Attachment Styles as Predictors of Stigma Tendency in Adults

The purpose of this study was to examine the association between attachment styles and stigma in adults. Participants were 361 adults (186 females and 175 males) aged between 18 and 69 (M=31.77, SD=9.45). Participants completed the measurement instruments for determining their stigmatizing tendencies and attachment styles. Study results showed that, stigma tendencies of people with the secure attachment style are lower for the discrimination and exclusion, prejudgment and psychological health dimensions, and are higher for people with the fearful attachment style for the discrimination and exclusion, labeling and psychological health dimensions. Preoccupied and dismissive attachment styles are also positively associated with prejudgment tendency. Finally, stigma tendencies of males are more likely to be higher than females for the discrimination and exclusion, labeling and psychological health dimensions. Because different attachment styles are related variously to the subscales of stigma in this study, interventions to decrease stigma of individuals can verge to enhancing the quality of mother-child interactions.


Introduction
According to the World Health Organization (World Health Organization, 2002), stigma and discrimination are main topics worldwide in improving the standards of mental health practices.Many technical developments and social transformations witnessed in daily life may negatively affect stigma tendency.Identifying the determinants of stigma tendency, which is the individual's internal and social area of influence, will have significant contributions in understanding the individual's relationships with the self and the environment, in solving possible questions, in the type of interference and in determining the practices.Hence, the purpose of this study is to examine the relationship between different attachment styles and the sub-dimensions of stigma.
When considered within the social context, the individual is a reflection of public opinion within the society he or she exists.From this point of view, the individual recognizes public opinions and adopts these public opinions, attitudes and behaviors into his or her own life.Although this offers the individual with a secure framework in adapting to society, it can also lead to negative outcomes, both for the individual and for the other people.One of these possible problems is stigma, which is defined as the individual's tendency to displaying the need to marginalize or exclude people who are perceived as different by society (Corrigan, Roe, & Tsang, 2011).Stigma is defined by Goffman (1963) as, perception based on evaluations which the majority of society find odd and which are made about an individual due to features such as a psychological illness, drug use or physical handicap.Goffman (1963), who examines stigma as a process, identifies the process as the communication between an individual's characteristics and the values recognized by society.According to Goffman (1963), stigma emerges when the individual's characteristics and the tendencies accepted by the public contradict with each other (King et al., 2007).The difference between what's normal and what is not even leads to discussions about whether or not the person is human (Green & Sobo, 2000).The perception considering the stigmatized individual as a 'non-human', avoidance of some characteristics of the stigmatized individual and the belief of having to mistreat this person introduces us the fact that stigma considerably affects social relationships, beyond personal perceptions.Due to the distinctive features of stigmatized individuals, it is believed that they carry a sign or trace that belittles or slanders them (Green & Sobo, 2000;Hebl & Dovidio, 2005).
If we are to define the stigma behavior under the three elements, emotion, thought and behavior, of the cognitive structure; the repetitive and negative thoughts of the social structure considering stigmatized individuals, the negative-affective responses resulting from confirming these prejudgments and the behaviors which are displayed due to these prejudgments (Overton & Medina, 2008;Peterson, Barnes, & Duncan, 2008).The terms stigma tendency, discrimination, marginalization and violation of human rights are usually related to each other.Discrimination is defined as individuals being deprived of certain rights and interests due to the stigma and prejudgments asserted by people or groups of society (Green & Sobo, 2000).On the other hand, age, gender, marital status, tribal background, sexual preferences, and lifestyle, along with discrimination, prejudgment and labeling are considered as the dimensions of stigma tendency (Yaman & Gungor, 2013).Goffman's (1963) typology of stigma consists of three dimensions, 'abominations of the body', 'blemishes of individual character' and 'tribal stigma'.While abominations of the body indicate negative attributions to physical appearance, blemishes of individual character EDUPIJ / VOLUME 5 / ISSUE 1 / SPRING / 2016 indicate negative attributions to the personality traits.Tribal stigma identifies negative attributions concerning a society with specific racial or religious features (Kus-Saillard, 2010).
Distinctive features that cause individuals or groups to be stigmatized can emerge according to time, culture or universal values.While leprosy is not considered a stigma factor today, it is known that diseases such as obesity (Puhl & Heuer, 2010) or HIV/AIDS (Earnshaw, Bogart, Dovidio, & Williams, 2013;Earnshaw, Smith, Chaudoir, Amico, & Copenhaver, 2013) may cause people to be stigmatized.The stigmatized individual may be evident in the social structure with a marginalized and weak self-structure (Meyer, 2003;Miller & Kaiser, 2001).This issue, which is related to the negative attitudes displayed towards the stigmatized person, is accepted by society during the socialization process and leads to an unhealthy social distance (Mueller et al., 2006).With regards to the relationship between stigma tendency and gender, gender is a crucial variable in shaping daily practices and general attitudes and perceptions.Both biologically and sociologically, the individual's gender can be the main determinant for stigma tendency (Topkaya, 2014;Hosgorur & Gecer, 2012).
Attachment is defined as the permanent results that an individual's experience with his or her caregiver during the initial years has on the way the individual approaches his or her own life and interpersonal relationships (Bowlby, 1973;Mikulincer & Shaver, 2007).According to Bowlby's (1973) attachment theory, the moment people are born they seek attachment to people surrounding them and intent to perpetuate relationships with their environment based on love and trust.The attachment which begins during childhood and expands towards adolescence and adulthood progresses from the caregiver/mother to the family, close friends, groups and societies (Bowlby, 1973).Attachment is generally observed when the individual feels afraid, becomes exhausted or feels ill, and the extent of the attachment behavior is related to the attitudes of the attachment figure.Thus, the sensitivity the attachment figure displays to the attaching figure leads to a lifelong observable behavior pattern (Ainsworth, Blehar, Waters, & Wall, 2015).Colleagues who preferred religious/political groups can be initial attachment figures for many people.According to the attachment theory (Bowlby, 1973), children use their infancy experiences and relationships with their mothers as models for all kinds of relationships in their future lives; which is the way the self-model and the others model develop.The self-model is the degree to which the individual sees him/herself as worth loving and precious.The other model is the way an individual perceives other people as reliable and ready to give love and care.Bartholomew and Horowitz (1991) define adult attachment dimensions as the two internal working models, 'the model of the self' and 'the model of others'.According to this model; there are four different types of attachment styles based on positive and negative opinions towards the self and other people.Attachment styles predict the quality of an individual's relationships for the further stages of life and aim at identifying data which explain how the individual enters, improves and perpetuates relationships with the environment.People with secure attachment are more likely to have positive attitudes towards themselves and other people.They also tend to find themselves worth loving and find other people acceptable and supportive.People with preoccupied attachment styles are more likely to have negative opinions about themselves but positive opinions about others.In other words, they find other people worth loving while finding themselves not.People with dismissive attachment individuals are more likely to be positive towards themselves, EDUPIJ / VOLUME 5 / ISSUE 1 / SPRING / 2016 but negative towards others.These individuals tend to give particular importance to autonomy need.Unlike the secure attachment style, people with fearful attachment style are more likely to find themselves unworthy of other peoples' love and support, and perceive other people as unacceptable and unreliable (Bartholomew & Shaver, 1998).The self and the others models are rather stable.Although we constantly change with regards to development, the change in the internal working models does not vary considerably.The fact that the past experiences with the attachment figure will determine future experiences to some extent, and the idea that the limitations of the experiences with the figure are predicted, prevent the change in the identified internal working models (Cassidy & Shaver, 2008).
Research has shown that secure attachment individuals display more constructive behaviors in coping with negative feelings during their social relationships (Kobak & Sceery, 1988).This may help them to display constructive, unprejudiced and collective attitudes towards the people around them.Studies examining attachment styles have found that while avoidant and fearful attachments serve as mediators for self-stigma and anxiety for seeking professional psychological help (Nam & Lee, 2015), avoidant and anxious attachment styles were observed to be related to stigma against psychological deficits (Vogel, Shechtman, & Wade, 2010).A study by Tamaki and Takahashi (2013) indicated that, when compared to secure and anxious attachment individuals, avoidant and fearful attachment individuals displayed low levels of social skill.Studies also suggest that attachment styles are correlated with the tendency to depression (Bifulco, Moran, Ball, & Lillie, 2002;Cooley, Van Buren, & Cole, 2010).Moreover, depression and anxiety symptoms are evident more in people with anxious attachment style (Mikulincer & Shaver, 2007), while anxious and avoidant attachment styles are evident in individuals in need of less social support (Shallcross, Frazier, & Anders, 2014).
Despite the developing and changing perception events, the tendency towards stigma continues to be a major problem for every society (Green & Sobo, 2000).It is valuable to examine the relationship between the attachment styles and stigma tendency of adults.Therefore, we predicted that attachment styles, and gender would show significant associations with stigma dimensions.Specifically, secure attachment would be negatively related to the dimensions of the stigma tendency.Contrary to secure attachment, fearful, preoccupied, and dismissive attachment styles would be positively correlated with the dimensions of stigma tendency.Finally, males would show more stigma tendency than females.

Methodology
This research is a quantitative study which examines the relationship between various attachment styles and stigma dimensions.Dependent variables of this study are dimensions of stigma namely, the discrimination and exclusion, labeling, psychological health and prejudgment.The independent variables of this study are gender, secure, fearful, preoccupied and dismissive attachment styles.
Participants of this study comprised of individuals attending various non-governmental organizations which serves to various age and socio-economic groups for social activities such as education, culture, art, etc. in the Ilkadım region in Samsun, Turkey.Samsun is an economically developed and most populous city in the central Black Sea region of Turkey.
Samsun is approximately 400 kilometers from Ankara, the capital of Turkey.The participants were informed about the study's purpose before it was conducted and they were asked to answer the questionnaire voluntarily.The study participants consisted of 361 (186 female and 175 male) adults whose ages ranged between 18 and 69 (M=31.77,SD=9.45).The educational level of the participants showed that 2.8% received only primary school education, 18.8% were from high schools, and 67.0%graduated from university, whilst 11.4% had obtained masters or doctorate degrees.
The study questionnaire consisted of a personal information form, stigma scale and relationship scales survey.Details on the psychometric properties of these scales are as follows: Personal Information Form: The participants were asked to indicate their age, gender and highest educational attainment level.
Stigma Scale: The Stigma Scale, developed by Yaman and Gungor (2013) in the context of Turkish culture, was used in the study to determine the psychological stigma levels of participants.The scale assesses stigma under four dimensions.The first subscale, 'discrimination and exclusion', aims at determining the discrimination and exclusion tendency regarded as a result and an indicator for stigma tendency.The second subscale, 'labeling', aims at assessing the tendency to label individuals with regards to the gender, marital status, age, tribal background, and sexual preferences of the person.The third subscale, 'psychological health', measures the stigma tendency towards individuals with psychological deficit.The fourth subscale, 'prejudgment', measures the stigma tendency towards individuals who have different characteristics such as the tendency for crime, world perspective, seniority, and lifestyle.The scale's construct validity was examined by the researchers via Exploratory Factor Analysis, and found that the scale consisted of the above mentioned four factors.The four factors explained 43.63% of the total variance.The first subscale (discrimination and exclusion) consists of six items with factor loadings ranged between .47 to .70; the second subscale (labeling) also consists of six items with factor loadings ranged between .52 to .60; the third subscale (psychological health) consists of five items with factor loadings ranged between .50 to .66; and the last subscale (prejudgment) consists of five items with factor loadings ranged between .47 to .64.
Participants answered each statement of the scale by marking Likert-type options ranging from Totally Disagree (1) to Totally Agree (5).The Cronbach alpha internal consistency coefficient of the scale reported by Yaman and Güngör (2013) was .84.While the scores that could be obtained from the scale's psychological health and prejudgment subdimensions ranged between 5 and 25, they ranged between 6 and 30 for the discrimination and exclusion and the labeling subscales.Higher scores in each subscales indicate higher level stigma tendency for that dimension.
Relationship Scales Questionnaire (RSQ): The Relationship Scales Questionnaire (RSQ), developed by Griffin and Bartholomew (1994), was used to determine attachment styles of the participants.The scale was adapted into Turkish culture by Sümer and Güngör (1999).The scale consisted of four subscales to measure the secure, fearful, preoccupied and dismissive attachment styles.RSQ is a seven-point Likert type self-report scale consisting of 17 items.The internal consistency coefficients of RSQ's sub-scales identified by Sumer and EDUPIJ / VOLUME 5 / ISSUE 1 / SPRING / 2016 Gungor (1999) range between .27 and .61.The scale's test-retest reliability range between .54 and .78.Studies in which the scale was used show that the reliability of the subscales were low.Griffin and Bartholomew (1994) state that the low Cronbach alpha values of the sub-scales may be due to the fact that different models are present in the same style and that the sub-scales were assessed with limited number of items.Although the reliability values of the subscales were low, it was observed that their construct validity were adequate (Griffin & Bartholomew, 1994;Sumer & Gungor, 1999).
Participants answered each statement of the scale and explained to what extent they agree by marking the options ranging from It totally doesn't describe me (1) to It totally describes me (7).The fifth item is used by two subscales both by being coded reversely and straightforwardly.Thus, questions 1, 4, 9, and 14 measure the fearful style, while questions 2, 5, 12, 13, and 16 measure the dismissive style, and questions 3, 7, 8, 10, and 17 measure the secure style, and finally, questions 5 (reverse scoring), 6, 11, and 15 measure the preoccupied style.The scores which reflect the four attachment styles are calculated by adding the items that measure the styles, and then dividing the total with the number of items in each subscale.Thus, the scores which can be obtained from the subscales range between 1 and 7.The subscale which the individual obtains the highest score denotes his or her attachment style (Sumer & Gungor, 1999).
The questionnaires were administered to participants under the presence and supervision of a study researcher.Participants completed the questionnaire in approximately 20 minutes.Pearson correlation analysis was conducted to determine the relationships between stigma (discrimination and exclusion, labeling, psychological health, and prejudgment) and the attachment styles (secure, fearful, preoccupied, and dismissive).Multiple regression analysis was conducted to determine to what extent stigma (discrimination and exclusion, labeling, psychological health, and prejudgment) was predicted by the attachment styles (secure, fearful, preoccupied, and dismissive).Before conducting the analyses, accuracy and compatibility of the data, whether or not any outliers were present in the dataset, and the statistical analyses assumptions were examined.In order to determine the outliers in the dataset, scores of the secure, fearful, preoccupied, and dismissive subscales of the attachment scale and the discrimination and exclusion, labeling, psychological health, and prejudgment subscales of the stigma scale were transformed into standardized z-values and individuals who obtained scores beyond +3 and -3 were excluded from the dataset, as suggested by Field (2013).A total of 11 outliers, four from the discrimination and exclusion subscale, one from the labeling subscale, one from the psychological health subscale, two from the prejudgment subscale, two from the secure attachment scale, and one from the fearful attachment scale were excluded from the dataset because they obtained scores other than these values.Field (2013) states that the Pearson correlation coefficient analysis has three assumptions.These are, the data being gathered in an interval scale, the data displaying normal distribution and a linear relationship exist between two variables.The first assumption is related to the research design.Because the sample size was large, based on the suggestions of Tabachnick and Fidell (2012), the second assumptions were examined with Histogram, normal Q-Q plot and Box-Plot graphics and the data were observed to be approximately normally distributed.The last assumption, the linearity, was evaluated by examining the scatter plots between the subscales of stigma and the subscales of EDUPIJ / VOLUME 5 / ISSUE 1 / SPRING / 2016 attachment.Inspections of the scatter plots revealed that the variables have linear relationships.Pearson correlation coefficient value can range between +1 and -1.Positive relationship between variables indicates that individuals who obtain a high score from one variable tend to obtain a high score from the other variable, whereas a negative relationship indicates that individuals who obtain a high score from one variable tend to obtain a low score from the other variable.Only the variables which resulted with significant relationships in the correlation analysis were included in the regression analysis.Accordingly, the multiple regression analysis has five assumptions; sample size, linearity, heteroscedasticity, normality of error distributions and multicollinearity.
Firstly, sufficient sample size was examined based on Green's (1991) formula and it was observed to be 109 for this study.Because the sample size of this study was 361, this assumption was confirmed.The linearity assumption was examined and confirmed before the correlation analysis, thus, it wasn't reexamined.Heteroscedasticity was tested by examining the scatter diagrams between the residuals of the independent variables and the dependent variables suitable for each regression analysis.The errors were observed to be normally distributed.Normality of errors was tested through the scatter diagrams of unstandardized residuals of each regression analysis, it was observed that error distributions were approximately normally distributed.Finally, VIF and tolerance values were examined to test multicollinearity.This assumption would be confirmed if the tolerance value is above .10and the VIF value is below 10 (Mooi & Sarstedt, 2011).VIF and tolerance values were observed to be between these ranges in all analyses.
The gender variable and attachment styles were included in the analyses as the dummy variables.Like the Pearson correlation analysis, the beta (β) coefficients of the regression analysis enable interpreting of the relationship, the explained total variance (R 2 ) indicates the level of the relationship between variables.Levine (2013) suggests that statistical analysis results should be provided with effect size estimates.Effect size indicates to what extent results are significant in practice.The effect size classification suggested by Cohen (1992) was used in interpreting the values of the correlation coefficients and the explained variance concerning the models.According to Cohen (1992)
On the other hand, there was a negative relationship between discrimination-exclusion and secure attachment (r = -.15,p< .01),but a low level positive relationship between fearful EDUPIJ / VOLUME 5 / ISSUE 1 / SPRING / 2016 (r = .18,p< .01)attachment and gender (r = .21,p< .01).There was also a negative relationship between labeling and secure attachment in low magnitude (r = -.12,p< .05),but there was positive relationship between fearful (r = .20,p< .01)attachment and gender (r = .14,p< .01) in small magnitude.There was a negative relationship between psychological health and secure attachment (r = -.14, p< .01),but there was a small positive relationship between fearful attachment (r = .16,p< .01)and gender (r = .11,p< .05).There was a negative relationship between prejudgment and secure attachment (r = -.14, p< .01),and a small positive relationship between fearful attachment (r = .14,p< .01),preoccupied attachment (r = .15,p< .01)and dismissive attachment (r = .16,p< .01).In order to determine to what extent the discrimination and exclusion, labeling, psychological health, and prejudgment subscales of stigma are predicted by the secure, fearful, preoccupied, and dismissive attachment styles and the gender variable, a series of regression analyses were conducted by discrimination and exclusion, labeling, psychological health, and prejudgment subscales of stigma as dependent variable and the secure, fearful, preoccupied, and dismissive attachment styles as well as the gender variable as the predictors.Multiple regression analysis results presented in Table 2. First multiple regression analysis showed that secure and fearful attachment styles and gender significantly predict discrimination and exclusion (F(3,357) = 12.673, p<.01, R = .310,R 2 = .096,R 2 adj = .088).This model has a medium effect size and the three variables explained the 9% of total variance.As seen in Table 2, the contribution of secure attachment to the total variance was significant (β = -.14, t = -2.546,p< .05)and it negatively predicted the discrimination and exclusion behavior.Similarly, the contributions of fearful attachment (β = .15,t = 2.712, p< .01)and gender (β = -.24,t = 4.648, p< .01) to the total variance were significant.Both variables positively predicted discrimination and exclusion behaviors.
The second multiple regression analysis revealed that secure and fearful attachment styles and gender significantly predict labeling (F(3,357) = 8.531, p< .01,R = .259,R 2 = .067,R 2 adj = .059).This model has a medium effect and its three variables explained the 7% of total variance.As shown in Table 2, the contribution of secure attachment to the total variance is not significant (β = -.08,t = -1.471,p> .05).Secure attachment did not significantly predict the labeling behavior.However, the contributions of fearful attachment (β = .18,t = 3.245, p< .01)and gender (β = .16,t = 3.146, p< .01) to the total variance was significant.Fearful attachment and gender positively predicted labeling.
The third multiple regression analysis showed that secure and fearful attachment style and gender significantly predicted psychological health (F(3,357) = 6.469, p< .01,R = .227,R 2 = .052,R 2 adj = .044).This model has a medium effect and the three variables explained the 5%oftotal variance.As seen in Table 2, the contribution of secure attachment to the explained variance was significant (β = -.12,t = -2.080,p< .05)and secure attachment negatively predicted psychological health behaviors.But the contributions of fearful attachment (β = .13,t = 2.371, p< .05)and gender (β = .13,t = 2.537, p< .05) to the explained variance are significant.Both variables positively predicted psychological health behaviors.
The final regression analysis showed that secure, preoccupied and dismissive attachment style significantly predicted prejudgment (F(4,356) = 7.711, p< .001,R = .282,R 2 = .080,R 2 adj = .069).Similar to the other three models, this model also had a medium effect size and the three variables contribution to the explained total variance was 8%.As seen in Table 2, the contribution of secure attachment to the total explained variance is significant (β = -.15,t = -2.742,p< .01)and secure attachment negatively predicted prejudgment behaviors.However, the contribution of fearful attachment to the total variance was not significant (β = .02,t = .342,p> .05).These results indicate that fearful attachment does not significantly predict prejudgment behaviors.Contributions of preoccupied attachment (β = EDUPIJ / VOLUME 5 / ISSUE 1 / SPRING / 2016 .18,t = 3.433, p< .01)and dismissive attachment (β = .18, t = 3.193, p< .01)style to the explained total variance were significant.Preoccupied and dismissive attachment styles positively predicted prejudgment.

Conclusion and Discussion
This study examined the relationship between the attachment styles and stigma tendency of adults.Study results indicated that secure attachment style negatively predicted the discrimination and exclusion, prejudgment and psychological health sub-dimensions of stigma.This study results are in line with previous studies, suggesting that individuals with secure attachment styles are more likely to display more positive social relationships (Kobak & Sceery, 1988;Riggs, Jacobovitz, & Hazen, 2002;Zhao et al., 2015).According to Riggs et al. (2002), people with secure attachment style are less neurotic, more extrovert, less anxious and more sincere and have higher coping skills than individuals with insecure attachment styles.Similarly, Zhao et al. (2015) found that secure attached adolescents have lower selfstigmatizing tendencies with regards to receiving psychological help than anxious and dismissive attached adolescents.People with secure attachment style have positive perceptions about themselves and other people.Secure attached individuals also have high self-confidence and self-esteem; find themselves worth loving and tend to perceive other people as acceptable (Bartholomew & Horowitz, 1991).Thus, secure attached individuals are expected to have low stigma tendencies concerning the discrimination, prejudgment and psychological health dimensions.
Study results also showed that people with fearful attachment style are higher for discrimination-exclusion, labeling and psychological health dimensions.This finding supports the adult attachment theory (Bartholomew & Horowitz, 1991).People with fearful attachment style tend to have high anxiety levels, low self-esteem, find other people unreliable and unacceptable (Bartholomew & Horowitz, 1991).Tendency for depression is also high in these individuals (Carnelley, Pietromonaco, & Jaffe, 1994).These studies suggest that fearful attached individuals are willing to enter significant relationships with other people, but avoid this due to the fear of being rejected and the feeling of distrust (Henderson, Bartholomew, & Dutton, 1997).Studies emphasize characteristics of fearful attached individuals such as low self-esteem and difficulty in initiating close relationships (Knox, 1999), distrust in the social environment (Howard & Medway, 2004;Murphy & Bates, 1997), negative stigma tendency towards mentally ill individuals (Zhao et al., 2015) and a criticizing attitudes towards the self (Blatt, 2004).Riggs, Vosvick, and Stallings (2007) underlined that fearful and anxious attached individuals with HIV positive virus have higher stigma levels than people with secure attachment style and that they have a more negative self-image about themselves.The need to hide one's mental illness can be a result of avoiding possible undesired events and the discontent in reminding the individual of the selfprejudgments.The fear of being identified can mostly result with an internal stigma (King et al., 2007).Results of these studies suggest that individuals who have negative opinions about themselves may be carrying and reflecting these opinions due to the environment.
Study results have also shown that preoccupied and dismissive attachment styles were positively correlated with prejudgment behaviors.People with preoccupied attachment tend to have negative opinions about themselves, but positive opinions about others.Dismissive attachment individuals have positive opinions about themselves, but negative opinions EDUPIJ / VOLUME 5 / ISSUE 1 / SPRING / 2016 about others (Bartholomew & Shaver, 1998).Nam and Lee (2015) suggest that there is a high relationship between preoccupied attachment and social stigma, and between fearful attachment and self-stigma, due to receiving psychological support.According to Nam and Lee (2015), preoccupied individuals are afraid of other people stigmatizing and judging them.Thus, people with preoccupied and dismissive attachment style can be expected to display prejudgment behaviors.In preoccupied attachment style, the negative self-perceptions of the individual leads to positive perceptions about others (Henderson et al., 1997).With regards to all these results, for the prejudgment dimension it can be stated that preoccupied and dismissive individuals, whose opinions about themselves and the environment are not based on reality, can be expected to display stigma tendency on individuals who they define as prone to crime and find as different in appearance and lifestyle.
This study also showed that males are more likely to have higher stigma tendencies than females with regards to the discrimination and exclusion, labeling and psychological health dimensions.These findings are in line with previous research findings examining gender differences (Topkaya, 2014;Hosgorur & Apikoglu, 2013).Topkaya (2014) reported that with regards to receiving psychological help, males are more likely to have higher stigma perceptions than females with respect to self-stigma and social stigma.A similar result was also obtained by Topkaya, Sahin, Dogru-Cabuker, and Okten (2015).These researchers found that, with respect to receiving psychological help, self-stigma is higher in males than in females.Males may have higher stigma tendency levels due to gender roles and cultural factors related to this.This study has some limitations.Firstly, this study was conducted on a limited number of adults residing in the central Black Sea region of Turkey.Thus, external validity of this study is low.The experiences of the adults in this region may differ from those of the adults who live in other regions of Turkey.In addition, educational level of the participants are higher than the average educational level throughout Turkey (Sahin, Barut, & Ersanli, 2013;Sahin, Barut, Ersanli, & Kumcagiz, 2014).Thus, the results of this study cannot be generalized to people with a low educational level.It could be beneficial to conduct future studies on adult samples residing in larger and different regions and who have different educational backgrounds.Secondly, cross-sectional research design was used in this study.Although cross-sectional studies provide information about the present condition of the studied samples, a causal relationship cannot be made.Thus, future experimental and longitudinal studies on the same sample will be beneficial.Thirdly, the information was collected using a self-report scale for adults.Self-report scales may lead to a number of common method bias such as social desirability (Podsakoff, MacKenzie, Lee, & Podsakoff, 2003).However, this common method bias was tried to be prevented by giving participants the opportunity to maintain identity confidentiality (Podsakoff et al., 2003).However, it may be beneficial to examine these variables with information gathered from different resources such as close friends and family members in future studies.
Improvement in social relationships results from lower tendencies towards stigma.One way of struggling with stigma based on psychological deficits may be to enter into social relationships with other individuals (Overton & Medina, 2008).Stigma tendency expands in societies where a fearful culture is dominant, rather than in societies where relationships are based on principles (Yaman & Gungor, 2013).This issue not only shapes an individual's behavior, but also undertakes the role of determining the relationships with the EDUPIJ / VOLUME 5 / ISSUE 1 / SPRING / 2016 environment.The marginalization and discrimination an individual displays to a person within the same society and who has a different shape, condition, or attitude, may evolve into a social acceptance tool.In order to struggle with stigma, a structure based on an accurate and sound information and communication network provided by institutions which offer education and health services, along with precautions against possible destructivedevastating effects is required.The effect of individuals' attachment styles on stigma tendencies should be taken into consideration when planning and implementing psychological counseling and guidance services, offered both at schools and social areas.
In conclusion, it was evident in this study that various attachment styles are related to the sub-dimensions of stigma in different ways.Interventions to decrease stigma tendencies of individuals can verge to enhancing the quality of mother-child interactions.

Notes
Corresponding author: NURSEL TOPKAYA , .10 correlation indicates low, .30correlation indicates medium, and .50 and above correlation indicates high effect size.For the R 2 value .02values indicate low, .13values indicate medium, and .26and above values indicate high effect size.Data collection was carried out between March-June, 2014.

Table 1 .
Relationship coefficients between variables, mean and standard deviation values

Table 2 .
Multiple regression analysis results in predicting discrimination and excluding, labeling, psychological health and prejudgment