The Mediating Role of Anxiety: The Relationship Between Anxious

Attachment Style and Willingness to Move out of Multi-Person

Dormitories in Chinese University Students

Shiyao Luo

Department of Behavioral Sciences, Vincennes University, Vincennes, IN 47591, U.S.A.

Keywords: Anxious Attachment Style, State Anxiety, Dormitory Life.

Abstract: Dormitory life has an important impact on university students, especially in China, where the total population

and the number of university students are large. In China, university students usually live in multi-person

dormitories based on social common sense and even social norms. Studies have shown that different

attachment styles are likely to have different preferences for accommodation, but there are still gaps in

research on how the various influence mechanisms work. This study analyzed the direct and indirect effects

of anxious attachment style on the willingness to move out of multi-person dormitories, mediated by anxiety

state. An online questionnaire was distributed to Chinese university students via Chinese social media. The

results show that the anxiety state is the important and significant influence mechanism in achieving the

correlation between anxious attachment style and moving-out willingness and was linked closely to both of

them. For the direct effect, there was a nearly significant positive correlation between the independent and

dependent variables. The results are consistent with the previous research on attachment and housing choices.

On this basis, the conclusion obtained in this study regarding the significant correlation of the anxiety state

with the variables and effect is important for revealing the way attachment styles work.

1 INTRODUCTION

Dormitory life is an important and integral part of

universities, especially in China. If students decide to

live in the multi-person dormitories on campus, i.e.,

not in a private room by themselves, it means that

they have to share their living space with others. In

these spaces, university students inevitably have to

interact with other people. Moreover, the situation

might be more intense in Chinese society. In China, it

is common sense that university students need to,

should, or can live in multi-person dormitories is

social common sense. Meanwhile, under the tight

policies in many universities, most students choose to

live in multi-person dormitories instead of

commuting when they enroll. Moreover, due to the

high population and increasing number of university

students, the multi-person dormitory environment

that students need to face may be even more severe.

According to the statistics from the Ministry of

Education of the People’s Republic of China (MOE)

in 2023, there were a total of 47.6319 million students

enrolled in various forms of higher education, an

increase of 1.0811 million from the previous year

alone, and a gross enrollment rate of 60.2% (Ministry

of Education of the People’s Republic of China,

2024).

Dormitory life can bring more things to handle for

university students. University students who live in

dormitories, no matter male or female, have higher

stress and social dysfunction levels than those who do

not live in dormitories, as the study by Molasaeidi

points out (Molasaeidi, 2014). These differences may

be explained through attachment theory to some

extent. Attachment theory was first proposed by

Bowlby, who noted that this infant attachment

behavior to the mother would bring them rich

emotional responses and teach them social adaptation

skills (Bowlby, 1997). In the study by Ainsworth et

al., this pattern of attachment was further

distinguished into secure, anxious, and avoidant

attachment styles through the Strange Situation,

expressing a pattern of exploration of the

environment influenced by the caregiver’s behavior

(Ainsworth, 2014). Individuals with an anxious

attachment style in adulthood are more likely to have

stronger emotional ups and downs and a desire for

return in romantic relationships, while those with

170

Luo, S.

The Mediating Role of Anxiety: The Relationship Between Anxious Attachment Style and Willingness to Move out of Multi-Person Dormitories in Chinese University Students.

DOI: 10.5220/0014110100004942

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 2nd International Conference on Applied Psychology and Marketing Management (APMM 2025), pages 170-176

ISBN: 978-989-758-791-7

avoidant attachment styles are more likely to avoid

intimacy (Hazan, 1987). In the last thirty years, as

more and more studies on adult attachment styles

have been conducted, new research has emerged that

can explain the relationship between attachment

styles and dormitory life. According to Lapsley and

Edgerton, insecurely attached individuals are less

adaptable to college, which may affect their

adjustment to residential life and their ability to

manage interpersonal interactions in multi-person

dormitories (Lapsley, 2002). Attachment styles do

affect the housing choice during university.

Compared to securely attached individuals,

avoidantly and anxiously attached individuals are

more likely to choose to live alone or with their

parents instead of sharing the space with others, as

pointed out in Sasso and Paladini’s study (Sasso and

Paladini, 2021). However, there is still a gap in the

field regarding the specific mechanism through how

insecure attachment styles lead to their choice of not

living with other students, including living in multi-

person dormitories.

This study focused on the relationship between

anxious attachment style and willingness to not live

in multi-person dormitories. It further confirmed the

hypothesis that individuals with higher anxious

attachment styles are more likely to have the

willingness to move out of multi-person dormitories

based on the social culture and atmosphere of the

research subjects, Chinese university students. The

independent variable was set as the score of anxious

attachment styles, while the dependent variable was

the willingness to move out of multi-person

dormitories. To study the influencing mechanism, an

intervening variable, the anxiety state of living in a

multi-person dormitory, was set up to investigate

whether the correlation between anxious attachment

style and the willingness to move out of a multi-

person dormitory is achieved by anxiety state

2 MATERIALS AND TOOLS

2.1 Samples

The questionnaire was imported into a well-known

Chinese platform and distributed through Chinese

social media to find college students currently

enrolled in universities in China. In addition to the

main questions, questions were added to count the

current accommodation situation and the grade of the

sample. The valid rate of the questionnaire was

calculated by excluding responses that were too short,

too random, or too repetitive.

2.2 Experience in Close Relationships

Scale- Chinese Short (ECR-CS)

This study used the Experience in Close

Relationships Scale- Chinese Short (ECR-CS) to

measure the attachment styles of the subjects. The

ECR-CS was first revised into a Chinese version by

Chinese scholars Li et al. in 2016, based on the

original ECR developed by Brennan, Clark, and

Shaver, and then revised again into a Chinese short

version by She et al. in 2014. The ECR-CS revised by

She et al. has good internal consistency (She et al.,

2015). This study indirectly cited the ECR-CS by She

et al. in the appendix of Dong’s article (Dong, 2021).

The advantage of ECR-CS is that it is suitable for the

measurement of adult attachment styles, conforms to

the Chinese context, and is short enough to ensure

more valid data can be collected in a limited time.

Considering that the subjects were Chinese university

students, which was different from the young and

middle-aged people in Dong's research, the author

replaced all the “spouses” with “partners” in the scale.

Also, the main question section indicated that

subjects could imagine future partners or important

others if they did not have romantic relationship

experience since some Chinese university students

might lack it.

2.3 State Anxiety Inventory (SAI)

Concerning the moderate variable, anxiety state, this

study used the State Anxiety Inventory (SAI) part of

the State-Trait Anxiety Inventory, STAI-Form

Y(STAI). The original English version was

developed by Spielberger et al., and the Chinese

version used by this questionnaire was developed by

Wang et al. in 1992 and is included in the Rating

Scale for Mental Health (Wang et al., 1999). The

reason for choosing the SAI instead of the Trait

Anxiety Inventory (TAI) or STAI is that the anxiety

experienced by individuals in the dormitory that

motivates them to want to move out of the dormitory

is more likely to be state anxiety than trait anxiety.

Meanwhile, the SAI is shorter than the STAI,

ensuring that the questionnaire has a high return rate

and is efficient.

2.4 Moving-Out Willingness

Considering the sociocultural context of China,

besides the dependent variable itself, which was the

willingness to move out of a university dormitory, the

subjects also had to answer the difficulty level of

moving out of the dormitories if they had ever had

The Mediating Role of Anxiety: The Relationship Between Anxious Attachment Style and Willingness to Move out of Multi-Person

Dormitories in Chinese University Students

171

thought about moving out (i.e., not choosing “I have

never considered about it, I have never had such

thoughts”). This was to make sure the difficulty of

moving out would not affect the willingness.

3 PROCEDURE

This study developed an online questionnaire that was

distributed online via Chinese social media, and a

total of 343 valid questionnaires were collected. The

questionnaire divided the subjects into three groups

through their campus housing situation and set

different question-skipping logic. This is intended to

allow subjects in different housing situations to

answer SAI scales with different modifiers, and

subjects who no longer live in university dormitories

do not need to answer the questions about their

willingness to move out.

Figure 1: Different questions that need to be answered by different groups (Photo/Picture credit: Original).

There were three sets of SAI with different

modifiers for three different groups (See Fig. 1.). The

differences were nearly only the modifiers of some

sentences to make the questions more relevant to the

situation of the subjects. For example, the stems for

SAI for Group 1 and Group 2, respectively,

“Considering my feelings about dormitory life and

staying in the dormitory” and “Thinking back to when

I lived in the university dormitory before;” for the

scale questions, one in Group 1 was “I am presently

worrying about living in the dormitory, feeling it may

be over possible misfortunes,” whereas the other was

“I worry about living in the dormitory, feeling it may

be over possible misfortunes.” It is worth noting that

the English version of the questionnaire shown in this

article is translated by the author for only

demonstration purposes. It was not used in formal

experiments and has not been reliability tested.

Because the ECR-CS and the three versions of the

SAI scale for different groups all had different extents

of changes in modifiers and question sentences,

Cronbach’s Alpha was used to measure the reliability

of those for scales. However, the SAI scale for Group

3 was not tested due to insufficient samples (N=1).

After data collection was completed, the

questionnaire data was exported from the online

questionnaire platform to IBM SPSS statistical

software. It was finally found that there was a total of

336 questionnaires in Group 1, while there were 4 and

3 in Group 2 and 3, respectively. Since the sample

sizes of Group 2 and 3 were too small, this study

mainly limited the scope of the data analysis to the

336 Chinese university students in Group 1 who were

still living in the school dormitories.

Before the formal analysis, the correlation and

significance between the willingness to move out and

the difficulty of moving out were first analyzed using

the Spearman correlation coefficient to ensure that the

difficulty of moving out would not affect the

individual's willingness to move out. Because the

original questionnaire included the option “I'm not

sure/I haven't looked into how difficult it is to move

out of the dormitory,” the sample that selected this

option was excluded from the analysis of the

correlation between willingness and difficulty. The

willingness to move out and the difficulty of moving

out were treated as 0-5 and 0-4, respectively. For the

main hypothesis, PROCESS Version 4.2 was used to

Group1:

Living in

university

multi-person

dormitories

SAI (Anxiety

state in

dormitories)

Moving out

willingness

[If ever thought

about moving

out] Moving out

difficulty level

ECR-CS

(Attachment

style)

Group2: Living

in university

multi-person

dormitories

SAI (Anxiety

state in

dormitories)

Moving out

willingness

[If ever

thought about

moving out]

Moving out

difficulty level

ECR-CS

(Attachment

style)

Group3: Living

in university

multi-person

dormitories

SAI (Anxiety

state in

dormitories)

Moving out

willingness

[If ever thought

about moving

out] Moving out

difficulty level

ECR-CS

(Attachment

style)

APMM 2025 - International Conference on Applied Psychology and Marketing Management

172

obtain the results of the mediation analysis on the

independent variable of the mean score of anxious

attachment style, the mediator variable of the mean

score of anxiety state, and the dependent variable of

the willingness to move out.

The main hypothesis was additionally verified from

the dimension of avoidant attachment, that is, X was

replaced by the mean value of the individual avoidant

attachment dimension measured by the ECR-CS, and

an additional multiple regression analysis was

performed. Meanwhile, a cross-tabulation analysis

was also conducted based on demographic questions

to determine differences in propensity to move out

between different grades.

4 RESULTS

After a total of 371 responses to this questionnaire,

questionnaires with obviously unreasonable answers

(those with evidently repeated and regular answers, as

well as those with answers that were too short) were

eliminated. A total of 343 valid questionnaires were

collected, for an effective rate of approximately

92.5%. Of these, 336 were from Group 1 (currently

living in a university multi-person dormitory) and

were used for the main data analysis. The grade

distribution of the 336 subjects ranged from

undergraduate freshmen to third-year graduates.

Among them, there were 135 undergraduate freshmen

(FR), 59 undergraduate sophomores (SO), 62

undergraduate juniors (JR), 49 undergraduate seniors

(SR), 5 first-year graduate students (G1), 15 second-

year graduate students (G2), and 11 third-year

graduate students (G3).

The reliability of the 20-item SAI for Group 1

(336 samples), using Cronbach’s Alpha, turned out to

be alpha=0.956, which is a high internal consistency

and consistent measurement of the topic of anxiety

states. The Group 2 version, which also had 20 items,

only yielded four valid responses, but was still

measured, and turned out with an alpha of 0.958. All

343 valid responses answered the 12-question ECR-

CS, yielding an alpha of 0.856, indicating good

internal consistency.

Table 1. Mean value and standard deviation of willingness

to move out and difficulty of moving out.

Mean

value

Standard

Deviation

Willingness to move out 1.32 1.225

Difficult

of movin

out 2.13 1.103

The willingness to move out and the difficulty of

moving out were treated as 0-5 and 0-4, respectively.

The mean value and standard deviation are shown in

Table 1. In the Spearman correlation and significance

test to examine the willingness to move out and the

difficulty of moving out, r=-0.003 and p=0.968 were

obtained. This shows that there is no significant

correlation between the willingness to move out and

the difficulty of moving out. Therefore, the difficulty

of moving out is not considered as a variable. The

dependent variable only considers the willingness to

move out.

4.1 Mediation Analysis

In a multiple regression analysis using PROCESS

v4.2 with a confidence level of 95%, the R² was

0.0265 in the meditator model part, which means that

the mean value of the individual anxious attachment

style scale (X) explained 2.65% of the meditator

variable, anxiety state (M). It turned out that

p=0.0027, which is less than 0.05, can be considered

statistically significant. The non-normalized

regression coefficient b is 0.0804, which indicates

that for every unit increase in the individual anxious

attachment style score, the anxiety state score

increases by about 0.0804.

With regard to the part of the dependent variable

model, it can be seen from R²=0.4791 that X and M

explained 47.91% of the dependent variable, the

willingness to move out (Y). It is evident that

individuals’ anxious attachment style and anxiety

state in the dormitory had a very strong explanatory

power on the willingness to move out of the

dormitory. Meanwhile, in the indirect effect, the

effect was 0.1045, which means that part of the effect

of X on Y is achieved through the anxiety state; and

the confidence interval is 95% CI [0.0314, 0.1759],

which indicates that it is significant. If the effect of M

on Y is considered alone, b=1.3008, p<0.0001, it can

be seen that the mediator variable, anxiety state, has

an extremely strong and significant effect on

increasing the willingness to move out. However,

when X acts on Y alone, b=0.0697, p=0.0675, is not

statistically significant but close to, and the effect is

not as strong as M on Y. This might mean that the

effect of anxious attachment style on willingness to

move out is mainly achieved by the mediator variable.

That is, it is more the anxious attachment style that

leads to an anxiety state, and the anxiety state leads to

a willingness to move out, rather than the anxiety

state acting as an influencing factor.

When X is replaced with individual avoidant

attachment style scores and analyzed again, the effect

The Mediating Role of Anxiety: The Relationship Between Anxious Attachment Style and Willingness to Move out of Multi-Person

Dormitories in Chinese University Students

173

of X on M, M on Y, and the mediation effect were

basically the same, but for the direct effect, which is,

the effect of X on Y alone, b=-0.0421, p=0.1662. In

other words, for every unit increased in avoidant

attachment style scores, the willingness to move out

decreased by 0.0421, but the result is not significant

(See Table 2).

Table 2. Direct and indirect effects in mediation analysis.

X→M M→Y X→Y indirect effect

X: Anxious attachment b=0.0804 b=1.3008 b=0.0697

=0.0675

0.1221

Y: Avoidant attachment b=0.0905 b=1.3493 b=-0.0421

p=0.1662

0.1045

4.2 Cross-Tabulation Analysis

In the cross-tabulation analysis of the differences in

the willingness to move out between different grades

based on the demographic questions, the comparison

of the percentages of options selected in each grade is

shown in the figure below (See Fig. 2.). It can be seen

that the largest number of people in FR and G3 chose

to have not considered moving out of the dormitory

at all, and the smallest number of people in FR

already had a complete plan.

Figure 2: Proportions of dormitory moving out willingness levels among students by grade (Photo/Picture credit: Original).

5 DISCUSSION

5.1 Anxious Attachment Style, Anxiety

State, and Moving-out Willingness

In mediation analysis, it was shown that the direct

effect of anxious attachment style to moving out

willingness was relatively low and close to significant

(b=0.0697, p=0.0675), while the indirect effect when

using anxiety state as mediator variable (the effect

was 0.10450, 95% CI [0.0314, 0.1759]), the effect of

independent variable to mediator variable (b=0.0804,

p=0.027), and the effect of mediator variable to

dependent variable (b=1.3008, p <0.0001) all had

high and significant effectiveness. First, consistent

with the hypotheses of this study, these results can

prove that anxiously attached individuals are more

likely to show anxiety state in multi-person

dormitories, and this kind of state will increase their

willingness to move out of the dormitories.

Meanwhile, the anxiety state was an extremely

important mediation in this indirect effect, and this

result was also statistically significant. However,

although anxious attachment style also had a direct

effect on moving-out willingness, it is not significant

enough.

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

FR SO JR SR G1 G2 G3 Total

Willingness=0 Willingness=1 Willingness=2

Willingness=3 Willingness=4 Willingness=5

APMM 2025 - International Conference on Applied Psychology and Marketing Management

174

This shows a situation that this mediation effect

was mainly achieved through the mediator variable,

anxiety state, which may reflect a certain

psychological mechanism in Chinese culture and

society: Although an individual is anxiously attached,

if they do not feel a more influential, intuitive, and

“obvious” anxiety state, they may not easily realize

their internal need to move out of the dormitory, nor

will they easily realize that moving out of the

dormitory is more in line with their personal needs. In

the Chinese culture, this may stem from an emphasis

on collectivism; “living in a dormitory” is seen as

something that everyone should do as a matter of

course and is harmless, while “behaving differently

from the group” brings cultural pressure and a sense

of shame to the individual. A prior study by Guess

points out that under collectivist values, individuals

are more cautious and want their decisions to conform

to social expectations (Guess, 2004). When this

culture reflects onto society, the individuals may

discover that the vast majority of people around them

are either viewing tolerating dormitory living as a

matter of course or just continuing tolerating it

reluctantly. At the same time, if individuals seek

feedback from the outside world, they find that

parents and teachers do not have a positive attitude

towards moving out of the dormitory. Especially

anxiously attached individuals might be more likely

to seek feedback from others, for example, Hepper

and Carnelley’s research shows that individuals with

a high-anxious attachment style are more eager for

feedback and more open to negative feedback

(Hepper and Carnelley, 2010). These are all possible

reasons that affect the result that individuals did not

evaluate their decision on dormitory through their

own personality and interpersonal characteristics

until they realized their distress in an anxiety state. In

the design stage of this study, three groups were

designed based on different accommodation

situations, but the final samples that were not

currently living in a dormitory only accounted for

about 2.04%. This data distribution can also prove to

some extent that not living in a shared

accommodation is not mainstream at all.

On the contrary, when using avoidant attachment

style as an independent variable, the result can only

support avoidantly attached individuals are also more

likely to have anxiety state in multi-person

dormitories, and this state can prompt them to

consider moving out. When the avoidant attachment

style directly affects the moving-out willingness, the

direction of this effect is negative, and the result is

very insignificant. This can prove that the effect of an

avoidant attachment style and an anxious one on the

moving-out willingness is different. It is worth noting

that, as shown in Table 1, the greatest effect value in

the two mediating analyses is the part where anxiety

leads to an increase in the willingness to move out.

This shows that regardless of attachment style,

anxiety is a strong factor influencing the willingness

to move out of the dormitory.

The results of the cross-analysis of demographic

questions and willingness to move out show that

undergraduate freshmen (FR) were the least likely to

consider moving out of the dormitories, but this

changed with increasing grade levels. As the grade

level increases, more and more individuals have a

stronger desire to move out of the dormitories, which

might be related to the anxiety state they experienced

in their residential life, or the gradual maturity of their

abilities as growing up. It is worth noting that the

sample size among the graduate student population

was very limited, so the proportions of willingness

levels among the three graduate student grades are of

limited reference value.

5.2 Limitation and Expectation

In terms of sample collection, the sample size of

different grades was not evenly distributed, which

might have a certain impact on the results of the

specific situations in different grades, such as

undergraduate freshmen seemed less likely to think

about moving out. The most significant limitation of

this study was that the cultural and social background

might have a strong impact on the subjects, for

example, there was a very limited size of subjects that

did not live in university multi-person dormitories.

Meanwhile, as mentioned earlier, collectivist values

might have an impact on the study. Subsequent

research should add culture-related questions to the

questionnaire to increase consideration of the impact

of cultural background on individual values and

decision-making patterns. At the same time, the

hypothesis also deserves cross-cultural research and

observation if the direct effect would be stronger.

This hypothesis also needs to be carried out in

societies with more diverse university

accommodation and commuting methods to assess

whether anxious attachment styles and anxious states

have an impact on different commuting methods, that

is, a further subdivision based on moving out of the

dormitory.

The Mediating Role of Anxiety: The Relationship Between Anxious Attachment Style and Willingness to Move out of Multi-Person

Dormitories in Chinese University Students

175

6 CONCLUSION

In this study of the correlation between anxious

attachment style and willingness to move out of

multi-person dormitories, in addition to the near-

significant correlation between the two variables, the

role of anxiety as a mediator variable was

significantly verified. The results show that anxiety

state is a very strong and powerful factor in the

achievement of the relationship between anxious

attachment style and moving-out willingness,

especially the impact of anxiety state on willingness

(b=1.3008, p <0.0001). This means that, if anxiously

attached individuals decide to live in multi-person

dormitories, they need to pay more attention to their

anxiety and find coping strategies to ensure that stress

does not affect their study life. In addition, they also

need to focus more on their own needs in daily life

and when making housing decisions, and consider the

high risk of anxiety state they may encounter. For

those anxiously attached individuals who are in

anxiety in multi-person dormitories, this study will be

a tool that helps them understand their situation and

reminds them to pay attention to their mental state.

For higher education institutions, such as universities,

this will be a reminder to provide more flexible

commuting policies and give more understanding to

diverse students. For the field of attachment theory

research, this can be an introduction to focusing on

the more specific influence pattern of anxious

attachment styles.

REFERENCES

Ministry of Education of the People's Republic of China,

Statistical bulletin on the development of national

education in 2023, Ministry of Education of the

People's Republic of China Official Website (2024)

Molasaeidi, A comparison of mental health in dormitory

and non-dormitory students, Indian J. Fundam. Appl.

Life Sci. 4, 328–334 (2014)

J. Bowlby, Attachment and loss: Vol. 1. Attachment,

Pimlico, London. (1997)

M.D.S. Ainsworth, Patterns of attachment: A psychological

study of the strange situation, Psychology Press, New

York. (2014)

C. Hazan, P. Shaver, Romantic love conceptualized as an

attachment process, J. Pers. Soc. Psychol. 52, 511–524

(1987)

L.K. Lapsley, D.J. Edgerton, Separation-individualization,

adult attachment style, and college adjustment, J.

Couns. Dev. 80, 484–492 (2002).

P.A. Sasso, A. Paladini, Around the house: A descriptive

study of attachment styles, living arrangement, and

involvement in undergraduate commuter students, J.

Campus Act. Pract. Scholar. 3, 16–29 (2021)

Z. She, Q. Sun, G. Jiang, Revision of the Experience in

Close Relationship Scale-Short Version (ECR-CS), in

Proceedings of the 18th National Academic Conf. on

Psychology: Psychology and Social Development,

Chinese Psychological Society Ed., 945–946 (2015)

S.Q. Dong, Intervention research on psychological distress

in middle-aged and young lymphoma patients based on

adult attachment theory, Master’s thesis, Zhengzhou

University (2021)

X.D. Wang, X.L. Wang, H. Ma, Manual of mental health

assessment scales (Revised Edition), Chinese J. Ment.

Health Press, Beijing. (1999)

C.D. Guess, Decision making in individualistic and

collectivistic cultures, Online Read. Psychol. Cult. 4, 1

(2004)

E.G. Hepper, K.B. Carnelley, Adult attachment and

feedback-seeking patterns in relationships and work,

Eur. J. Soc. Psychol. 40, 448–464 (2010)

APMM 2025 - International Conference on Applied Psychology and Marketing Management

176