Psychometric evaluation of the Online Education Student Satisfaction Scale: a translation and validation study among Chinese nursing students

Abstract

Objective

To translate the Online Education Student Satisfaction Scale and evaluate its reliability and validity among Chinese nursing students.

Background

As online education continues to expand rapidly, understanding student satisfaction is essential for improving educational quality. However, there is a notable lack of assessment tools specifically designed to evaluate the satisfaction of nursing students in online learning environments in China.

Design

A cross-sectional study design.

Method

The Chinese version of OESSS adopted Brislin’s translation model and conducted expert consultations to validate the face validity and testing of the translation version. Reliability and validity were tested using exploratory factor analysis, confirmatory factor analysis, and internal consistency reliability.

Results

The Chinese version of the 28-item scale demonstrated good psychometric properties. Exploratory factor analysis identified five factors, accounting for 66.623% of the total variance. Confirmatory factor analysis showed excellent fit indices: CMIN/DF = 1.185, RMSEA = 0.025, NFI = 0.917, IFI = 0.986, TLI = 0.984, and CFI = 0.986. Reliability measures were also robust, with a total Cronbach’s α coefficient of 0.918, McDonald’s Omega of 0.910, and test–retest reliability of 0.916.

Conclusion

The Chinese version of Online Education Student Satisfaction Scale (C-OESSS) exhibits strong psychometric properties and is well-suited for measuring online education satisfaction among nursing students in China.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12912-025-03864-6.

Introduction

With the rapid development of information technology, the Internet has become an integral part of the global education system [1]. Especially in the post-epidemic era, driven by the network era, the new education model combining the Internet and traditional education has developed rapidly [2]. This model, as a flexible, convenient, and personalized learning approach, has permeated various educational fields, significantly transforming traditional educational methods. Online teaching has maintained rapid development, gradually forming a scientifically structured and well-regulated independent teaching system. It has become a “natural component and important support of higher education.” [3] Online education not only provides learning opportunities across regions and time zones but also offers students space for self-directed learning, interactive communication, and personalized education.

Medical education, as an important discipline ensuring people’s life and health safety, undertakes the key tasks of cultivating high-quality medical talents, promoting medical progress, and improving people’s health levels [4]. Nurses are important members of the medical team [5], responsible for providing basic care to patients, implementing treatment plans, and conducting health education. Nursing education offers systematic knowledge and skills training to nursing students, helping nurses master professional knowledge, clinical skills, and humanistic care, thereby enhancing the quality of nursing services and ensuring that patients receive timely, effective and high-quality care [6]. Nowadays, online education has been widely used. However, problems such as ‘insufficient interaction, outdated concepts and poor results’ still occur, which may reduce students’ motivation and participation in learning and affect their overall satisfaction with online education [7]. Therefore, how to effectively evaluate students’ learning experience and satisfaction with online education has become an urgent problem to be solved. The satisfaction with online education not only influences students’ learning motivation, academic performance, and course completion rates [8], but also has a profound impact on the ongoing optimization and development of educational models [3]. This is particularly important in the field of nursing education, where students must not only acquire solid theoretical knowledge but also develop strong clinical skills [9]. Whether online education can effectively meet their specific educational needs and expectations warrants further in-depth exploration.

Currently, there is a lack of assessment tools to measure student satisfaction with online education [10], particularly research instruments specifically designed for nursing students [11]. This makes it challenging to gain a systematic and comprehensive understanding of nursing students’ true experiences and satisfaction with online learning. Therefore, having a scientifically sound and reliable assessment tool to measure nursing students’ satisfaction with online education is of utmost importance. In 2024, Turkish scholar Harmancı Seren successfully developed the Online Education Student Satisfaction Scale (OESSS), demonstrating strong psychometric properties among 525 nursing students in Turkey [12].

The purpose of this study was to translate the OESSS into Chinese and assess its reliability and validity among Chinese nursing students. Through this research, we aim to provide robust support for the quality assessment of online education, promote the optimization of online education practices in nursing education, and ensure that it better meets the educational needs and expectations of nursing students in the digital age.

Methods

Design

The study comprised two main stages: cross-cultural adaptation and psychometric evaluation, with the latter conducted using a cross-sectional design.

Participants and samples

From January to March 2025, nursing students from three medical universities in Liaoning Province were surveyed using a convenience sampling method. The inclusion criteria were as follows: (1) full-time enrolled nursing students; (2) prior or current experience with online learning; (3) informed consent and voluntary participation in the study. The exclusion criteria included: (1) students currently participating in other research projects; (2) students on leave of absence.

The sample size was estimated using Kendall’s method, which recommends a sample size of 5 to 10 times the number of questionnaire items [13]. Based on a preliminary calculation and an anticipated inefficiency rate of 20%, and given that the scale used in this study contains 28 items, the required sample size ranged from 175 to 350 participants. Furthermore, to meet the minimum sample size requirements for exploratory factor analysis ≥ 100 participants and confirmatory factor analysis ≥ 200 participants [14], a total of 536 nursing students were ultimately recruited.

Measuring instrument

(1) General Information Questionnaire: it was developed by the research team and included items on gender, age, education level, year of study, interest in the nursing major, and the number of online courses completed so far.

(2) Chinese version of Online Education Student Satisfaction Scale(C-OESSS): it consists of 28 items across 5 dimensions: EDUCATION PROGRAM (items 1 ~ 12), TECHNICAL INFRASTRUCTURE (items 13 ~ 16), FLEXIBILITY OF THE EDUCATIONAL ENVIRONMENT (items 17 ~ 19), INSTRUCTOR (items 20 ~ 25), and EDUCATOR-STUDENT INTERACTION (items 26 ~ 28). It uses a 5-point Likert scale, with responses ranging from 1 “totally disagree” to 5 “totally agree”. The total score ranges from 28 to 140, with higher scores indicating greater satisfaction with online education.

Formal investigation & quality control

This study was conducted using the China Questionnaire Star platform (https://www.wjx.cn/) in accordance with the ethical principles of medical research. The researcher uploaded the survey instrument to the platform and, in collaboration with the faculty advisor, distributed the questionnaire via WeChat to nursing student groups across all years of study. Participation was entirely voluntary and anonymous. Prior to accessing the questionnaire, participants were presented with an electronic informed consent form on the first page of the survey. The consent form clearly explained the purpose of the study, procedures involved, potential risks and benefits, assurance of data confidentiality, voluntary participation, and the right to withdraw at any time without penalty. Participants were required to read the information and confirm their willingness to participate by selecting the “I agree to participate” option. Only those who provided active electronic consent were allowed to proceed to the survey questions. To ensure the quality and integrity of the data, each IP address was restricted to a single submission, and all questions were set as mandatory to prevent incomplete responses. After data collection, the responses were reviewed individually. Questionnaires were excluded if they had a completion time of less than 60 s or showed response patterns (e.g., identical choices throughout). In total, 600 questionnaires were distributed, and 536 valid responses were collected, yielding an effective response rate of 89.33%, which is adequate for factor analysis [15].

Translation and cross-cultural adaptation

The research team contacted Prof. Harmancı Seren by email to obtain authorization for the use and adaptation of the original scale. Following the Brislin model [16], the English version of the scale was translated into Chinese using the “forward translation ─ back translation” method. During the cross-cultural adaptation process, particular attention was paid to achieving semantic, idiomatic, experiential, and conceptual equivalence to ensure that the scale items were culturally appropriate and retained their original meaning in the target context [17]. The overall translation and adaptation procedure is illustrated in Fig. 1.

Fig. 1.

OESSS’s scale translation and cross-cultural adaptation process

Data analysis

Statistical analyses were performed using IBM SPSS 27.0 and AMOS 28.0. Qualitative data were described using frequency and percentage (%), while quantitative data conforming to a normal distribution were presented as mean ± standard deviation (Mean ± SD). Quantitative data not following a normal distribution were expressed as interquartile range (IQR). When the skewness and kurtosis scores were between − 1 ~ + 1, the data were deemed to be regularly distributed [18]. The critical ratio (CR) and correlation coefficient methods were employed for scale item analysis. Reliability was evaluated using Cronbach’s α coefficient and test-retest reliability. Validity was assessed through content validity, structural validity, convergent validity, and discriminant validity. Statistical significance was set at P < 0.05.

Item analysis

• (i)Critical ratio method: the total scores of the 536 scales were sorted from high to low by putting the first 27% in the high group and the next 27% in the low group, and t-tests were performed on the entries for two independent samples, and the entries were retained if the critical ratio was > 3 and the difference was statistically significant (P < 0.05) and deleted if the difference was the other way round [19].
• (ii)Correlation Coefficient Method: The Pearson correlation coefficient was used. Items were retained if the correlation coefficient between the item score and the total score was greater than 0.40 and the difference was statistically significant (P < 0.05). Items were deleted if the correlation coefficient was lower or the difference was not statistically significant [20].
• (iii)Cronbach’s α Coefficient Method: The total Cronbach’s α coefficient t of the scale was calculated, and the change in the coefficient was assessed after removing each item. If the Cronbach’s α coefficient after removing an item was greater than 0.5 compared to the coefficient before removal, the item was considered for deletion [21]. Otherwise, the item was retained.

Validity analysis

The data of 536 cases were randomly divided into two parts, where sample 1 (236 cases) was used for exploratory factor analysis (EFA) and sample 2 (300 cases) was used for confirmatory factor analysis (CFA).

• (i)Structural validity: EFA was conducted using Bartlett’s test of sphericity, with a statistically significant χ² value, and a Kaiser-Meyer-Olkin (KMO) measure greater than 0.80 [15]. Orthogonal rotation was applied using principal component analysis (PCA) with varimax rotation. Factor extraction was based on eigenvalues greater than 1. Factor loadings ≥ 0.5 were used as the inclusion criterion [22], and items that did not meet this threshold or had high cross-loadings were considered for removal.

  CFA was conducted using Sample 2, and the model was estimated using robust maximum likelihood estimation. Model fit was evaluated using several indices: Chi-square/degree of freedom (CMIN/DF) < 3, Root Mean Square Error of Approximation (RMSEA) < 0.05, and Comparative Fit Index (CFI), Normed Fit Index (NFI), Incremental Fit Index (IFI), Tucker-Lewis Index (TLI) and Comparative fit index(CFI)>0.90 [23]. A good model fit is indicated by values meeting these thresholds.

• (ii) Content validity: Nine experts were invited to assess the relevance of the scale items and research concepts, using a 4-point scale: 1 = irrelevant, 2 = weakly relevant, 3 = strongly relevant, and 4 = very relevant. The content validity of the scale was evaluated using the Content Validity Index (CVI). The item-level Content Validity Index (I-CVI) was considered acceptable if > 0.78, and the average scale Content Validity Index (S-CVI/Ave) was considered acceptable if > 0.90 [24].
• (iii) Convergent Validity: Composite Reliability (CR) and Average Variance Extracted (AVE) are commonly used indicators for assessing convergent validity in structural equation modeling. It is generally accepted that latent variables exhibit good convergent validity when CR > 0.7 and AVE > 0.5 [6].
• (iv)Discriminant Validity: Discriminant validity was assessed using the Fornell-Larcker criterion. According to this criterion, discriminant validity is established if the square root of the AVE for each latent variable is greater than its correlation with any other latent variable [25]. This indicates that the measurement of that latent variable effectively discriminates between different latent variables.

Reliability analysis

(i)Internal consistency

The internal consistency of the scale was assessed by calculating Cronbach’s α coefficient and McDonald’s Omega. Coefficient of ≥ 0.70 was considered indicative of good internal consistency [26].

(ii) Test-retest reliability: Previous research suggests that a sample size of 30 to 50 participants is typically sufficient for assessing test-retest reliability [27, 28]. Therefore, this study selected a sample of 40 students for the second measurement, which was conducted after a two-week interval. If the correlation coefficient between the two measurements exceeded 0.70 [18], the scale was deemed to exhibit stable test-retest reliability.

Results

Pre-survey

In January 2025, a total of 30 nursing students from three medical colleges in Liaoning Province, China, were selected as pilot study participants using convenience sampling. After being informed of the purpose and significance of the study, all participants signed informed consent forms. The pilot results indicated that the scale had clear themes, a well-structured format, and coherent logic. No difficulties in semantic comprehension were reported. On average, participants took approximately three minutes to complete the questionnaire. As no major issues were identified, no revisions were made, and the Chinese version of the OESSS scale was finalized for use in the main study.

Basic characteristics of the participants

A total of 536 nursing students participated in this study, with ages ranging from 17 to 34 years (21.31 ± 2.734). The sample was predominantly female (84.0%) and consisted mostly of students with a bachelor’s degree (51.3%), with sophomores comprising the largest group (45.5%). Detailed information is presented in (Table 1).

Table 1.

Basic characteristics of the participants(n = 536)

Variables		Frequency(n = 536)	Percentage%
Gender	Males	86	16.0
	Females	450	84.0
Education level	College degree	240	51.3
	Bachelor’s degree	275	44.8
	Postgraduate degree	21	3.9
Year of study	First	117	21.8
	Second	244	45.5
	Third	156	29.1
	Fourth	19	3.6
Residency	Rural	274	51.1
	Urban	262	48.9
Interest in the nursing major	Yes	473	11.8
	No	63	88.2
The number of online courses completed so far	≦ 1	28	5.2
	> 2	508	94.8

Item analysis

All items follow a normal distribution. A comparison of the scores between the high and low groups in the C-OESSS revealed a CR ranging from 7.948 to 19.828, indicating good differentiation among the items. The Pearson correlation coefficients for each item with the total score ranged from 0.415 to 0.711, indicating good homogeneity of the scale. Additionally, the Cronbach’s α coefficient for each item after deletion was lower than 0.5 compared to its value before deletion, with the overall Cronbach’s α coefficient remaining between 0.912 and 0.917, not exceeding the total Cronbach’s α of 0.918. Therefore, all items were retained (Table 2).

Table 2.

C-OESSS of skewness, kurtosis, pearson correlation coefficient, critical ratio(n = 536)

Items	Mean (SD)	Skewness/Kurtosis	Pearson correlation coefficient	Critical ratio	Cronbach’s α coefficient after deletion of item
EP1	3.00(0.81)	0.055/0.192	0.684**	16.305***	0.913
EP2	3.02(0.87)	-0.047/0.097	0.654**	14.642***	0.913
EP3	2.98 (0.85)	0.164/-0.009	0.587**	13.803***	0.914
EP4	3.05(0.82)	0.067/0.008	0.711**	19.282***	0.912
EP5	3.00(0.84)	0.117/-0.045	0.692**	18.603***	0.912
EP6	3.03(0.82)	0.246/0.051	0.624**	14.702***	0.914
EP7	3.00(0.86)	0.056/-0.047	0.669**	16.997***	0.913
EP8	3.02(0.83)	0.157/0.313	0.657**	15.421***	0.913
EP9	3.02(0.86)	0.084/0.192	0.660**	16.302***	0.913
EP10	3.02(0.83)	0.036/-0.041	0.624**	14.559***	0.914
EP11	2.99(0.84)	-0.064/-0.078	0.609**	14.839***	0.914
EP12	2.99(0.87)	0.022/-0.130	0.638**	15.493***	0.913
EP13	3.00(0.87)	0.041/-0.097	0.475**	10.317***	0.916
TI14	3.04(0.82)	-0.044/-0.076	0.463**	9.156***	0.916
TI15	2.99(0.81)	-0.007/-0.039	0.510**	11.013***	0.916
TI16	3.01(0.87)	0.040/-0.082	0.443**	10.057***	0.917
TI17	3.03(0.85)	-0.024/-0.055	0.415**	8.579***	0.917
FO18	3.06(0.87)	-0.137/0.116	0.421**	8.559***	0.917
FO19	3.04(0.86)	-0.208/-0.061	0.460**	8.719***	0.916
IN20	3.04(0.86)	0.084/-0.200	0.558**	12.725***	0.915
IN21	3.01(0.84)	-0.105/0.086	0.550**	11.95***	0.915
IN22	3.01(0.86)	-0.011/-0.081	0.582**	12.718***	0.914
IN23	2.98(0.85)	0.046/-0.076	0.504**	10.462***	0.916
IN24	2.98(0.88)	-0.095/-0.149	0.609**	14.249***	0.914
IN25	3.01(0.87)	-0.005/-0.114	0.503**	11.39***	0.916
ES26	3.01(0.84)	-0.021/0.002	0.426**	7.948***	0.917
ES27	3.00(0.82)	0.056/-0.024	0.434**	8.975***	0.917
ES28	3.00(0.83)	-0.003/0.045	0.440**	8.705***	0.917

Note: **p < 0.01; ***P < 0.001

Exploratory factor analysis (EFA)

EFA was conducted on Sample 1, with the results indicating suitability for factor analysis: KMO = 0.888, Bartlett’s test of sphericity χ² = 3604.967 (p < 0.001). Five factors with eigenvalues greater than 1 were extracted using principal component analysis and the maximum variance method, explaining a cumulative variance of 66.623% (Fig. 2). Unexpectedly, item 13 was reallocated from the “EDUCATION PROGRAM” dimension to the “FLEXIBILITY OF THE EDUCATIONAL ENVIRONMENT” dimension. However, the remaining dimensions and item allocations were consistent with the original scale. All factor loadings exceeded 0.5, and there were no instances of double loading (Table 3).

Fig. 2.

Five-factor scree plot of the C-OESSS (n = 236)

Table 3.

Factor loadings and cumulative variance contribution of C-OESSS(n = 236)

KMO values	0.888	Bartlett’s test	Chi-squared/DF	3604.967/378	P < 0.001
Characteristic root	7.4816	4.0012	2.8438	2.3429	1.9848
Cumulative variance	26.010%	40.716%	50.749%	58.722%	66.623%
Items	Factor1	Factor2	Factor3	Factor4	Factor5
EP7	0.817	-	-	-	-
EP11	0.811	-	-	-	-
EP5	0.805	-	-	-	-
EP10	0.801	-	-	-	-
EP3	0.793	-	-	-	-
EP6	0.773	-	-	-	-
EP2	0.772	-	-	-	-
EP4	0.761	-	-	-	-
EP12	0.759	-	-	-	-
EP8	0.744	-	-	-	-
EP9	0.741	-	-	-	--
EP1	0.737	-	-	-	-
IN20	-	0.838	-	-	-
IN25	-	0.828	-	-	-
IN24	-	0.824	-	-	-
IN23	-	0.820	-	-	-
IN22	-	0.812	-	-	-
IN21	-	0.787	-	-	-
TI17	-	-	0.845	-	-
TI16	-	-	0.830	-	-
TI14	-	-	0.825	-	-
TI15	-	-	0.824	-	-
FO18	-	-	-	0.866	-
EP13	-	-	-	0.859	-
FO19	-	-	-	0.829	-
ES28	-	-	-	-	0.854
ES26	-	-	-	-	0.846
ES27	-	-	-	-	0.841

Confirmatory factor analysis (CFA)

CFA was conducted on Sample 2, and the model fit indices showed the following results: CMIN/DF = 1.185, RMSEA = 0.025, NFI = 0.917, IFI = 0.986, TLI = 0.984, and CFI = 0.986 (Fig. 3).

Fig. 3.

C-OESSS standardized 5-factor structural model (n = 300)

Content validity

The CVI for the C-OESSS at the item level ranged from 0.88 to 1, with S-CVI/Ave of 0.95 at the scale level.

Convergent and discriminant validity

The CR for the five dimensions of the C-OESSS were 0.753 ~ 0.941, all exceeding 0.700. The AVE values were 0.504 ~ 0.598, all greater than 0.500. Additionally, the square root of the AVE for each dimension was greater than the corresponding correlation coefficients between the dimensions (Table 4).

Table 4.

Convergent and discriminant validity of the C-OESSS

Convergent validity							Discriminant validity
Items	Std. estimate	S.E.	P	AVE	CR	Cronbach’s α coefficient	Factors	F1	F2	F3	F4	F5
EP1	0.817			0.575	0.941	0.941	F1	0.785
EP2	0.765	0.067	***
EP3	0.668	0.068	***
EP4	0.809	0.061	***
EP5	0.796	0.063	***
EP6	0.718	0.064	***
EP7	0.799	0.064	***
EP8	0.765	0.064	***
EP9	0.778	0.064	***
EP10	0.717	0.064	***
EP11	0.707	0.065	***
EP12	0.747	0.066	***
TI13	0.812			0.598	0.856	0.768	F2	0.373***	0.773
TI14	0.744	0.063	***
TI15	0.788	0.063	***
TI16	0.748	0.069	***
FO17	0.689			0.504	0.753	0.700	F3	0.446***	0.502***	0.709
FO18	0.696	0.110	***
FO19	0.743	0.114	***
IN20	0.769			0.574	0.889	0.893	F4	0.445***	0.554***	0.456***	0.757
IN21	0.746	0.07	***
IN22	0.819	0.072	***
IN23	0.714	0.073	***
IN24	0.764	0.075	***
IN25	0.73	0.075	***
ES26	0.761			0.557	0.790	0.795	F5	0.413***	0.494***	0.611***	0.473***	0.746
ES27	0.713	0.083	***
ES28	0.764	0.086	***

NOTE: *** P < 0.001; Bold represents the square root value of AVE

Reliability analysis

The Cronbach’s α coefficient of the C-OESSS total scale was 0.918, the McDonald’s Omega coefficient was 0.910, and the test-retest reliability coefficient was 0.916. In addition, the Cronbach’s α coefficient of the five dimensions is 0.941, 0.768, 0.700, 0.893 and 0.795 respectively (Table 4).

Discussion

Translation

The introduction of a translated scale necessitates a rigorous evaluation of its equivalence to the original version to ensure its validity and applicability in the new cultural context. Among various types of equivalence, content equivalence is particularly critical—it requires that each item on the translated scale not only retains the original meaning but also aligns appropriately with the cultural norms, values, and expressions of the target population [29]. This ensures that respondents can interpret and respond to each item as intended. To achieve this, a systematic and well-managed translation and back-translation process is essential. This typically involves multiple bilingual experts independently translating the original scale into the target language, followed by a synthesis and reconciliation of the translations. Subsequently, a separate set of experts conducts a back-translation into the original language, which is then compared with the original scale to identify and resolve any discrepancies. This iterative process helps to preserve both the semantic and conceptual equivalence of the scale, reducing the risk of misinterpretation or cultural bias. Ultimately, meticulous management of these procedures enhances the reliability and cross-cultural validity of the translated instrument.

Cross-cultural adaptation

In the research on psychometric feature testing of assessment tools, the review by an expert panel is necessary. Based on the research published by Dr. Valmi et al. On Translation, adaptation and validation of instruments or scales for use in cross-cultural health care research: a clear and user-friendly guideline [30], the number of expert panel members should be between 6 and 10. We finally selected 9 experts based on the themes of the scale to determine whether items should be included, modified or omitted [31]. Based on the experts’ suggestions, in order to better adapt to the cultural background and language of China, we have made minor adjustments to some projects. The specific situation is as follows:

First, the word “reflect” in “In online lessons, I can reflect on what I have learned (EP₁₁)” was replaced with “review.” In Western educational culture, “reflect” generally refers to deep thinking, self-assessment, and critical analysis of one’s learning process, emphasizing personal insights and reflective thought [12]. In contrast, Chinese educational culture places a stronger emphasis on reviewing and summarizing during the learning process. Particularly in test-oriented education, reviewing and revising knowledge points are common learning behaviors [32]. Therefore, the term “review” is more easily understood and accurately reflects the process of revisiting and consolidating knowledge in the Chinese context. In addition, in Chinese education, students commonly use the word “review,” while “reflect” is more often associated with academic writing and differs from the language habits of students in their daily learning activities. Using “review” makes it easier for participants to grasp the intent of the question, thereby enhancing the reliability and validity of the scale. Secondly, in the phrase “I can easily convey the technical problems I experience in the online education environment (EP₁₄),” the word “convey” is replaced with “feedback.” According to experts, “convey” typically refers to the act of transferring or expressing information, focusing on the process of transmitting information from one party to another [33]. In contrast, “feedback” refers to providing a response to an action or situation, emphasizing both the reception and the response. It is not just about conveying information but also about offering a response that can lead to improvements. This term is often used to enhance interactions during a process. In Chinese education, the term “feedback” is commonly used, particularly when addressing problem resolution and improvement. Students are often asked to provide feedback on technical issues so that technical support or relevant personnel can respond promptly. In contrast, the term “convey” is more abstract and less intuitive than “feedback,” which could lead to confusion or misinterpretation among participants. Therefore, “feedback” aligns better with Chinese linguistic and cultural norms. Thirdly, we were surprised to find that “I monitor my learning progress in online classes (EP₁₃)” was categorized under the dimension of “FLEXIBILITY OF THE EDUCATIONAL ENVIRONMENT.” This may be attributed to the fact that, from a cultural perspective, the rapid development of online education in China has increasingly emphasized the flexibility and autonomy of learning [34]. In recent years, online education environments have increasingly emphasized the need for students to adjust their learning progress based on their personal time and learning needs, contrasting with traditional education systems, which are characterized by teacher-driven and fixed curricula [35]. In China, especially within online learning, students are required to be more flexible and take greater control over their learning progress [36]. As such, the behavior of “monitoring progress” aligns more closely with the definition of the “flexibility of the educational environment” dimension. The act of self-regulation, where students monitor their own progress to adapt to individual learning needs, reflects the core concept of flexibility [37]. Therefore, categorizing this item under the “FLEXIBILITY OF THE EDUCATIONAL ENVIRONMENT” dimension more accurately captures the characteristics of student self-directed learning in online education. Secondly, from an expression standpoint, “monitor my learning progress” not only implies control over the learning content but also emphasizes flexibility and autonomy. In an online education environment, students often need to manage their learning progress within an irregular schedule and learning mode, allowing them to make adjustments according to their individual circumstances. Consequently, the content of this item aligns more closely with the definition of a “flexible educational environment,” highlighting students’ initiative and adaptability in online learning. In contrast, the traditional “EDUCATION PROGRAM” dimension focuses on the structuring of course content and teaching objectives, emphasizing the fixed nature of teaching plans and syllabi. Therefore, categorizing the original entry, “monitor my learning progress,” within the “EDUCATION PROGRAM” dimension may not fully capture its emphasis on flexibility and autonomy in learning. The phenomenon of structural changes in scale dimensions is not unique to this study; rather, it is commonly observed in cross-cultural scale validation research. For example, a similar issue was reported in a study by Gao Ziyun et al. [38], titled “Reliability and Validity of the Chinese Version of the Self-Directed Learning Instrument in Chinese Nursing Students,” which involved 975 nursing students in China. During the localization process, certain structural adjustments were made to the original scale, highlighting the influence of cultural differences, linguistic nuances, and respondents’ cognitive styles on the structural stability of the instrument. Therefore, moderate structural modifications are both acceptable and necessary in cross-cultural adaptation studies, provided that the revised instrument undergoes rigorous reliability and validity testing, and that the theoretical rationale for the changes is clearly established. This ensures the appropriateness and scientific rigor of the scale within the new cultural context.

Good discrimination, reliability and validity of C-OESSS items

Scale discrimination can be assessed using the critical ratio method, correlation coefficient method, and discriminant validity [39]. The results from the critical ratio method indicated that the differences in scores between the high and low groups were statistically significant (all P < 0.001), demonstrating that the items effectively discriminated between high and low scores. The correlation coefficient method revealed that the Pearson correlation coefficients between the scores of each item and the total score exceeded the minimum standard, indicating sufficient homogeneity among the items. Additionally, the results from the Fornell-Larcker standardized test showed that the square root of AVE was greater than the correlation coefficients between the latent variable and other latent variables, confirming that the C-OESSS effectively discriminates between different latent variables. Reliability refers to the consistency of results when repeated measurements are made on the same subject using the same method [40]. The total Cronbach’s α coefficient and McDonald’s Omega of the C-OESSS both exceeded 0.7, indicating that the scale demonstrates good reliability and stability. Additionally, the test-retest reliability exceeded the minimum standardized value, confirming that the C-OESSS provides consistent and credible results in measuring nursing students’ satisfaction with online education over time. Validity refers to the degree to which the measurement instrument accurately reflects the intended research concepts, ensuring the authenticity and precision of the instrument [25]. Both EFA and CFA supported the construct validity of the C-OESSS. The KMO value of 0.76 and the significant Bartlett’s Test of Sphericity indicated that the data were suitable for factor analysis. The 5-factor model explained 66.623% of the total variance, reflecting the multidimensional nature of nursing students’ satisfaction with online education. CFA results showed a good model fit, with CMIN/DF < 3, RMSEA < 0.05, and NFI, IFI, TLI, and CFI all > 0.9, indicating an excellent fit. Nine experts assessed the content of the scale, and several items were modified, resulting in I-CVI and S-CVI values that met expectations. These results confirm that the scale items align with the experiences of Chinese nursing students and effectively capture the nuances of online education satisfaction within the Chinese cultural context. This high content validity is crucial for ensuring that the scale accurately reflects the challenges faced by Chinese nursing students in online education satisfaction.

In conclusion, the results of the cross-sectional study confirmed the validity and applicability of the C-OESSS as a tool for assessing Chinese nursing students’ satisfaction with online education. The study demonstrated that the scale accurately captures nursing students’ satisfaction within online education environments, supported by high reliability and validity.

The application value and significance of the C-OESSS scale

With the rapid development of online education within China’s higher education system, student satisfaction has emerged as a critical indicator for evaluating the quality of online instruction and optimizing instructional design [41]. The introduction and localization of the OESSS are of significant importance for systematically assessing Chinese university nursing students’ satisfaction in online learning environments. The scale comprehensively covers the rationality of course design, the effectiveness of technical support, the flexibility of learning modes, the instructional support provided by teachers, and the quality of interaction between instructors and students, enabling a systematic assessment of students’ satisfaction across various aspects of online learning [12].

Currently, there is a lack of a standardized, psychometrically sound instrument in China that is broadly applicable across diverse academic disciplines and institutional contexts for measuring student satisfaction with online education [42]. Through the scientific introduction, translation, cultural adaptation, and validation of the OESSS, a reliable assessment tool can be established for educational researchers and administrators.

Specifically, the Chinese version of Online Education Student Satisfaction Scale for nursing students can be applied in multiple contexts. For instance, nursing educators and administrators may use it to evaluate the effectiveness of online courses, identify strengths and weaknesses in teaching quality, and inform curriculum improvements. The scale may also serve as a monitoring tool in blended learning environments or during large-scale transitions to online education. In practice, it can be administered at the end of a course, and subscale scores can provide detailed feedback on different aspects of student satisfaction, such as instructional design, interaction, and technical support. Moreover, repeated administration over time can help track changes in satisfaction and evaluate the impact of educational interventions.

Limitations

Although the C-OESSS demonstrated strong reliability in this study, several limitations must be acknowledged. First, the cross-sectional design and convenience sampling method may have introduced sampling bias, limiting the generalizability of the findings to a broader population of nursing students. To enhance the robustness of the scale, future research should consider conducting multicenter studies across different regions and cultural contexts, which would provide a more comprehensive evaluation of the scale’s reliability and validity. Second, during the EFA stage, one item was adjusted. However, factor loadings may fluctuate based on factors such as region, sample size, and respondent characteristics. Therefore, further research is needed to validate the appropriateness of this item adjustment.

Conclusion

Despite these limitations, this study confirmed that the C-OESSS exhibits strong psychometric properties and is well-suited for measuring online education satisfaction among nursing students in China. The C-OESSS offers a scientifically sound and reliable instrument for educational researchers, educators, and administrators to assess nursing students’ satisfaction with online learning. This tool not only aids educators in understanding the specific challenges and needs nursing students face in online education but also provides valuable insights for optimizing and improving online education curricula, thereby contributing to the enhancement of education quality.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

CMIN/DF

Chi-square/Degree of freedom

RMSEA

Root-mean-square error of approximation

CFI

Comparative fit index

NFI

Normed fit index

IFI

Incremental fit index

TLI

Tucker-Lewis index

S-CVI/Ave

The average scale Content Validity Index

I-CVI

Item-level content validity index

KMO

The Kaiser–Meyer–Olkin

EFA

Exploratory factor analysis

CFA

Confirmatory factor analysis

CR

Critical ration

AVE

Average variance extracted

RMR

Root mean square residual

Author contributions

Chuang Li: Conceptualization, Data curation, Investigation, Methodology, Project administration, Software, Supervision, Validation, Writing – original draft, Youbei Lin, Hong ye Guo, Haixia Zhao, Jinyan Xiong, Lan Zhang: Investigation, Methodology, Resources. Xuyang Xiao: Funding acquisition, Methodology, Resources, Supervision, Visualization, Writing-review & editing.

Funding

This research was supported by the 2024 Liaoning Provincial Science and Technology Joint Program (2024-MSLH-168). We sincerely appreciate the funding provided.

Data availability

Experimental data from this study are available from the first author under reasonable request. The specific items of the Chinese version of the OESSS scale can be found in Supplementary Material 1. However, authorization must be obtained by contacting the first author, Chuang Li, before use.

Declarations

Ethical statement

The study was approved by the Ethics Review Committee of Jinzhou Medical University (JZMULL2025012) and adhered to their ethical guidelines. Informed consent was obtained from all participating students to ensure their confidentiality and anonymity. The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. Besides, there was no personal or professional relationship between the researchers and the participants. Participation was voluntary and anonymous.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.