Psychometric Testing of the Chinese version of the Nursing Home Adjustment Scale: Classical theory Test and Item Response Theory

doi:10.21203/rs.3.rs-1732152/v1

Download PDF

Research Article

Psychometric Testing of the Chinese version of the Nursing Home Adjustment Scale: Classical theory Test and Item Response Theory

https://doi.org/10.21203/rs.3.rs-1732152/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

China has the largest number of older people in the world, thus implying a considerable potential demand for relocation to long-term care facilities. It is critical to developing effective measurement instruments for assessing the adaptation to nursing homes among older adults in China. The nursing home adjustment scale (NHAS) was designed for assessing nursing home adjustment and has been translated into Chinese recently. However, the evidence for the psychometric properties of the Chinese version is insufficient, with the lack of validation in less developed regions of China and item response theory (IRT)-based analysis.

Methods

The English version of the NHAS was translated into Chinese according to the standard procedures. A total of 761 older adults who lived in 21 nursing homes in Hunan Province, China, were investigated from July 2021 to December 2021. The psychometric characteristics of the Chinese version of the NHAS (C-NHAS) were examined based on the classical test theory (CTT) and item response theory (IRT).

Results

The overall Cronbach’s α coefficient of the C-NHAS was 0.913. The 4-factor solution (CFI = 0.958, TLI = 0.951, and RMSEA = 0.049) was regarded as applicable to the C-NHAS: depressed mood, emotional stress, relationship development, and sense of belonging. Rasch analysis showed the C-NHAS was multidimensional. The information-weighted fit statistic (infit) and outlier-sensitive fit statistic (outfit) mean square (MNSQ) values of the subscales of C-NHAS were 0.64–1.42 and 0.64–1.46, indicating all items showed an acceptable model fit. No items displayed substantial differential item functioning (DIF) across gender. In the Wright map, the difficulty of items matched most individuals’ ability, while a small number of individuals with the low latent trait had no items for matching.

Conclusions

In summary, the C-NHAS is a reliable and valid instrument that can assist in measuring nursing home adjustment among Chinese seniors. Rasch analysis results support the multidimensional and DIF-free across gender of the C-NHAS. The difficulty of the NHAS may restrict the measurement of adjustment, thus further development and testing of a brief version assessing adjustment in the nursing home population is worth considering.

Nursing home adjustment

Long-term care

Psychometric testing

Validation

Classical theory test

Item response theory

Rasch analysis

As life expectancy increases, the number of older people is growing inevitably, and the aging of the population has become a global trend. China has also been experiencing a rapid pace of the population aging process and has the largest older population in the world. According to China’s Seventh National Population Census in 2020, older people (≥ 60 years old) constitute around 18.70% of the total population, representing a 5.44% increase compared to the rate in 2010 [1]. The growing number of older people is a major challenge for the country's social security system and public services such as health care; meanwhile, healthy aging, proposed by the World Health Organization (WHO) as a process of maintaining functional ability to enable wellbeing in older age has gradually become a world consensus [2, 3].

Older adults typically experience a decline or loss of self-care ability due to chronic diseases, age-related frailty, etc. In addition, with the rapid process of industrialization, urbanization, and educational expansion in contemporary society, the family structure is experiencing a dramatic transition from the conventional nuclear family model to the pluralistic and decentralized family model, gradually reducing the amount of care that older people receive in their home [4–6]. A previous study based on the Chinese Longitudinal Healthy Longevity Survey (CLHLS) showed that as the population size of the Chinese oldest-old continues to expand, the demand for long-term care will increase accordingly [7]. It is foreseeable that more long-term care facilities will be generated and increasing numbers of older people, especially those who are disabled, empty nests or living alone will choose to reside in them for receiving professional care services. For older people, transitioning from their familiar homes to nursing homes is a profoundly influential life event with uncertainty and stress. Because it not only means the conversion of aging from home care to institutional care but also represents the mutual adaptation of older people to the new living environment, such as the disruption of the old social relationships and the formation of new ones [8]. In addition, several studies have reported the emergence of negative emotions such as anxiety, depression, loneliness, apprehension, and increased confusion in older people when relocating to long-term care facilities [9–11].

Nursing home adjustment can be defined as a process of overcoming discomfort, regulating negative emotional and behavioral reactions, and eventually forming a new concept of collective life and self-orientation [12, 13]. According to the transition process framework presented by Brandburg, the transition process consists of four consecutive components: initial reaction, transitional influences, adjustment, and acceptance [14]. Among them, the adjustment process is more likely to be intervened and critical in preventing adverse transitional influences associated with relocation stress. Identifying the predictors of nursing home adjustment helps to provide effective interventions, and to further promote better acceptance of older adults residing in the nursing home. As a foundation, it is imperative to develop appropriate measurement instruments. To directly assess nursing home adjustment, the nursing home adjustment scale (NHAS) was designed based on the Korean language by Lee in 2007 [15]. As one of the most used measurement instruments in this field, it has been utilized in some studies among older people residing in nursing homes in South Korea [12, 16–18]. In 2015 and 2019, the original scale was translated, validated, and finalized in both English and Chinese versions respectively [19, 20].

To date, researchers have primarily employed classic test theory (CTT) to assess the psychometric properties of the entire NHAS [15, 19, 20], which ignored the characteristics of individual responses to a single item. In contrast, as a typical representation of modern testing theory (MTT), item response theory (IRT) has been applied for reanalysis in numerous psychometric instruments in recent years [21–23]. Unlike CTT, IRT assesses the properties of each item and thus does not rely on a combination of items and allows for generalizing testing results across samples and items [24, 25]. According to IRT, a series of IRT models that can assist in determining the psychometric properties of instruments have been developed. In IRT models, there are three decisive parameters being assumed to influence the probability of achieving a specific score from the individual and the accuracy of the measurement; and correspond to the following three models: the one-parameter (1PL; ability) or Rasch model, two-parameter logistic model (2PL logistic; ability and discrimination), and the three-parameter (3PL Logistic; ability, discrimination, and guessing). In addition, the latent trait of the subject was based on the parameters of each item. Depending on the assumptions, the probabilistic relationship between response behaviors and the latent trait can also be described for each item using a monotonously increasing function, the so-called item response function (IRF) [26, 27]. Despite the advantages in certain aspects, there are still no studies to examine the psychometric properties of the NHAS based on IRT. Of note, compared to CTT, IRT also has shortcomings, for example, the limitation in the exploration of the factor structure of the entire scale [28, 29]. Therefore, a combination of the two methods can provide a more comprehensive evaluation of the psychometric properties of an instrument.

The objective of this study is to psychometrically test the Chinese version of the Nursing Home Adjustment Scale (C-NHAS) using CTT and IRT among older people living in long-term care facilities in central China, and to provide further evidence on the improvement of the applicability for the Chinese older adults.

2.1 Translation procedure

We contacted the author of the original scale by email and have obtained permission from the original author to translate the English version of the NHAS. According to standardized international guidelines [30, 31], this study conducted an independent translation of the original English version of the NHAS into Chinese to ensure robust linguistic equivalency. First, after obtaining permission from the original author, two bilingual translators whose mother tongue was Chinese conducted independent forward translations of the NHAS scale from English to Chinese. One translator had a doctoral degree in social medicine and was knowledgeable about psychological terminology. The other was a nurse practitioner with a master's degree. Then the two translated versions were checked and synthesized into an interim Chinese version by the first author and a professional translator. Second, the interim Chinese version was independently backward translated by two native English translators who were proficient in Chinese and had no knowledge of the original English version of the NHAS before translation. Third, an expert panel including a language professional, a psychiatrist, a public health professional, and a clinical nurse specialist reviewed all the translated and original NHAS items to investigate the aspects of cross-cultural equivalency and consistency until a pre-final version of the Chinese NHAS was generated. Fourth, after the original author reviewed and approved the backward translated version, a pilot test was conducted using the pre-final version in 32 residents living in nursing homes in China to confirm its clarity of Chinese culture-adjusted expression and readability of contents. This translation procedure did not delete or add any items from the original version, only adding a few words or making small grammatical changes to adapt it to the Chinese culture.

2.2 Participants and process

From July 2021 to December 2021, with the support of the Civil Affairs Bureau in central China, 761 residents from 21 nursing homes located in three cities (Changsha-Zhuzhou-Xiangtan urban agglomeration, CZX) in Hunan Province, China were recruited by a non-probability, convenience purposive sampling procedure [32]. Each resident who agreed to participate was invited for an interview in a private room in the nursing home. Trained research staff then read each statement of the questionnaire aloud and the residents were asked to respond verbally. Their responses were recorded objectively to complete the questionnaires. Finally, a total of 761 participants (response rate of 95.5%) living in these nursing homes were investigated. Two weeks later, 38 (5.0%) participants were randomly selected and completed the questionnaires again to evaluate test-retest reliability. All participants meeting the following explicit criteria were included in this study: (1) ≥ 60 years of age; (2) voluntarily agreed to participate; (3) ability to read or speak Chinese and (4) healthy enough to participate in a conversation. The exclusion criteria were as follows: (1) cognitive impairment or mental retardation (MMSE < 18; the Mini-Mental State Examination[33]); (2) history of major psychiatric disorders or substance abuse. Written or oral informed consent was obtained from all participants. This study was performed in agreement with the Declaration of Helsinki ethics requirements and was approved by the Ethics Committee of the Faculty of Central South University (XYGW-2021-64).

2.3 Measures

2.3.1 The Nursing Home Adjustment Scale (NHAS)

The original version of the NHAS was initially published in Korean to measure the adjustment of nursing home residents, then the English version was generated by the author and confirmed to be reliable and validated (Cronbach’s α = 0.770)[15, 19]. This scale consists of 23 items which have been divided into five dimensions: emotional distress, difficulty fitting in, relationship development, acceptance of new residence, and depressed mood. Each item is rated on a 5-point Likert scale from 0 = strongly disagree to 5 = strongly agree (some items are scored reversely), providing a total score of the NHAS from 23 to 115, with 115 representing the best nursing home adjustment.

2.3.2 The 15-item Geriatric Depression Scale (GDS-15)

The GDS-15 is a short form of the 30-item GDS made for the identification of depression in older individuals [34, 35]. This scale consists of 15 binary questions (several items are reverse scored) in which participants are asked to answer how they felt over the past week (1 = Yes, 0 = No); higher total scores indicate more severe depressive symptoms. It has been translated into Chinese and both validity and reliability were validated as satisfactory among Chinese older people in previous studies with a good internal consistency (Cronbach’s α = 0.745) [36–38].

2.3.3 The General Self-Efficacy Scale (GSES)

The GSES was developed by Schwarzer et al. and has been translated into Chinese to measure an individual’s self-efficacy in coping with daily hassles or stressful events[39, 40]. It consists of 10 items, each of which is rated on a 4-point Likert scale that ranges from 1 (not at all) to 4 (exactly true), and a higher score indicates higher self-efficacy of the individual. The Chinese version of the GSES showed sufficient reliability (Cronbach's α was 0.920) and factorial validity [41, 42].

2.3.4 Demographic and Sociological questionnaire

The questionnaire was used to collect residents’ demographic information regarding age, gender, educational background, race, length of stay, income sources, and activities of daily living (ADL). The ADL function was measured by the Chinese version of the Modified Barthel Index (MBI-C), which has shown to be robust in validity and reliability across the original and Chinese populations [43].

2.4 Statistical analyses

2.4.1 Descriptive statistics

The continuous variable (e.g., age, item scores) was presented as mean ± standard deviation (SD), and categorical variables (e.g., gender, marital status, education) were expressed using frequency and percentage. In addition, the skewness and kurtosis of the item scores in the C-NHAS were also calculated.

2.4.2 CTT analysis

2.4.2.1 Reliability

Internal consistency of the NHAS was assessed by Cronbach’s α coefficient (criterion value ≥ 0.70) and item-total correlation (corrected for overlap; criterion value, Spearman correlation coefficient (r_s) ≥ 0.30) [44, 45]. With a subsample of 38 participants who was retook 2 weeks later, the intraclass correlation (ICC) was used to represent the test-retest reliability for the NHAS; a value greater than 0.70 of ICC indicates good test-retest reliability [46].

2.4.2.2 Validity

Factorial validity was evaluated by exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). The total sample was divided into two sets (sample A, n = 380 and sample B, n = 381) randomly to ensure that EFA and CFA were conducted independently. Before conducting EFA, the Kaiser-Meyer-Olkin Test (KMO) and Bartlett's test of sphericity were performed to assess correlation matrix and sample adequacy which were recommended to evaluate the suitability for factor analysis. The value of the KMO index greater than 0.7 is considered acceptable, and the value of Bartlett's test of sphericity less than 0.05 is considered significant [47]. For EFA, the principal component analysis (PCA) method with varimax rotation was applied to determine the scale's factor structure; a scree plot with eigenvalues > 1 and factor loadings > 0.40 were extracted and analyzed [48]. Then the factor structure obtained from EFA was tested via CFA. Modification indices were evaluated to obtain a stronger model. The model was rearranged by adding covariance between error terms. For CFA, the mean- and variance-adjusted weighted least squares (WLSMV) estimator was employed, as it is superior to other estimation methods for Likert-type rating scales [49]. To choose the optimal model from a series of alternative factor solutions, several available models were compared based on the following model fit indices: the chi-square test of model fit (P > 0.05), comparative fit index (CFI), Tucker Lewis Index (TLI), root mean square error of approximation (RMSEA), and the Akaike information criterion (AIC). The values of CFI and TLI greater than 0.90 are acceptable, while a value less than 0.08 is considered acceptable for RMSEA. The value for the AIC should be smaller to obtain the most parsimonious model fit [50, 51]. As for convergent validity of the NHAS, since the factor “depressed mood” is a construct of the original NHAS scale [19] and the association of self-efficiency with nursing home adjustment was reported in previous studies [16, 52, 53], we examined the r_s between the total and subscale scores of the NHAS and those of the GDS-15 and the GSES. A value of r_s of less than 0.1 represents a weak correlation, correlations between 0.30 and 0.50 represent a small correlation; between 0.30 and 0.50 represent moderate correlation, and more than 0.5 represent a strong correlation.

2.4.3 IRT analysis

Once the factorial structure of the scale was determined, the matrix of single raw scores for each subject of the entire scale and subsequently of each subscale were analyzed using the Rasch model [54, 55]. Opting for the Rasch rating scale model, the unidimensionality, item difficulty, item validity, item, and person separation index, and item and person separation reliability of the NHAS was investigated. Since the unidimensionality of latent traits is an important prerequisite for Rasch analysis, a PCA of standardized residuals, with a combination of criteria (variance explained by the measures to be greater than 40%, unexplained variance in the first factor to be less than 15%, and eigenvalue of the first construct to be less than 3.0) was employed for a dimension test of the scale [56]. The item difficulty was presented using the log-odds unit (i.e., logit, a higher logit represents a more difficult item). Item validity was examined using information-weighted fit statistic (infit) mean square (MNSQ) and outlier-sensitive fit statistic (outfit) MNSQ with a recommended range between 0.5 and 1.5 suggested a good fit; the closer to 1, the better the fit [57]. The point-measure correlation (PT-measure) indicated the correlation between individual performance on a single item and that on the whole scale, with a value ranging from 0.40 to 0.80 is acceptable. The person reliability index (PRI) and item reliability index (IRI) were used to indicate the consistency of the scale and person; a value of > 0.7 indicates an acceptable model fit. The person separation index (PSI) and item separation index (ISI) were used to estimate the spread of items and persons, and to verify whether the scale was sensitive enough to distinguish different levels of sample in this study; a value of > 2.5 indicates good model fit [58]. A person-item map was generated to reflect the joint distribution of the latent trait and the item difficulties. An alignment between items and persons and a normal distribution of persons in this map is considered acceptable for the quality of the scale. In addition, differential item functioning (DIF) in the Rasch models was used to test measurement invariance across gender at the item level. DIF contrast greater than 0.5 logits and the absolute value of a T value greater than 2 indicated substantial DIF [59].

SPSS 26.0 software (IBM Corporation, Armonk, NY, USA) and Mplus 8.3 software (Muthén & Muthén, Los Angeles, CA, USA) were used to conduct CTT analyses. Rasch analyses were carried out via Winsteps® software version 3.72.3 (https://www.winsteps.com).

3.1 Sociodemographic characteristics

A total of 761 individuals living in 21 nursing homes in Hunan Province, China participated in this study. The Han Chinese population accounted for the majority of participants (n = 730; 95.9%). The mean age of the participants was 78.3 years (SD = 9.3). More than half of the participants were females (n = 469; 61.6%) and widowed (n = 511; 67.1%). The specific characteristics of the participants were summarized in Table 1.

Table 1

Characteristics of nursing home participants (N = 761)
Characteristics	Frequency (%)	Mean ± SD
Age (years)		78.3 ± 9.3
60~	159 (20.9)
70~	218 (28.6)
80~	305 (40.1)
90~	79 (10.4)
Gender
Male	292 (38.4)
Female	469 (61.6)
Race
Han ethnic	730 (95.9)
Ethnic minority	31 (4.1)
Marital status
Unmarried	43 (5.7)
Married	191 (25.1)
Divorced	16 (2.1)
Widowed	511 (67.1)
Educational background
Illiteracy	187 (24.6)
Elementary school	223 (29.3)
Middle school	157 (20.6)
High school	109 (14.3)
College and above	85 (11.2)
Length of stay (months)
< 3	63 (8.3)
3~	122 (16.0)
6~	199 (26.2)
12~	377 (49.5)
ADL (MBI)		71.8 ± 26.1
≤ 20 (Totally dependent)	52 (6.8)
25–45 (Very dependent)	90 (11.8)
50–70 (Partially dependent)	129 (17.0)
75–95 (Minimally dependent)	257 (33.8)
100 (Independent)	233 (30.6)

Abbreviations: ADL, Activities of Daily Living; MBI, Modified Barthel Index; SD, Standard Deviation.

3.2 CTT analysis

3.2.1 Item properties

Table 2 illustrates the psychometric properties of all the C-NHAS items. For the distributional properties (mean, SD, skewness, and kurtosis), all the items were not extremely skewed, with the mean of each item ranging from 3.286 to 3.860. The item properties in the factor loadings (0.490–0.779) and corrected item-total correlation (0.454–0.751) were also presented.

Table 2

The item properties and internal consistency data of the C-NHAS.
Items#	Mean (SD)	Skewness/ Kurtosis	Factor loading^a	Corrected Item-total correlation	Test–retest reliability (ICC)	Cronbach’ α if item deleted
#1 I feel close with my roommates and/or neighbors	3.621(0.817)	-0.785/0.295	0.540	0.454	0.840	0.912
#2 I become angry easily	3.714(1.180)	-0.628/-0.623	0.729	0.506	0.791	0.912
#3 I get annoyed with trivial things often	3.531(0.980)	-0.536/-0.453	0.589	0.478	0.825	0.912
#4 I do not like to participate in group activities with others	3.442(0.973)	-0.269/-0.725	0.549	0.624	0.740	0.909
#5 This is not my home but just a temporary residence	3.305(0.997)	-0.175/-0.744	0.631	0.596	0.783	0.910
#6 I am not friendly with the other residents	3.443(0.918)	-0.732/-0.172	0.605	0.574	0.788	0.910
#7 Living here makes me feel at ease	3.488(0.884)	-0.597/-0.385	0.499	0.626	0.812	0.909
#8 I want to make friends here	3.476(0.969)	-0.296/-0.677	0.589	0.612	0.853	0.909
#9 I feel living here boring	3.404(0.985)	-0.501/-0.560	0.685	0.526	0.754	0.911
#10 I try to help other residents	3.286(0.930)	-0.373/-0.801	0.535	0.515	0.739	0.911
#11 I am satisfied with living here	3.571(0.963)	-0.579/-0.251	0.550	0.655	0.814	0.908
#12 Daily living has no meaning for me	3.690(1.256)	-0.659/-0.571	0.525	0.629	0.732	0.909
#13 It’s difficult to get along with other residents	3.374(0.892)	-0.472/-0.514	0.631	0.619	0.825	0.909
#14 I don’t want to live here, although it’s impossible to return home	3.411(0.971)	-0.268/-0.657	0.518	0.650	0.726	0.908
#15 My life has value in this world	3.606(0.925)	-1.060/0.665	0.490	0.572	0.821	0.910
#16 I want to have a good life here	3.542(0.923)	-0.535/-0.368	0.653	0.509	0.762	0.911
#17 I have accepted and confirmed to live here	3.414(1.009)	-0.088/-1.094	0.735	0.613	0.728	0.909
#18 I have several close friends here	3.521(0.873)	-0.633/-0.026	0.537	0.632	0.833	0.909
#19 I feel lonely most of the time	3.587(1.145)	-0.479/-0.773	0.730	0.460	0.774	0.913
#20 It is painful for me to think about my children (family)	3.860(1.671)	-0.944/-0.905	0.566	0.601	0.809	0.913
#21 I feel like crying most of the time	3.667(1.024)	-0.561/-0.490	0.777	0.740	0.816	0.906
#22 I often shed tears for no reason	3.762(1.018)	-0.697/-0.164	0.730	0.713	0.814	0.907
#23 I often think about my previous home and that makes me sad	3.646(1.121)	-0.488/-0.832	0.779	0.751	0.822	0.906
Cronbach’s α of the whole scale = 0.913

Abbreviations: SD, Standard Deviation; ICC, Intraclass Correlation Coefficient.

Note:^aBased on exploratory factor analysis.

3.2.2 Reliability

The overall Cronbach’s α coefficient of the C-NHAS was 0.913 and the Cronbach’s α coefficients if item deleted ranged from 0.906 to 0.913, indicating good internal consistency reliability of the whole scale. The test-retest reliability for the total scale was also good, with an ICC within the range of 0.726–0.853 (Table 2).

3.2.3 Construct validity

There was no significant difference in the sociodemographic characteristics of sample A and sample B (see Additional file 1). In sample A, the KMO and Bartlett’s test of sphericity suggested the appropriateness for factor analysis (KMO = 0.887; χ² = 9173.380, P < 0.001). Thus, an EFA was conducted (principal components extraction with varimax rotation), and four factors were extracted with eigenvalues of 4.66 to 2.85, which together explained 62.09% of the overall variance (Table 3). The 4-factor model proposed by the EFA was different but thematically consistent with the original 5-factor model. The Cronbach’s α coefficient of each factor structure still showed satisfactory internal consistency (0.826–0.872). For the subsequent CFA in sample B, the 4-factor model and two previous 5-factor models presented by Xiao et al. [20] and Lee et al. [19] were examined respectively. As shown in Table 4, the 4-factor model of the C-NHAS indicated the relatively better model fitness and was supported by the CFA fit indices (CFI = 0.915, TLI = 0.906, and RMSEA = 0.060). Thus, the 4-factor structure of the C-NHAS was considered the most appropriate in this study. In the final 4-factor model (Fig. 1), the original two factors: difficulty fitting in and relationship development, were grouped into one dimension and named “relationship development”; the names of other factors were also minorly modified. In addition, considering the length of stay may act as a potential systematic error affecting mood changes and relationship development, two pairs of residual correlations (#2 with #19 and #8 with #13) were set to improve the model fitness (CFI = 0.958, TLI = 0.951, and RMSEA = 0.049).

Table 3

Exploratory factor analysis of the C-NHAS.
Items# (Cronbach’s α/no. of items)	Factor loadings
Items# (Cronbach’s α/no. of items)	F1	F2	F3	F4
Factor 1. Relationship development (0.872/8)
#16 I want to have a good life here	0.821
#6 I am not friendly with the other residents	0.740
#1 I feel close with my roommates and/or neighbors	0.712
#10 I try to help other residents	0.709
#4 I do not like to participate in group activities with others	0.657
#13 It’s difficult to get along with other residents	0.614
#8 I want to make friends here	0.613
#15 My life has value in this world	0.608
Factor 2. Depressed mood (0.858/5)
#21 I feel like crying most of the time		0.799
#23 I often think about my previous home and that makes me sad		0.790
#22 I often shed tears for no reason		0.771
#20 It is painful for me to think about my children(family)		0.694
#12 Daily living has no meaning for me		0.642
Factor 3. Sense of Belonging (0.826/6)
#17 I have accepted and confirmed to live here			0.804
#5 This is not my home but just a temporary residence			0.741
#18 I have several close friends here			0.622
#11 I am satisfied with living here			0.591
#14 I don’t want to live here, although it’s impossible to return home			0.586
#7 Living here makes me feel at ease			0.548
Factor 4. Emotional stress (0.841/4)
#19 I feel lonely most of the time				0.843
#2 I become angry easily				0.831
#9 I feel living here boring				0.787
#3 I get annoyed with trivial things often				0.731

Note: KMO = 0.887, Bartlett’s test was significant (χ² = 5867.088, df = 253, P＜0.001); Factor loadings less than 0.5 was omitted.

Table 4

Confirmatory factor analysis of the C-NHAS.
Model	χ²	Df	CFI	TLI	RMSEA (90% CI)	AIC
4-factor model	1271.074	224	0.915	0.906	0.060 (0.056, 0.064)	52611.472
5-factor model^a	1536.466	220	0.889	0.874	0.072 (0.068, 0.076)	54216.654
5-factor model^b	1360.676	220	0.907	0.901	0.063 (0.059, 0.067)	53087.836
Adjusted 4-factor model^c	872.548	222	0.958	0.951	0.049 (0.045, 0.053)	52074.677

Abbreviations: Df, Degree of freedom; CFI, Comparative Fit Index; TLI, Tucker Lewis Index; RMSEA, Root Mean Square Error of Approximation; AIC, Akaike Information Criterion.

Note: ^aBased on previous study conducted by Lee et al.; ^b Based on previous study conducted by Xiao et al.; ^cAdjusted via setting two pairs of residual correlations (#2 with #19 and #8 with #13).

Regarding convergent validity (Table 5), as expected, the C-NHAS showed substantial positive and negative correlation coefficients with GSES scores (r_s = 0.440; P < 0.01) and GDS-15 scores (r_s = -0.416; P < 0.01). Especially, stronger correlations were observed between the “relationship development” subscale and GSES scores, and between the “depressed mood” subscale and GDS-15 scores.

Table 5

Convergent validity of the C-NHAS.
C-NHAS subscales	GDS-15	GSES
Relationship development	-0.102*	0.497**
Depressed mood	-0.522**	0.153*
Emotional stress	-0.309**	0.118*
Sense of Belonging	-0.098	0.241**
Total scores	-0.416**	0.440**

Abbreviations: GDS-15, The 15-item Geriatric Depression Scale; GSES, The General Self-Efficacy Scale; C-NHAS, the Chinese version of the Nursing Home Adjustment Scale.

Note: *P < 0.05; **P < 0.01 (bilateral).

3.3 IRT/Rasch analysis

3.3.1 Unidimensionality

Results of the PCA of standardized residuals confirmed the presence of multi-dimensionality in the entire C-NHAS (eigenvalue of the first construct > 3.0), while for the three subscales of the C-NHAS, the percentage of variance explained by the measures ranged from 53.5–60.7%, the unexplained variance (10.4–14.1%) in the first construct was all below the criterion of 15%, and the eigenvalue of the first construct ranged from 1.1 to 1.5 (Table 6). These findings supported that the C-NHAS subscales were unidimensionality and absence of item dependence, thus all subscales were suitable for conducting Rasch analysis.

Table 6

Characteristics of the C-NHAS and its subscales based on Rasch analysis.
	C-NHAS	C-NHAS Subscales
		Emotional stress	Depressed mood	Relationship development	Sense of Belonging
Person separation index	3.31	2.75	2.68	2.59	2.55
Person reliability index	0.92	0.88	0.74	0.83	0.80
Item separation index	3.52	3.17	3.41	2.67	3.23
Item reliability index	0.96	0.93	0.98	0.87	0.95
Variance explained by measures (%)	42.3	58.4	60.7	53.5	57.9
unexplained variance in the first factor (%)	11.6	12.7	10.4	14.1	11.9
Eigenvalue of the first construct	4.3	1.2	1.1	1.5	1.4

Abbreviations: C-NHAS, the Chinese version of the Nursing Home Adjustment Scale.

3.3.2 Model Fitness and Difficulty

The infit and outfit MNSQ values of the C-NHAS were 0.64–1.42 and 0.64–1.46, indicating all items showed acceptable fitness to the Rasch model, and no items needed to be dropped in the subsequent analyses. Likewise, the PT-measure of each item exhibited a good correlation between individual performance in a certain item and that in the whole scale. The correlation values tend to be high, ranging from 0.45 to 0.71, and were all close to the expectations of the model. Besides, the C-NHAS showed moderate difficulty, with the difficulty of the items ranging from − 0.55 to 0.44 logits (Table 7).

Table 7

Rasch analysis for each item of the C-NHAS.
		Infit		Outfit		PT-Measure		DIF contrast across gender
Item#	Difficulty	MNSQ	ZSTD	MNSQ	ZSTD	CORR	EXP	Measure	T
Factor 1. Relationship development
#6	-0.21	0.75	-1.50	0.84	-0.80	0.63	0.57	0.09	1.11
#16	-0.22	0.91	-1.10	0.91	-0.90	0.52	0.57	-0.07	-0.86
#1	-0.55	0.92	-2.00	0.94	-1.30	0.45	0.52	-0.06	-0.67
#10	0.23	0.75	-1.90	0.78	-1.30	0.64	0.58	-0.07	-0.91
#4	0.31	0.81	-1.10	0.86	-1.70	0.61	0.58	0.00	0.00
#13	-0.55	1.28	1.60	1.25	0.90	0.58	0.57	0.02	0.27
#8	0.27	0.84	-1.40	0.83	-1.70	0.61	0.57	-0.02	-0.28
#15	-0.42	0.93	-0.20	0.92	-0.60	0.54	0.58	0.00	0.00
Factor 2. Depressed mood
#21	-0.17	0.77	-1.90	0.76	-2.10	0.70	0.65	0.00	0.00
#23	0.19	0.99	1.60	1.05	2.40	0.51	0.58	-0.07	-0.88
#22	-0.20	0.86	0.10	0.83	-0.60	0.58	0.54	-0.07	-0.80
#20	0.38	0.70	-0.60	0.67	-1.70	0.60	0.59	-0.16	-1.92
#12	0.06	0.77	-2.20	0.81	-0.60	0.63	0.57	-0.02	-0.25
Factor 3. Sense of Belonging
#17	-0.04	0.73	-2.10	0.73	-1.90	0.66	0.58	0.14	1.66
#5	0.44	0.84	-0.90	0.90	0.80	0.61	0.59	0.05	0.63
#18	-0.18	0.64	-1.80	0.64	-1.30	0.63	0.57	0.05	0.59
#11	0.12	1.16	1.50	1.21	1.80	0.62	0.61	0.00	0.00
#14	0.12	0.87	-2.20	0.86	-1.40	0.54	0.60	-0.02	-0.30
#7	0.25	0.81	-0.40	0.76	-1.60	0.56	0.58	0.00	0.00
Factor 4. Emotional stress
#19	0.00	1.18	-1.30	1.12	-0.70	0.66	0.55	-0.09	-1.02
#2	0.08	1.42	1.10	1.46	1.40	0.53	0.56	0.41	2.10
#9	0.00	0.91	0.20	0.92	0.60	0.71	0.56	0.00	0.00
#3	0.08	1.09	1.70	1.18	2.10	0.49	0.57	-0.28	-2.08

Abbreviations: MNSQ, Mean Square; ZSTD, Z-Standard; CORR, Correlation; EXP, Expected; PT-measure, The point-measure correlation; DIF, differential item functioning.

3.3.3 Reliability

The reliability and separation index values for person and item are presented in Table 6. For the whole C-NHAS, the item reliability index was 0.96 and the person reliability index was 0.92, and the results of each subscale of the C-NHAS were all stable, indicating that either item and response difficulties or the latent score for persons were all consistent. In addition, the ISI and PSI values were more than the criterion (2.5), indicating the good item and person discrimination.

3.3.4 Person-Item Map

A person-item map (i.e., the Wright map) was generated to explore the compatibility between participant ability and item difficulty. As shown in Fig. 2, the distributions of item difficulty and person ability were plotted in a logit ruler, the former is on the right side of the map, and the latter is on the left. The latent trait of individuals achieved approximatively normal distribution and the distribution of item difficulty was also concentrated, with the mean difference between them no greater than 0.5 logits. There were not extremely difficult or easy items. The difficulty of items matched most individuals’ ability, indicating it can properly distinguish the differences in the person's latent trait. However, there exist still a small number of individuals with the low latent trait that have no items for matching.

3.3.5 DIF

Based on an ordinal logistic regression model, the differences in the item parameters for gender groups were tested. As shown in Table 7, according to the values of DIF contrast and T values for each item, no items displayed substantial DIF between males and females. Noteworthy, although all the values of DIF contrast < 0.5, two items (#2 and #3) showed relatively higher DIF contrast values with T values > 2.

The NHAS was first translated from Korean to Chinese and validated among the older adults in Fujian Province, China in 2019 by Xiao et al. [20]. After that, this scale was employed in only one study conducted in Jiangsu Province, China [60]. Fujian and Jiangsu Province are both located in the southeast coastal region of China and are generally considered to be economically developed. China has vast territories; the unevenness of development levels in different regions of China is worthy of consideration. Specifically, the older people who reside in economically developed areas tend to possess higher socioeconomic status (SES), and higher SES has been recognized to be associated with social support, social participation, self-rated health, and psychological distress among older people [61, 62]. On the other hand, there are also differences in the dominant types of nursing homes in different regions. As being expected to provide comparatively higher levels of service and amenities, more for-profit nursing homes may be available in economically developed areas. In previous studies, these social- and economic-related factors have been demonstrated to be associated with nursing home adjustment in older adults [13, 63]. Thus, for the C-NHAS, the evidence on its applicability to the Chinese older people is insufficient. In this study, we carried out an independent translation based on the English version and examined its psychometric properties in a sample of nursing home residents located in an underdeveloped region of China as a compliment. Besides, to compensate for the shortcomings of CTT, we first combined the CTT and an IRT model to give a comprehensive assessment of the C-NHAS.

According to the results of CTT, the internal consistency of the C-NHAS was good, with a Cronbach’s α coefficient of 0.913 for the whole scale. For each item, when a single item is deleted, the Cronbach’s α coefficient did not change obviously, ranging from 0.906 to 0.913. No individual item or subscale significantly affected the overall internal consistency of the scale, showing that none of these items or subscales from the C-NHAS should be specifically dropped. Since the repeatability of this scale has not been tested for the Chinese version, we examined the test-retest reliability via ICC as a compliment, and the result also showed a good performance on the Chinese version. The reliability of our study was consistent with the previous study conducted by Xiao et al. (Cronbach’s α: 0.87) [20] and the Korean version (Cronbach’s α: 0.837) [15], but slightly stronger than the English version (Cronbach’s α: 0.77) [19]. The variation may be explained by the similarity and diversity of cultural backgrounds, indicating the applicability of the NHAS in East Asian cultures. In terms of structural validity, the results of EFA and CFA suggested that the factor structure for the 23 items was grouped into four factors and the 4-factor solution showed a relatively better model fit. This 4-factor structure is similar to the existing Chinese version translated by Xiao et al.[20], combining the two factors “acceptance of new residence” and “feeling at home” into a new factor called “sense of belonging”. Compared to the 5-factor structure presented in the original scale, it is consistent thematically and reflects a linear progression of adaptation. Unlike in the original scale, the three factors, “relationship development”, “difficulty fitting in” and “acceptance of new residence” were recategorized as two dimensions: “relationship development” and “sense of belonging”. This finding on factor structure emphasized the coherence of nursing home adjustment, suggesting that the factor “difficulty fitting in” represents the psychological characteristics of both “relationship development” and “sense of belonging”. In line with the conclusions from current qualitative studies, three themes from the personal level: “resilience of the older person”, “interpersonal connections and relationships” and “this is my new home” helps to shape a healthy transition for older people [64, 65]. In addition, as anticipated, the convergent validity of the C-NHAS with the GDS-15 and GSES was in the expected direction, adding the evidence regarding the factorial validity of the NHAS.

As supplementary, Rasch analysis provided additional evidence for the widespread applications of the NHAS in China. Based on the results from Rasch analysis, these findings strengthen the evidence that the NHAS has a multidimensional structure. For each structure explored through EFA, a good fit was also observed in the Rasch model, with the infit MNSQ and outfit MNSQ all between 0.5 and 1.5. The PRI and IRI, which are similar to the Cronbach’s α, indicated good consistency in the rank order of both item-response difficulty and individual’s NHAS scores. In addition, two items (#2 and #3) in the NHAS showed relatively higher DIF contrast values with T values > 2 (but not significantly), which indicated that the NHAS was equally effective across genders and probably not biased toward demographic characteristics. However, evidence on the gender invariance of measurements is still inadequate. Studies based on CTT are required for further confirmation. In the Wright Map, the difficulty of the items can be matched to most individuals' latent traits. The results of PSI and ISI also confirmed the NHAS to be effective in separating various levels of nursing home adjustment. Noteworthy, with a small number of unmatched individuals, the difficulty of the entire NHAS is likely to be slightly higher for older people, indicating that a brief version with fewer items is worth being developed.

This study has some limitations that should be considered. First, participants were recruited based on convenience sampling in one province of mainland China, which may limit the generalizability of the study results. Future studies are still needed to examine the psychometric properties of this scale with a more diverse sample. Second, some potential biases that need to be recognized equally. Although older people with cognitive impairment or dementia were excluded from our study, prevalent cognitive decline in older adults may response accuracy. Also, data collection mainly relied on the interview, and the common bias in self-report (e.g., social desirability and recall bias) could confound the findings. Third, the strength of evidence for convergent validity remains weak due to the lack of other appropriate Chinese measurement instruments for measuring nursing home adjustment, indicating that developing the corresponding measurement instruments based on the Chinese cultural context should be considered. Finally, our study showed relatively better applicability in East Asia based on the psychometric characteristics of three versions (Korean, Chinese, and English) of the NHAS. However, the strength of this evidence is quite weak. Thus, it is warranted to explore the measurement invariance or DIF issues across cultures for the NHAS to gain a clearer understanding in the subsequent study.

In summary, the C-NHAS is a reliable and valid instrument that can assist in measuring nursing home adjustment among Chinese older people. Rasch analysis results support the multidimensional and DIF-free across gender of the C-NHAS. The difficulty of the NHAS may restrict the measurement of adjustment, thus further development and testing of a brief version assessing adjustment in the nursing home population is worth considering.

NHAS: Nursing Home Adjustment Scale; C-NHAS: Chinese version of the Nursing Home Adjustment Scale; CTT: Classical Test Theory; IRT: Item Response Theory; GDS-15; the 15-item Geriatric Depression Scale; GSES: General Self-Efficacy Scale; ADL: Activities of Daily Living (ADL); ICC: Intraclass correlation; EFA: Exploratory Factor Analysis; CFA: Confirmatory Factor Analysis; KMO: Kaiser-Meyer-Olkin Test

PCA: Principal Component Analysis; CFI: Comparative Fit Index; TLI: Tucker Lewis Index; RMSEA: Root Mean Square Error of Approximation; AIC: Akaike Information Criterion; infit MNSQ: information-weighted fit statistic (infit) mean square; outfit MNSQ: outlier-sensitive fit statistic (outfit) mean square; PRI: Person Reliability Index; IRI: Item Reliability Index; PSI: Person Separation Index; ISI: Item Separation Index; DIF: Differential Item Functioning.

Ethics approval and consent to participate

This study protocol was approved by the Ethics Committee of the Faculty of Xiangya School of Public Health, Central South University (XYGW-2021-64). The oral informed consent was approved and confirmed as equally valid as written informed consent by the Ethics Committee of the Faculty of Xiangya School of Public Health, Central South University (XYGW-2021-64). Written or oral informed consent was obtained from all participants. All methods were performed in accordance with the relevant guidelines and regulations.

Consent for publication

Not applicable.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing interests

The authors declare that they have no competing interests.

Funding

Not applicable.

Authors’ contributions

YC-G, SJ-L, and HL-X conceived and designed this study. SJ-L, YC-G and Z-H contributed to the translation of the instrument. YC-G, FX-Z, SY-L and Z-H were responsible for data collection and inputting into the software. YC-G, FX-Z and SY-L conducted the data analysis. YC-G contributed to the interpretation of results and drafting of the manuscript. HL-X, SJ-L and YC-G revised the manuscript critically for important intellectual content. All authors read and approved the final manuscript.

Acknowledgments

We sincerely appreciate G.E. Lee’s support for our research. We thank all the participants and investigators for their efforts.

Main Data of the Seventh National Population Census [http://www.stats.gov.cn/english/PressRelease/202105/t20210510_1817185.html]
Beard JR, Officer A, de Carvalho IA, Sadana R, Pot AM, Michel JP, Lloyd-Sherlock P, Epping-Jordan JE, Peeters G, Mahanani WR et al: The World report on ageing and health: a policy framework for healthy ageing. Lancet (London, England) 2016, 387(10033):2145–2154.
Rudnicka E, Napierała P, Podfigurna A, Męczekalski B, Smolarczyk R, Grymowicz M: The World Health Organization (WHO) approach to healthy ageing. Maturitas 2020, 139:6–11.
Sharma R: The Family and Family Structure Classification Redefined for the Current Times. Journal of Family Medicine and Primary Care 2013, 2:306.
Roberto KA, Blieszner R: Diverse Family Structures and the Care of Older Persons. Canadian journal on aging = La revue canadienne du vieillissement 2015, 34(3):305–320.
Moral-Fernández L, Frías-Osuna A, Moreno-Cámara S, Palomino-Moral PA, Del-Pino-Casado R: The start of caring for an elderly dependent family member: a qualitative metasynthesis. BMC geriatrics 2018, 18(1):228.
Zhang L, Shen S, Guo Y, Fang Y: Forecasting Future Demand of Nursing Staff for the Oldest-Old in China by 2025 Based on Markov Model. International journal of health policy and management 2021.
Plattner L, Brandstötter C, Paal P: [Loneliness in nursing homes-Experience and measures for amelioration: A literature review]. Zeitschrift fur Gerontologie und Geriatrie 2022, 55(1):5–10.
Drageset J, Dysvik E, Espehaug B, Natvig GK, Furnes B: Suffering and mental health among older people living in nursing homes-a mixed-methods study. PeerJ 2015, 3:e1120.
Ellis JM: Psychological transition into a residential care facility: older people's experiences. Journal of advanced nursing 2010, 66(5):1159–1168.
Wu CS, Rong JR: Relocation experiences of the elderly to a long-term care facility in Taiwan: a qualitative study. BMC geriatrics 2020, 20(1):280.
Lee GE: Predictors of adjustment to nursing home life of elderly residents: a cross-sectional survey. International journal of nursing studies 2010, 47(8):957–964.
Yong B, Lin R, Xiao H: Factors associated with nursing home adjustment in older adults: A systematic review. International journal of nursing studies 2021, 113:103790.
Brandburg GL: Making the transition to nursing home life: a framework to help older adults adapt to the long-term care environment. Journal of gerontological nursing 2007, 33(6):50–56.
Lee GE: [Scale development of free nursing home-adjustment for the elderly]. Taehan Kanho Hakhoe chi 2007, 37(5):736–743.
Choi JY, Sok SR: Factors influencing the adaptation to skilled nursing facilities among older Korean adults. International journal of nursing practice 2015, 21(2):184–191.
Park S, Sok SR: Relation Modeling of Factors Influencing Life Satisfaction and Adaptation of Korean Older Adults in Long-Term Care Facilities. International journal of environmental research and public health 2020, 17(1).
Yoon JY: Relationships among person-centered care, nursing home adjustment, and life satisfaction: a cross-sectional survey study. International psychogeriatrics 2018, 30(10):1519–1530.
Lee GE, Yoon JY, Bowers BJ: Nursing home adjustment scale: a psychometric study of an English version. Quality of life research: an international journal of quality of life aspects of treatment, care and rehabilitation 2015, 24(4):993–998.
Xiao H, Yong B, Liu X, Lin Y: A psychometric evaluation of the Chinese version of the Nursing Home Adjustment Scale. Quality of life research: an international journal of quality of life aspects of treatment, care and rehabilitation 2019, 28(9):2535–2542.
Luo H, Andersson B, Tang JYM, Wong GHY: Applying Item Response Theory Analysis to the Montreal Cognitive Assessment in a Low-Education Older Population. Assessment 2020, 27(7):1416–1428.
Silverstein MW, Petri JM, Kramer LB, Weathers FW: An item response theory analysis of the PTSD checklist for DSM-5: Implications for DSM-5 and ICD-11. Journal of anxiety disorders 2020, 70:102190.
Martin M, Kosinski M, Bjorner JB, Ware JE, Jr., Maclean R, Li T: Item response theory methods can improve the measurement of physical function by combining the modified health assessment questionnaire and the SF-36 physical function scale. Quality of life research: an international journal of quality of life aspects of treatment, care and rehabilitation 2007, 16(4):647–660.
Lord FM: Applications of Item Response Theory to Practical Testing Problems: LAWRENCE ERLBAUM ASSCCIAATES; 1980.
Edelen MO, Reeve BB: Applying item response theory (IRT) modeling to questionnaire development, evaluation, and refinement. Quality of life research: an international journal of quality of life aspects of treatment, care and rehabilitation 2007, 16 Suppl 1:5–18.
De Ayala RJ: Journal of Educational Measurement 1993, 30(1):84–87.
Thomas ML: The value of item response theory in clinical assessment: a review. Assessment 2011, 18(3):291–307.
Prieto L, Alonso J, Lamarca R: Classical Test Theory versus Rasch analysis for quality of life questionnaire reduction. Health and quality of life outcomes 2003, 1:27.
Thomas ML: Advances in applications of item response theory to clinical assessment. Psychological assessment 2019, 31(12):1442–1455.
Beaton DE, Bombardier C, Guillemin F, Ferraz MB: Guidelines for the process of cross-cultural adaptation of self-report measures. Spine 2000, 25(24):3186–3191.
Sousa VD, Rojjanasrirat W: Translation, adaptation and validation of instruments or scales for use in cross-cultural health care research: a clear and user-friendly guideline. Journal of evaluation in clinical practice 2011, 17(2):268–274.
Jager J, Putnick DL, Bornstein MH: II. MORE THAN JUST CONVENIENT: THE SCIENTIFIC MERITS OF HOMOGENEOUS CONVENIENCE SAMPLES. Monographs of the Society for Research in Child Development 2017, 82(2):13–30.
Katzman R, Zhang MY, Ouang Ya Q, Wang ZY, Liu WT, Yu E, Wong SC, Salmon DP, Grant I: A Chinese version of the Mini-Mental State Examination; impact of illiteracy in a Shanghai dementia survey. Journal of clinical epidemiology 1988, 41(10):971–978.
Yesavage JA, Brink TL, Rose TL, Lum O, Huang V, Adey M, Leirer VO: Development and validation of a geriatric depression screening scale: a preliminary report. Journal of psychiatric research 1982, 17(1):37–49.
Yesavage JA, Sheikh JI: 9/Geriatric Depression Scale (GDS). Clinical Gerontologist 1986, 5(1–2):165–173.
Zhang C, Zhang H, Zhao M, Chen C, Li Z, Liu D, Zhao Y, Yao Y: Psychometric properties and modification of the 15-item geriatric depression scale among Chinese oldest-old and centenarians: a mixed-methods study. BMC geriatrics 2022, 22(1):144.
Zhang C, Zhang H, Zhao M, Liu D, Zhao Y, Yao Y: Assessment of Geriatric Depression Scale's Applicability in Longevous Persons based on Classical Test and Item Response Theory. Journal of affective disorders 2020, 274:610–616.
Lee H-cB, Chiu HFK, Kwok WY, Leung C-m, Kwong PK, Chung DWS: Chinese elderly and the GDS Short Form A preliminary study. Clinical Gerontologist 1994, 14:37–42.
Zhang J, Schwarzer R: Measuring optimistic self-beliefs: A Chinese adaptation of the General Self-Efficacy Scale. Psychologia: An International Journal of Psychology in the Orient 1995, 38:174–181.
Schwarzer R, Jerusalem M, Weinman J, Wright S, Johnston M: Generalized Self-Efficacy Scale. Measures in Health Psychology: A User's Portfolio Causal and control beliefs Windsor 1995.
Cheung SK, Sun SY: Assessment of optimistic self-beliefs: further validation of the Chinese version of the General Self-Efficacy Scale. Psychological reports 1999, 85(3 Pt 2):1221–1224.
Leung DY, Leung AY: Factor structure and gender invariance of the Chinese General Self-Efficacy Scale among soon-to-be-aged adults. Journal of advanced nursing 2011, 67(6):1383–1392.
Leung SO, Chan CC, Shah S: Development of a Chinese version of the Modified Barthel Index– validity and reliability. Clinical rehabilitation 2007, 21(10):912–922.
Aaronson N, Alonso J, Burnam A, Lohr KN, Patrick DL, Perrin E, Stein RE: Assessing health status and quality-of-life instruments: attributes and review criteria. Quality of life research: an international journal of quality of life aspects of treatment, care and rehabilitation 2002, 11(3):193–205.
Hobart JC, Cano SJ, Zajicek JP, Thompson AJ: Rating scales as outcome measures for clinical trials in neurology: problems, solutions, and recommendations. The Lancet Neurology 2007, 6(12):1094–1105.
Bartko JJ: The intraclass correlation coefficient as a measure of reliability. Psychological reports 1966, 19(1):3–11.
Fabrigar LR, Wegener DT, MacCallum RC, Strahan EJ: Evaluating the use of exploratory factor analysis in psychological research. Psychological Methods 1999, 4(3):272–299.
Streiner DL, Norman GR, Cairney J: Health measurement scales: A practical guide to their development and use, 5th ed. New York, NY, US: Oxford University Press; 2015.
Beauducel A, Herzberg PY: On the Performance of Maximum Likelihood Versus Means and Variance Adjusted Weighted Least Squares Estimation in CFA. Structural Equation Modeling: A Multidisciplinary Journal 2006, 13(2):186–203.
Hu Lt, Bentler PM: Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal 1999, 6(1):1–55.
Hooper D, Coughlan JP, Mullen MR: Structural equation modelling: guidelines for determining model fit. In: 2008.
Johnson BD, Stone GL, Altmaier EM, Berdahl LD: The relationship of demographic factors, locus of control and self-efficacy to successful nursing home adjustment. The Gerontologist 1998, 38(2):209–216.
Lee H, Lee HY, Lee J: [Predictors of facility adaptation in nursing home residents]. Journal of Korean Academy of Nursing 2009, 39(2):177–185.
Andrich D: A rating formulation for ordered response categories. Psychometrika 1978, 43(4):561–573.
Andersen EB: The Rating Scale Model. In: Handbook of Modern Item Response Theory. Edited by van der Linden WJ, Hambleton RK. New York, NY: Springer New York; 1997: 67–84.
Linacre JM: A user’s guide to Winsteps-Ministep: Rasch-model computer programs. Program manual 3.68.0. WINSTEPS. Chicago, Illinois; 2009.
Smith AB, Rush R, Fallowfield LJ, Velikova G, Sharpe M: Rasch fit statistics and sample size considerations for polytomous data. BMC medical research methodology 2008, 8:33.
Tennant A, Conaghan PG: The Rasch measurement model in rheumatology: what is it and why use it? When should it be applied, and what should one look for in a Rasch paper? Arthritis and rheumatism 2007, 57(8):1358–1362.
Shih C-L, Wang W-C: Differential Item Functioning Detection Using the Multiple Indicators, Multiple Causes Method with a Pure Short Anchor. Applied Psychological Measurement 2009, 33(3):184–199.
Sun C, Yu Y, Li X, Cui Y, Ding Y, Zhu S, Li X, Chen S, Zhou R: The factors of adaptation to nursing homes in mainland China: a cross-sectional study. BMC geriatrics 2020, 20(1):517.
Zhou S, Gao L, Liu F, Tian W, Jin Y, Zheng ZJ: Socioeconomic status and depressive symptoms in older people with the mediation role of social support: A population-based longitudinal study. International journal of methods in psychiatric research 2021, 30(4):e1894.
Achdut N, Sarid O: Socio-economic status, self-rated health and mental health: the mediation effect of social participation on early-late midlife and older adults. Israel journal of health policy research 2020, 9(1):4.
Brownie S, Horstmanshof L, Garbutt R: Factors that impact residents' transition and psychological adjustment to long-term aged care: a systematic literature review. International journal of nursing studies 2014, 51(12):1654–1666.
Fitzpatrick JM, Tzouvara V: Facilitators and inhibitors of transition for older people who have relocated to a long-term care facility: A systematic review. Health & social care in the community 2019, 27(3):e57-e81.
Freeman S, Bishop K, Spirgiene L, Koopmans E, Botelho FC, Fyfe T, Xiong B, Patchett S, MacLeod M: Factors affecting residents transition from long term care facilities to the community: a scoping review. BMC health services research 2017, 17(1):689.

No competing interests reported.

Additionalfile1.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Psychometric Testing of the Chinese version of the Nursing Home Adjustment Scale: Classical theory Test and Item Response Theory

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

1 Background

2 Methods

2.1 Translation procedure

2.2 Participants and process

2.3 Measures

2.3.1 The Nursing Home Adjustment Scale (NHAS)

2.3.2 The 15-item Geriatric Depression Scale (GDS-15)

2.3.3 The General Self-Efficacy Scale (GSES)

2.3.4 Demographic and Sociological questionnaire

2.4 Statistical analyses

2.4.1 Descriptive statistics

2.4.2 CTT analysis

2.4.2.1 Reliability

2.4.2.2 Validity

2.4.3 IRT analysis

3 Results

3.1 Sociodemographic characteristics

3.2 CTT analysis

3.2.1 Item properties

3.2.2 Reliability

3.2.3 Construct validity

3.3 IRT/Rasch analysis

3.3.1 Unidimensionality

3.3.2 Model Fitness and Difficulty

3.3.3 Reliability

3.3.4 Person-Item Map

3.3.5 DIF

4 Discussion

5 Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1