Relationship between Sense of Security and Views on Love and Marriage among Undergraduates-Mediating Role of Adult Attachment

Objective: To explore the characteristics of adult attachment, sense of security and views on marriage and love in undergraduates, and analyze the relationship among the above three variables. 

Methods: Five hundred and five undergraduates were collected from Guangdong Province, China by stratified random sampling method. They were investigated with Adult Attachment Scale (AAS-1996), Sense of Security Questionnaire (SSQ) and Questionnaire of Undergraduates' View on Marriage and Love (VMLUQ). 

Result: 1. The total score of VMLUQ and the scores of erotic choice and marital tendency in this group were above the average, and the total score of SSQ and scores of 3 dimensions of AAS and the other five factors of VMLUQ were in the medium level. 2. It had an entirely mediating effect of intimity when interpersonal security was taken as a predictor to the total score of VMLUQ, and the ratio was 61.29%. It had an entirely mediating effect of intimity when interpersonal security was taken as a predictor to sexual choice, and the ratio was 65.96%. It had an entirely mediating effect of intimity when interpersonal security was taken as a predictor to marital tendency, and the ratio was 66.23%. It had an entirely mediating effect of intimity when the sense of control and confirmation was taken as a predictor to the marital tendency, and the ratio was 93.01%. It had an entirely mediating effect of intimity when the total score of SSQ was taken as a predictor to the marital tendency, and the ratio was 73.35%. It had an entirely mediating effect of intimity when interpersonal security was taken as a predictor to marital value, and the ratio was 34.98%. It had a masking effect of anxiety when interpersonal security was taken as a predictor to the score of sexual choice. 

Conclusion: The sense of security has a significant effect on the views on marriage and love by adult attachment in undergraduates.

Views of marriage and love refer to an individual's basic views on love and marriage, including views on love, mate selection, heartbreak, and sexuality, which deeply reflects an individual's outlook on life and values, and has a subtle influence on them [1]. As a social consciousness, Views on marriage and love reflect the cultural characteristics of society and has a strong dependence on social living conditions. Therefore, different societies have different views on marriage and love. Currently, the views on marriage and love among college students are generally positive and healthy [2]. However, influenced by social transformation, lack of family and school education, popularization of new media, and certain Western ideologies (such as postmodernism), college students have encountered some problems in their views on marriage and love, mainly manifested as complex love motives, utilitarian marriage goals, weakened marriage responsibilities, lack of rationality in choosing marriage partners, conformity in love behavior, overly casual attitude towards divorce, tolerance towards extramarital affairs and extramarital sexual behavior, and lack of sexual knowledge [3-6]. The above situation is not conducive to the physical and mental health and family happiness of college students, nor is it conducive to the harmonious development of society [7]. Sense of security is a premonition of potential danger, as well as an individual's sense of strength/powerlessness in responding to danger, mainly manifested as a sense of certainty and controllability [8]. Sense of security is one of the determining factors for mental health. For people who lack a sense of security, due to the fact that everything is more easily explained in a dangerous way, they have a strong sense of inferiority, distrust, and hostility, making it difficult for them to approach others and resulting in a sense of loneliness. For people with a high sense of security, because everything is more easily explained in a safe way, they have full confidence, trust, hope, cooperation, and empathy, are easy to approach others, and have a sense of belonging and intimacy.    Adult attachment is an individual's tendency to seek objects that provide stable and secure physiological and psychological support for them, which is a lasting emotional connection between an individual and their current partner, with the following two characteristics. The first is closeness. Individuals seek and attempt to maintain closeness to the attachment object, and separation from it can be painful. The second is safety. When faced with difficulties, individuals will seek comfort and help from their attachment, using them as a safe base [9]. Adult attachment is an "internal working mode" formed on the basis of early parent-child relationship experience, which largely operates through unconscious automated processing. It reflects an individual's expectations of others, organizes and structures their experiences, and determines their way of life [10-11]. When individuals encounter stressful situations, they will activate their attachment system, guide themselves to process and interpret information with specific tendencies, adopt specific behaviors, and form specific interpersonal relationship patterns [11-14]. In this sense, adult attachment is closely related to multiple factors that affect the quality of marriage and relationships, and the most important one is that it influences an individual's beliefs and expectations towards the relationships (i.e., their views on marriage and love) [15].     In summary, from a definition perspective, there is a close relationship between sense of security, adult attachment, and views on marriage and love. Sense of security is the psychological foundation of adult attachment. Views on marriage and love are constrained by adult attachment. As a relatively enduring interpersonal relationship among adults, both love and marriage partners should be able to provide individuals with a stable sense of security and make them willing to approach. The sense of security affects an individual's interpretation of everything, including the explanation of the phenomenon of marriage and love. In other words, the sense of security affects individuals’ views on marriage and love. From the origin perspective, views on marriage and love is an outward psychological phenomenon that belongs to the outcome variable; Adult attachment belongs to the deep and early onset distal psychological factors that affect individual behavioral performance and form expectations for behavioral outcomes, while sense of security belongs to the late onset proximal psychological factors. It can be said that sense of security is more mediated by adult attachment. On the other hand, empirical studies have shown a significant correlation between adult attachment and sense of security, as well as between adult attachment and views on marriage and love [16-17]. There have been no reports on whether there is a further relationship (such as moderating or mediating effects) between sense of security, adult attachment, and views on marriage and love.

Based on the above analysis, this study intends to use college students as an example to explore the mediating effect of adult attachment in the relationship between sense of security and views on marriage and love.

Objects

A stratified random sampling is used to select 570 undergraduates in Guangdong Province, China, and 505 valid questionnaires are collected, with an effective rate of 88.6%. Among them, there are 174 boys (34.5%) and 331 girls (65.5%); 115 freshmen (22.8%), 123 sophomores (24.4%), 130 juniors (25.7%), and 137 seniors (27.1%); 296 in urban areas (58.6%), 209 in rural areas (41.4%), 246 in humanities (48.7%), 157 in science (31.1%), 53 in engineering (10.5%), 34 in arts (6.7%), and 15 in sports (3.0%); 162 only children (32.1%) and 343 non only children (67.9%); 23 economically affluent families (4.6%), 399 good families (79.0%), and 83 poor families (16.4%); 470 (93.1%) of their parents had normal marital relationships, 7 (1.4%) had separated parents, and 28 (5.5%) had divorced parents.

Tools

Undergraduates' Views on Marriage and Love Questionnaire, VMLUQ

Compiled by Su Hong (2006) [18] , there are a total of 34 items, divided into 7 factors: sexual choices (SC), marital roles (MR), marital autonomy (MA), marital tendencies (MT), romantic motivations (RM), marital loyalty (ML), and marital values (MV). The Likert 5-point rating method is used to score from 1 to 5 points corresponding to "completely disagree" to "very agree". The higher the score, the more conservative the views on marriage and love. In this study, the Cronbach’s α coefficients of 7 factors ranged from 0.665 to 0.770, and the Cronbach’s coefficient of entire questionnaire is 0.841.

Adult Attachment Scale, AAS

Compiled by Collins et al. (1990) based on type measurement, revised by Wu Weili (1996) [19] into Chinese version, AAS consists of 18 items, divided into 3 subscales: intimacy (IN), dependence (DE), and anxiety (AN). The Likert 5-point rating method is used, with "1" representing "completely disagree" to "5" representing "completely agree". In this study, the Cronbach’s α coefficient of full scale is 0.837; and the Cronbach’s α coefficient of each subscale ranges from 0.762 to 0.809.

Sense of Security Questionnaire, SSQ

Compiled by Congzhong et al (2004) [8] based on Maslow's Safety Insecurity Scale (S-I), there are a total of 16 items, divided into two factors including interpersonal security (IS) and sense of determination and control (SDC). The Likert 5-point scoring method is used to score from 1 to 5 points corresponding to “very consistent” to “very inconsistent”. The higher the score, the stronger the sense of security. In this study, the Cronbach’s α coefficients of entire scale and two factors are 0.8103, 0.7816, and 0.8322, respectively.

Data processing

The parallel input method is used for data entry. After entering and reviewing with Excel, statistical analysis is conducted with SPSS 20.0 statistical software. Quantitative data are analyzed by independent sample t-test, Pearson correlation analysis, and multiple linear regression analysis. Counting data are analyzed by χ2 tests.

Testing for Common Method Bias

Due to the use of multiple scales in this study, there may be common methodological biases. Harman's One-factor Test [20] is used, which involves conducting unrotated principal component analysis on all variables simultaneously. The result of principal component analysis without rotation shows that there are 15 factors with eigenvalues greater than 1, and the first factor explains a variation of 26.35%, which is less than the critical standard of 40%, indicating that there is no significant common method bias in this study.

Correlation analysis of various variables

As showed in Table 1, there is a significant pairwise correlation between the total score of IS, IN and VMLUQ; IS, IN and SC; IS, AN and SC; IS, IN and MT; SDC, IN and MT, OS, IN and MT, as well as IS, IN and MV (all P<0>

Table 1: Descriptive statistics and correlation analysis results of variable (n-=505).

Notes: *p Less than 0.05, **p Less than 0.01.

The mediating role of adult attachment in the relationship between sense of security and views on marriage and love.

The mediating role of intimacy in the relationship between interpersonal security and the total score of VMLUQ

According to the mediation effect test method proposed by Wen Zhonglin et al. [21], the total score of VMLUQ is used as dependent variable (Y), the total score of IS is used as independent variable (X), and the mediator variable (M) that needs to be verified is the total score of IN. The specific steps are as follows. In the first step, regression analysis is performed with VMLUQ total score as the dependent variable and IS as independent variable to obtain the regression coefficient c; In the step 2, total score of IN is used as the dependent variable and IS total score as the predictive variable to conduct regression analysis and obtain the regression coefficient a; In step three, taking VMLUQ as the dependent variable and IS and IN total score as the predictive variables together, perform regression analysis to obtain regression coefficients b and c. The results are shown in Table 2.

Table 2: Test of the mediating effect of intimacy in the relationship between interpersonal security and total score of VMLUQ

According to Table 2, in the first step, IS positively predicted the total score of VMLUQ with a regression coefficient of 0.105, indicating a significant progression to the second step; In the second step, IS predicts IN positively with a regression coefficient of 0.346, and the regression coefficient significantly enters the third step; In the step three, after controlling for the impact of IN on the total score of VMLUQ, the regression coefficient between IS and the total score of VMLUQ is 0.041, which is not significant. That is, after adding the mediator variable IN between the dependent variable VMLUQ total score and the independent variable IS, the regression coefficient between UVMLQ total score and IS is not significant. It can be seen that the mediating effect of IN in the relationship between IS and the total score of UVMLQ is a complete mediating effect, with the mediating effect accounting for 61.29% of the total effect (i.e. effect m= ab/c = (0.346) × (0.186)/(0.105)×100%=61.29%).

The mediating role of intimacy between interpersonal security and sexual choice

The same method as 2.2.1 [21] is used to analyze the possible mediating effect of intimacy between interpersonal security and sexual choice. In this round of testing, the dependent variable Y is sexual choice (SC), the independent variable X is interpersonal security (IS), and the mediating variable M that needs to be verified is closeness (IN). The results are shown in Table 3.

Table 3: Test of the mediating effect of intimacy between interpersonal security and sexual choice

From Table 2, it can be seen that in the first step, IS positively predicts SC with a regression coefficient of 0.096, and the regression coefficient significantly enters the second step; In the second step, IS predicts IN positively with a regression coefficient of 0.346, and the regression coefficient significantly enters the third step; In the step three, after controlling for the impact of IN on SC, the regression coefficient between IS and SC is 0.033, which is not significant. That is, after adding the mediator variable IN between the dependent variable SC and the independent variable IS, the regression coefficient between SC and IS is not significant. From this, it can be seen that the mediating effect of IN in the relationship between IS and SC is a complete mediating effect, with the mediating effect accounting for 65.96% of the total effect (i.e. effect m=ab/c=(0.346) × (0.183)/(0.096) × 100%=65.96%).

The mediating role of anxiety between interpersonal security and sexual choice

The same method as 2.2.1 [21] is used to analyze the possible mediating effect of anxiety (AN) between interpersonal security (IS) and sexual choice (SC). In this round of testing, the dependent variable Y is sexual choice (SC), the independent variable X is interpersonal security (IS), and the mediator variable M that needs to be verified is anxiety (AN). The results are shown in Table 4.

Table 4: Test of the mediating effect of anxiety between interpersonal security and sexual choice

According to Table 4, in the first step, IS predicts SC positively with a regression coefficient of 0.096, indicating a significant progression to the second step; In the second step, IS positively predicts AN with a regression coefficient of (-0.309), and the regression coefficient significantly enters the third step; In the step three, after controlling for the impact of AN on SC, IS can still predict SC positively with a regression coefficient of 0.058, and a * b has a different sign with c '. It can be inferred that the role of anxiety in relationship between IS and SC is a masking effect.

The mediating role of intimacy between interpersonal security and marital tendency

The same method as 2.2.1 [21] is used to analyze the possible mediating effect of intimacy (IN) between interpersonal security (IS) and marital tendency (MT). In this round of testing, the dependent variable Y is marital tendency (MT), the independent variable X is interpersonal security (IS), and the mediator variable M that needs to be verified is intimacy (IN). The results are shown in Table 5.

Table 5: Test of the mediating effect of intimacy between interpersonal security and marital tendency

According to Table 5, in the first step, IS positively predicts MT with a regression coefficient of 0.221, indicating a significant progression to the second step; In the second step, IS predicts IN positively with a regression coefficient of 0.346, and the regression coefficient significantly enters the third step; In the step three, after controlling for the impact of IN on MT, the regression coefficient between IS and MT is not significant. That is, after adding the mediating variable IN, the regression coefficient between IS and MT is not significant. It can be seen that the mediating effect of IN in the relationship between IS and MT is a complete mediating effect, with the mediating effect accounting for 66.23% of the total effect (i.e. effect m=ab/c=(0.346)× (0.423)/(0.221) × 100%=66.23%). 

The mediating role of intimacy in the relationship between sense of determination and control and marital tendency

The same method as 2.2.1 [21] is used to analyze the potential mediating effect of intimacy (IN) in the relationship between sense of determination and control (SDC) and marital tendency (MT). In this round of testing, the dependent variable Y is marital tendency (MT), the independent variable X is the sense of determination and control (SDC), and the mediator variable M that needs to be verified is intimacy (IN). The results are shown in Table 6.

Table 6: Test of the mediating effect of intimacy in the relationship between sense of determination and control and marital tendency

According to Table 6, in the first step, SDC positively predicts MT with a regression coefficient of 0.099, and the regression coefficient significantly enters the second step; In the second step, SDC positively predicts IN with a regression coefficient of 0.206, and the regression coefficient significantly enters the third step; In the step three, after controlling for the impact of IN on MT, the regression coefficient of SDC on MT is not significant, that is, after adding the mediator variable IN between the dependent variable MT and the independent variable SDC, the regression coefficient between MT and SDC is not significant. It can be seen that the mediating effect of IN in the relationship between SDC and MT is a complete mediating effect, with the mediating effect accounting for 93.01% of the total effect (i.e. effect m=ab/c=(0.206)×(0.447)/(0.099)×100%=93.01%).

The mediating role of intimacy between overall security and marital tendency

The same method as 2.2.1 [21] is used to analyze the possible mediating effect of intimacy (IN) between overall security (OS) and marital tendency (MT). In this round of testing, the dependent variable Y is marital tendency (MT), the independent variable X is total score of SQ (OS), and the mediating variable M that needs to be verified is closeness (IN). The results are shown in Table 7.

Table 7: Test of the mediating effect of intimacy between overall security and marital tendency

According to Table 7, in the first step, OS positively predicts MT with a regression coefficient of 0.173, and the regression coefficient significantly enters the second step; In the second step, OS predicts IN positively with a regression coefficient of 0.299, and the regression coefficient significantly enters the third step; In the step three, after controlling for the impact of IN on MT, the regression coefficient between OS and MT is not significant, that is, after adding the mediator variable IN between the dependent variable MT and the independent variable OS, the regression coefficient between MT and OS is not significant. It can be seen that the mediating effect of IN in the relationship between OS and MT is a complete mediating effect, with the mediating effect accounting for 75.35% of the total effect (i.e. effect m=ab/c=(0.299)×(0.436)/(0.173)×100%=75.35%).

The mediating role of intimacy between interpersonal security and marital value

The same method as 2.2.1 [21] is used to analyze the possible mediating effect of intimacy (IN) between interpersonal security (IS) and marital value (MV). In this round of testing, the dependent variable Y is marital value (MV), the independent variable X is interpersonal security (IS), and the mediator variable M that needs to be verified is intimacy (IN). The results are shown in Table 8.

Table 8: Test of the mediating effect of intimacy between interpersonal security and marital value

According to Table 8, in the first step, IS negatively predicts MV with a regression coefficient of (-0.090), and the regression coefficient significantly enters the second step; In the second step, IS predicts IN positively with a regression coefficient of 0.346, and the regression coefficient significantly enters the third step; In the Step three, after controlling for the impact of IN on MV, the regression coefficient of IS on MV is not significant. That is to say, after adding the mediating variable IN between the dependent variable MV and the independent variable IS, the regression coefficient between MV and IS is not significant. It can be seen that the mediating effect of IN in the relationship between IS and MT is a complete mediating effect, with the mediating effect accounting for 34.98% of the total effect (i.e. effect m=ab/c=(0.346)× (0.091)/(0.090) × 100% = 34.98%).

This group of students scored at a moderate level in both the total score and two dimensions of SSQ, slightly lower than the research results of Wang Yan [22] and Zhang E [23]. There may be regional and generational differences in the sense of security among college students.

The students in this group scored at a moderate level in all three dimensions of AAS. Among them, 283 were classified as safe type (56.2%), 148 as preemptive type (29.3%), 44 as refusal type (8.7%), and 30 as fear type (6.1%). Compared with the results of Tang Furong [24], the proportion of safe type has slightly increased, while the proportion of non-safe type has slightly decreased. It is suggested that the adult attachment characteristics of college students may vary over time. Although both the post-80s and post-90s generations are only children, the parenting style of post-90s is more scientific, the parent-child relationship is more rational, and the upbringing environment (family and social environment) is more open-minded and inclusive [25].

This group scored above average on the two factors of sexual choice and marital orientation in VMLUQ. The total score of VMLUQ and the scores of the other five factors were all at a moderate level, which is inconsistent with previous research results [3-6]. The current college students hold a moderate view on marriage and love, which is neither conservative nor radical, influenced by traditional concepts and has modern characteristics. They hope to have a stable marital relationship, view emotions as a necessary foundation for sex and sex as an important element of marriage. It is worth noting that college students score at a moderate level in terms of their views on marriage roles and love motivation, indicating that they no longer emphasize the traditional marriage role positioning of strong men and weak women, and male lead female secondary. Instead, they tend to prefer modern marriage relationships that are gender equal, respectful and mutually supportive. They are not stubborn and can moderately borrow the opinions of others. They are not eager to fall in love during their college years, do not use it as a pastime or a tool to satisfy curiosity, and do not adhere to the dogma of "consistency". However, their attitude towards premarital sex (especially for girls) is more tolerant. It can be seen that college students tend to have rational views on marriage and love, but lack sufficient judgment ability for some trendy marriage and love concepts or phenomena (such as premarital sex), which to some extent also reflects their post materialistic tendencies in values [26].

This study indicates that intimacy plays a complete mediating role between interpersonal security and the overall score of marriage and love, interpersonal security and sexual choices, interpersonal security and marital tendency, sense of determination and control and marital tendency, interpersonal security and marital values, as well as overall security and marital tendency; anxiety plays a masking role between interpersonal security and sexual choices. That is to say, the stronger the sense of interpersonal security among college students, the higher their level of intimacy. Only by stimulating a high level of desire for intimacy can individuals form more conservative views of marriage and love. This is because college students with a stronger sense of interpersonal security are more likely to establish trust and mutual assistance relationships with close partners, believe that they can grasp the development direction of marital relationships, and believe more in the value of marriage, yearning for a single and stable marital relationship. On the other hand, whether an individual wishes to get married also depends on the desire for closeness stimulated by their own sense of security. When he/she has a high sense of security, it can stimulate a higher desire for closeness. Only by stimulating a high level of closeness can an individual develop an intention to get married. When individuals have a high sense of determination and control over their environment (people and things in life), they have the confidence to master and handle these people and things. Only then can they develop a sense of reliability and trust in their romantic partners, stimulate their desire for closeness, and ultimately generate the intention to get married. In addition, interpersonal security has no direct impact on sexual choice, but instead anxiety is the direct influencing factor of sexual choice. This indicates that regardless of the strength of interpersonal security, individuals may adopt an avant-garde attitude towards sexual behavior, and the key lies in the level of attachment anxiety. High anxiety individuals always worry about being abandoned in interpersonal relationships (especially between genders), and they are more inclined to adopt conservative sexual attitudes and behaviors to protect themselves; While low anxiety individuals are less worried about being abandoned in interpersonal relationships, especially between genders. They are more willing to adopt an open sexual attitude and behavior to showcase their attractiveness and create intimate romantic relationships.