much more strict and protective for children and late and mid adolescents. Recent studies, however, showed that there is a shift toward granting children and adolescents more autonomy in Turkish families (Kağıtçıbaşı, 1996, a;b; Taylor & Oskay, 1995). Kağıtçıbaşı emphasized that families' age expectations for autonomous decision making in Turkey are as similar to the expectations within the Western culture (Steward, Bond, Deeds & Chung, 1999).

The other general finding, which is in accordance with the first one is about parental feelings and limitations: Yes 2 answers decreased with increasing age of the adolescent. This item reflects negative feelings and limitations of the parents in response to the adolescents' decisional autonomy. The decrease of the Yes 2 answers, thus, fits the developmental pattern of increasing adolescent decisional autonomy in the family. Like the answers to the adolescent choice and normality questions, responses of the mothers and fathers to the parental feeling questions were very similar, and, for a number of issues rather different from the responses of the adolescents. Especially in the following issues a strong difference was found: chores, bedtime, private, sweets, clothes, money, friends, and homework. In all these domains the frequencies of the yes answers of the mothers and fathers were much higher. Interestingly, the parents thus reported that they impose more limitations in these areas than adolescents perceived. This is a very interesting finding, in combination with the findings re yes1 and yes4; The adolescents report more autonomy than the parents say and less limitations than the parents impose.

Another striking result of the study is that the Yes 3 (frequent arguments) responses decreased with age in the adolescent as well as the parental perspectives. Parents reported that they have more frequent conflicts with early adolescents (13-14 years of age) than with 15-16 and 17-18 year olds. As the adolescents get older conflicts become less frequent. For the two younger age groups parents tend to perceive more conflict than the adolescents. For the17-18 years old adolescents, the views of the three groups are comparable. The higher incidence of conflicts in early adolescence might be attributed to the processes of puberty. The rapid physical maturation also affects the emotions and the social development of the growing person. From a cognitive perspective, the more frequent conflicts in early adolescence may also be attributed to the power of hypothetical reasoning which enables the young person to contemplate and articulate alternatives. All these factors might explain why early adolescence is more conflictuous than late adolescence. Furthermore, when autonomy is more or less achieved in late adolescence and the battle for autonomy is over, the necessity for conflicts also diminishes.

Social learning theorists claim that adolescents have to experience explicit exposure to conflicts as a means of problem solving together with the intermitted reinforcement obtained when parents yield to their own conflicting demands. Smetana (1988) suggested that parent-adolescent conflicts reflect the parties' different role in the family and their attempts to coordinate conflicting social cognitive perspectives. The high frequency of conflict answers for the 13-14 year olds supports these theoretical conceptualizations.

Especially for issues such as chores, bed-time and language mothers and adolescents tend to differ while adolescents' and fathers' responses seem to be more similar. This could be explained by the mothers' higher level of involvement in the adolescents' daily lives at home.

In general, the results thus supported the first two hypotheses of the study which stated that decisional autonomy will increase by age. This has become evident from the increase of Yes 1 and Yes 4 responses and the decrease of Yes 2 and Yes 3 responses. This age trend was evident in the perspectives of both the adolescents and the parents.

Another objective of this research was to study gender differences in decisional autonomy. MANOVA analyses were applied for all Yes answers. Except Yes 4 answers (normality), no significant differences between genders were obtained for the adolescents' perspectives. Frequency distributions of Yes answers in different issues for boys and girls showed that in Yes 4 answers only in go out, time in and sex issues, boys more often tend to think that it is normal to decide for themselves. This fits with cultural expectations of being male and has also been found in similar research in Italy and the Netherlands (Zani et al., 2001).

From the parental perspectives, both mothers and fathers reported more conflicts with their sons, especially for the following issues: bed time, language, alcohol, crime and homework. Not only the frequency but also the gender difference with regard to homework is striking. Almost half of all the parents reported often having conflicts with their sons about homework. For the daughters these frequencies amounted to about 18%. This seems parallel to the general tendency of Turkish girls to be more successful at school. In the traditional conceptualizations, it has long been suggested that boys have stronger autonomy needs and they are permitted to be more autonomous by their parents than girls. More recently it has been suggested that emotional autonomy during early adolescence could be greater among girls. Furthermore, there appeared to be no effect of gender on adolescents' perceptions of power and cohesion in their families (Feldman & Quatman, 1988). (NERMİN BURALARDA HARKE'NİN KUŞKULARI VARDI ÇIKARDIM.) So, our third hypothesis was not supported by the results. It could be that daughters are under the influence of higher autonomy expectations of their parents, which is the opposite of our initial hypothesis. In general the gender differences are not so big but the significant ones seem to favor higher autonomy among girls. It could well be that in the process of the changing family and changing self in a changing culture as Kagitcibası has proposed, women tend to change more than man. Especially the finding that mothers reported more use of norm supported autonomy with their daugthers than with their sons also fits this explanation.

Results with regard to constellations of answers per issue, indicating patterns of family decision making, showed that there is an increase in norm supported compromise (the YYNY constellation), especially between early and mid adolescence, as well as an increase in norm supported autonomy (the YNNY constellation). This latter constellation showed differences between all three age groups. Accepted parental authority (the NYNN constellation) decreased by age, while no age related change was found in norm supported conflict (the YYYY constellation). Both the increasing and the decreasing age trends reflect the healthy development of adolescents. If the Turkish families were strictly authoritarian such a decrease by age wouldn't be observed. This finding seems to be in accordance with Kagıtcibasi's formulations. While the adolescents become more autonomous, they also want to keep their relatedness with the parents. In Kagıtcibasi's conceptualization the new family parenting style is authoritarian; the child-rearing orientation is control- and obedience-oriented and the self is relational-autonomous, in general manifesting both autonomy and relatedness (Kagıtcibasi, 2000). These explanations fit very well with the findings with the constellations in this study.

As a conclusion the decisional autonomy increases with age. Parental autonomy expectations are lower than the adolescents' expectations, they seem to lag behind. The fathers' and mothers' perspectives are similar rather than being different. Minor gender differences were found favoring higher decisional autonomy of girls, while less conflicts with girls were reported. Cross-cultural comparisons seem very helpful to get a better picture of adolescent decisional autonomy within its cultural context.

Recebido em 29/09/2006.

Aceito para publicação em 05/01/2007.

This study has been conducted as a research project of Ankara University, Center for Research on Child Culture (CRCC)

• Basow, S. (1992). Gender stereotypes and roles Brooks Gunn: Pasific Grove.
• Bosma, H.; Jackson, S.; Djurre, H. Zani, B.; Ciconani, E.; Xerril.; Honess, T., & Charman, L. (1996). Who has the final say? Decisions on adolescent behavior within the family. Journal of Adolescence, 19, 277-291.
• Collins, W. A. (1990). Parent-child relationships in the transition to adolescence: Continuity and change in interaction, affect and cognition. In R. Montemayor, G. Adams & T. Gullotta (Eds.), Advances in Adolescent Development from childhood to adolescence: A Transitional period? (pp. 85-106). Newsbury Park, CA: Sage.
• Crittenden, P. M. (1990). Toward a concept of autonomy in adolescents with disability. Children's Health Care, 19, 162-168.
• Durkin, K. (1995). Developmental Social Psychology. From infancy to old age. Oxford: Blackwell.
• Ekºi, A. (1990). Çocuk, genç, anabalar (Children, youth, parents) Ystanbul: Remzi.
• Erikson, E. (1950). The problems of ego identity. Journal of American Psychiatric Association 4, 56-121.
• Feldman, S. S., & Quatman S. (1988). Factors influencing age expectations for adolescent autonomy: A study of early adolescents and parents. Journal of Early Adolescence, 8, 325-343.
• Goodnow, J. (1983). Parents ideas about development and parenting. A review of issues and recent work. In M. Lamp; B. Rogoff & A. Brown (Eds.), Advances in Developmental Psychology (pp. 193-297). Hillsdale, NJ: Erlbaum.
• Goodnow, J. J., & Collins, W. A. (1990). Development according to parents: The nature, sources and consequences of parents' ideas Hillsdale, NJ: LEA.
• Hill, J. P., & Holmbeck, G. N. (1986). Attachment and autonomy during adolescence. In G. J. Whitehurst (Ed.), Annals of Child Development (pp. 145-189). JAI Press.
• Hortaçsu, N. (1989). Targets of communication during adolescence. Journal of Adolescence, 12,  253-263.
• Imamoglu, E.  O. (1997). An Interdependence model of human development. In. C. Kagytcibasi (Ed.), Growth and Progress in Cross Cultural Psychology (pp.138-145), Lisse, Netherlands: Swets & Zeitlinger.
• Kagytcibasi, C. (1996b). The autonomous-relational self: A new synthesis. European Psychologist, 1 (3), 180-186.
• Kagytcibasi, C. (2000). Cultural mediation of autonomy-relatedness in adolescence, Paper presented at 7th EARA Conference, Jena, May 31-June 4 -2000.
• Kaðýtçýbaþý, Ç. (1996a). Family and human development across cultures Lawrence N. Jersey: Erlbaum Associates.
• Kazgan, G. (2002). Kuptepe gençlik araptyrmasy(Kuptepe youth study). Ystanbul: Ystanbul Bilgi Üniversitesi.
• Lecompte, W., & Lecompte, G. (1973). Generational attribution in Turkish and American youth. Journal of Cross Cultural Psychology, 4, 175-191.
• Maccoby, E. E. (1984). Socialization and developmental change. Child Development, 55, 317 - 328.
• Muuss, R. E. (1996). Theories of Adolescence. N.Y.: McGraw-Hill.
• Osterweil, Z., & Nogano, K. (1991). Maternal views on autonomy. Journal of Cross Cultural Psychology, 22, 362-375.
• Sessa, F. M. & Steinberg, L. (1991). Family structure and development of autonomy during adolescence. Journal of Early Adolescence, 11, 38-55.
• Smetana, J. G. (1988). Adolescents and parents' conceptions of moral and social rules. Child Development, 59, 321-335.
• Smetana, J. G. (1993). Understanding of social rules. In M. Bennett, (Ed.), The development of social cognition: The child as psychologist (pp.111-141). New York: Guilford Press.
• Steinberg, L. (1990). Interdependency in the family: Autonomy, conflict and harmony in the parent-adolescent relationship. In. S. Feldman & G. Elliot (Eds.), At the Treshhold; The Developing Adolescence (pp. 225-276). Cambridge MA; Harvard Univ. Press.
• Steinberg, L. (2002). Adolescence (6th ed.). Boston: McGraw Hill College. 
• Steward, S. M.; Bond, M. H.; Deeds, O., & Chung, S. F. (1999). Intergenerational patterns of values and autonomy: Expectations in cultures of relatedness and separateness. Cross Cultural Psychology, 30 (5), 578-593.
• Taylor, R. D., & Oskay, G. (1995). Identity formation in Turkish and American adolescents. Cross Cultural Psychology, 26 (1), 8-22.
• Youniss, J., & Smollar, J. (1990). Self through relationship development. In H. Bosma & S. Jackson (Eds.), Coping and self-concept in adolescence (pp.129-148). Berlin: Springer Verlag.
• Zani, B., Bosma, H.A., Zijsling, D.H., & Honess, T.M. (2001). Family context and the development of adolescent decision-making. In J. E. Nurmi (Ed.), Navigating through adolescence: European perspectives (pp. 251-278). New York: Routledge Falmer Press.

2

The capitals represent the Yes or No answers to the four questions. The labels for the constellations are summarizing statements. They do not directly refer to psychological constructs.

Publication Dates

• Publication in this collection
  30 Oct 2007
• Date of issue
  Dec 2006

History

• Accepted
  05 Jan 2007
• Received
  29 Sept 2006