Fathers in Turkey : Paternity characteristics , gender role , communication skills

Objective of this study is to examine the correlation the quality of paternity, gender roles and communication skills of fathers. The scores in the scale of supporting developmental tasks were used in order to determine the quality of paternity. The other data collection tools were the BEM sex role ınventory and the communication skills ınventory. The study included a multiple regression analysis for predicting the quality of paternity. The multiple regression analysis suggested that the scores in the “behavioral”, “emotional” and “cognitive” sub-dimensions of the communication skills inventory and those in the “femininity” and “masculinity” sub-scales of the BEM sex role inventory can account for the scores in the scale of supporting developmental tasks at a rate of 31%. It was concluded that those fathers who display feminine traits and those who communicate with their children through an “emotional” or “behavioral” communication language support their children’s process of development in a more effective way.


INTRODUCTION
Individuals can succeed in social relationships to the extent that they recognize and express their own feelings and opinions, pay attention to the thoughts and feelings of the others and adjust their behaviors according to this context (Signe ve Van Schaik, 2000).
Identifying and transmitting emotions is an important factor for communication.Accepting the feelings of others is a significant social skill in especially establishing close and satisfying relationships (Shapiro, 2010).Children develop these important social skills through the experiences they gain within the family life.The effective relationship between the children and their parents enable them to succeed in life and constitute positive social interactions.Moreover, they begin to assume social gender roles through these interaction activities.
The demand for female labor force has increased owing to the changing economic conditions ever since the ındustrial revolution.Women's higher educational status and the increasing number of working mothers have led to a need to redefine feminine and masculine roles and responsibilities.The specification of sex roles has been mainly shaped by the expectations of every particular society.Each society has defined certain feminine and masculine roles in accordance with its own culture (Winstead and Derlega, 1993).Redefined feminine and masculine roles have resulted in a change in the family structures, too.
The changes in family structure have taken place in the form of certain phases.Throughout the transition phases, individuals have made an attempt to adapt themselves to their new positions within their families and social circles.The quality of marriage has played a key role in this adaptation (Minsel et al., 1999).Men have started to lose their monopoly of being the group who maintained the family; however, they have not fully assumed the role of raising their children yet.Women's roles in maintaining the family have started to be accepted whereas the responsibilities of men for child care have started to increase (Colombo, 2008;Deutsch and Saxon, 1998;Drobeck, 1998;Lamb et al., 1987;Riggs, 1997;Wentworth and Chell, 2001;Zimmerman, 2000).
In recent years, increased research on gender roles and family relations.Studies on parenting roles found that a transformation has been experienced in paternal responsibilities.It was observed that fathers in western world are more interested in domestic tasks and more emotional than those in the old days (McMahon, 1999).Paternity has gained importance as a topic that people are interested and discuss (Lupton and Barclay, 1997).
Paternity characteristics (Renk at al., 2010), responsibilities in child care and education (Ünüvar et al., 2010), communication skills and parenting competencies (Ünüvar, 2010(Ünüvar, , Fears, 2010) ) have been observed by various researchers.The researches cover topics such as the self-efficacy of parents and parental roles.
In some researches, parents' self-efficacy perception is found to be associated with how they evaluate themselves in parenting and how they observe their children's developmental status (Coleman and Karraker 1998;Jones and Prinz 2005;Shumow and Lomax 2002).
In other words, the parents who found the developmental status of their children satisfactory, evaluate themselves as qualified at parenting.The effective participation of both mother and father to child care enables children to observe and get different gender roles (Rossi, 1984) and simplifies gaining sexual identity (Garrett, 1992).
Studies on parenting place a particular emphasis on the importance of parents to child education.Family is where a child goes through his/her first social experiences.Parents, in turn, are the first and the most long-standing teachers of their children.Therefore, it is necessary for parents and individuals that make up the immediate surroundings of a child to be a part of the educational process (Arnas and Yasar, 2008;Dinc, 2011;Gulay and Akman, 2009).
Parents prove to the most crucial element in providing pre-school children with an environmental support for their development (Clerkin et al., 2007).Pearson states that socially effective environments mediate the share of values and emotions and makes it easier for social roles to be performed by satisfying requirements (Kaner, 2010).The stronger and better the communication is between parents and a child, the more positive process of development a child will experience (Ustun, 2010).It was reported that substance abuse and negative behaviors get decreased when parents adopt a supportive and emotional communication in their interaction with their children (Brody and Ge, 2001).
Paternal participation in child care and child rearing has a positive influence on a child's cognitive development, social skills, internal locus of control and level of empathy (Fagan and Inglesias, 1999).An active paternal participation in children's lives enables fathers to learn about their requirements and how to satisfy these needs in a more efficient manner (Ehrensaft, 1990).
Paternal participation leads to increased sense of self on the part of their children and develops their skills in controlling what is going on around them.(Pleck and Masciadrelli 2004).Children with a close interaction with their fathers will have a healthier psychology and higher academic achievement (Flouri and Buchanon, 2004).
Paternal participation in child care and child rearing is mostly associated with the perception of sex role (Coverman, 1985;Radin and Golsmith, 1989).Traditional roles have been undergoing a transformation.All the same, child care and child rearing are still regarded as exclusively feminine roles in many cultures.
Whereas, it is among the findings of the aforementioned researches (Clerkin et al., 2007;Fagan and Inglesias, 1999;Kaner, 2010;Üstün, 2010;Ehrensaft, 1990;Pleck, 2004) that the effective participation of fathers to child care and child rearing has positive effects on children.For that reason, this study aims to analyze the paternity characteristics, gender role and communication skills of fathers who have children at the 3 to 6 age group.In line with this purpose, the subobjectives of this research are to analyze if the paternity characteristics of fathers vary depending on the social gender role they own; and if paternity characteristics can be predicted with gender role and communication skills.

METHODOLOGY
The present study is based on a descriptive and quantitative design.The population of the study is comprised of 540 fathers who were grown, and live in 35 different cities in Turkey.The data is collected from regions with different characteristics (economic, geographical, social, cultural) and for this purpose, the research is based on service regions table of the Ministry of Education.This table divided the cities in Turkey in three groups according to their similar characteristics such as geographic location, economic and social development, fulfillment of their service needs (meb.gov.tr, 2013).
The table lists the categories as Region 1, Region 2, Region 3; Region 1 representing the top region.As per this classification, 53% of the cities in Turkey are in Region 1, 27% of them are in Region 2 and 20% of the cities are listed under Region 3.Among the data of this research, 61% of the fathers have been grown up and are living in cities in Region 1, 22% of them have been grown up and are living in cities in Region 2 and 17% of the fathers have been grown up and are living in cities in Region 3. In total, the data is collected from 35 cities. (Cities in Region 3: Ağrı, Bitlis, Hakkari, Kars, Nearly 70% of the fathers included within the study group are aged between 31 and 40.Furthermore, 86% of them became a father for the first time when they were aged between 26 and 30 (Table 1).According to the statistics released by Turkish Statistical Institute TUİK (2010), 93.3% of the men in Turkey get married before they are 30.Therefore, national data support the finding that 86% of the subjects became a father for the first time when they were aged between 26 and 30.
Thirty-three percent of the subjects in the population are primary school graduates; 22% of them high school graduates; and the remaining 46% university graduates.Whereas 31% of the fathers (168) have a working wife, 69% of them (372) are married with a housewife.Twenty-two percent of the fathers have one child; 47% of them two children; 19% of them three children; and the remaining 12% four or more children.Fifty-eight percent of the subjects (315) were born and raised in a village or town, but the remaining 42% of them (225) in a city.According to the BEM Sex Role Inventory, 18% of the fathers have masculine traits; 18% of them feminine traits; 38% of them androgen traits; and 25% of them undifferentiated sex traits (Table 1).

Data collection tools
In this research, Communication Skills Inventory, The BEM Sex Role Inventory and The Inventory of Supporting Developmental Tasks are used as data collection tools.Various characteristics of these scales are described below:

Communication skills inventory:
The inventory is developed by Balcı and finalized by the studies of Ersanlı and Balcı (1998).Communication skills inventory is composed of likert-type 45 questions.The inventory consists of three sub-scales that measure communication skills in "mental", "emotional" and "behavioral" aspects.The items are graded as "Always=5", "Usually=4", "Sometimes=3", "Rarely=2" and "Never=1".While the maximum score gained from the scale can be 225, the minimum score is 45.For each sub-scale, the maximum score is 75 and minimum score is 15.If the individual has the highest score in one of the subscales, than it can be argued that the individual's communication skills are good at that sub-scale.For the whole scale, high score represents the high communication skill of the individual.Within the development process of the scale, the Cronbach Alpha coefficient which is applied to determine internal consistency is found as 72.The test-retest method is used for reliability testing and Cronbach Alpha reliability coefficient is found as 68.As a result of the bisection method, Cronbach Alpha reliability coefficient is found as 64.The study conducted by Gürşimşek et al. (2008) in order to determine the internal reliability of communication skills inventory found Cronbach Alpha coefficient as 72.The correlation of subscales and total communication skills are 83, 73 and 82 respectively.Validity coefficient of the scale is 70.

The BEM sex role ınventory:
The inventory was developed by Bem in 1974.It was adapted to the Turkish community by Kavuncu (1987), who found the test-retest reliability of the inventory to be 0.75 for femininity and 0.89 for masculinity.The scales of femininity and masculinity contain 20 items severally, amounting to 40 items in total.The items are listed randomly in the form of one single scale.Individuals rate themselves on each item, on a scale from 1 (never or almost never true) to 7 (almost always true).Two particular scores are obtained from the scales of femininity and masculinity.The median score defines whether an individual is feminine, masculine, androgynous or undifferentiated.An individual is defined as androgynous in the event that his/her score of femininity is higher than the median of femininity and his/her score of masculinity is higher than the median of masculinity; as masculine in the event that his/her score of femininity is lower than the median of femininity and his/her score of masculinity is higher than the median of masculinity; as feminine in the event that his/her score of femininity is higher than the median of femininity and his/her score of masculinity is lower than the median of masculinity; and as undifferentiated in the event that his/her scores of both femininity and masculinity are lower than the medians of femininity and masculinity.Dokmen (1999) analyzed the psychometric characteristics of the Turkish Version of the Scales of femininity (F) and masculinity (M) in the BEM sex role ınventory.The analysis concluded that the internal consistency reliability of F and M is as follows: alpha coefficient of K is 0.73 and its split-half reliability is 0.76 (N=989).
Alpha coefficient of E is 0.75 and its split-half reliability is 0.75 (N=989).According to the scores obtained by all the subjects (N=1762) in the scales of K and E, the median of K is 111 (the average score being 5.55) and the median of E is 104 (the average score being 5.20).In the end, it was proposed that 111 (5.55) and 104 (5.20) should be used as medians for K and E respectively and they should be accepted as reliable norms in studies.The present study defines the feminine, masculine, androgynous or undifferentiated traits in consideration of the medians specified by Dokmen (1999) for K and E.

The scale of supporting developmental tasks:
The scale was developed by Unuvar and Sahin (2011) in consideration of the psycho-social developmental characteristics and developmental tasks specified by Erikson and Havighurst.An explanatory factor Ünüvar 707 analysis and confirmatory factor analysis were conducted for the construct validity of the scale, which yielded a five-dimensional structure consisting of 17 items.The dimensions are as follows: supporting language development, spending time in a qualified way, supporting action development, supporting emotional development and supporting self-care skills.The first six items of the scale constitute the first dimension; 7th to 10th items the second dimension; 11th to 13th items the third dimension; 14th and 15th items the fourth dimension; and 16th and 17th items the fifth dimension.Cronbach's alpha of the inventory was estimated to be 0.84; the reliability of the first dimension to be 0.81; that of the second dimension to be 0.70; that of the third dimension to be 0.73; that of the fourth dimension 0.86; and that of the fifth dimension to be 0.77.Furthermore, the test-retest reliability of the inventory was found to be 0.86.Getting high scores in the whole inventory indicates that the child is well supported in the process of development whereas low scores suggest that he/she is not supported properly in the process in question.The minimum and maximum scores that can be obtained in the inventory are 17 and 119 respectively.
The reliability test of this research that are applied on scale scores, the Cronbach Alpha reliability of "communication skills inventory" is found as 73; for the sub-scales, the Cronbach Alpha reliability is found as 65 for mental aspect; 61 for emotional aspect; 74 for behavioral aspect.For the "BEM sex role inventory", Cronbach Alpha reliability is determined as 71 for femininity sub-scale and 75 for masculinity sub-scale (N=540).Cronbach Alpha reliability of "The scale of supporting developmental tasks" is found as 85.
Cronbach Alpha reliability is found as 79 for "supporting language development"; 75 for "spending time in a qualified way"; 75 for "supporting action development"; 79 for "supporting emotional development" and 75 for "supporting self-care skills".

Data analysis
The data is analyzed by using statistical package for social sciences (SPSS) program and the impact of fathers' gender role on paternity characteristics is tested by ANOVA.The Levene test that was implemented before ANOVA statistics pointed out that the group variances are not distributed homogeneously and there are significant differences between them.For that reason, Welch test which is the alternative of F test and additionally among the robust tests Tamhane's T2 are used (Sipahi et al., 2006).Multiple regression analysis is applied to observe if paternity characteristics can be predicted with gender role and communication skills.Assumptions on the compatibility of the data for multiple regression analysis have been tested, and results are explained in the findings section.

FINDINGS Do the fathers' paternity characteristics vary according to their gender roles?
"The scale of supporting developmental tasks" scores are used as an indicator of paternity characteristics.Table 3.The correlation coefficient of the scores in "the scale of supporting developmental tasks", the "behavioral", "emotional" and "mental" sub-dimensions of the communication skills ınventory, and the "femininity" and "masculinity" sub-scales of the BEM sex role ınventory.according to the fathers' gender role, it has been found that the paternity characteristics of the fathers showing "femininity" and "androgynous" gender roles are significantly higher than the fathers showing "undifferentiated" gender role.

Paternity characteristics can be predicted with gender role and communication skills?
Firstly, the study made an analysis into the correlation among the quality of paternity, sex traits and communication skills.The study used the scores in the inventory of supporting developmental tasks as an indicator of the quality of paternity.It also investigated the correlation between the scores the fathers got in the inventory of supporting developmental tasks and their scores of femininity and masculinity in the BEM Sex Role Inventory as well as the scores in the "Behavioral", "Emotional" and "Mental" sub-scales of the communication skills inventory.The results are presented in Table 3.
There is a positive correlation between the scores in The ınventory of supporting developmental tasks, in the "behavioral", "emotional" and "mental" sub-dimensions of the communication skills ınventory and in the "femininity" and "masculinity" sub-scales of the BEM sex role ınventory"; and the correlation has a significance of 0.01 (Table 3).The correlation coefficients vary between 0.16 and 0.51 (<0.70), which suggests that no multiple correlation exists (Sipahi et al., 2006;Buyukozturk, 2010).Other indicators of the lack of a multiple correlation are presented in the following sections of the study.
The study investigated whether the scores in the ınventory of supporting developmental tasks can be predicted through sex roles and communication skills.To do so, the first thing the study did was to determine whether the data were suitable for a multiple regression analysis.Therefore, the study made an analysis into whether scatter plot and error terms are subject to a normal distribution or not.The study presented the findings in Figure 1.
In the scatter matrix, dependent and independent variables are linear and error terms distribute in a normal manner (Error terms are on the line in the diagonal) (Table 4).In the light of the findings, the data were found to be suitable for a multiple regression analysis.The results of the analysis are presented in Table 4.
In view of the findings 1-R 2 = 0.69 >20 and the biggest VIF=2.238<10, it is concluded that no multiple correlation exists between the variables (Sipahi et al., 2006;Buyukozturk, 2010).The multiple regression analysis suggests that the scores in the "behavioral", "emotional" and "mental" sub-dimensions of the communication skills inventory and those in the "femininity" and "masculinity" sub-scales of the BEM sex role inventory can account  for the scores in the inventory of supporting developmental tasks at a rate of 31%.The finding has a significance of 0.01.According to standardized regression coefficients, the order of importance to predicting the fathers' scores in the inventory of supporting developmental tasks is as follows: the score in "emotional" communication (β=366), the score in "behavioral" communication (β=244) and the score in "femininity" (β=101).It is observed that the fathers' scores in "mental" communication (β=049) and "masculinity (β=-0.29) are not a significant predictor of their scores in the inventory of supporting developmental tasks (p>0.05).In the light of these data, it can be concluded that their scores in the inventory of supporting developmental tasks can be predicted in a significantly meaningful way by their scores in "emotional" communication, "behavioral" communication and "femininity".Below is the mathematical model on predicting the scores in the inventory of supporting developmental tasks in accordance with the results of the regression analysis.

RESULTS AND DISCUSSION
The findings of this research argue that the fathers who show "femininity" and "androgynous" gender roles have high paternity characteristics.Gender role orientation covers the psychological and sociologic processes that include femininity and masculinity roles and responsibilities determined by women and men in the society (Cinamon and Rich, 2002).
The paternity characteristics in this research are defined by the inventory of supporting developmental tasks.The items in this inventory covers topics related to the fathers' participation to child care and education.As child care and education are interpreted as feminine practices, the fathers who show "femininity" and "androgynous" gender roles seems to have higher scores in terms of paternity characteristics.Tzuriel (1984) states that the individuals who show "androgynous" gender roles are more flexible, and are more adaptive to unexpected situations.Unexpected situations frequently happen during child care and rearing.The fact that "androgynous" fathers have high paternity characteristics may be based on their adaptability to unexpected situations.
The research conducted with married couples (Gunter and Gunter, 1990) highlights that the fathers who show "femininity" and "androgynous" gender roles tend to do more housework which is considered to be feminine.Individuals act in accordance with their perceived sex role rather than their actual sex (Athenstaedt et al., 2002).
Therefore, it is not surprising that those fathers who have feminine traits are more successful in child care and supporting their development, two tasks that are stereotypically considered as feminine.The findings about the predictors of quality of paternity the following variables were mentioned: "emotional" and "behavioral" communication skills and "feminine" traits.The positive correlation between the dependent and independent variables suggests that a significant increase is experienced in the scores in the inventory of supporting developmental tasks when fathers use emotional or behavioral communication skills and display feminine traits.
Furthermore, those fathers who use "emotional" and "behavioral" communication skills succeed better in child care and support as well as establishing more qualified interaction with their children.It has been reported in the literature that child development is affected in a positive way when parents use an emotional and supportive communication language (Brody and Ge, 2001;Ustun,2010).
The researches conducted in the recent years indicate that fathers have a higher impact on the well-being of children (England and Folbre, 2002;Palkovitz, 2002;Lamb and Lewis 2010).The children who are raised in a family environment where father-child communication and participation is effective, have developed positive behaviors and personality traits (Fagan and Iglesias, 1999;Lamb and Tamis-Lemoda, 2004).
These results points out that the number and quality of the studies and parent training programs should be increased in order to improve the quality of time spend between the father and children.The training programs should highlight the importance of communication with children and active participation to child development.The effects of these programs should be tested with follow-up researches.

Figure 1 .
Figure 1.Dependent and independent variables, scatter matrix and distribution of error terms.

Table 1 .
The characteristics of the study group.

Table 2 .
Tamhane's T2 results for the paternity characteristics score according to fathers' gender role.

Table 4 .
The results of the multiple regression analysis regarding the quality of parenting.