www.ijcer.net Mediator Role of the Need for Social Approval in the Relationship between Perfectionism and Codependency: A Structural Equation Modeling Study

The purpose of the study is to test the model developed in relation to the mediator role of the need for social approval in the relationship between perfectionism and codependency. The study was conducted on a total of 188 students (144 females and 44 males) attending the Education Faculty of Amasya University. In the study, Spann-Fischer Codependency Scale, Need for Social Approval Scale and Frost Multidimensional Perfectionism Scale were used. The theoretical model proposed in the study was tested with the structural equation model. In the analysis of the collected data, LISREL 8.51 was used. The goodness-of-fit indices calculated for the model were found to be good. The goodness-of-fit indices found for the structural model are as follows: RMSEA=0.07, Standardized RMR=0.08, NNFI=0.90, CFI=0.91 and IFI=0.91. It was found that the structural model constructed to reveal the mediator role of the need for social approval in the relationship between perfectionism and codependency has a good fit and the need for social approval was found to have a partial mediator role in the model.


Introduction
Codependency is a subject which has been investigated only in recent years in Turkey. While it is a subject widely studied in the field of health in recent years (Ançel, 2012;Ançel, Yuva and Öztuna, 2012;Ölçüm and Büyükkayacı Duman, 2017;Öz, 1998;Özdemir and Buzlu, 2019), there is a limited amount of research carried out on it in the field of education (Reisoğlu, Yazıcı and Aydın, 2018;Tanhan and Mukba, 2014). Ançel (2017) stressed that the prevalence of codependency in Turkey is not known as the subject has not been researched widely enough. When the international literature is reviewed, it is seen that a large number of studies have focused on the subject (Hillborg, 1995;Martsolf, Sedlak and Doheny, 2000;Chang, 2010;Reyome and Ward, 2007;Parker, Faulk and Lobello, 2003;Chang, 2018;Bereza and Isaeva, 2018;Bacon et al., 2018;Zielinski et al., 2019). Beattie (2012) defined the characteristics of codependency as follows: Codependent people strive for other people's contentment rather than their own contentment. While they try to give something to other people throughout their lives, they feel sad as they are not given anything. These people who tend to devote themselves to others have feelings of pity and guilt towards people with problems. They also tend to be self-sacrificing.
When we look at the history of the concept of codependency, it is seen the concept came into question after alcohol dependence was seen as a physical illness in the 1930s. Since the 1950s, it has become a notable concept with the recognition of the fact that some other members of the family were also affected from the person addicted to alcohol. After 1980s, studies on the concept started to increase. It has also been shown that there is also a codependency relationship between patients with chronic disease and their caregivers (Ançel, 2017). Therefore it can be said that codependency was initially used to illustrate dysfunctional relationship between the alcoholic person and his/her spouse, but was subsequently used to define dysfunctional relationships (Lindley, Giordano and Hammer, 1999). IJCER (International Journal of Contemporary Educational Research) Ançel (2012) defined codependency as the pathological relationship in which the interdependence between the individual who is dependent on the care of the other and the caregiver is supported, increased and maintained. As a result of this relationship, the individual feels approved and strong. Tanhan and Mukba (2014) describe one person's developing a relationship dependent on another person due to his/her personality traits and behavioural patterns in interpersonal relationships as codependency. Interpersonal attitudes specific to codependency were defined as "diseased dependency" by Karen Horney (Crothers and Warren, 1996). Despite the difficulties in making a precise definition of codependency, codependent people are defined as those who care too much on what is happening around and lose their own feelings and thoughts while trying to control other people's lives (Lindley, Giordano and Hammer, 1999). DesRoches (2000) pointed out that codependency is not only a psychological concept but also a life-threatening, destructive and health-impairing disease. Ançel (2017) stated that having codependent behaviours is not a disease. He stated that it is a new form of a person traditionally considered to be good, self-sacrificing and considerate of others more than himself/herself. Codependency can be explained by schemes of self-sacrifice and emotional deprivation from among the schema domains involved in the schema therapy. Young, Klosko and Weishaar (2009) stated that one's belief in the self-sacrifice scheme is as follows: The individual believes that if he / she meets all the needs of other people and does not care about his / her needs, others will accept him / her and will not feel loveless. The emotional deprivation scheme is also seen in people who have the self-sacrifice scheme. These people focus more on meeting others' needs rather than their own. Hollabaugh (1995) stated that codependency can be explained through concepts from the social psychology literature. The concepts of self-presentation and impression management explain how people try to influence others' perceptions (Goffman, 1959;Vohs, Baumeister and Ciarocco, 2005;Murphy, 2007;Schlenker, 2012). Vohs, Baumeister, and Ciarocco (2005) stated that one of the most basic skills in social life is that the individual presents himself/herself pozitively to other people. On the basis of these explanations, it can be said that codependency seems to be related to the social approval. The need for social approval is defined as a concept in which the opinions of others are considered important. People having a strong need for social approval are those who try to make a positive impression on others (Karaşar, 2014). Horney (2019) defined the need for neurotic love and approval as the need to keep others pleasant and to be loved and approved by them. This need includes acting in accordance with the expectations of others, and seeing others' wishes and thoughts as superior to their own. It also includes fear of self-disclosure and fear of hostility of others. These explanations regarding the need for social approval are similar to the characteristics of codependency. Codependent people feel the need for approval (Yates and McDaniel, 1994;Beattie, 2012;Tanhan and Mukba, 2014;Friel and Friel, 2010).
Particularly in eastern societies, making sacrifices, altruism and devotion are characteristics strongly approved. Such characteristics are reinforced by the society (Ançel, 2017). In Turkey, characteristics such as making sacrifices and codependency are culturally promoted from early ages (Ançel, 2012). In addition to these characteristics, a great importance is attached to what other people think about what we do and say in our culture (Karaşar, 2014). The concern felt about what others think about what we do and say is generally accompanied by codependency in our culture. Tanhan and Mukba (2014) pointed out that characteristics such as suppressing their feelings, need for approval and lack of self-confidence are commonly seen in Turkey. Yates and McDaniel (1994) stated that symptoms such as suppressing their problems and feelings, perfectionism, low self-esteem, fear, anxiety and depression are seen in these individuals. At the same time, self-neglect of these individuals may result in various health problems. Ançel (2017) stated that the need for empowerment in codependent people is met with the approval and appreciation of the other person. With the approval the individual receives, he/she increases his/her self-esteem. Codependent people value others' thoughts more than their own. They have behaviors to please others. They have expectations to be approved by people (DesRoches, 2000).
Codependency also seems to be related to perfectionism. Yates and McDaniel (1994) and Ançel (2017) argue that one of the characteristics of codependent people is perfectionism. Codependent people hope to do everything perfectly. They are trying to show that they are good enough for others. The codependent individual in the role of the saviour shows obsessive thoughts and behaviours. Together with perfectionism, being obsessive ensures the continuity of controlling behaviours. Codependents striving to control events avoid making mistakes (Ançel, 2017). Yates and McDaniel (1994) stated that one of the main characteristics of codependency on which experts agree is perfectionism and the need to control other people.
Perfectionism also appears to be related to the need for social approval. According to Antony and Swinson (2009), perfectionist individuals are those who experience extreme anxiety about other people's thoughts, and these people have a high need for approval. There is a limited amount of research on this subject in the literature (Hewitt and Flett, 1991;Sherry, 2002;Karaşar and Öğülmüş, 2016a). In a study by Hewitt and Flett (1991), it was found that socially prescribed perfectionism was related to the social approval seeking sub-dimension of the Irrational Beliefs Test. In their research on young adults, Karaşar and Öğülmüş (2016a) found that self-directed perfectionism predicted the need for social approval. The findings of these studies show that the need for social approval and perfectionism are related.
Given the delineations above, it seems to be important to explain the relationships between codependency, perfectionism and the need for social approval. Although there is a great deal of research on the subject in the field of health, the number of such studies in the field of education is highly limited (Tanhan and Mukba, 2014;Reisoğlu et al, 2018). Especially the fact that the teaching profession is based on self-sacrifice and devotion can cause tendency towards codependency. Teachers can develop codependency by putting the needs of their students and parents ahead of their own needs. In this connection, Koehler (2010) stated that teachers' overload and fatigue are the greatest evidence for codependency. He also stated that teachers tend to overwork for the satisfaction of other people. It is also emphasized that the existence of external reward systems to evaluate teachers may also affect this. Considering the negative consequences of codependency, it can be said that the studies on pre-service teachers may be important in terms of preventing future problems in their professional lives. For this reason, investigation of the subject in relation to pre-service teachers will help fill an important void in the literature. In this regard, the aim of the this study is to investigate the mediator role of the need for social approval in the relationship between codependency and perfectionism.

Research Model
In the current study, the quantitative and relational model was used to investigate the relationships between variables. In the testing of the model developed in the study, the structural equation model was employed.

Study Group
The study group of this research is comprised of 188 students (144 females and 44 females) attending the Education Faculty in Amasya University. Random sampling method was used in the research. In the research, pre-service teachers were informed about the study. After giving information, scales were applied in classrooms.

Data Collection Tools
In the study, the Spann-Fischer Codependency Scale, the Need for Social Approval Scale and the Frost Multidimensional Perfectionism Scale were used to collect data.

Spann-Fischer Codependency Scale:
The Turkish adaptation studies of the scale developed by Fischer, Spann and Crawford (1991) were carried out by Tanhan and Mukba (2014). It is a six-point Likert type scale. High scores taken from the scale show that codependency is high. The internal consistency coefficient of the scale measuring the characteristics of codependency such as experiencing difficulty in expressing feelings and overfocusing on external world was found to be α=0.65. The common variance of the scale was found to be 48.69%. In this study, cronbach alpha coefficient was found .73 for codependency. In this research, the goodness-of-fit indices found for the codependency are as follows: x²/sd=232.06/103=2.23, RMSEA=0.08, Standardized RMR=0.08.
The Need for Social Approval Scale: It was developed by Karaşar and Öğülmüş (2016b). This scale consisted of 25 items. High scores taken from the scale show that the need for social approval is high. Sub-dimensions of the need for social approval are (i) sensitivity to others' judgements, (ii) leaving a positive impression and (iii) social withdrawal. The scale items are responded on a five-point Likert scale ranging from "Strongly Agree" to "Strongly Disagree". The internal consistency coefficients calculated for the sub-dimensions of the scale were found to be varying between .80 and .83. The internal consistency coefficient calculated for the whole scale is .90. In order to test the convergent validity of the Need for Social Approval Scale, its correlation with the "The Fear of Being Negatively Evaluated Scale Short Form" was checked and the correlation was found.79. The Need for Social Approval Scale's pretest-posttest reliability was found to be .90. In this study, cronbach alpha coefficient was found .92 for social approval. In this research, the goodness-of-fit indices found for the need for social approval are as follows: x²/sd= 523.96/272= 1.92, RMSEA=0.07, Standardized RMR=0.06, NNFI=0.96, NFI=0.93, CFI=0.96, IFI=0.96.

Frost Multidimensional Perfectionism Scale:
Turkish adaptation studies of the scale developed by Frost, Marten, Lahart and Rosenblate, (1991) into Turkish were conducted by Özbay and Mısırlı Taşdemir (2003). The scale consists of 35 items. The response options to the scale items range from "Strongly Disagree" to "Strongly Agree". The scale was found to be consisted of 6 sub-dimensions explaining 47.8% of the variance. The reliability and validity studies of the scale were conducted on university students by Kağan (2011). As a result of the confirmatory factor analysis, it was concluded that the original six-factor structure of the scale is also valid for the Turkish sampling. Internal consistency coefficient was found to be .91. The internal consistency coefficients of the sub-scales were found to be ranging from .64 to .94. The fifteen-day classroom total correlation coefficient was found to be .82 for total scores. Kağan (2011) stated that order sub-dimension in the Turkish sampling should be evaluated independently and that the score taken from this sub-dimension should be evaluated separately from the general scores of the scale. Therefore, the order sub-dimension was discarded from the current study and the other five sub-dimensions were included in the analyses. In this study, cronbach alpha coefficient was found .89 for perfectionism.Cronbach alpha coefficient was found .70 for order subdimension. In this research, the goodness-of-fit indices found for the perfectionism are as follows: x²/sd=1175.43/545= 2.15, RMSEA=0.07, NNFI= 0.92, Standardized RMR=0.08, CFI=0.93, IFI=0.93.

Data Analysis
In the study, the theoretical model developed to explain the mediator role of the need for social approval in the relationship between pre-service teachers perfectionism and codependency was tested with the structural equation modelling. In the analysis of the data, SPSS 20 and LİSREL 8.51 program packages were used.

Findings
In this section, firstly the correlations between the variables are presented. Then the structural equation modelling and analysis results related to the mediator role of the need for social approval in the relationship between perfectionism and codependency are presented. In Table 1, the correlation between variables and descriptive statistics are presented. As can be seen in Table 1, there is a medium and positive correlation between the need for social approval and codependency. A positive and medium correlation was found between perfectionism and codependency. Moreover, there is a positive and medium correlation between perfectionism and the need for social approval. When the model presented in Figure 1 and the mediator role of the need for social approval in the relationship between perfectionism and codependency were tested, the relationships established between perfectionism and the need for social approval (t= 6.63), between the need for social approval and codependency (t= 2.95) and between perfectionism and codependency (t= 2.72) were found to be statistically significant. Chi-square value for the scale was found to be significant (χ 2 =532.77, sd= 249, p = .00). When the ratio of the Chi-square value to degree of freedom (χ 2 /sd = 2.13) is below 3, it shows a perfect fit. When the goodness-of-fit indices of the model were calculated, they were found to be as follows: RMSEA=.07, Standardized RMR = 0.08, NNFI= .90, CFI=.91 and IFI=.91. When the modification suggestions for the model are examined, it can be said that there is no modification that can make a significant contribution to the χ 2 value. These values obtained as a result of the analysis show that the model has a good fit. When the RMSEA value is ≤.06, it indicates a good fit (Hu and Bentler, 1999) and when it is ≤.08, it indicates an acceptable fit (Sümer, 2000). Thus, the RMSEA value found in the current study can be argued to be an acceptable value. According to Karaca (2013), when RMR value is <.05, it indicates a perfect fit, when it is <.08, it indicates a good fit, when it is <.10, it indicates an acceptable fit. Thus, the RMR value found in the current study can be said to be an acceptable value. Sümer (2000) stated that when NNFI value is ≥.95, it indicates a perfect fit, when it is .90-.94, it indicates an acceptable fit. Thus, the NNFI value found in the current study can be said to be an acceptable fit. When IFI value is ≥.90, it indicates an adequate fit (Bryne,1998;Hoyle and Panter, 1995;Ullman, 2001;cited in Şimşek, 2007). Thus, the IFI value found in the current study can be said to be an acceptable value.  Table 2, it is seen that the need for social approval and perfectionism variables positively affect codependency. The need for social approval is seen to positively affect perfectionism.

Discussion and Suggestions
In the current study, the need for social approval was taken as the mediator variable and its relationship with the pre-service teachers' perfectionism and codependency was investigated. When the model and the mediator role of the need for social approval in the relationship between perfectionism and codependency were tested, it was found that the correlations between perfectionism and the need for social approval, between the need for social approval and codependency and between perfectionism and codependency were found to be statistically significant. The goodness-of-fit indices found in the current study show that the structural model established has a good fit. The results of the current study show that with increasing perfectionism, the need for social approval also increases, and with increasing need for social approval, codependency is strengthened. At the same time, these results show that there are direct and indirect interactions between perfectionism and codependency.
The higher the level of individuals' perfectionism is, the greater their need for social approval is. As a result, they may exhibit codependency by ignoring their needs while trying to please other people. There are various studies focusing on the relationship between perfectionism and the need for social approval (Hewitt and Flett, 1991;Sherry, 2002, Karaşar andÖğülmüş, 2016a). Antony and Swinson (2009) stated that perfectionist individuals are individuals who have excessive anxiety about the thoughts of other people and that these people may need high levels of approval. In the current study, perfectionism was found to be directly and indirectly correlated with codependency. Yates and McDaniel (1994) and Ançel (2017) stated that one of the characteristics of codependent people is perfectionism. People with a high tendency towards perfectionism may strive to make a positive impression in the eyes of other people. This may cause them to make an excessive effort to become a better person. At the same time, they are more likely to exhibit obsessive behaviours as they are more interested in others' becoming perfect and focused on other people's shortcomings.
A direct correlation was found between the need for social approval and codependency. In the literature, it has been stated that codependent people have a higher need for social approval (Beattie, 2012;Tanhan and Mukba, 2014;Yates and McDaniel, 1994). The need for social approval and codependency are similar concepts. Through the influence of culture, the need for social approval and codependency may be increased. People with a high need for social approval and who want to be perceived as a good person in the eyes of others may be directed to the exhibition of codependent behaviours in order to create a positive impression and avoid negative evaluations. Codependency involves the role of the saviour who neglects himself/herself to help other people. The reasons behind this behaviour may be being remembered as a good person, pleasing others, and creating a positive impression. These factors related to the need for social approval indicate a strong need for social approval in the emergence of codependency. Culturally, the valuing others' thoughts, efforts to avoid criticism and efforts to make a good impression are common in Turkish society. Individuals who need a high level of social approval to be seen as a good person may exhibit codependency behaviours in order to be more accepted by the society without being aware of this. Therefore, they can forget their own desires and needs while living for others. It can be said that codependency can be seen at higher levels especially in those who grow up in a cultural structure where one's caring about oneself more than others is seen to be negative.
The findings of the current study should be evaluated on the basis of various limitations. The first limitation of the study is that the study was conducted on the students attending the Education Faculty of Amasya University. The model constructed in the current study can be tested with broader samplings from different universities. In light of the findings of the current study, psycho-education programs can be developed for pre-service teachers to better understand codependency. At the same time, future research can explore codependency in relation to some other concepts such as self-esteem, depression, anxiety and health problems among university students. Besides pre-service teachers, in-service teachers should also be included in such studies. The number of female students in the research is more than the number of male students. This is a limitation of the research. In future studies, this subject can be studied in larger sample groups.Finally, more comprehensive studies can be conducted to explain the reasons for codependency considering various variables related to culture and family.