Perceptions of parental behavior with regard to parents ’ gender and respondents ’ age and gender

The findings of the research into the perceived parental behaviour provide contrasting data as to the existence and the nature of differences in the perception of parental behaviour based on parents’ gender and respondents’ gender. The purpose of the present study is to examine the differences in the perceived parental behaviour in adolescents with regard to parents’ gender and respondents’ age and gender. The study included 466 respondents (262 girls and 204 boys), in middle to late adolescence, divided into four sub-groups according to their age. The respondents were asked to fill in the Parental Bonding Instrument (PBI) questionnaire which measures the care and overprotection in mothers and fathers respectively. The obtained findings show the existence of significant difference based on the parents’ gender for both subscales: both maternal care and maternal overprotection were estimated as higher. Observing the differences by respondents’ gender on the whole sample, only one significant difference is found: paternal overprotection was estimated as higher by girls. The differences by age as observed within gender groups are completely disparate for girl and boy groups. The best insight into the differences is obtained through analysis by gender, for groups relatively homogenous in terms of their age (for the first three groups the only significant difference appears in the paternal overprotection subscale; the difference disappears in the subgroup of the oldest respondents’, while the differences between the perception of maternal and paternal care are of significance here). One particularly important finding for future research into rearing behaviour is the fact that the perception of parental behaviour changes over the period of adolescence differently for boys and girls; therefore, the analysis including perceived parental behaviour should be performed for subgroups by gender, which are as homogenous as possible in terms of their age.

The study of parental behaviour stems from the theory of attachment.Bowlby (1969, 1973, 1980, according to Stefanović-Stanojević, 2005) defined the attachment as a strong affectionate bond between the child and the caretaker, which is built in the first months of life, based on their interaction.Ainsworth, Blehar, Waters, andWalls (1978, Rikhye et al., 2008)

identified 3
Corresponding author: jela.ops@gmail.comtypes of attachments by observing the interactions of a child and a mother under controlled laboratory conditions ('the strange situation').Maternal behaviour was considered in relation to her care for the child, and responsiveness, that is the willingness to respond to child's needs.Parker, Tupling, and Brown (1979) reviewed the attachment literature (Ainsworh, Bell, & Stayton, XXXX ;Bowlby, 1969;Rutter, 1972) and several factor analytic studies (e.g., Raskin, Boothe, Reatig, Sculterbrandt, & Odle, 1971;Roe & Siegelman, 1963;Schaefer, 1965), in order to identify the key characteristics of paternal behaviour that affects the parent-child relationship.Two factors were identified: care and psychological control or overprotection.Based on these findings, Parker et al., (1979) developed the Parental Bonding Instrument (PBI) as a self-report questionnaire completed by offspring who report on their childhood experiences of these two dimensions of parental behaviours and attitudes (for father and mother separately).The Care factor has two poles, measuring parental warmth, affection, and involvement versus coldness, rejection, detachment, and aloofness, respectively.The second factor, Overprotection, is also bipolar, and measures psychological control, infringement, and imposition versus encouragement of autonomy and selfsufficiency.
Therefore, it is a matter of subjectively perceived behavior, rather than parental behavior that is assessed by an independent investigator.The child's perception of their parents has been studied since 1894 (Goldin, 1969in Arrindell, Gerlsma, Vandereycken, Hageman, & Deaseleire, 1998).During the 1930s, it was recognized that this perception could be as important as the data collected from parents themselves, and in some cases even more important than real behavior of parents (Parker, 1984;Lewinsohn & Rosenbaum, 1987;Richman & Flaherty, 1987;in Arrindell et al., 1998).Agreement on parental behavior is not great between parents and children and between siblings themselves, one would rather say that there is a gap, at least in families with a member who has problems with psychological functions (Gerlsma, Snijders, van Dujin, & Emmelkamp, 1997).The study of perceived parental behavior has the advantage in allowing a person to express their perception of parental behavior through their own cultural and individual filters, thus avoiding the possibility of false interpretation of the meaning of a caregiver's behavior (Rohner, Khaleque, & Cournoyer, 2009).

Gender effects in perception of parental behaviour
Some research into the perception of parental behaviour and its association with the adaptive and maladaptive personality traits include the data on the differences in respect to parents' and/or respondents' gender (unfortunately, many studies do not quote such data, as rearing behaviour was not the variable examined (e.g.Perry, Silvera, Neilands, Rosenvinge, & Hanssen, 2008;Tata, Fox, & Cooper, 2001) some studies treat the rearing behaviour of a mother and father as one variable (e.g.Carter, Sbrocco, Lewis, & Friedman, 2001;Wiseman et al., 2006;Perry et al., 2008).We would like to stress that we are quoting only the studies that relied on Parental Bonding Instrument PBI (Parker et al., 1979).The study by Rihkye et al., (2008) comprising respondents ages 18-65 (with and without history of maltreatment) presents the findings that show the existence of differences in how subjects described bonding with their female parent in comparison to their male parent.Mean scores for maternal care were significantly higher than those for paternal care; the mean score for maternal overprotection was also somewhat higher than the mean score for paternal overprotection.Second gender effects analysis of the same study examined differences related to the gender of the subject within each group.Significant gender differences emerged only for the parental care factor in the maltreatment group.Men with a history of maltreatment reported receipt of greater maternal care than did women with similar histories of maltreatment.In contrast, women in the maltreatment group reported receipt of greater care from their fathers than did men in the same group.No gender differences were found in the reporting of parental overprotection in the maltreatment group.Similarly, no gender effects were found in the control group for parental care and overprotection.There are no gender differences in the perception of parental behaviour in the study by Carter et al., (2001) on African American and European American college students (mean of age 19.24 for European and 19.85 for African American students).In the study by Sunday et al., (2008) which seeks to compare the abused and nonabused children (ages between 12-18) no gender differences were found in the perception of rearing behaviour of mother and father in the non-abused group.All the cited studies relied on the Parental Bonding Instrument (PBI, Parker et al., 1979) and were conducted in the USA.The study by Sunday et al., (2008) points out that gender differences in the relationships between non-abused adolescents and parents have been more widely studied (e.g., Starrels, 1994;Steinberg, 1987;according to Sunday et al., 2008).A review by Russell and Saebel (1997), according to Sunday et al., 2008), however, concluded that while some gender differences were found in individual studies, no clear, definitive patterns emerged to indicate that father-daughter, mother-daughter, father-son, or mother-son relationships were fundamentally different across studies.
On the other hand, there are studies indicating the existence of gender differences, both for parents' and respondents' gender.One striking example is the paper by Rigby, Slee, & Martin (2007).Their study included Australian students (n=1432) aged 12-16 years, who completed the PBI questionnaire, among other forms.The findings show gender differences were evident for three of four parental bonding measures (in contrast with the findings of studies by Rihkye et al., 2008;and Carter et al., 2001).Girls rated perceived care from mother and from father as greater, and also perceived father control as greater.Further analysis indicated that for both girls and boys mothers were seen as more caring than fathers.(This is the difference based on the parent gender, in the favour of maternal care as shown in the study by Rihkye et al., (2008); while Rigby et al. did not obtain any differences for the dimension Overprotection).For girls, mothers and fathers were not seen as differing in parental control.However, boys saw their mothers as more controlling than fathers.The authors go on to conclude that the existence of substantial gender differences requires consideration of the association of these variables with other variables separately for boys and girls.
In conclusion, even the studies in which the same instrument was used (PBI, Parental Bonding Instrument), show somewhat clashing results when it comes to gender differences.The seemingly undisputable difference is the one concerning the parent gender and the care dimension: mothers are perceived as more caring than fathers.This means that there is no justification for the observation of maternal and paternal care as one dimension -parental care, which is sometimes done by researchers (e.g.like in the above mentioned study by Carter et al., (2001), which can be the reason why non-significant gender differences were obtained (for respondents' gender in this case).It remains unclear if the differences in the perception of maternal overprotection and paternal overprotection exist.Another question that remains unanswered is: are there any differences in the perception of rearing behaviour based on the respondent's gender?This perplexity could not be easily attributed to different ages of the respondents in different studies as there is evidence of long-term stability of the PBI scores over time (Wilhelm, Niven, Parker, & Hadzi-Pavlovic, 2005).Still, it is worth noticing that the study by Wilhelm et al. is longitudinal, while vast majority of studies on parental behaviour relying on PBI or other instruments are cross-sectional.It should also be noted that the respondents were in their late adolescence (mean age 23 years) in the early stages of the study.
Before we go on to present the goals and hypothesis of this study we would like to point out that in this country (Serbia) there were no studies relying on PBI, but there are studies using EMBU (PBI and EMBU are the two most commonly used instruments in the examination of parental behaviour; their subscales have similar measure subjects (Arindell et al., 1998).Todorović (2005) conducted a study on secondary school students using the EMBU and obtained significant gender differences in the perception of the emotional warmth displayed by the mother: girls rated mothers as warmer than boys.Regarding the other dimension measured by the EMBU (inconsistency, rejection, overprotection) there were no gender differences.There is no data on the differences based on parents' gender.
One earlier study on local population (using several different questionnaires) came up with the following: 'regarding the gender, girls more frequently perceived warm-permissive practices in their mothers, than boys (…).At the same time, boys more frequently perceived cold-restrictive practices in both parents and cold-restrictive practices in fathers compared to girls.(Genc, 1994, pg. 21-22) THE PRESENT STUDY The goal of the present study is to establish the differences in the perceptions of rearing behaviour based on parents' gender and respondents' gender.Namely, (1) we wish to determine the existence of any significant differences in the perceptions of maternal and paternal care respectively; as well as the perception of maternal and paternal overprotection (based on the parent gender).Further (2) we will seek to determine if there are gender differences in the perceptions of maternal and paternal care and overprotection (based on the respondent's gender).Based on the existing findings it can be assumed that the age of respondents is related to the perception of parental behaviour so (2a) we will look for differences in boys and girls within the 4 age groups, two of which are in middle adolescence, and remaining two in late adolescence (establishing age differences within groups by gender).In addition, we expect the boys and girls within the same age groups to differ in their perceptions of parental behaviour (2b -determining gender differences within the same age group).We expect to obtain significant differences in the perceptions based on parent gender, which would among other things mean that these variables are to be treated separately (e.g.maternal care and paternal care in contrast to parental care as one value).We also expect to obtain significant differences in the perception of parental behaviour based on the respondents' gender.Furthermore, we expect to obtain significant differences in respect to age and gender -gender differences within groups and between groups of boys and girls of the same age; that is, we assume that the perception of parental behaviour changes over the period of adolescence differently for boys and girls.

Method
Sample: The sample comprised secondary school students attending grammar school and Art School in Niš, as well as students from the Faculty of Philosophy and the Medical Faculty in Niš (the Art School students were chosen among secondary school students so that there would be a greater number of respondents living away from home during their schooling in this population as it is the case with university students).Adolescents participating in the research were in their middle adolescence (the second and forth grade of secondary school) and late adolescence (the second and fourth year of university).Descriptive data on the sample is shown in Table 1.Measures: Parental Bonding Instrument, PBI (Parker et al., 1979), measures basic parenting styles as perceived by a child.The measure is retrospective, which means that respondents over 16 years of age respond to statements based on their memories of parental behaviour in the fi rst 16 years of their lives.Participants respond to the same items for mother and father respectively.There are two subscales: care and overprotection.Care subscale consists of 12 items (e.g.(my mother/my father) "Enjoyed talking things over with me", "Frequently smiled at me", "Did not seem to understand what I needed or wanted'), while Overprotection has 13 items (e.g.(my mother/my father) "Did not want me to grow up", "Tried to control everything I did", "Invaded my privacy", and participants respond by circling a number on a four-degree scale, ranging from 0 = very unlike to 3 = very like (my mother, or my father)).

Results
First, differences between dimensions of care and overprotection coming from a mother and a father were analysed (i.e. the differences according to parent's gender); the paired samples t-test was used.Descriptive data for these dimensions and t-test results are shown in the Table 2.The next step was to determine differences on PBI dimensions in relation to the gender of respondents, i.e. a "child"; the independent-samples t-test was used.The results are given in the Table 3. Accordingly, t-values were computed using separate estimates of variance.
The differences within gender and age groups of respondents were also determined.The results of the analysis of variance for boys of different age are given in the Table 4; the results for girls are presented in the Table 5. Post-hoc test showed that maternal care gradually rises with the age of boys, hence in the second year of university (age mean 20.11) maternal care is at its peak and is significantly different compared to two lower age periods, but not different from the care as estimated by the oldest respondents (the differences appearing in the fourth year of university studies are of borderline significance compared to the second and fourth year of secondary school, but there are still no significant differences which might have been obtained had the subsamples been larger).
When it comes to paternal care its values seem to decline slightly with the age of respondents.Significant differences are present in the oldest group of respondents where the perceived care is significantly lower compared to other groups.Post-hoc test was performed, and based on the results it can be concluded that the estimations of maternal care in girls have the opposite tendency to those of boys -this type of care declines with the age and it is at its lowest in the fourth year of university (the estimations given by second grade female students at secondary schools and second year university female students are significantly higher compared to the estimations given by the fourth year female university students; however, the differences in estimations given by the fourth year female students at secondary school are not significant in comparison to those given by the oldest respondents, however the same trend can be found -the higher maternal care is estimated.A larger subsample should be provided in order to obtain more conclusive findings.Paternal overprotection is differently perceived by girls of different adolescence periods: the post-hoc test shows that middle adolescent girls (the second and fourth year of secondary school) estimate higher paternal overprotection than late adolescent girls (the second and fourth year of university studies).Total differences in the paternal care and maternal overprotection are not significant.
Finally, the differences in relation to gender within the age groups were also determined; the t-test results are given in the Tables 6, 7, 8, and 9. Tables 6, 7, and 8 show only one significant difference for the first three age groups with regard to respondents' gender -it is the difference in the perception of paternal overprotection.Girls perceived their fathers as more overprotective compared to boys.In table 9 with the oldest age group, the results go to show that the difference in the perception of paternal overprotection no longer exists; significant differences are those in the perception of care of both parents in the way that maternal care is perceived as higher by boys, while paternal care is rated as greater by girls.

DISCUSSION
The differences in the perception of parental behaviour with regard to parents ' gender In this study the significant differences emerged in the perception of parental behaviour of a mother and a father.Adolescents perceived maternal care as higher, as well as maternal overprotection, compared to those of a father.In other words, for the adolescents (coming from south-east Serbia) mother is the figure with more prominent parental behaviour.
Based on the findings of earlier studies, we can assume that the perception of maternal care as higher was expected.Regarding control, that is overprotection, there is insufficient data, but we can make a verifiable assumption that this was the dimension under the greater influence of culture from which the respondents came from.It is possible that non-significant differences were obtained in environments in which both the parents and the children think that controlling child's behaviour is the duty of both parents (and conversely -significant differences were obtained in environments where it is believed that the whole rearing of the child is the duty of a the mother).It is also possible that differences exist within the same culture based on the socio-economic background, primarily concerning the education level of parents and financial status of the family (e.g.Zotović, Mihić, & Petrović, 2007).Regarding our sample, the great majority of respondents opted for the neutral category of financial status (namely, a small number of respondents stated their financial status as low or high so it was impossible to analyze differences).Educational differences between parents were analyzed, but no significant differences were discovered.

The differences in the perception of parental behaviour with regard to respondents' gender
Only one significant difference emerged and that is in the dimension paternal overprotection which was rated as significantly higher by girls compared to boys.Research done abroad came up with the same findings (Rigby et al., 2007), unlike the studies done in Serbia (Todorović, 2005;Genc, 1994).It is curios that the absence of differences in care, where we would expect the girls to rate the care of both parents as higher as in Rigby et al., 2007 or just rate maternal care as higher (Todorović, 2005;Genc, 1994) compared to boys.Since there is a possibility that this discrepancy in findings concerning differences by gender is in fact a result of age differences among the respondents, analyses by age were performed within gender groups, followed by the analyses by gender within the age groups.

a. Age differences within groups by gender
Within the boy group significant age differences appear in the perception of maternal care and paternal care.The differences in the estimations of the two dimension go in opposite directions -maternal care is perceived as higher in late adolescence (in comparison to ratings by adolescents in the middle adolescence); while paternal care is perceived as lower.
Within the girl group differences occur in the perceptions of maternal care and paternal overprotection.Girls' estimates of maternal care are completely different from those of boys -the girls perceive maternal care as declining.Paternal overprotection declines starting from the transition from middle to late adolescence.
Before we try to come to a conclusion based on these findings, it is worth examining the differences between the boys and girls belonging to the same age group.

b. Gender differences within the same age group
For the first three age groups the only significant difference is in the boys' and girls' perception of paternal overprotection: girls perceive their fathers as more overprotective, as can be seen in the overall gender differences (higher rating of maternal care by girls has not yet appeared).These groups include respondents in middle adolescence and those in the early stages of late adolescence.However, in the oldest age group (late adolescence, mean of age 23.7) the nature of differences of the perceptions of parental behaviour made by boys and girls changes completely: girls rate maternal care as lower, and boys of this age rate maternal care as higher.Significant differences are also in the perceptions of paternal care: they are the opposite compared to perceptions of maternal care -girls now perceive care as higher, while boys rate it lower.

On the differences in perceptions of parental behaviour with regard to respondents' age and gender
We believe that the most important finding of the present study is the presentation of the differences in the results due to the different treatment of the sample.If we return to overall gender differences (paternal overprotection is only significant, higher scores obtained in girls) we notice that failing to consider age differences even within the same age (adolescence) concealed the differences which can be found when analysing results by gender within the same age (2b, in the oldest age group, differences in the perception of paternal overprotection disappear, but they appear in the maternal care and paternal care).Essentially the same, but less clear results are obtained through the analysis of age differences, within groups by gender (2a).We can assume that the perception of parental behaviour over the course of adolescence is not so stable as it appears to be after late adolescence and during adulthood (Wilhelm et al., 2005).The differences in the results of the study into perceived rearing behaviour can be accounted for by the differences in the respondents' age (especially if adolescents are in question), as well as by neglecting gender differences.

Limitations
The main limitation of the present study comes from the fact that this is a cross-sectional survey, that is, it included boys and girls of different ages coming from different families.Obtaining clearer findings takes conducting a longitudinal study and/or the survey of the children from the same family.In addition, it should be noted that this research was performed on the sample of adolescents coming from the middle socio-economic stratum, so there are very few cases of extreme categories of financial status and educational background of parents.

Implications
In the estimation of parental behaviour of the father and the mother, significant differences occur, so we believe it is necessary to treat these variables separately (instead of using it as a combined measure).Not keeping the two apart could lead to unreliable findings in cases where the association of the perceived rearing behaviour with different personality traits of 'children' is examined.More frequent presentation of the data on the differences in the perception of maternal and paternal care and overprotection in these studies would help make a complete picture concerning the variants of differences that exist therein, and establish which variables they are linked to.
In the studies of the perceived parental behaviour conducted on adolescents, it is necessary to consider not just the gender, but also the age of adolescents.If the age group is treated as an integral subsample, it should be as homogenous as possible in terms of the respondents' age, as the perception of parental behaviour changes over the course of adolescence and probably becomes stable only after the period of late adolescence.

Table 1 .
Descriptive data of the sample

Table 2 .
Mean scores and standard deviations on PBI dimension for the whole sample, with the comparative results by parents' gender

Table 3 .
Overall differences by respondents' gender Indicates that the variances of scores for boys and girls differed significantly (p<.05) by Levene`s test. *

Table 4 .
Mean scores and standard deviations on PBI dimension for boys with the test of significance of differences for different age groups

Table 5 .
Mean scores and standard deviations on PBI dimension for girls with the test of the significance of differences for different age groups

Table 6 .
PBI dimensions by gender, for the first age group (age mean 16.6), with comparisons

Table 7 .
PBI dimensions by gender, for the second age group (age mean 18.4), with comparisons

Table 8 .
PBI dimensions by gender, for the third age group (age mean 20.11), with comparisons

Table 9 .
PBI dimensions by gender, for the fourth age group (age mean 23.7), with comparisons