How does social media addiction affect decisional procrastination? Mediation role of work-family life balance

How does media addiction In this study, we examined the mediating role of work-family life conflict in the effect of employees ’ social media addiction on decisional procrastination. Gender, age, marital status, education, and the sector employed were used as control variables. The data were obtained using the social media addiction scale-adult form, work-family life conflict scale, decisional procrastination scale, and the personal information form in the online medium. We analyzed the responses of the 400 participants in the SPSS program and applied descriptive statistics, correlation, regression, mediation, and moderation analyses. As a result of the research, it was found that the increase in social media addiction has a positive effect on work-to-family conflict (WtoF) and family-to-work conflict (FtoW), and decisional procrastination. We concluded that WtoF and FtoW mediate the impact of social media addiction on decisional procrastination. Also, those living with families have a moderation role in the impact of work-family conflict on decisional procrastination.


INTRODUCTION
With the development of information technologies, social media has become a platform used by almost all society segments. Its influence is seen in the social, cultural, and economic fields. Social media is used for many purposes related to our daily life, such as communication, access to information, and entertainment. It is seen that both internet and social media usage is increasing rapidly in the world. The number of people using the internet worldwide reached 4.54 billion, with an increase of 7% (298 million new users) compared to January 2019. There were 3.80 billion social media users worldwide in January 2020, with an increase of more than 9% (321 million new users) from January 2019. According to data of Turkey's population in January 2020 (TUIK, 2020), 64% (54 million) seem to use social media. This figure has been realized with an increase of 4.2% (2.2 million) after April 2019 (Kemp, 2020).
The increasing use of social media has caused social media addiction and has affected our daily lives. As with any addiction, social media addiction has the effect of disrupting our personal life and the balance of family and work life. The deteriorating work and family life balance negatively affect people's psychology and causes decisional procrastination to increase. In this research, we investigated the mediating role of work-to-family conflict (WtoF) and family-to-work conflict (FtoW) in the influence of social media addiction on decisional procrastination. We discovered that the rise in addiction to social media positively impacts WtoF, FtoW, and decisional procrastination. We found that WtoF and FtoW play a mediating role in social media addiction's impact on decisional procrastination. Besides, gender and living with family (LwF) variables were tested as moderation variables in the effect of social media addiction, WtoF, and FtoW on decisional procrastination, were found significant.

Social Media and Addiction
As a dimension of information technologies that transform societies globally, social media is used for many different purposes and needs, such as communicating in daily life, creating agendas and public opinion, obtaining information, socializing, and entertainment (Dolgun, 2016). Today, information technologies used by individuals to create the identities they want for themselves and reach the identities they think are the most suitable for them have created a surreal world with social media environments (Ugur and Bilici, 1998: 494). As a mass communication tool, social media has become a part of communities' daily lives beyond just individual use.
It has been the determinant of both social and political agendas. Social media is a commonly used term for online tools and websites that create mutual interaction by allowing users to share information, thoughts, interests, and knowledge (Sayimer, 2008). Social media is an application area that enables sharing of information, different thoughts and experiences, and rapidly embeds the Internet into our lives (Solmaz et al., 2013). The most important difference between social media from traditional one is that anyone can create, comment, and contribute to social media content (Scott, 2010).
The number of studies investigating social media addiction and its effects on various cases, excluding internet addiction, is increasing (Aktan, 2018;Gur et al., 2018;Kirik et al., 2015;Unlu, 2018). Individuals' use of social media in a way that disrupts their daily work will harm individuals physically, cognitively, and psychologically. This condition, which causes psychopathological symptoms in individuals, is defined as social media addiction (Fox and Moreland, 2015;Tutgun-Unal, 2015;Unlu, 2018). Studies show that social media addiction has many negative effects on children, young and advanced adults, and the elderly (Balci and Baloglu, 2018;Safak and Kahraman, 2019;Soner and Yilmaz, 2018).

Work-Family Life Balance
Work-family life balance is individuals' ability to manage multiple life roles in their lives in harmony (Izki, 2019). In another definition, work-family life balance increases life satisfaction by distributing limited resources such as time, energy, and opportunity that individuals have among all living spaces in a balanced way (Temel, 2005). Work-family life balance occurs when employees do not restrict their work and family life in one of the other living areas (Turgut, 2011). Work-family life balance is interpreted differently by employees and employers. Employees consider it a priority to obtain sufficient time and income to fulfill family responsibilities and meet their families' care and needs. On the other hand, employers aim to create a corporate culture that will enable employees to focus on their jobs (Parlak Kul, 2016). Studies on work-family life balance issues, negatively affecting both family, work, and personal lives of individuals if not realized, have started to take place in the literature.

Decisional Procrastination
Decision-making refers to actions taken as a result of thinking, discussion, and calculation in individuals' actions (Onaran, 1971). In the literature there are different definitions for decisional procrastination. Ellis and Knaus (1977) divided procrastination into five types: general procrastination, compulsive procrastination, decision-making procrastination, neurotic procrastination, and academic procrastination. Generally, decisional procrastination is defined as a behavioral tendency or a personality trait to delay in making decisions or doing a job (Milgram et al., 1998). In this context, decisional procrastination is defined as the inability to make appropriate decisions even in insignificant situations regarding various experiences (Effert and Ferrari, 1989). In the literature, procrastination has been examined in three dimensions: cognitive, emotional, and behavioral (Eksi et al., 2019). According to Sarioglu (2011), cognitive procrastination is the incompatibility between individuals' goals and their effort in achieving these goals. The emotional dimension is that individuals feel emotions such as distress, anxiety, and shame when they do not do what they are supposed to do (Binder, 2000;Uzun Ozer et al., 2012). The behavioral dimension is defined as postponing the work that needs to be done without any reason, holding the tasks until the last moment, and putting forward other activities that are more enjoyable for the person (Aydogan, 2008;Ozdemir, 2012).
Studies in the literature have shown that the relationship between social media addiction and procrastination behavior primarily focuses on academic and general procrastination tendency (Eksi et al., 2019;Gur et al., 2018;Tekin, 2019;Yakut and Kuru, 2020). There is no study in the national literature on the relationship between social media addiction and the tendency to decisional procrastination. However, in the studies of Lavoie and Pychyl (2001) and Susilawati (2019), it was found that there is a positive correlation between these two factors. Resources are also very limited in terms of social media and work-family life balance. In a study conducted in South Korea in 2019, the positive relationships between social media use and WtoF and FtoW were found (Choi et al., 2019). Voydanoff (2005) reveals the negative impact of social media addiction on work-life balance. Although there are no studies focusing directly on the relationship between work-family balance and decisional procrastination in the reviewed literature, supportive relationships were found in some studies. In this sense, it has been understood that children's decision-making behaviors who are raised in more democratic and healthy families are timely and consistent in both their youth and adulthood (Tatlilioglu, 2014;Ummet et al., 2016). Therefore, in our study, we examined the relationships among working individuals' social media addiction level, work-family life balance, and decisional procrastination. It is expected that the study will fill an important gap in the literature.

RESEARCH MODEL AND HYPOTHESIS
The conceptual model of the study is as shown in Figure 1. The graph in Figure 1 shows the relationship between social media addiction, decisional procrastination, and work-family life balance. Accordingly, it is assumed that work-family conflict and family-work conflict have mediation effects in the effect of social media addiction on decisional procrastination. Also, it is thought that those living with their families have a moderation effect on the impact of social media addiction and work-family conflict on decisional procrastination.

Study Design, Participants, and Procedure
It is aimed to understand independent variation variables among people in Turkey at a given time, we planned to conduct a cross-sectional study. We used the snowball method to reach the participants in the study. Snowball sampling method is defined as explaining different phenomena by reaching out from person to person (Aziz, 2011;Baltaci, 2018;Ogel, 1999). We used the convenience sampling process to evaluate the sample and used the survey as a means of data collection. Survey studies are cross-sectional studies that include quantitative data collection using questionnaires on the tendencies, attitudes, or opinions of the sample representing the universe (Creswell and Poth, 2016;Kocabas, 2016). Quantitative and causal designs were used in this research. Variables were measured at a certain point in time using crosssectional data. This study concept is suited for our objective because we did not seek to generalize variables, but rather to analyze the pattern of cause and effect and to see their extent.
All participants of the study were in Istanbul. Participants were reached online using the google form. The research started on 10 February 2020 and was completed on 15 March 2020. We stated in the explanation section that the research was conducted only with the employees and detailed information about the aims of the study to the participants. After the consent of the participants was obtained, they were asked to answer the questions. We ensured that the participants were able to complete the research questions within the time they wanted after they started answering the research questions. Data were ensured to be confidential and anonymous. The research was performed according to the Helsinki declaration criteria.

Data Analysis
After downloading the data from the online medium, they were first transferred to MS Excel and then to IBM SPSS 25 for editing (George and Mallery, 2010). By carrying out descriptive analysis, we evaluated demographic variables with means and standard deviations. Factor analysis was done to ensure the validity of the construct. To test the hypotheses, the first is the correlation between dependent and independent variables, and the regression analysis for the effect was performed. PROCESS-Macro (Hayes and Rockwood, 2020) and simple slope tests for two-way interactions (Dawson, 2014) were conducted for mediation and moderation analysis. At α> 90 %, statistical importance was set.

Measures
In the study, we first used a personal information form to determine participants' socio-demographic characteristics. Subsequently, the social media addiction scale adult form, work-family life balance scale, and decisional procrastination scale were used. This data collection form, which includes a personal information form and scales, was created over Google forms, and was implemented online.

Personal information form
A personal information form was created that provides access to information such as age, gender, marital status, and educational status of the individuals participating in the study, as well as with whom they live, the business sectors they work, the purpose of using social media and the most frequently used social media platform.

Social media addiction scale adult form
Social media addiction scale was used to measure the social media addiction levels of the participants. The scale, developed by Sahin and Yagci (2017), is a five-point Likerttype scale consisting of 20 items in total. The participants rated each of the five-point Likert-style items between 1 "not suitable for me" and 5 "very suitable for me". 5 th and 11 th items are scored in reverse. The highest score can be obtained from the scale is 100, and the higher the score, the more the individual perceives himself as a social media addict. Cronbach's alpha value was found as .879 in the reliability study. Based on these data, it is possible to say that the scale is valid and reliable.

Work-family life balance scale
In the study, we used the work-family life balance scale to measure working individuals' work-family life balance (Apaydin, 2011). The 5-point Likert-type rating scale was used in the measurement tool. Ratings are classified as (5) strongly agree, (4) strongly agree, (3) slightly agree, (2) slightly agree, (1) totally disagree. As a result of the factor analysis performed during the scale development, items 1, 7, 10, 12, 13, and 14 were removed to simplify and have a better structure. As a result, a scale consisting of 11 items and 3 sub-dimensions as "work-family conflict", "family-work conflict," and "familywork balance" was obtained. In the study, only "work-family conflict" and "family-work conflict" sub-factors were applied. Tabachnick and Fidell (2013) state that the Kurtosis values between ±1.5 in their study show a normal distribution of the data. Based on this, the value of Kurtosis in the work-family conflict sub-factor is -1.098; the Kurtosis value of the family business conflict sub-factor was found to be -.991. According to these data, it is seen that the sub-factors of the work-family life balance scale show a normal distribution.

Decisional procrastination scale
The decisional procrastination scale, which is a sub-scale of the conflict coping behaviors scale developed by Mann (1982), was adapted into a single form with the studies of Balkis (2015). The decisional procrastination scale is in a 5point Likert style and consists of five items. Individuals themselves; 1 evaluates as "completely wrong", 5 as "completely true". High scores from the scale indicate the high decisional procrastination tendency of the participants (Eksi and Dilmac, 2010). It was determined that Kurtosis value of this scale was -.672, and data showed normal distribution. In reliability analysis, Cronbach's alpha value was found as .844.

FINDINGS
In Table 1, the answers of 400 participants who answered the questions thoroughly were analyzed. Participants: 51.2% of them were women, and 48.8% were men; the average age is 31; 48.3% were single, 51.7% were married; it was determined that 2.3% were primary school graduates, 2% were secondary school graduates, 10% were high school graduates, 11% were college graduates, and 74.8% were university graduates. Also, 73.3% of them live with their family; it has been determined that 37.8% of them work in the public sector and 62.3% in the private sector. It was found that 16.5% used social media to spend their free time, 24% to have fun, 51.5% to access information, 4.3% to share their personal presentations, and 3.8% to meet new people.
In the factor analysis of the scale's questions (Table 2), the KMO value was 0.845, and the cumulative variance was 74.788%. This ratio is sufficient to use scale questions. In the reliability study, Cronbach's alpha value was found to be 882. Among the sub-factors, WtoF Cronbach's alpha value is 913; FtoW Cronbach's alpha value was 832.
In Table 3, significant relationships between variables and factors were examined. A low-level negative correlation between decisional procrastination and age variable (r=-.11, p<.05), a moderate positive correlation with social media addiction (r=.51, p<.01), a moderate positive relationship (r=.45, p<.01) with work-to-family, and a moderate positive relationship (r=.43, p<.01) with family-to-work were found. We discovered a low-level negative relationship between familyto-work and education level (r=-.24, p<.01), a moderate positive relationship (r=39, p<.01) with social media addiction, and a moderately positive (r=.42, p<.01) significant relationship between work-to-family. A low-level positive relationship between work-to-family and gender (r=.13, p<.05), a low-level positive relation (r=.15, p<.01) with the sector of work, and a moderate positive relation (r=.39, p<.01) with social media addiction were found. A low-level positive correlation between LwF and age variable (r = 11, p<.05), a lowlevel positive correlation with marital status (r=12, p<.05) and a low-level negative correlation (r=-.15, p<.01) with education level were discovered. There was a low-level negative correlation between the sector employed and the education variable (r=-.23, p<.01). There was a low-level positive correlation between education and gender variables (r=.11,   p>.05), and a moderate negative relationship between education and age variables (r=-.34, p<.01). A low-level negative correlation was found between marital status and gender variables (r=-.11, p<.05).
The effects of independent variables on the decisional procrastination dependent variable were shown in Table 5.
We showed the impacts of independent variables of gender, age, marital status, education, sector, living with a family, WtoF, and FtoW on decisional procrastination in model 4. Accordingly, while the impact of age was negative and significant (B=-0.016, p<.01), WtoF was positive and significant (B=0.298, p<.001) and FtoW was positive and significant (B=0.270, p<.001) on the decisional procrastination. In model 5, the social media addiction variable was also included in the analysis in addition to the variables in model 4. In model 5, regarding decisional procrastination, age related negatively (B=0.012, p<.05), whereas social media addiction (B=0.495, p<.001), WtoF (B=0.214, p<.001) and FtoW (B=0.189, p<.001) related positively. In addition to previous model, three interaction variables were included in model interaction. We tested the effects of interaction variables on decisional procrastination together with other variables. We found that social media addiction and age (B=-0.225, p<.01), WtoF and LwF (B=0.168, p<.05) and FtoW and gender (B=0.171, p<.05) interaction variables to have a statistically significant effect on the decisional procrastination.
It has been observed that the social media addiction variable, which is the independent variable in our model, has a significant effect on the dependent variable WtoF, and FtoW in Table 4 and model 1-2. The social media addiction variable was also found to have a significant impact on decisional   Table 5 and model 4. We found that WtoF and FtoW variables' impacts were statistically significant on decisional procrastination. Ultimately, social media addiction was added to the previous variables and analyzed as predictors of decisional procrastination in model 5. As seen in model 5, social media addiction, WtoF, and FtoW variables were statistically meaningful (p<.001).
The social media addiction variable continued its effect on the decisional procrastination dependent variable through the mediator variables WtoF and FtoW, as seen in Table 6. WtoF had a positive and significant effect as a mediator in the effect of social media addiction on decisional procrastination (=.0992, SE=.0241, 95% CI [.0553, .1498]). FtoW had a positive and significant effect as a mediator in the effect of social media addiction on decisional procrastination (=.0711, SE=.0219, 95% CI [.0298, .1163]).

H1:
WtoF has a mediation effect in the impact of social media addiction on decisional procrastination. Accepted.
H2: FtoW has a mediation effect in the impact of social media addiction on decisional procrastination. Accepted.
The SMAXGender interaction variable's effect on the decisional procrastination (DP) dependent variable was significant (B=-0.225, p<.01), as shown in Table 5 model interaction. As a result, there was a statistically significant difference between males and females in the impact of social media addiction (SMA) on decisional procrastination. As shown in Figure 2, females' decisional procrastination increases at a slower rate than males' as SMA rises. In other words, while males' SMA rises, it becomes more resistant than females to decrease decisional procrastination.

H3:
Gender has a moderation effect in the impact of social media addiction on decisional procrastination. Accepted. Table 5 model interaction, we found the impact of the WtoFXLwF interaction variable on the decisional procrastination dependent variable to be significant (B=0.168, p<.05). Accordingly, there was a statistically significant difference in the impact of WtoF on decisional procrastination between those who were LwF and living without family. As seen in Figure 3, we understood that as the WtoF of those who were LwF increases, decisional procrastination increases more than those who were living without family. In other words, while the WtoF of those who were LwF increase, their decrease in decisional procrastination becomes more resistant than those who were living without family.

As seen in
H4: LwF has a moderation effect in the impact of WtoF on decisional procrastination. Accepted. Table 5 model interaction, the FtoWXGender interaction variable significantly impacted the decisional procrastination dependent variable (B=0.171, p<.05). As a result, there was a statistically significant difference between males and females in the effect of FtoW (FtoW) on decisional procrastination. As seen in Figure 4, females' decisional procrastination increases faster than males' as their FtoW increases. In other words, as the conflict with family-to-work rises, females' decisional procrastination increases, making them more prone to it than males.

As shown in
H5: Gender has a moderation effect in the impact of FtoW on decisional procrastination. Accepted.

DISCUSSION
In this study, we examined the mediating role of workfamily life conflict in the effect of employees' social media addiction on decisional procrastination. The research was conducted with 400 participants working in Istanbul. According to the results of the research, it was determined that five hypotheses were supported. In the study, we found that social media addiction affects WtoF and FtoW. Zivnuska et al. (2019) revealed a similar result in their study and stated that social media addiction negatively affected the work-family life balance. A study conducted in South Korea found that social media addiction was both affected by work-family and familywork conflict and that it affected them (Choi et al., 2019). In another study conducted with 230 women in Jakarta, it was concluded that the use of social media affects the work-life balance, and the relationship between the two affects women's subjective well-being (Dwianti, 2020). In the study, no statistically significant difference was found in the social tolerance and social communication subscales of the social media addiction scale according to the gender variable. In other studies, it has been revealed that gender is not effective on social media addiction (Aktan, 2018;Balci et al., 2020).
Our study found a positive relationship between social media addiction and decisional procrastination, and we determined that as social media addiction increased, decisional procrastination increased. In the literature, it has been observed that social media addiction and academic procrastination, and general procrastination tendency are the subjects, but there is no adequate study focusing on its relationship with decisional procrastination (Eksi et al., 2019;Gurultu, 2016;Tekin, 2019;Yurdakos and Bicer, 2019). Some studies revealed that social media addiction positively affected decisional procrastination, and there was a positive correlation between social media addiction and decisional procrastination (Aneta et al., 2016;Susilawati, 2019). In the study conducted by Lavoie and Pychyl (2001: 439), a positive relationship was found between internet use, which is common than social media addiction, and decisional procrastination behavior. We found a negative relationship between age and decisional procrastination; as individuals get older, their decisional procrastination decreases. In their study, Eksi and Dilmac (2010) found a significant relationship between students' decisional procrastination and their ages. Contrary to the study of Eksi and Dilmac (2010), Balkis (2006) found that the ages of prospective teachers were not statistically significant on decisional procrastination. It is thought that such differences arise between the results due to the different age ranges of the participants selected for the studies.
We observed that the FtoW and WtoF positively mediate the effect of social media addiction on decisional procrastination. Based on this, we understood that social media's use positively impacted FtoW and WtoF, and FtoW and WtoF had a positive effect on decisional procrastination. It was found that women experience WtoF more than men. In the literature, some studies found that the effect of work-life on family life does not differ according to the gender of the individuals, WtoF does not differ between women and men, and that there is no negative effect of work on the family (Carikci and Celikkol, 2009;Ozmete and Eker, 2012). Contrary to this, as a result of the studies conducted by Topgul (2016) and Tremblay (2012), it was revealed that women experience WtoF more than men. In our study, we found a statistically significant difference in the WtoF subscale in favor of female participants.
Our study determined that employees working in the private sector experience more WtoF than those working in the public sector. In a study, it was found that the sector worked in did not have a direct effect on the work-family life balance. However, it has been stated that the intention of leaving the job of the private sector employees is higher than the one working in the public sector (Gercek et al., 2015). In the private sector, the intense workload and insufficient job security compared to the public sector can increase the WtoF and the willingness to leave the job. In a study by Dogan et al. (2017), it was stated that there was no difference between public and private sector employees regarding the WtoF.
In the study, we found that WtoF affects decisional procrastination positively. Likewise, we observed that as individuals' FtoW increases, decisional procrastination also rises. In the literature, it was found that there were supportive relationships and that the decisional procrastination behaviors of children raised in democratic and healthy families were exhibited at the desired level and form (Tatlilioglu, 2014;Ummet et al., 2016). However, WtoF affects decisional procrastination more than FtoW, according to our study. Based on this, we understood that decisional procrastination increases negativity in working individuals' work-family life balance. Probably, when the work-family life balance is disturbed, individuals' psychology and healthy thinking abilities can be impaired. According to the education variable, when we examined the FtoW, a significant negative relationship we found between them. In other words, as the level of education increases, FtoW decreases. In this aspect, the study conducted differs from some study findings in the literature. Ozafsarlioglu and Kilic (2013) stated in their research that the education factor does not affect WtoF, while Izki (2019) found that as the level of education increases, the WtoF increases.

CONCLUSIONS AND SOME IMPLICATIONS
In the study, we examined the effect of social media addiction levels of working individuals on work-family life balance and decisional procrastination. We found that the WtoF, FtoW, and decisional procrastination of the participants increased with the increase of social media addiction. Additionally, WtoF and FtoW were found to have a partial mediation role in the effect of social media addiction on decisional procrastination. We understood that gender plays a moderating role in the impact of social media addiction and FtoW on decisional procrastination. Also, we observed the moderation effect of living with the family in the impact of the WtoF on decisional procrastination.
We determined that social media addiction had a negative effect on both work and family life and their balance. Another finding is that the increase in social media addiction disrupted the work-family life balance, and this deterioration further increased decisional procrastination. As a result, unconscious and uncontrolled use of social media adversely affect daily lives, disrupting work-family life balance and increasing decisional procrastination by damaging individual psychology and healthy thinking skills.
The uncontrolled and unconscious use of social media, which is one of the new communication channels, has led to the spread of social media addiction in the daily lives of individuals. This addiction mainly leads to the deterioration of work and family life balance and physical and psychological health problems. Therefore, policies that will prevent excessive use of technology and social media should be supported. However, there is a need for positive discrimination policies for disadvantaged women in work and family life balance. Supporting working women in maintaining the balance between work and family life will contribute to the development of family welfare and social structure. Ultimately, planning and implementing awareness-raising activities about the risks of social media addiction, developing protective and preventive studies on the use of applications where social media addiction is seen higher, and prioritizing the work-family life balance by social policymakers and implementers will be beneficial for the welfare of the society.