Fan et al.

BMC Psychiatry (2024) 24:594 BMC Psychiatry

https://doi.org/10.1186/s12888-024-05991-x

RESEARCH Open Access

Self-concealment scale (SCS) in middle-aged

people: psychometric features and cross-age
equivalence test
Zhiguang Fan1,3, Xiaoli Shi2, Xiangbao Huang1 and Li Li3*

Abstract
Background Individuals’ concealment of negative information and privacy may lead to impaired social interactions
and threatened health conditions. This study aimed to investigate the effectiveness of the Self-Concealment Scale
(SCS) in the middle-aged Chinese population and to examine the equivalence of the SCS among different age groups.
Methods The current research adopted the SCS, Distress Disclosure Index (DDI), Revised Cheek and Buss Shyness
Scale (RCBS), Social Interaction Anxiety Scale (SIAS), Social Phobia Scale (SPS), UCLA Loneliness Scale (ULS-8), and
Kessler Psychological Distress Scale (K10) to survey 1124 middle-aged people. To explore the factor structure of the
SCS, the study employed exploratory factor analysis and confirmatory factor analysis. The reliability of the SCS was
measured based on Cronbach’s α coefficients, McDonald’s Omega coefficients, and split-half reliability. Correlation
analysis was applied to examine the relationship between SCS and RCBS, SIAS, SPS, ULS-8, and K10. Moreover, this
study recruited 1458 emerging adults and 1104 older adults to identify the cross-age invariance of the SCS.
Results Exploratory factor analysis of the middle aged adults’ data supported a single factor model. The factor
loadings of SCS items ranged from 0.62 to 0.73, the commonality ranged from 0.39 to 0.53, and the single-factor
model fitted well. The scale showed a Cronbach’s α coefficient value of 0.895, McDonald’s Omega coefficient of 0.893,
and a split-half reliability coefficient value of 0.861. In addition, the SCS demonstrated invariance in emerging adults,
middle-aged adults, and older adults. Further analysis showed that the scores of the SCS (F = 3.55, p = 0.029) among
emerging adults (M = 26.43, SD = 7.96) were significantly higher than among middle-aged (M = 25.77, SD = 7.79), and
older cohort (M = 25.69, SD = 7.91).
Conclusion The SCS revealed favorable psychometric characteristics among the middle-aged Chinese population.
The degree of self-concealment among emerging adults was higher than that of middle-aged and older people.
Keywords Middle-aged people, Self-concealment, Reliability, Validity, Equivalence

*Correspondence:
Li Li
[email protected]
1
Department of Psychology, Shaoxing University, Zhejiang
312000, People’s Republic of China
2
School of Education, Jilin International Studies University,
Changchun 130117, People’s Republic of China
3
School of Economics and Management, Jilin Engineering Normal
University, Changchun 130052, People’s Republic of China

© The Author(s) 2024. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0
International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you
give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the
licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation
or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://
creativecommons.org/licenses/by-nc-nd/4.0/.
Fan et al. BMC Psychiatry (2024) 24:594 Page 2 of 10

Introduction Although the SCS has been employed to explore the

Self-concealment is a common and well-known human characteristics, detrimental effects, and the influencing
experience that anyone who has ever suffered from factors of self-concealment among middle-aged Chinese
embarrassment, unpleasantness, or pain is aware of it. adults [12, 13], the psychometric properties of the SCS
However, individuals significantly differ in their feeling of in the Chinese middle-aged population are still specu-
self-concealment. This displays as different levels of con- lative. The original version of the scale was developed
cealment of personal information [1]. Self-concealment mainly based on middle-aged adults in different occupa-
is perceived as a relatively stable personality trait in that tions, such as nurses, physical therapists, social workers,
individuals are consciously inclined to conceal painful and volunteers in social service agencies [3]. Two exist-
memories, traumatic experiences, or negative emotions ing studies on the psychometrics of the Chinese version
[2]. What constitutes self-concealment is made up of dif- of the SCS, however, were only conducted in adolescent
ferent components, including a tendency to keep secrets, and older adult populations, not involving middle-aged
reluctance to share agonizing details with others, and adults.
concerns about privacy breaches [3]. Middle-aged people normally represent the nucleus of
Researchers, hitherto, have developed several scales to a family and the main financial contributor. Not only do
assess the construct of self-concealment. For instance, they deal with the pressure of work, but also they bear the
Larson and Chastain developed the 10-item Self-Con- family responsibilities of raising and educating children
cealment Scale (SCS) based on a systematic elaboration and supporting parents. This group of people frequently
of the concept of self-concealment and related theories suffer from mounting pressure from work and family,
[2]. It was found that the SCS could serve as a valid tool especially in the modern industrialized society [14]. As
for assessing an individual’s general self-concealment emphasized by psychologists, people’s disclosure of their
tendency due to its favorable psychometric charac- negative feelings, depressed thoughts, and intense sorrow
teristics. Moreover, the study further highlighted that is with the intention of accumulating social support and
although self-concealment and self-disclosure are cor- reducing psychological distress [15]. However, middle-
related to a certain extent, they do not share an identi- aged adults may resort to hide painful feelings and nega-
cal psychological structure. In ensuing related studies, tive messages with the purpose of protecting self-esteem
researchers also examined the validity of the SCS several and avoiding disappointment, contempt, and irritation
times among people characterized by diverse countries, from others [16]. In addition, individuals who fear pub-
ages, and occupations [7]. Scholars have adopted the SCS lic exposure of painful matters may lead to negative effect
to measure self-concealment in their studies among Japa- on others or impedance in problem-solving may tend to
nese, Nigerian, Asian, and African-American populations conceal themselves [17].
[4–7]. At present, the SCS has become the most widely Therefore, this study, while following the psychometric
used scale for evaluating self-concealment. requirements, was designed to examine the applicabil-
Although the SCS is developed based on U.S. residents, ity of the Chinese version of the SCS in the middle-aged
it also possesses certain cross-cultural applicability. The population, to provide a valid tool for processing related
other language versions of the SCS still demonstrated studies. Furthermore, as is apparent in available litera-
good reliability and validity, i.e., the Turkish version ture, no studies have examined the cross-age consistency
and the Chilean version [8, 9]. In addition, prospec- of the SCS. The measured equivalence of the scale in
tive quantitative and cross-cultural comparative stud- different cohorts is a prerequisite for variance analyses.
ies of self-concealment among Chinese people require a Therefore, the study further examined the cross-cohort
valid assessment tool. To this end, Chinese scholars have consistency of the SCS in emerging adults, middle-aged
attempted to translate the English version of the SCS into adults, and older adults, analyzing age differences.
Chinese and to conduct reliability and validity tests. In To examine the criterion validity of the SCS, the cur-
particular, Wang revised the Chinese version of the SCS rent study included the Distress Disclosure Index (DDI),
for the first time in a population of secondary school stu- Kessler Psychological Distress Scale (K10), Revised
dents [10], which concluded that the Chinese version was Cheek and Buss Shyness Scale (RCBS), Short Forms
consistent with the original scale in terms of the number of the Social Interaction Anxiety Scale (SIAS), and the
of items and the dimensions, and demonstrated favorable Social Interaction Anxiety Scale (SIAS). Self-conceal-
internal consistency reliability and split-half reliability. ment can induce negative emotional experiences that
Moreover, Fan et al. re-verified the validity of the SCS in directly or indirectly threaten an individual’s health and
the Chinese older population [11]. The above two studies well-being [18]. Cruddas et al. found that self-conceal-
indicate that the SCS is applicable in the Chinese social ment was significantly and positively associated with
context. low self-disclosure, negative social comparison, anxiety,
depression, and stresss [19]. In addition, the conservation
Fan et al. BMC Psychiatry (2024) 24:594 Page 3 of 10

of resources theory (COR) suggests that resources are the population aged 35 to 64 into middle-aged people,
highly valued in human survival and development and and the population aged 65 and above into older people.
that people are inclined to strive to retain and maintain Recruitment of middle-aged and older people was car-
existing resources and continue to acquire and cultivate ried out using convenience sampling and snowball sam-
new resources [20]. If individuals fail to effectively pre- pling. The survey was conducted by systematically trained
vent resource depletion or obtain beneficial resources students who visited their home communities or rural
promptly, a loss spiral will be formed and poses a seri- areas during holidays. The research was based on a one-
ous threat to mental health [21]. Self-concealment is to-one basis and participants completed the responses
a stressor that can lead to ongoing resource depletion. independently after obtaining informed consent. In addi-
Concurrently, self-concealment also impedes the acquisi- tion, at the end of the survey, the respondents would rec-
tion of resources by diminishing an individual’s willing- ommend potential participants to expand the scope of
ness to express and reducing the opportunity to acquire the survey. It’s worth noting that the data for older adults
assistance and obtain social support [22]. Therefore, self- in this study consisted of two parts. First, data from older
concealment may be significantly and negatively associ- adults obtained in the previous study was included, total-
ated with self-disclosure and mental health status. ing 1085 people [11], which means that the current study
Self-concealment has a destructive effect on the main- included former research data. Second, in this survey, 19
tenance and development of interpersonal and intimate respondents who are over 65 years old were not deleted
relationships [23]. According to the Social Anxiety Cog- but added to the data for the older cohort. The discrep-
nitive Model, individuals with high social anxiety tend to ancy from previous studies on self-concealment among
anticipate others’ negative responses towards themselves older Chinese people is that the current study examined
and internalized self-protection behavior patterns i.e. the effectiveness of SCS in the middle-aged population,
concealment and avoidance, experiencing intense lone- and the equivalence and variance of SCS in the three
liness [24]. It is evident that self-concealment manifests populations of emerging, middle-aged, and older adults.
itself as a predictor of social anxiety, social phobia, loneli- Previous findings underpin the actualization of the cur-
ness, and low self-esteem [25]. In Akdoğan et al.‘s study, rent study. Therefore, there are significant differences
it was found that self-concealment and loneliness play a between the two studies in terms of the targeted popula-
mediating role in the effect of low self-esteem on well- tion and corresponding conclusions.
being [26]. Thus, self-concealment may be significantly In the study, convenience sampling was adopted to col-
and positively associated with social anxiety, social pho- lect data from emerging adults by distributing question-
bia, loneliness, and shyness. naires to undergraduate and postgraduate students in
In the wake of the discussion mentioned above, this three universities. The respondents consisted of under-
study hypothesized that the SCS would exhibit favorable graduate and graduate students majoring in medicine,
reliability and validity in the middle-aged Chinese popu- nursing, education, and language-related fields. Before
lation, which can be used as a valid instrument to assess starting the survey, the investigators explained in detail
self-concealment tendency. Moreover, it further hypoth- the purpose, confidentiality, anonymity, and voluntary
esized that the SCS total score would be significantly nature of the survey to different age groups. Only after
positively correlated with the K10, RCBS, SIAS, SPS, and obtaining informed consent from the respondents did
ULS-8 total scores and that the SCS total score would the survey begin. The study followed the Declaration of
be significantly negatively correlated with the DDI total Helsinki and was approved by the Ethics Committee of
score. Another hypothesis was that the SCS had mea- Jilin International Studies University (approval num-
surement equivalence in diverse age groups. ber: JY202211003). There was a total of 3686 valid data
obtained for the study (see Table 1). Among them, there
Method were 1458 emerging adults, 1124 middle-aged people,
Procedure and participants and 1104 older people.
No consensus has been reached by different researchers
on the criteria used to classify the age ranges of emerging Instrument
adults, middle-aged people, and older people. In a review Self-concealment scale (SCS)
study, it was noted that the majority of researchers cited There are 10 items comprising the SCS, which is struc-
35 to 65 years of age and 40 to 60 years of age as the two tured as a one-dimensional scale [3]. Item example “If
most widely used criteria for classifying the age range of I shared all my secrets with my friends, they’d like me
middle-aged people [27]. Moreover, the National Bureau less.” The scale is rated on a 5-point scale ranging from
of Statistics of China classifies the population aged 65 1 “strongly disagree” to 5 “strongly agree”. The scale is
and above as older adults. Therefore, this study divides rated on a 5-point scale from 1 “strongly disagree” to
the population aged 34 and below into emerging adults, 5 “strongly agree”. All items are positively scored. The
Fan et al. BMC Psychiatry (2024) 24:594 Page 4 of 10

Table 1 Socio-demographic characteristics

Emerging Adults Middle-aged Adults Older Adults
(N=1458) (N=1124) (N=1104)
N % N % N %
Gender
Male 434 29.77 523 46.53 532 48.19
Female 1024 70.23 601 53.47 572 51.81
Age (Mean, SD) 21.74 (3.10) 46.68 (4.54) 74.05 (5.49)
Learning Phase
Graduate student 469 32.17 - - - -
Undergraduate 989 67.83 - - - -
Education
Bachelor’s degree & above - - 217 19.31 44 3.99
College - - 148 13.17 39 3.53
High school or technical secondary school - - 285 25.36 139 12.59
Junior high school - - 331 29.45 281 25.45
Elementary school - - 143 12.72 601 54.44
Marital status
Unmarried 1421 97.46 40 3.56 9 0.82
Married 37 2.54 1072 89.77 984 89.13
Divorced or widowed 0 0 12 1.07 111 10.05
Ethnic group
Han majority 1323 90.74 1009 89.77 1004 90.94
Ethnic minority 135 9.26 115 10.23 100 9.06
Only child or not
Only Child 741 50.82 287 25.53 171 15.49
Non-only Child 717 49.18 837 74.47 933 84.51

higher the total score, the higher the tendency of the Short forms of the social interaction anxiety scale (SIAS)
individual to hide his/her negative feelings, thoughts, and Social interaction anxiety and social phobia are the two
information from others. dimensions that make up social anxiety. The SIAS con-
sists of 6 items scored on a 5-point scale [30]. Take “Ner-
Distress disclosure index (DDI) vous mixing with people when don’t know well.” as an
To use the DDI to assess an individual’s propensity to example, each item is positively scored. The higher the
conceal psychological distress [28]. The scale consists of total score, the higher the level of anxiety experienced
12 items and is scored on a 5-point scale. An example of by the individual in social interactions. The Cronbach’s α
an item is “When I am in a bad mood, I talk about it with coefficient for the scale in this study was 0.89.
my friends.” where items 2, 4, 5, 8, 9, and 10 are reverse
scored. The higher the total score, the stronger the indi- Short forms of the social phobia scale (SPS)
vidual’s willingness to reveal psychological distress. The The social phobia is one of the dimensions that make up
Cronbach’s alpha coefficient for the scale in this study social anxiety. The SPS consists of 6 items scored on a
was 0.83. 5-point scale [30]. An item example is “Worry head will
shake or nod in front of other people”. Each item is posi-
Revised cheek and buss shyness scale (RCBS) tively scored. The higher the total score, the higher the
The level of shyness in individuals was measured by the individual’s fear of social interaction. The Cronbach’s α
RCBS [29]. The scale consists of 13 items and is scored coefficient for the scale in this study was 0.90.
on a 5-point scale. An example of an item is “If I shared
all my secrets with my friends, they’d like me less.” Items UCLA loneliness scale (ULS-8)
3, 6, 9, and 12 are reverse scored, and the rest are positive The ULS-8 is a shortened version of the ULS-20, and the
scored. The higher the total score, the higher the level of Chinese version of the ULS-8 has demonstrated good
shyness experienced by the individual in interpersonal reliability and validity [31]. The ULS-8 consists of 8 items
interactions. The Cronbach’s α coefficient for the scale in and is scored on a 4-point scale. An item example is
this study was 0.84. “People are around me but not with me.” Among them,
items 3 and 6 are reverse-scoring questions. The higher
the total score, the stronger the perceived loneliness of
Fan et al. BMC Psychiatry (2024) 24:594 Page 5 of 10

the individual. The Cronbach’s α coefficient for the scale Since there is a correlation among the items, this study
in this study was 0.79. conducted oblique rotation using the Promax method
(oblique rotation) [35].
Kessler psychological distress scale (K10) For validity testing, the study conducted confirmatory
The K10 is widely recognized as a tool for assessing the factor analysis, and criterion-related validity. In con-
mental health status of people of all ages [32]. The scale firmatory analysis being estimated by maximum likeli-
consists of 10 items rated on a 5-point scale. For example, hood, the criterion for model fitting well include that χ2/
the item “About how often did you feel so restless you df < 3, RMSEA < 0.08, SRMR < 0.05, CFI、IFI、TLI > 0.90,
could not sit still?” is positively scored. The higher the PNFI、PCFI > 0.50 [36]. As to the reliability test, Cron-
total score, the more psychological distress the individual bach’s α coefficient value, McDonald’s Omega coefficient,
experiences and the worse the mental health status. The and the split-half reliability value of the SCS were cal-
Cronbach’s α coefficient for the scale in this study was culated. If these are greater than 0.70, the SCS has good
0.93. reliability [37]. To measure the equivalence of SCS across
gender and age groups, the study compared four nested
Statistical analysis models: Configural Invariance model (M1), Weak Invari-
To examine the differentiation of the items, the study ance model (M2), Strong Invariance model (M3), and
undertook an independent samples t-test to analyze the Strict Invariance model (M4) [38]. Compare the differ-
differences between the items in the high and low sub- ence in CFI and RMSEA between M2 and M1, M3 and
groups. In this case, the top 27% of the total score was M2, and M4 and M3 in turn. If both ∆CFI and ∆RMSEA
defined as a high subgroup, and the bottom 27% was are less than 0.01, then equivalence holds.
defined as a low subgroup. Moreover, to examine the
homogeneity of the items, the study utilized the item- Results
total correlation and Cronbach’s α coefficient test. If the Item analysis
correlation coefficient between the items and the total The results of the item analysis showed (see Table 2)
score is lower than 0.40, or if there is an increase in the that the scores of the items were significantly higher in
Cronbach’s α coefficient value of the scale after the dele- the high subgroups than in the low subgroups (t = 13.67–
tion of an item, then it means that the homogeneity is 18.93), and the values of the correlation coefficients of
poor and should be deleted [33]. Based on previous stud- the items with the total scores ranged from 0.68 to 0.75.
ies, the data were randomly divided into two groups, one Furthermore, the Cronbach’s α coefficient value for the
group of data for exploratory factor analysis (n = 562) and SCS was 0.895. After deleting any of the items, the Cron-
the other group of data for confirmatory factor analysis bach’s α coefficient values for the scale ranged from 0.882
(n = 562) [34]. In exploratory factor analysis, the maxi- to 0.887. During the item analysis phase, all data met the
mum likelihood method was employed for factor extrac- criteria for inclusion and none of the items were deleted.
tion. Statisticians recommend the orthogonal rotation
method when it is assumed or found that the items are Exploratory factor analysis
not correlated, while oblique rotation methods should Before conducting exploratory factor analysis, the KMO
be used when there is a correlation between the items. value of SCS was examined in the study. The results

Table 2 Item analysis results of SCS

Item Sample Low High t value Correlation Coefficients Cronbach’s α coefficient test
(N=1124) subgroup(N=303) subgroup(N=303)
M SD M SD M SD
Item 1 2.67 1.15 1.55 0.67 3.66 0.84 33.99*** 0.74*** 0.884
Item 2 2.32 0.99 1.52 0.62 3.09 0.97 23.72*** 0.68*** 0.887
Item 3 2.81 1.16 1.76 0.84 3.71 0.87 28.19*** 0.69*** 0.887
Item 4 2.36 1.03 1.49 0.58 3.19 0.96 26.41*** 0.70*** 0.886
Item 5 2.72 1.13 1.67 0.80 3.62 0.82 29.66*** 0.69*** 0.887
Item 6 2.53 1.09 1.52 0.56 3.52 0.88 33.19*** 0.75*** 0.883
Item 7 2.64 1.11 1.62 0.71 3.61 0.85 31.17*** 0.74*** 0.883
Item 8 2.68 1.11 1.64 0.67 3.68 0.82 33.65*** 0.75*** 0.882
Item 9 2.45 1.05 1.50 0.58 3.38 0.92 30.23*** 0.75*** 0.882
Item 10 2.59 1.10 1.63 0.72 3.47 0.96 26.70*** 0.69*** 0.887
Note. ***p<0.001
Abbreviations: M, Mean; SD, Standard Deviation
Fan et al. BMC Psychiatry (2024) 24:594 Page 6 of 10

Table 3 Results of exploratory factor analysis of SCS (N=1124)

Items Factor loading Communality
1. I have an important secret that I haven’t shared with anyone. 0.65 0.42
2. If I shared all my secrets with my friends, they’d like me less. 0.62 0.39
3. There are lots of things about me that I keep to myself. 0.63 0.40
4. Some of my secrets have really tormented me. 0.63 0.40
5. When something bad happens to me, I tend to keep it to myself. 0.69 0.48
6. I’m often afraid I’ll reveal something I don’t want to. 0.71 0.51
7. Telling a secret often backfires and I wish I hadn’t told it. 0.71 0.50
8. I have a secret that is so private I would lie if anybody asked me about it. 0.73 0.53
9. My secrets are too embarrassing to share with others. 0.71 0.51
10. I have negative thoughts about myself that I never share with anyone. 0.66 0.44

Table 4 Criterion validity test for SCS (N=1124)

1 2 3 4 5 6 7
1.SCS -
2.DDI 0.30*** -
3.RCBS 0.45*** -0.30*** -
4.SIAS 0.54*** -0.26*** 0.69*** -
5.SPS 0.51*** -0.28*** 0.71*** 0.85*** -
6.ULS-8 0.39*** -0.34*** 0.56*** 0.58*** 0.61*** -
7.K10 0.55*** -0.31*** 0.57*** 0.67*** 0.70*** 0.66*** -
Mean 25.76 38.75 31.77 11.39 15.21 15.55 21.89
SD 7.83 7.60 8.23 4.18 5.80 4.19 7.70
Note. **p<0.01;
Abbreviations: SCS, Self-Concealment Scale; DDI, Distress Disclosure Index; RCBS, Revised Cheek and Buss Shyness Scale; SIAS, Social Interaction Anxiety Scale; SPS,
Social Phobia Scale; ULS-8, UCLA Loneliness scale; K10, Kessler Psychological Distress Scale; SD, standard deviation

showed that the KMO value was 0.920 and Bartlett’s to 0.55), except for the significant negative correlation
sphericity test value was 2340.01 (df = 45, p < 0.001). The between SCS and DDI (r=-0.30).
results indicated that the data met the prerequisites for
exploratory factor analysis of SCS. In further explor- Reliability test of SCS
atory factor analysis (see Table 3), only one factor with The Cronbach’s α coefficient value, McDonald’s Omega
an eigenvalue greater than 1 explained 51.01% of the total coefficient, and split-half reliability were included as
variance. The results of the scree plot showed that it was indicators of SCS reliability. The results showed that the
appropriate to extract one factor. In addition, the factor Cronbach’s α coefficient value of SCS was 0.895, McDon-
loading values for all items ranged from 0.62 to 0.73 and ald’s Omega coefficient was 0.893, and the split-half reli-
the commonality ranged from 0.39 to 0.53 (see Table 2). ability value was 0.861, all of which were higher than the
The factor loading values for all items were above 0.40 standard of 0.70.
and the commonality was above 0.30 [39]. Therefore, no
items were removed from the exploratory factor analysis Equivalence test and analysis of variance of SCS in different
summary. age groups
In the four models constructed in this study, the Con-
Confirmatory factor analysis figural Invariance model (M1), Weak Invariance model
The results of the study showed that the fit indices of (M2), Strong Invariance model (M3), and Strict Invari-
the one-factor model were χ2/df = 3.194, SRMR = 0.031, ance model (M4), all the fit indices were good, except
RMSEA = 0.063, CFI = 0.971, IFI = 0.971, TLI = 0.959, for the χ2/df which was higher than 3 (see Table 5). This
PNFI = 0.682, PCFI = 0.690. result fulfills the prerequisites for performing the test
of equivalence. In addition, ∆CFI (-0.008 to -0.001) and
The criterion-related validity test of SCS ∆RMSEA (-0.002 to 0.003) were less than 0.01 in the
Pearson correlation analysis was used to examine the comparisons between M2 and M1, M3 and M2, and M4
criterion-related validity of the SCS. The results showed and M3. All four models were valid, and SCS was equiva-
(see Table 4) that SCS was significantly and positively lent in different age groups.
correlated with all validity criteria instruments (r = 0.39 On this basis, the study examined age differences
in self-concealment. The results showed significant
Fan et al. BMC Psychiatry (2024) 24:594 Page 7 of 10

Table 5 Results of equivalence test of SCS

Model χ2 χ2/df CFI IFI TLI SRMR RMSEA ∆CFI ∆RMSEA
(90%CI)
M1 713.39 7.93 0.962 0.963 0.944 0.036 0.043
(0.040 ∼ 0.046)
M2 747.44 6.92 0.961 0.962 0.952 0.038 0.040 -0.001 -0.003
(0.037 ∼ 0.043)
M3 899.10 7.02 0.953 0.964 0.951 0.038 0.040 -0.008 0.0003
(0.038 ∼ 0.043)
M4 1018.16 6.44 0.948 0.948 0.956 0.041 0.038 -0.005 -0.002
(0.036 ∼ 0.041)
Abbreviations: M1, Configural Invariance model; M2, Weak Invariance model; M3, Strong Invariance model; M4, Strict Invariance model

variations in self-concealment among emerging adults Working Model proposed by Larson et al., it is stated
(M = 26.43, SD = 7.96), middle-aged adults (M = 25.77, that self-concealment is a complex trait-like motivational
SD = 7.79), and older adults (M = 25.69, SD = 7.91) construct and insecure attachments, trauma incidence,
(F = 3.55, p = 0.029). Post hoc test analysis revealed that and social-evaluative concerns and social-evaluative
self-concealment was significantly higher in emerging concerns are its potential precursors [2]. Therefore, it is
adults than in middle-aged (p = 0.034) and older peo- unclear whether a unidimensional construct is optimal
ple (p = 0.018), while the difference in self-concealment for the SCS and whether the SCS reflects the full range
between middle-aged and older people was not signifi- of self-concealment. Moreover, the scale only assessed
cant (p = 0.810). the individual’s propensity for self-concealment and did
not address the antecedent and outcome variables of
Discussion self-concealment. More empirical evidence is needed to
The objectives of this study were to analyze the psycho- validate the dimensional structure of the SCS more pru-
metric characteristics of the Self-Concealment Scale dently in the future. Concurrently, researchers may also
(SCS) using data collected from a population of middle- consider revising the SCS again by incorporating the lat-
aged Chinese adults. In addition, the study was con- est research findings on self-concealment. By modifying,
ducted to analyze the cross-age invariance of the SCS by deleting, and adding items, the scale items correspond
testing the equivalence of the scale across three samples: more to the characteristics of contemporary middle-aged
emerging adults, middle-aged adults, and older adults. adults and also respond more comprehensively to the
The results of the analysis of the reliability, validity, and definition of self-concealment.
equivalence of the SCS were satisfactory. The findings
suggest that the Chinese version of the SCS is concise, Criterion validity analysis
reliable, and valid, which can be used to assess self-con- The results of the criterion validity analyses demon-
cealment tendencies in middle-aged adults. strated that self-concealment can affect an individual’s
social interactions and psychological well-being. Specifi-
Factor structure analysis cally, individuals with high self-concealment tended to
The study first examined the quality of the SCS items. have higher levels of shyness, loneliness, social anxiety,
Item analyses showed that all items measured the same and social phobia, were less likely to reveal their distress
construct and discriminated well between subjects with to others, and had worse mental health. Self-conceal-
different levels of self-concealment. In this regard, the ment has been found to lead to negative outcomes such
study retained all the items for factor analysis. The find- as increased psychological distress, deteriorated men-
ings of the exploratory factor analysis and the confirma- tal health, and disrupted interpersonal interactions in
tory factor analysis supported the idea that the SCS is a most previous studies [40–42]. Moreover, in the Work-
unidimensional construct. The SCS was found to have ing Model of self-concealment, it is also pointed out that
a unidimensional structure in the English version, the the most direct harm caused by self-concealment is the
Turkish version, and the Chinese version of the scale [3, impact on the individual’s interpersonal behavior and
8, 10]. In terms of the way the scale is divided into dimen- mental health [2]. As the result of this study, the Working
sions and the number of entries, the present study con- Model is validated and expanded. In addition, grounded
cluded that it is consistent with previous studies. in resource preservation theory, research has viewed self-
However, notably, in Larson and Chastain’s definition concealment as a stressor that is considered to have a
of self-concealment, it is considered not as a unidimen- destructive effect on an individual’s social support, which
sional structure, but as a multidimensional structure can result in both the depletion of resources and the
consisting of three components [3]. Furthermore, in the acquisition of effective resources [43]. The preliminary
Fan et al. BMC Psychiatry (2024) 24:594 Page 8 of 10

validation of this viewpoint by the findings of the present self-concealment to explain its harms, which provides
study contributes to a better understanding of how self- new perspectives for related research. Notably, this study
concealment affects mental health from the perspective is the first to examine cross-cohort consistency in SCS
of resource preservation theory. among emerging adults, middle-aged adults, and older
adults, which enriches and adds to the existing literature.
Cross-group invariance analysis
The present study is the first to examine the measure- Strengths
ment equivalence of the SCS across age groups. The fact The present study encompasses practical value. Self-
that the scale has equivalence across cohorts is a prereq- concealment is prevalent for all ages and negatively
uisite for the analysis of variance between group com- affects individuals’ wellbeing in the aspect of psychol-
parisons. The study found that SCS has measurement ogy, emotion, and behavior. The SCS represents an effec-
equivalence across young, middle-aged, and older age tive tool for assessing self-concealment in middle-aged
groups. On this basis, the study further examined age dif- adults, which can be applied to various areas of scientific
ferences in self-concealment. It was found that middle- research, mental health education, and psychotherapy.
aged and older adults had a significantly higher tendency For concealing certain specific information, self-conceal-
to conceal than emerging adults. Previous studies have ment is significantly higher in middle-aged adults than in
found that individuals of different ages do not conceal other age groups [44]. In this regard, the SCS is suitable
different types of information to the same extent [44]. to identify and evaluate the self-concealment tendency
For instance, O’Connor et al. found that individuals’ con- of middle-aged people and to screen key populations
cealment of COVID-19 information was decreasing with in need of intervention. Furthermore, there is evidence
increasing age, and their attitudes toward concealment that self-concealment adversely affects the establish-
were more negative [45]. The present study reached dis- ment of the counseling relationship and therapeutic out-
crepant conclusions. The reason for this may be analyzed comes in psychotherapy [49]. In both family therapy and
because the present study mainly examined individuals’ couples therapy, it is emphasized that visitors should be
general tendency to self-conceal rather than conceal- encouraged to disclose secrets with the aim of reduc-
ment of specific information. Compared to emerging ing self-concealment, thereby promoting intimacy [50].
adults, middle-aged and older adults are more concerned Therefore, psychotherapists can employ the scale revised
and worried about the potential negative consequences in this study to evaluate therapeutic outcomes.
of information disclosure, fearing that it will affect their
image [46]. Therefore, middle-aged and older adults have Limitations and future research
a higher propensity for self-concealment. Although the Chinese version of the Self-Concealment
Some theoretical significance is attached to this study. Scale (SCS) was found to encompass favorable applica-
The study analyzed the reliability and validity of the SCS bility in the middle-aged population in this study. Never-
in a middle-aged Chinese population, which provided theless, some limitations may impact the reliability of the
empirical evidence for evaluating the psychometric char- study’s findings. First, the study did not adopt a strictly
acteristics of the Chinese version of the scale. At present, random sampling approach, but rather convenience sam-
when researchers revise the different language versions pling and snowball sampling. Limitations in the research
of the SCS, it is mainly done based on the youth popu- design could reduce the representativeness of the sam-
lation. As can be seen from the analysis of the existing ples and the accuracy of the findings. In prospective stud-
literature, this study is the first to examine the validity ies, on the one hand, a more rigorous sampling approach
of the SCS in a non-Western middle-aged population. could be adopted; on the other hand, further testing of
In addition, the findings enrich the literature on self- the validity of SCS could be considered among partici-
concealment. Previous studies have mainly explored the pants with different demographic characteristics, such
negative effects of self-concealment on emerging adults as different places of residence, economic incomes, occu-
and older adults, while less attention has been paid to pations, health statuses, ethnicities, and countries. Sec-
middle-aged adults [47, 48]. Moreover, the study’s analy- ondly, the study adopted a cross-sectional questionnaire
sis of the criterion validity not only helps to better under- which could not reflect the stability and validity of the
stand the effects of self-concealment on the interpersonal scale across time.
and psychological health of middle-aged adults but also Third, in the criterion validity test, we focused on the
is an effective expansion of the theory of resource pres- effects of self-concealment on individuals’ interpersonal
ervation. The study views self-concealment as a stressor and psychological well-being. It has been noted that self-
that can lead to the depletion of individual resources and concealment is strongly associated with attachments,
impede the acquisition of valuable resources. The study trauma incidence, social-evaluative concerns, emo-
introduces resource preservation theory to the field of tion regulation, mindfulness, psychological flexibility,
Fan et al. BMC Psychiatry (2024) 24:594 Page 9 of 10

on study design and participated in data analysis. ZF wrote the main content
help-seeking attitudes, and negative health behaviors [2]. of the manuscript up. ZF and XS contributed to manuscript preparation and
To this end, a more comprehensive examination of the revision. LL contributes to the request for funding of the study. All authors
antecedent and outcome variables of SCS is required in read and approved the final manuscript.

future research. Fourth, the SCS was only used to mea- Funding
sure an individual’s general propensity for self-conceal- This study was under the Key Research and Development Program of the
ment and did not address what kind of information and Jilin Provincial Department of Science and Technology (No. 20230203104SF).
Social science project of Cultural and Tourism Department of Jilin Province
content was being concealed. For instance, individuals (2024KT046).
may not conceal the information about sexual orienta-
tion, abuse experiences, illness, and negative evaluations Data availability
The data are not publicly available due to privacy or ethical restrictions. If there
of self to the same degree. Therefore, in ensuing research, is a reasonable request, it can be obtained from the corresponding author on
self-concealment scales could be specifically designed to reasonable request.
assess specific information.
Declarations
Conclusion
Ethics approval and consent to participate
As shown in the obvious study results, the Chinese ver- This study followed the Helsinki Declaration, and the research was
sion of the Self-Concealment Scale (SCS) was identified approved by the Ethics Committee of Jilin International Studies University
to be a reliable and valid scale for assessing the ten- (JY202211003). All participants received informed consent to participate in
the study.
dency to conceal painful information and negative feel-
ings among middle-aged adults. Self-concealment is Consent for publication
able to undermine interpersonal interactions, increase Not applicable.

loneliness, and threaten mental health status. In addi- Competing interests

tion, the SCS demonstrates measurement equivalence The authors declare no competing interests.
across age groups, facilitating the comparison of age dif-
Received: 22 January 2024 / Accepted: 29 July 2024
ferences. Considering that self-concealment is higher in
middle-aged and older adults than in emerging adults, it
is pivotal to improve the assessment, intervention, and
research on self-concealment in those adults to minimize
References
its potential harm. Therefore, the Chinese version of the 1. Love M, Farber B. Honesty in psychotherapy: results of an online survey
SCS is suitable for large-scale investigations and studies comparing high vs. low self-concealers. Psychother Res. 2018;29:1–14.
of self-concealment in middle-aged adults. 2. Larson DG, Chastain RL, Hoyt WT, Ayzenberg R. Self-concealment: integrative
review and working model. J Soc Clin Psychol. 2015;34:705–14.
Abbreviations 3. Larson D, Chastain R, Self-Concealment. Conceptualization, measurement,
SCS Self-Concealment Scale and Health implications. J Soc Clin Psychol. 1990;9:439–55.
SD Standard Deviance 4. Omori M. Japanese college students’ attitudes toward professional psycho-
DDI Distress Disclosure Index logical services: the role of cultural self-construal and self-concealment.
ULS-8 UCLA Loneliness Scale Psychol Rep. 2007;100:387–99.
BIPQ Brief Illness Perception Questionnaire 5. Chinweuba DC, Ifeagwazi CM, Chinweuba AU, Chukwuorji JC. Does self-
RCBS Revised Cheek and Buss Shyness Scale concealment and self-compassion differentially influence substance use for
SIAS Social Interaction Anxiety Scale male versus female adolescents? J Subst Use. 2023;28:117–22.
SPS Social Phobia Scale 6. Mendoza H, Tully EC, Goodnight B, Gray J, Masuda A. The indirect effect of
ULS-8 UCLA Loneliness Scale self-concealment on distress through psychological inflexibility in Asian
K10 Kessler Psychological Distress scale American, black American, and White American college students. Pers Indiv
EFA Exploratory Factor Analysis Differ. 2018;126:93–8.
CFA Confirmatory Factor Analysis 7. Wallace BC, Constantine MG. Africentric Cultural values, psychological
RMSEA Root Mean Square Error of Approximation help-seeking attitudes, and Self-Concealment in African American College
CFI Comparative Fit Index Students. J Black Psychol. 2005;31:369–85.
IFI Incremental Fit Index 8. Deniz M, Cok F. Psychometric properties and Adaptation of the self-conceal-
TLI Tucker-lewis Index ment scale to the Turkish adolescents. lköretim Online. 2010;9:424–32.
PNFI Parsimonious Normed Fit Index 9. Salinas Oñate N, Baeza Rivera MJ, Salinas Rehbein B, Escandón Nagel N,
Pcfi Parsimonious Comparative Fit Index Escobar BA. Validación De La versión Adaptada De La Escala De Auto-Oculta-
SRMR Standardized Root Mean Square Residual miento De Larson & Chastain en universitarios chilenos. Revista Iberoameri-
M1 Configural Invariance Model cana De diagnóstico Y evaluación psicológica. 2022;1:51–65.
M2 Weak Invariance model 10. Wang C. Self-concealment tendency of middle school students: factor struc-
M3 Strong Invariance model ture and development characteristics. J Appl Psychol. 2002;:15–7.
M4 Strict Invariance model 11. Fan Z, Shi X, Hu C, Zhu L, Wang Z. Reliability and validity of self-concealment
scale in Chinese older adults. Psychol Res Behav Manag. 2023;16:4341–52.
Acknowledgements 12. Qin Q-Y, Li S-W, Li W, Li Y-R. The mediating role of cognitive emotion regula-
We thank the participants who helped us with this study. tion in the relationship between self-concealment and quality of life among
breast cancer chemotherapy patients. J Clin Psychol. 2023.
Author contributions 13. Sease TB, Perkins DR, Sandoz EK, Sudduth H. Automatic thoughts: under-
ZF, XS, XH, and LL conceived and designed the study. ZF, XS, and XH are standing the precursors of self-concealment within the psychological flex-
responsible for data collection. ZF, XS, XH, and LL provided statistical advice ibility framework. J Context Behav Sci. 2021;22:68–73.
Fan et al. BMC Psychiatry (2024) 24:594 Page 10 of 10

14. Nan C, Tsai C-TL. The relationships between Leisure Time, Perceived stress and 34. Fung S-F, Fung ALC. Development and evaluation of the psychometric prop-
subjective well-being in Young and Middle-aged working adults in China. erties of a brief parenting scale (PS-7) for the parents of adolescents. PLoS
Leisure Sci. 2022;0:1–19. ONE. 2020;15:e0228287.
15. Guo L. The delayed, durable effect of expressive writing on depression, anxi- 35. Worthington RL, Whittaker TA. Scale Development Research: a content analy-
ety and stress: a meta-analytic review of studies with long-term follow-ups. Br sis and recommendations for best practices. Couns Psychol. 2006;34:806–38.
J Clin Psychol. 2023;62:272–97. 36. Wolf MG, McNeish D. Dynamic: an R Package for deriving dynamic fit Index
16. Davis CG. Self-concealment, secrecy, and guilt. J Pers. 2023. Cutoffs for Factor Analysis. Multivar Behav Res. 2023;58:189–94.
17. Denmark AB, Hess E, Becker M. College Students’ reasons for concealing 37. Viladrich C, Angulo-Brunet A, Doval E. A journey around alpha and omega to
suicidal ideation. J Coll Student Psychother. 2012. estimate internal consistency reliability. Anales De Psicología. 2017;33:755–82.
18. Williams CJ, Cropley M. The relationship between perfectionism and engage- 38. Dimitrov DM. Testing for factorial invariance in the context of construct
ment in preventive health behaviours: the mediating role of self-conceal- validation. Meas Evaluation Couns Dev. 2010;43:121–49.
ment. J Health Psychol. 2014;19:1211–21. 39. Carpenter S. Ten steps in Scale Development and Reporting: A Guide for
19. Cruddas S, Gilbert P, Mcewan K. The relationship between Self-Concealment Researchers. Communication Methods Measures. 2018;12:25–44.
and Disclosure, early experiences, attachment, and Social Comparison. Int J 40. Carvalho SA, Carvalho F, Fonseca L, Santos G, Castilho P. Beyond the centrality
Cogn Therapy. 2012;5:28–37. of shame: how self-concealment and fear of receiving Compassion from
20. Hobfoll SE. The influence of culture, community, and the nested-self in the others Impact Psychological suffering in Transgender adults. J Homosex.
stress process: advancing conservation of resources theory. Appl Psychology: 2022;0:1–19.
Int Rev. 2001;50:337–70. 41. Kealy D, Rice SM. Dispositional connectedness and depressive symptoms
21. Wang C. Self-concealment tendency of middle school students: factor struc- among help-seeking outpatients: the mediating role of self-concealment. Eur
ture and development characteristics. J Appl Psychol. 2002;2:15–7. +7. J Psychiatry. 2020;34:219–22.
22. ten Brummelhuis LL, Bakker AB. A resource perspective on the work–home 42. Uysal A, Lin HL, Bush AL. The reciprocal cycle of self-concealment and trust in
interface: the work–home resources model. Am Psychol. 2012;67:545–56. romantic relationships. Eur J Social Psychol. 2012;42:844–51.
23. Sease TB, Perkins DR, Yang Y. The influence of emotion dysregulation on the 43. Hobfoll SE, Halbesleben J, Neveu J-P, Westman M. Conservation of
relationship between emotional abuse and self-concealment among roman- resources in the organizational context: the reality of resources and their
tic partners. Journal of Aggression, Maltreatment & Trauma; 2021. consequences. Annual Rev Organizational Psychol Organizational Behav.
24. Leigh E, Clark DM. Understanding social anxiety disorder in adolescents and 2018;5:103–28.
improving treatment outcomes: applying the cognitive model of Clark and 44. Chasteen A, Bashir N, Gallucci C, Visekruna A. Age and antiaging technique
Wells (1995). Clin Child Fam Psychol Rev. 2018;21:388–414. influence reactions to Age Concealment. Journals Gerontol Ser B Psychol Sci
25. Hagger BF, Riley GA. The social consequences of stigma-related self-conceal- Social Sci. 2011;66:719–24.
ment after acquired brain injury. Neuropsychol Rehabil. 2019;29:1129–48. 45. O’Connor AM, Evans AD. Dishonesty during a pandemic: the concealment of
26. Akdoğan R, Çimşir E. Linking inferiority feelings to subjective happiness: COVID-19 information. J Health Psychol. 2022;27:236–45.
self-concealment and loneliness as serial mediators. Pers Indiv Differ. 46. Costa PA, Alberto Ribeiro-Gonçalves J, Gomes G, Romeu I. Coming out expe-
2019;149:14–20. riences and Disclosure gap in three age cohorts of Portuguese cisgender
27. Merriam S. Middle Age: a review of the literature and its implications for sexual minority men. Sex Res Soc Policy. 2022;19:1662–673.
Educational intervention. Adult Educ. 1978;29:39–54. 47. Smetana J, Robinson J, Bourne SV, Wainryb C. I didn’t want to, but then I told:
28. Li X. Research on the relationship between adult attachment, self-disclosure adolescents’ narratives regarding disclosure, concealment, and lying. Dev
and depression of College Students. Master. Hebei Normal University; 2009. Psychol. 2019;55:403–14.
29. Xiang B, Ren L, Zhou Y, Liu J. Psychometric Properties of Cheek and Buss Shy- 48. Lou Y, Liu J, Carr D. To tell or not: Chinese older adults’ preferences for disclos-
ness Scale in Chinese College Students. Chin J Clin Psychol. 2018;26:4. ing or concealing Serious illness diagnoses. Gerontologist. 2023;63:1049–59.
30. Zhong J, Liu J, Xu G, Zheng H, Bowker J, Coplan RJ. Measurement invariance 49. Drinane JM, Owen J, Tao KW. Cultural concealment and therapy outcomes. J
of two different short forms of Social Interaction anxiety scale (SIAS) and Couns Psychol. 2018;65:239–46.
Social Phobia Scale (SPS) in Chinese and US samples. Eur J Psychol Assess. 50. Deslypere E, Rober P. Family Secrecy in Family Therapy Practice: an explorative
2023;39:14–20. Focus Group Study. Fam Process. 2020;59:52–65.
31. Xu S, Qiu D, Hahne J, Zhao M, Hu M. Psychometric properties of the short-
form UCLA loneliness scale (ULS-8) among Chinese adolescents. Med
(Baltim). 2018;97:e12373. Publisher’s Note
32. Ren Q, Li Y, Chen D-G. Measurement invariance of the Kessler Psychological Springer Nature remains neutral with regard to jurisdictional claims in
Distress Scale (K10) among children of Chinese rural-to-urban migrant work- published maps and institutional affiliations.
ers. Brain Behav. 2021;11:e2417.
33. Gehlbach H, Brinkworth ME. Measure twice, cut down error: a process for
enhancing the validity of survey scales. Rev Gen Psychol. 2011;15:380–7.