Validation of the Mental Health First Aid Knowledge Questionnaire for use among Chinese non-mental health nurses Authors: Li Li, Huize Zhang, Shurong Lu, Xiaolan Ma, Hong Tan, Yipei Zhang, Chunyan Xie, Yamin Li Published: 2025 ABSTRACT ObjectiveThe Mental Health First Aid Knowledge Questionnaire (MHFAKQ) is a tool used in English-speaking countries for assessing the levels of knowledge about how to support individuals with mental health problems among non-mental health professionals. This study aimed to evaluate the reliability and validity of the MHFAKQ among Chinese non-mental health nurses.MethodsFollowing the Sousa Translation Model (or the best practices in cross-cultural adaptation of research instruments), the MHFAKQ was translated into Chinese, back-translated, culturally adapted, and pre-tested. The instrument in the Chinese language was tested among a convenience sample of non-mental health nurses (n = 425) from a non-psychiatric department of a general hospital in China. Internal consistency and test–retest reliability were assessed by Cronbach’s alpha and intraclass correlation coefficient (ICC), respectively. Validity was evaluated using scale content validity and structural validity tests.ResultsThe Chinese version of the MHFAKQ comprised 16 items within a single dimension. The content validity index (I-CVI) for individual items ranged from 0.875 to 1.000, with an average content validity index (S-CVI) of 0.992. Exploratory factor analysis identified one factor with an eigenvalue of 8.367, explaining 52.29% of the variance. The questionnaire demonstrated excellent internal consistency (Cronbach’s α = 0.939), test-retest reliability (0.822), and split-half reliability (0.940).ConclusionThe Chinese version of the MHFAKQ shows reasonable reliability and validity to be applied among Chinese non-mental health nurses.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40359-025-03624-w. Introduction Mental health issues are escalating globally, becoming one of the most remarkable public health challenges worldwide [1, 2]. The prevalence of mental disorders is often higher among hospital patients than in the general population, with nurses among the healthcare professionals who interact most frequently with these patients [3, 4]. Mental Health First Aid (MHFA), introduced by Australian scholars in 2000 [5], is defined as the assistance provided to someone experiencing a mental health problem, a worsening mental health condition, or a mental health crisis (e.g., risk of suicide) until professional help is available or the crisis is resolved [6, 7]. MHFA is akin to cardiopulmonary resuscitation (CPR), only that it deals with offering early psychosocial intervention to prevent further deterioration of mental health [8–10]. Given the current shortage of mental health professionals, both in relative and absolute terms, nurses play a crucial role in identifying symptoms of mental disorders and offering necessary mental health support [11, 12]. With the development of China’s social economy, the public’s demand for healthcare services is increasing, and the scale of general hospitals is expanding. However, the number of nurses in China is difficult to meet the increasing demand, and the limited number of nurses makes them face moderate to high levels of work stress [13, 14]. Some studies have shown that the bed-to-nurse ratio in China is at a low level of 1:0.4–0.8 [15], moreover, 86.9% of nurses work more than 40 h per week as stipulated in the labor law, and 39% of nurses work more than 50 h per week [16]. Long working hours, heavy workload, and shift-related fatigue make nurses in a state of physical and mental overload for a long time [13, 14]. Therefore, in such a high-pressure environment, it is even more important for nurses to have knowledge and skills related to MHFA, which will not only help nurses assist patients facing mental health problems or crises more effectively, but also to maintain their own mental well-being [6]. At the same time, the Chinese healthcare system faces systemic limitations in mental health service provision. The ratio of mental health professionals to patients remains imbalanced, and psychiatric resources are unevenly distributed across regions, creating significant disparities in care [17]. Against this backdrop, nurses—particularly those working in non-psychiatric departments—are increasingly expected to take on responsibilities in the prevention, early identification, and management of mental health problems among hospitalized patients. WHO points out that when there is a shortage of mental health resources, the task of preventing and intervening in patients’ mental health problems can be appropriately shifted to clinical healthcare workers, especially nurses [18]. However, many non-mental health nurses lack adequate MHFA knowledge and skills, despite their frontline role in patient care [19]. Given that mental health disorders are often insidious in onset and can result in serious adverse outcomes that affect patients’ health, doctor–patient relationships, and even hospital credibility [20, 21], it is of great practical importance to investigate the current status of MHFA knowledge and skills among non-mental health nurses in China. Such research can inform the development of locally tailored strategies to strengthen nurses’ capacity in this domain. The Mental Health First Aid Knowledge Questionnaire (MHFAKQ), developed by Nicola J. Reavley at the University of Melbourne [11] and based on the MHFA Handbook [6], has demonstrated strong reliability in Australian surveys, and shows the instrument has robust validity in the English language. It covers 16 items, to assess the level of MHFA knowledge among the general public [11]. The higher the score, the better the level of MHFA knowledge is represented. To assess the effectiveness of interventions for training non-mental health nurses in managing mental health disorders or crises, locally validated tools are essential.This is to help managers evaluate the level of MHFA knowledge for nurses in non-psychiatric units and better understand their needs for training planning. Therefore, it’s important to validate it for the Chinese. This study involved adapting the English version of the MHFAKQ into Chinese and evaluating its reliability and validity through a survey conducted among non-mental health nurses. The goal was to establish a reliable tool for assessing the MHFA knowledge of non-mental health nurses in China. Materials and methods Introduction to the MHFAKQ The Mental Health First Aid Knowledge Questionnaire (MHFAKQ), developed by Nicola J. Reavley and colleagues at the University of Melbourne, is a one-dimensional instrument designed to assess MHFA knowledge among the general public. The questionnaire comprises 16 items, each with right and wrong response options, with correct responses scoring 1 point and incorrect responses scoring 0 points. A higher total score indicates a greater level of knowledge in MHFA. The MHFAKQ has been widely utilized in public health services across Australia to evaluate the effectiveness of mental health training programs [11]. In this study, we adapted the MHFAKQ into Chinese to measure the MHFA knowledge among non-mental health nurses in China. The adaptation process involved forward translation, back-translation, and cultural adaptation to ensure the relevance and accuracy of the questionnaire in the Chinese context. After obtaining permission from the original authors, we conducted a pilot test with non-mental health nurses at a university hospital in Xinjiang Province to validate the structural integrity of the Chinese version. Exploratory factor analysis (EFA) was performed to assess the questionnaire’s structural validity, while internal consistency and test-retest reliability were measured to confirm its reliability. Research design This study followed a three-phase cross-sectional design. First is translation and adaptation, the original English version of the MHFAKQ, translated into simplified Chinese, followed by a back-translation process to ensure accuracy. Cultural adaptation was carried out to align the questionnaire with the local context. Second is content validation in which expert consultation and pilot testing were conducted to evaluate the content validity of the translated questionnaire. This phase involved feedback from mental health professionals and non-mental health nurses to refine the questionnaire. The third is reliability testing, a cross-sectional survey conducted using convenience sampling, during which EFA was performed on the valid data to evaluate the reliability and structural validity of the Chinese version of the MHFAKQ. Subjects This study used convenience sampling to recruit clinical registered nurses from the non-psychiatric rooms of a university hospital affiliated with a university in City A, Xinjiang Province, Western China, between April and May 2022. The hospital, founded in 1954, is the largest medical teaching hospital in Xinjiang, with 2,371 registered nurses and 2,700 open beds [22]. Inclusion criteria are as follows: (1) Non-mental health nurses who have been working for ≥ 1 year in a tertiary A equivalent hospital, with a certificate of nurse practitioner; and (2) Willingness to participate in the survey. Meanwhile, the following are the exclusion criteria: (1) nurses on rotations, internships, and further training; (2) temporarily absent from work due to maternity leave, sick leave, or for other reasons; and 3) non-mental health nurses in the mental health profession (i.e., nurses in psychiatric-psychological units). Sampling method and sample size calculation For measurement, this study used a convenience sampling method to select non-psychiatric clinical registered nurses from a university-affiliated hospital in City A, Xinjiang Province, Western China.Determination of the sample size: according to the standards related to the quality testing of questionnaires, the sample content should be 5–10 times the number of entries [23, 24] with the minimum requirement for EFA being at least 150 cases [25]. Given that our questionnaire contains 16 items, the minimum requirement was 80–160 participants and 10%−20% of invalid questionnaires were considered at the same time, The required sample size for EFA was 89–200 [26]. Additionally, in order to assess the test-retest reliability of the instrument, a subsample of 49 participants was randomly selected for a follow-up survey two weeks after the initial assessment. The final sample of 425 participants ensured robust psychometric evaluation, improved data quality, and allowed for a small margin of missing or invalid responses without compromising analytical rigor. For measurement, this study used a convenience sampling method to select non-psychiatric clinical registered nurses from a university-affiliated hospital in City A, Xinjiang Province, Western China. For measurement, this study used a convenience sampling method to select non-psychiatric clinical registered nurses from a university-affiliated hospital in City A, Xinjiang Province, Western China. Determination of the sample size: according to the standards related to the quality testing of questionnaires, the sample content should be 5–10 times the number of entries [23, 24] with the minimum requirement for EFA being at least 150 cases [25]. Given that our questionnaire contains 16 items, the minimum requirement was 80–160 participants and 10%−20% of invalid questionnaires were considered at the same time, The required sample size for EFA was 89–200 [26]. Additionally, in order to assess the test-retest reliability of the instrument, a subsample of 49 participants was randomly selected for a follow-up survey two weeks after the initial assessment. The final sample of 425 participants ensured robust psychometric evaluation, improved data quality, and allowed for a small margin of missing or invalid responses without compromising analytical rigor. Determination of the sample size: according to the standards related to the quality testing of questionnaires, the sample content should be 5–10 times the number of entries [23, 24] with the minimum requirement for EFA being at least 150 cases [25]. Given that our questionnaire contains 16 items, the minimum requirement was 80–160 participants and 10%−20% of invalid questionnaires were considered at the same time, The required sample size for EFA was 89–200 [26]. Additionally, in order to assess the test-retest reliability of the instrument, a subsample of 49 participants was randomly selected for a follow-up survey two weeks after the initial assessment. The final sample of 425 participants ensured robust psychometric evaluation, improved data quality, and allowed for a small margin of missing or invalid responses without compromising analytical rigor. Research tools The General Information Questionnaire includes section, gender, ethnicity, age, length of service, education, and title.The Chinese version of the MHFAKQ was adapted into Chinese for a total of 16 entries, which were scored in terms of right and wrong, with correct answers worth 1 point and wrong answers worth 0 points, with a maximum score of 16 and a minimum score of 0. The General Information Questionnaire includes section, gender, ethnicity, age, length of service, education, and title. The General Information Questionnaire includes section, gender, ethnicity, age, length of service, education, and title. The Chinese version of the MHFAKQ was adapted into Chinese for a total of 16 entries, which were scored in terms of right and wrong, with correct answers worth 1 point and wrong answers worth 0 points, with a maximum score of 16 and a minimum score of 0. The Chinese version of the MHFAKQ was adapted into Chinese for a total of 16 entries, which were scored in terms of right and wrong, with correct answers worth 1 point and wrong answers worth 0 points, with a maximum score of 16 and a minimum score of 0. Questionnaire evaluation Item analysis. The correlation coefficient method and the extreme group test were used to evaluate the quality of individual items and refine the questionnaire items accordingly. Item analysis. The correlation coefficient method and the extreme group test were used to evaluate the quality of individual items and refine the questionnaire items accordingly. Item analysis. The correlation coefficient method and the extreme group test were used to evaluate the quality of individual items and refine the questionnaire items accordingly. The correlation coefficient method was used to analyze the entries of the Chinese version of the MHFAKQ to explore the correlation between each entry score and the total score of the questionnaire. The extreme group test was used to screen the entries, which was based on the principle of ranking the total scores and selecting the top 27% of the total scores from high to low as the high group and the bottom 27% of the total scores from low to high as the low group to form two independent samples for the t-test to test the difference between the two groups of subjects in terms of the individual entries of the two groups of study subjects on each entry [25]. If statistical significance exists, the entry is proven to have better discrimination and quality; conversely, the entry is considered for deletion. (2)Reliability test: Reliability is evaluated through internal consistency and test-retest reliability. (2)Reliability test: Reliability is evaluated through internal consistency and test-retest reliability. Reliability test: Reliability is evaluated through internal consistency and test-retest reliability. Internal consistency reflects the degree of correlation between items that reflect different aspects of the same concept. The internal consistency of the questionnaire was determined by calculating the Cronbach’ α coefficient, which is greater than or equal to 0.7. If the entry-to-entry correlation coefficient is less than 0.3 and deleting these entries would increase the alpha coefficient of the entire questionnaire by 0.1 or more, then, these entries are considered non-homogeneous and are excluded [27]. The retest interval of 2 weeks was chosen for repeat test reliability because subjects were unlikely to recall their responses after this interval [28]. A test-retest reliability ICC value of 0.40–0.59 indicates moderate test-retest reliability of the questionnaire, and an ICC value of 0.60–0.79 indicates high test-retest reliability [29]. (3)Validity test: the scale was evaluated using content validity and structural validity tests. (3)Validity test: the scale was evaluated using content validity and structural validity tests. Validity test: the scale was evaluated using content validity and structural validity tests. Content validity tests whether or not the items that make up the questionnaire include all aspects the study is attempting to measure, and they are a subjective indicator of expert evaluation [25]. In this study, the content validity was measured using an expert assessment method to analyze the relevance of each entry in the translated version of the questionnaire to the dimension to which it belongs, and each entry was scored one by one on a 4-point Likert scale, with each entry scored from “1” = not relevant, to “4” = most relevant. The analysis calculates item-level content validity index (ICVI) and the scale-level content validity index (S-CVI/Ave, average of all I-CVIs) to determine content validity [25]. The I-CVI score is used to determine the relevance and clarity of each entry and is calculated by dividing the number of experts who rated the entry as a 3 or 4 by the number of content experts [30]. If the I-CVI score is above 0.79, the entry is considered appropriate; if the score is between 0.70 and 0.79, it indicates that the entry needs revision; if the score is below 0.70, it indicates that the entry should be deleted [30]. S-CVI/Ave is calculated by dividing the sum of the I-CVI scores by the total number of entries. I-CVI values of 0.78 or higher and S-CVI/Ave values of 0.9 or higher are generally considered standardized [25]. Detailed information on the experts is provided in Table 1. Table 1Basic information on the panel of expertsnumbergendersageYears of occupational experienceEducationProfessional titleoccupationWork department1Female4526Bachelor’s degreeSeniorNurseClinical nursing2Male5127Master’s degreeSeniorDoctorClinical care3Female4523PhDVice-seniorTeacherNursing education4Female4321PhDVice-seniorTeacherClinical care5Female4523Master’s degreeVice-seniorNurseNursing education6Female5230Bachelor’s degreeSeniorNurseClinical nursing7Female5535Bachelor’s degreeSeniorNurseClinical nursing8Female4014Bachelor’s degreeIntermediateNurseClinical nursing Table 1Basic information on the panel of expertsnumbergendersageYears of occupational experienceEducationProfessional titleoccupationWork department1Female4526Bachelor’s degreeSeniorNurseClinical nursing2Male5127Master’s degreeSeniorDoctorClinical care3Female4523PhDVice-seniorTeacherNursing education4Female4321PhDVice-seniorTeacherClinical care5Female4523Master’s degreeVice-seniorNurseNursing education6Female5230Bachelor’s degreeSeniorNurseClinical nursing7Female5535Bachelor’s degreeSeniorNurseClinical nursing8Female4014Bachelor’s degreeIntermediateNurseClinical nursing Basic information on the panel of experts Construct validity refers to the actual measurement, whether the theoretical conception sets up the questionnaire’s entries at the time of the design [25]. This study analyzed the structural validity of the MHFAKQ through EFA and assessed it through PCA and maximum variance orthogonal rotations [31]. The specific steps are as follows:The first step measured Bartlett’s Test of Sphericity and KMO, a measure of sampling fitness. The purpose was to determine whether the Chinese version of the MHFAKQ was suitable for factor analysis [32]. The KMO value is between 0 and 1. The larger the value, the better the factor analysis. If the KMO value is less than 0.5, it is unsuitable for factor analysis; a value greater than 0.7 is suitable and greater than 0.9 indicates the most effective factor analysis. In addition, the use of factor analysis is supported if Bartlett’s test of sphericity reaches statistical significance (P < 0.05) [25].The second step determined the number of common factors. Under the premise of suitability for factor analysis, PCA was selected, with factor loading greater than 0.3 and eigen root values of extracted common factors greater than 1 as the judgment criteria [33]. The public factors were extracted, rotated, and transformed for the best analysis. A questionnaire is considered to have good structural validity if the extracted metric explains more than 40% of the variance and each entry loads sufficiently on the corresponding factor (> or = 0.4) [34]. The first step measured Bartlett’s Test of Sphericity and KMO, a measure of sampling fitness. The purpose was to determine whether the Chinese version of the MHFAKQ was suitable for factor analysis [32]. The KMO value is between 0 and 1. The larger the value, the better the factor analysis. If the KMO value is less than 0.5, it is unsuitable for factor analysis; a value greater than 0.7 is suitable and greater than 0.9 indicates the most effective factor analysis. In addition, the use of factor analysis is supported if Bartlett’s test of sphericity reaches statistical significance (P < 0.05) [25]. The first step measured Bartlett’s Test of Sphericity and KMO, a measure of sampling fitness. The purpose was to determine whether the Chinese version of the MHFAKQ was suitable for factor analysis [32]. The KMO value is between 0 and 1. The larger the value, the better the factor analysis. If the KMO value is less than 0.5, it is unsuitable for factor analysis; a value greater than 0.7 is suitable and greater than 0.9 indicates the most effective factor analysis. In addition, the use of factor analysis is supported if Bartlett’s test of sphericity reaches statistical significance (P < 0.05) [25]. The second step determined the number of common factors. Under the premise of suitability for factor analysis, PCA was selected, with factor loading greater than 0.3 and eigen root values of extracted common factors greater than 1 as the judgment criteria [33]. The public factors were extracted, rotated, and transformed for the best analysis. A questionnaire is considered to have good structural validity if the extracted metric explains more than 40% of the variance and each entry loads sufficiently on the corresponding factor (> or = 0.4) [34]. The second step determined the number of common factors. Under the premise of suitability for factor analysis, PCA was selected, with factor loading greater than 0.3 and eigen root values of extracted common factors greater than 1 as the judgment criteria [33]. The public factors were extracted, rotated, and transformed for the best analysis. A questionnaire is considered to have good structural validity if the extracted metric explains more than 40% of the variance and each entry loads sufficiently on the corresponding factor (> or = 0.4) [34]. Steps in the study (i.e., formation of the Chinese version of the MHFAKQ) In forming the Chinese version of the MHFAKQ, the basic principles of cross-cultural debugging were consistently followed. Stage 1: Obtaining authorization to translate. Consent and authorization from the original author of the MHFAKQ were obtained via email. Using the Sousa translation model [35], The questionnaire was translated from English to Chinese. The translation process in this study includes forward translation, synthesis, back translation, and synthesis.① Forward translation: Two bilingual doctoral students (Nos.1 and 2), native speakers of Chinese, and have experience translating medical questionnaires, independently translated the original English text into Chinese, resulting in Translators A and B. This process minimized the individual idiosyncrasies introduced when only two independent translators were used, thus improving the quality of translation.② Synthesis: The researcher herself synthesized the two versions of the original translation, separately listed all the inconsistencies, and reached a consensus after discussing the appropriateness of the wording and grammar with the two original translators together and then generated the Chinese Translation C of the initial translation of the MHFAKQ.③ Back-translation: Translation C was independently translated by two other bilingual masters and doctoral students who had studied abroad and had not been exposed to the original questionnaire (No. 3 is a master’s degree student majoring in English, and 4 is a doctoral student in the UK majoring in psychological care) into English, forming back-translated versions of Translations D and E. This step clarifies the words and sentences used in the translation.④ Synthesis: The researcher herself compared the back-translated versions of translation D, translated the original English version, discussed the inconsistencies with the back-translators, and then consolidated them to form the back-translated version of Translation F. We also communicated with the original author by e-mail.⑤ Re-synthesis: The four translators and the researcher herself worked together to compare and weigh the translated expressions word by word, discuss all the ambiguities, and after reaching a consensus, the initial translated Chinese version of Translation C was further revised to form the second consolidated Chinese version of Translation G. ① Forward translation: Two bilingual doctoral students (Nos.1 and 2), native speakers of Chinese, and have experience translating medical questionnaires, independently translated the original English text into Chinese, resulting in Translators A and B. This process minimized the individual idiosyncrasies introduced when only two independent translators were used, thus improving the quality of translation. ① Forward translation: Two bilingual doctoral students (Nos.1 and 2), native speakers of Chinese, and have experience translating medical questionnaires, independently translated the original English text into Chinese, resulting in Translators A and B. This process minimized the individual idiosyncrasies introduced when only two independent translators were used, thus improving the quality of translation. ② Synthesis: The researcher herself synthesized the two versions of the original translation, separately listed all the inconsistencies, and reached a consensus after discussing the appropriateness of the wording and grammar with the two original translators together and then generated the Chinese Translation C of the initial translation of the MHFAKQ. ② Synthesis: The researcher herself synthesized the two versions of the original translation, separately listed all the inconsistencies, and reached a consensus after discussing the appropriateness of the wording and grammar with the two original translators together and then generated the Chinese Translation C of the initial translation of the MHFAKQ. ③ Back-translation: Translation C was independently translated by two other bilingual masters and doctoral students who had studied abroad and had not been exposed to the original questionnaire (No. 3 is a master’s degree student majoring in English, and 4 is a doctoral student in the UK majoring in psychological care) into English, forming back-translated versions of Translations D and E. This step clarifies the words and sentences used in the translation. ③ Back-translation: Translation C was independently translated by two other bilingual masters and doctoral students who had studied abroad and had not been exposed to the original questionnaire (No. 3 is a master’s degree student majoring in English, and 4 is a doctoral student in the UK majoring in psychological care) into English, forming back-translated versions of Translations D and E. This step clarifies the words and sentences used in the translation. ④ Synthesis: The researcher herself compared the back-translated versions of translation D, translated the original English version, discussed the inconsistencies with the back-translators, and then consolidated them to form the back-translated version of Translation F. We also communicated with the original author by e-mail. ④ Synthesis: The researcher herself compared the back-translated versions of translation D, translated the original English version, discussed the inconsistencies with the back-translators, and then consolidated them to form the back-translated version of Translation F. We also communicated with the original author by e-mail. ⑤ Re-synthesis: The four translators and the researcher herself worked together to compare and weigh the translated expressions word by word, discuss all the ambiguities, and after reaching a consensus, the initial translated Chinese version of Translation C was further revised to form the second consolidated Chinese version of Translation G. ⑤ Re-synthesis: The four translators and the researcher herself worked together to compare and weigh the translated expressions word by word, discuss all the ambiguities, and after reaching a consensus, the initial translated Chinese version of Translation C was further revised to form the second consolidated Chinese version of Translation G. Stage 2: Cultural debugging and pre-survey.① A cultural debugging team of eight experts were invited, including one clinical psychologist, three head nurses and one nurse, and one psychologist and two nursing administrators with rich working experience. The group of experts commented on the semantics and cultural equivalence of the translated second Chinese version of the translated manuscript G to form a culturally adjusted version of the third Chinese version of the “MHFAKQ” that was consistent with the cultural background of our country. Eight experts examined the validity of the questionnaire’s content using a 4-point Likert scale. The validity of the questionnaire was tested by eight experts using a 4-point Likert scale from 1 (not relevant) to 4 (very relevant). Moreover, the CVI of the questionnaire was estimated by calculating the average CVI of all items. Basic information on experts is shown in Table 1.② Pre-testing: The purpose of the pre-survey was to determine whether potential problems remained after the questionnaire had been correctly translated, synthesized, back-translated, and revised by the expert group. Through cognitive analysis of a pre-small sample, the Chinese version of the third edition after culture debugging was further modified and adjusted form the Chinese trial version of the “Mental Health First Aid Knowledge Questionnaire”. Finalize the trial revision before finalizing the official questionnaire in preparation for the official survey. ① A cultural debugging team of eight experts were invited, including one clinical psychologist, three head nurses and one nurse, and one psychologist and two nursing administrators with rich working experience. The group of experts commented on the semantics and cultural equivalence of the translated second Chinese version of the translated manuscript G to form a culturally adjusted version of the third Chinese version of the “MHFAKQ” that was consistent with the cultural background of our country. Eight experts examined the validity of the questionnaire’s content using a 4-point Likert scale. The validity of the questionnaire was tested by eight experts using a 4-point Likert scale from 1 (not relevant) to 4 (very relevant). Moreover, the CVI of the questionnaire was estimated by calculating the average CVI of all items. Basic information on experts is shown in Table 1. ① A cultural debugging team of eight experts were invited, including one clinical psychologist, three head nurses and one nurse, and one psychologist and two nursing administrators with rich working experience. The group of experts commented on the semantics and cultural equivalence of the translated second Chinese version of the translated manuscript G to form a culturally adjusted version of the third Chinese version of the “MHFAKQ” that was consistent with the cultural background of our country. Eight experts examined the validity of the questionnaire’s content using a 4-point Likert scale. The validity of the questionnaire was tested by eight experts using a 4-point Likert scale from 1 (not relevant) to 4 (very relevant). Moreover, the CVI of the questionnaire was estimated by calculating the average CVI of all items. Basic information on experts is shown in Table 1. ② Pre-testing: The purpose of the pre-survey was to determine whether potential problems remained after the questionnaire had been correctly translated, synthesized, back-translated, and revised by the expert group. Through cognitive analysis of a pre-small sample, the Chinese version of the third edition after culture debugging was further modified and adjusted form the Chinese trial version of the “Mental Health First Aid Knowledge Questionnaire”. Finalize the trial revision before finalizing the official questionnaire in preparation for the official survey. ② Pre-testing: The purpose of the pre-survey was to determine whether potential problems remained after the questionnaire had been correctly translated, synthesized, back-translated, and revised by the expert group. Through cognitive analysis of a pre-small sample, the Chinese version of the third edition after culture debugging was further modified and adjusted form the Chinese trial version of the “Mental Health First Aid Knowledge Questionnaire”. Finalize the trial revision before finalizing the official questionnaire in preparation for the official survey. The translated MHFAKQ was pretested by 49 randomly selected non-psychiatric clinical registered nurses from a university hospital in Xinjiang Province from April to May 2022 using a convenience sampling method. The questionnaire was collected through WeChat by forwarding the link to the Questionnaire Star platform. The purpose and significance of this survey were explained to the participants before the survey. Additionally, the questionnaire was filled out after obtaining their informed consent, allowing the nurses to evaluate the clarity of the entries to answer these entries with a simple dichotomous choice (clear or unclear) and to revise again for entries that were not clearly expressed. To test the test-retest reliability, these 49 participants were contacted by telephone two weeks after the end of this survey and asked to complete the questionnaire again. Stage 3: The questionnaire’s reliability and validity test:Finally, a convenience sampling method was used to test the reliability and validity of non-psychiatric clinical registered nurses from a university hospital in Xinjiang Province who met the inclusion and exclusion criteria through the Questionnaire Star platform with a third tentative Chinese version of the MHFAKQ to determine the reliability and validity of the final Chinese version of the MHFAKQ. This questionnaire completion was preceded by a brief introduction of the study’s purpose and significance, stating the data confidentiality and anonymity. After excluding invalid questionnaires, a total of 425 valid data were obtained. The process is shown in Fig. 1. Finally, a convenience sampling method was used to test the reliability and validity of non-psychiatric clinical registered nurses from a university hospital in Xinjiang Province who met the inclusion and exclusion criteria through the Questionnaire Star platform with a third tentative Chinese version of the MHFAKQ to determine the reliability and validity of the final Chinese version of the MHFAKQ. This questionnaire completion was preceded by a brief introduction of the study’s purpose and significance, stating the data confidentiality and anonymity. After excluding invalid questionnaires, a total of 425 valid data were obtained. The process is shown in Fig. 1. Finally, a convenience sampling method was used to test the reliability and validity of non-psychiatric clinical registered nurses from a university hospital in Xinjiang Province who met the inclusion and exclusion criteria through the Questionnaire Star platform with a third tentative Chinese version of the MHFAKQ to determine the reliability and validity of the final Chinese version of the MHFAKQ. This questionnaire completion was preceded by a brief introduction of the study’s purpose and significance, stating the data confidentiality and anonymity. After excluding invalid questionnaires, a total of 425 valid data were obtained. The process is shown in Fig. 1. Fig. 1Translation of the MHFAKQ Translation of the MHFAKQ Data collection ①Format: The researcher used Questionnaire Star and Wechat to collect data. ②Recruitment method: The researcher recruited respondents via phone and WeChat. The researcher contacted coworkers and classmates, and then coworkers and classmates contacted their coworkers and classmates and recruited participants online through the convenience sampling method. ③Consent procedures: The researcher screened the eligible participants based on the inclusion and exclusion criteria. The purpose of the study was explained to them. After obtaining written informed consent from the participants, the latter was asked to complete the questionnaire, and items that the participants did not understand were explained to them. Questionnaire Star: China’s leading online questionnaire platform, supporting a variety of question types, real-time data aggregation and statistical analysis, widely used in academic research, market research and other fields. WeChat: China’s mainstream multi-functional social application, both instant messaging, social sharing and other functions, is a digital communication tool with a large user base in China. Data analysis SPSS version 28.0 (IBM, Armonk, NY, USA) was used for statistical analysis. Means, standard deviations, total correlations of items, frequencies, and component ratios were used for descriptive statistics of questionnaire items and sample characteristics. The correlation coefficient method and extreme value method were used for item analysis. Reliability analysis was performed using Cronbach’s α coefficient and test-retest reliability to assess internal and external consistency. Validity analysis was performed using I⁃CVI, S⁃CVI, and EFA. The significance level was set at P < 0.05. Quality control ① The researcher strictly followed the conditions and requirements of scale adapted into Chinese and the standard international scale introduction procedure to adapt into Chinese and corrected the scale so that the Chinese version of the MHFAKQ is equivalent to the original scale. ② The translators and back-translators were proficient in both languages. To ensure the scientific rigor of the scale adaptation process, back-translators had no prior exposure to the original English version, minimizing potential bias in the translation. The scale Chinese questionnaires were filled out anonymously strictly in accordance with the criteria of the selection of research subjects. The data were exported via the Questionnaire Star platform. Questionnaires that did not meet the quality requirements are deleted, such as all the chosen entries have the same option, or the answer time is less than 60 s. Then, the data are analyzed. Ethical clearance This study was approved by the Medical Ethics Committee of the First Affiliated Hospital of Xinjiang Medical University (No. K202204-10), with ethical approval from the Ethics Committee for Investigative and Behavioral Research of Central South University (No. E202224). The researcher thoroughly explained the purpose of the study to all participants and obtained their informed consent whether or not to participate. The privacy and anonymity of the participants were fully protected by forwarding the questionnaire star link through WeChat to collect data. All participants were informed that the data would be kept strictly confidential. This research received grants from Major Scientific and Technological Projects in Hunan Province (No. 2020SK2085). Results Translation of the mental health first aid knowledge questionnaire This study strictly carried out localized translation following the questionnaire adapted into Chinese process while ensuring the format of the original questionnaire entries as much as possible. Most of the entries were directly translated from the original English into Chinese. In the correct translation of Entry 2, “Australia” was translated as “Chinese ”. In Entry 15, “espresso” was culturally adapted to “strong tea”. After back translation, these two items and others retained their original meaning. Pre-survey results General information of the pre-survey respondents: A total of 49 nurses were surveyed, 48 female and 1 male. The age primarily ranged between 30 and 39 accounting for 40.82%. The proportion of work experience ≥ 16 years is 38.78%. Bachelor’s degree holders account for the majority, and Table 2 presents the basic situation. General information of the pre-survey respondents: A total of 49 nurses were surveyed, 48 female and 1 male. The age primarily ranged between 30 and 39 accounting for 40.82%. The proportion of work experience ≥ 16 years is 38.78%. Bachelor’s degree holders account for the majority, and Table 2 presents the basic situation. General information of the pre-survey respondents: A total of 49 nurses were surveyed, 48 female and 1 male. The age primarily ranged between 30 and 39 accounting for 40.82%. The proportion of work experience ≥ 16 years is 38.78%. Bachelor’s degree holders account for the majority, and Table 2 presents the basic situation. Table 2Basic information of pre-survey respondents (n = 49)Demographic variablesFrequency (n)Percentage (%)Work departmentPediatric24.08Gynecology and obstetrics1020.41Emergency and outpatient714.29Internal medicine1428.57Operating room and anesthesia36.12Surgical department816.33Intensive care unit510.2GenderMale12.04Female4897.96Age group (year)< 301020.4130–392040.82≥ 401938.78Years of occupational experience1–3816.334–10612.2411–151632.65≥ 161938.78EthnicityHan3775.51Others1224.49EducationAssociate’s degree or below1326.53Bachelor’s degree2551.02Master’s degree or above1122.45Professional titlePrimary2551.02Intermediate1530.61Senior918.37 Table 2Basic information of pre-survey respondents (n = 49)Demographic variablesFrequency (n)Percentage (%)Work departmentPediatric24.08Gynecology and obstetrics1020.41Emergency and outpatient714.29Internal medicine1428.57Operating room and anesthesia36.12Surgical department816.33Intensive care unit510.2GenderMale12.04Female4897.96Age group (year)< 301020.4130–392040.82≥ 401938.78Years of occupational experience1–3816.334–10612.2411–151632.65≥ 161938.78EthnicityHan3775.51Others1224.49EducationAssociate’s degree or below1326.53Bachelor’s degree2551.02Master’s degree or above1122.45Professional titlePrimary2551.02Intermediate1530.61Senior918.37 Basic information of pre-survey respondents (n = 49) (2)Results of cognitive-semantic analysis: Forty-nine pre-survey respondents completed all entries in 2–3 min, and 47 indicated that the instructions and entries of the questionnaire were clear and easy to understand; therefore, no changes were made to the questionnaire. (2)Results of cognitive-semantic analysis: Forty-nine pre-survey respondents completed all entries in 2–3 min, and 47 indicated that the instructions and entries of the questionnaire were clear and easy to understand; therefore, no changes were made to the questionnaire. Results of cognitive-semantic analysis: Forty-nine pre-survey respondents completed all entries in 2–3 min, and 47 indicated that the instructions and entries of the questionnaire were clear and easy to understand; therefore, no changes were made to the questionnaire. Reliabilities and validities General information of formal survey respondents: General information of formal survey respondents: General information of formal survey respondents: A total of 436 questionnaires were collected, of which 11 were excluded due to irregular responses. Thus, 425 questionnaires were available for analysis. In terms of the distribution of the classification of the department, the participants were mostly in internal medicine, with a total of 109, accounting for 25.65%. The proportion of women was 95.53%; the proportion of 30–39 years was 61.41%. The proportion of those with ≥ 16 years of service was 138, accounting for 32.47%. The educational sample holding “undergraduate” degrees accounted for 72.24%, and the proportion of junior titles was 70.59%. Table 3 shows the details. Table 3Characteristics of participants (n = 425)Demographic variablesFrequency (n)Percentage (%)Work departmentPediatric133.06Gynecology and obstetrics4310.12Emergency and outpatient5813.65Internal medicine10925.65Operating room and anesthesia133.06Surgical department8319.53Diagnosis and subsidiary266.12Oncology/infection/hemodialysis286.59Intensive care unit5212.24GenderMale194.47Female40695.53Age group (year)< 306715.7630–3926161.41≥ 409722.82Years of occupational experience1–3327.534–1012128.4711–1513431.53≥ 1613832.47EthnicityHan30170.82Others12429.18EducationAssociate’s degree or below11226.35Bachelor’s degree30772.24Master’s degree or above61.41Professional titlePrimary30070.59Intermediate9522.35Senior276.35 Table 3Characteristics of participants (n = 425)Demographic variablesFrequency (n)Percentage (%)Work departmentPediatric133.06Gynecology and obstetrics4310.12Emergency and outpatient5813.65Internal medicine10925.65Operating room and anesthesia133.06Surgical department8319.53Diagnosis and subsidiary266.12Oncology/infection/hemodialysis286.59Intensive care unit5212.24GenderMale194.47Female40695.53Age group (year)< 306715.7630–3926161.41≥ 409722.82Years of occupational experience1–3327.534–1012128.4711–1513431.53≥ 1613832.47EthnicityHan30170.82Others12429.18EducationAssociate’s degree or below11226.35Bachelor’s degree30772.24Master’s degree or above61.41Professional titlePrimary30070.59Intermediate9522.35Senior276.35 Characteristics of participants (n = 425) (2)Item analysis is primarily used to evaluate the quality of individual items. This study used the correlation coefficient method and the extreme group test for this analysis. (2)Item analysis is primarily used to evaluate the quality of individual items. This study used the correlation coefficient method and the extreme group test for this analysis. Item analysis is primarily used to evaluate the quality of individual items. This study used the correlation coefficient method and the extreme group test for this analysis. Correlation coefficient method: the homogeneity test is used as an alternative method for censoring individual question items and is also often used to eliminate items unrelated to the study topic. Homogeneity primarily measures the consistency of all the questions within the questionnaire to determine whether all the items can measure the trait being measured. Additionally, the method typically used is the correlation coefficient method, which evaluates the quality of the items by calculating the degree of correlation between the scores of individual items and the total score of the questionnaire. A correlation coefficient of r greater than 0.3 in absolute value and P < 0.01 is considered a criterion for good quality questionnaire items [25]. The Pearson correlation coefficient values of the total score of the Chinese version of the MHFAKQ and the scores of the entries in this study ranged from 0.646 to 0.779 (P < 0.01) and were therefore retained (see Table 4). Table 4Pearson’s correlation coefficient between the score of each entry of the questionnaire and the total score (r)Items R P Items R P Q10.779**<0.001Q90.741**<0.001Q20.727**<0.001Q100.729**<0.001Q30.646**<0.001Q110.727**<0.001Q40.725**<0.001Q120.751**<0.001Q50.758**<0.001Q130.708**<0.001Q60.708**<0.001Q140.702**<0.001Q70.668**<0.001Q150.730**<0.001Q80.751**<0.001Q160.709**<0.001*P < 0.05; **P < 0.01 Table 4Pearson’s correlation coefficient between the score of each entry of the questionnaire and the total score (r)Items R P Items R P Q10.779**<0.001Q90.741**<0.001Q20.727**<0.001Q100.729**<0.001Q30.646**<0.001Q110.727**<0.001Q40.725**<0.001Q120.751**<0.001Q50.758**<0.001Q130.708**<0.001Q60.708**<0.001Q140.702**<0.001Q70.668**<0.001Q150.730**<0.001Q80.751**<0.001Q160.709**<0.001*P < 0.05; **P < 0.01 Pearson’s correlation coefficient between the score of each entry of the questionnaire and the total score (r) *P < 0.05; **P < 0.01 Extreme value method: the identification of the items in this questionnaire was tested using the extreme value test (also known as the critical ratio method), which aims to determine the validity and appropriateness of each item. It is based on the principle of first summing the analyzed items and then dividing them into high and low scores (bounded by the 27% and 73% quartiles) [25]. The independent samples t-test was used to compare the high and low grouping scores (t is the decision value CR for each entry). If the result presents significance and none of the 95% confidence intervals include 0, it means that the entries have a better discriminatory degree and should be retained; otherwise, it means that the entries are unable to differentiate the information, and the treatment of deletion should be considered. In this study, the Chinese version of MHFAK score of 425 clinical nurses was (8.06 ± 5.71), and the nurses presented a moderate level of knowledge about mental health first aid. The total scores of the Chinese version of the MHFAKQ of 425 non-mental health nurses were ranked from low to high, and the lowest score of 3 was categorized as the low subgroup (the first 27%). The highest score of 14 was categorized as the high subgroup (the second 27%), which constituted two independent samples [36]. Then, a t-test was conducted to determine whether the difference between high and low subgroups in terms of the entries was significant or not as seen in Table 4. For the extreme group test, participants were ranked by their total MHFAKQ scores. The lowest 27%(n = 134) had total scores ≤ 3, while the highest 27%(n = 137) had total scores ≥ 14. The differences between high and low subgroups were significant (P < 0.05) for all entries (CR > 3) (see Table 5). Table 5Extreme group test results for each entry in the question paperItemslow clusteringhigh clusteringt(CR) P 95%CI for difference(n = 134)(n = 137)lower limitupper limitQ10.02 ± 0.150.92 ± 0.27−33.531<0.001−0.95−0.845Q20.02 ± 0.150.80 ± 0.40−21.239<0.001−0.853−0.708Q30.13 ± 0.340.88 ± 0.32−18.546<0.001−0.828−0.669Q40.16 ± 0.370.98 ± 0.15−23.794<0.001−0.881−0.747Q50.13 ± 0.330.97 ± 0.17−26.326<0.001−0.907−0.781Q60.12 ± 0.330.98 ± 0.15−28.094<0.001−0.919−0.799Q70.08 ± 0.280.84 ± 0.37−19.128<0.001−0.835−0.679Q80.13 ± 0.330.97 ± 0.17−26.326<0.001−0.907−0.781Q90.06 ± 0.240.94 ± 0.24−30.682<0.001−0.938−0.825Q100.10 ± 0.310.94 ± 0.24−25.224<0.001−0.902−0.772Q110.04 ± 0.190.86 ± 0.35−24.17<0.001−0.891−0.757Q120.01 ± 0.120.91 ± 0.28−33.707<0.001−0.95−0.845Q130.13 ± 0.330.96 ± 0.19−25.471<0.001−0.901−0.772Q140.17 ± 0.380.96 ± 0.21−21.272<0.001−0.857−0.712Q150.05 ± 0.220.93 ± 0.26−29.61<0.001−0.933−0.817Q160.15 ± 0.360.99 ± 0.12−25.904<0.001−0.9−0.773 Table 5Extreme group test results for each entry in the question paperItemslow clusteringhigh clusteringt(CR) P 95%CI for difference(n = 134)(n = 137)lower limitupper limitQ10.02 ± 0.150.92 ± 0.27−33.531<0.001−0.95−0.845Q20.02 ± 0.150.80 ± 0.40−21.239<0.001−0.853−0.708Q30.13 ± 0.340.88 ± 0.32−18.546<0.001−0.828−0.669Q40.16 ± 0.370.98 ± 0.15−23.794<0.001−0.881−0.747Q50.13 ± 0.330.97 ± 0.17−26.326<0.001−0.907−0.781Q60.12 ± 0.330.98 ± 0.15−28.094<0.001−0.919−0.799Q70.08 ± 0.280.84 ± 0.37−19.128<0.001−0.835−0.679Q80.13 ± 0.330.97 ± 0.17−26.326<0.001−0.907−0.781Q90.06 ± 0.240.94 ± 0.24−30.682<0.001−0.938−0.825Q100.10 ± 0.310.94 ± 0.24−25.224<0.001−0.902−0.772Q110.04 ± 0.190.86 ± 0.35−24.17<0.001−0.891−0.757Q120.01 ± 0.120.91 ± 0.28−33.707<0.001−0.95−0.845Q130.13 ± 0.330.96 ± 0.19−25.471<0.001−0.901−0.772Q140.17 ± 0.380.96 ± 0.21−21.272<0.001−0.857−0.712Q150.05 ± 0.220.93 ± 0.26−29.61<0.001−0.933−0.817Q160.15 ± 0.360.99 ± 0.12−25.904<0.001−0.9−0.773 Extreme group test results for each entry in the question paper (3)Reliability analysis: The results of the internal consistency test of the questionnaire show that the Cronbach’s α coefficient of the Chinese version of the MHFAKQ is 0.939, which is greater than 0.7; the folding coefficient is 0.940, which indicates that the questionnaire has higher reliability. (3)Reliability analysis: The results of the internal consistency test of the questionnaire show that the Cronbach’s α coefficient of the Chinese version of the MHFAKQ is 0.939, which is greater than 0.7; the folding coefficient is 0.940, which indicates that the questionnaire has higher reliability. Reliability analysis: The results of the internal consistency test of the questionnaire show that the Cronbach’s α coefficient of the Chinese version of the MHFAKQ is 0.939, which is greater than 0.7; the folding coefficient is 0.940, which indicates that the questionnaire has higher reliability. Test-retest reliability: The questionnaire was retested for reliability before being formally distributed. A pre-survey was conducted first, in which 49 questionnaires were distributed. The questionnaires were modified and deleted by organizing and analyzing the data collected back to enhance completeness in the questionnaire content. The Chinese version of the MHFAKQ was distributed to 49 non-mental health nurses, and the questionnaire was redistributed to these 49 participants after a 2-week (14-day) interval, and the correlation coefficient of the total scores of the two times was calculated as the re-test reliability, and the results showed that the re-test reliability r was 0.822, suggesting that the questionnaire had high and good reliability. (4)Validity analysis: The CVI was measured using the expert evaluation method, which contains the following two levels: the entry-level and the overall level of the questionnaire. The results of this study showed that the Chinese version of the MHFAKQ was scored by a panel of eight experts, with I-CVI of 0.875–1.000 and S-CVI of 0.992, suggesting that the content validity of the Chinese version of the MHFAKQ was good (see Table 6).Table 6Results of content validity analysisItemsNumber of entry evaluation scoresI-CVIS-CVI1 score2 scores3 scores4 scoresQ1000810.992Q200081Q300081Q400081Q500081Q600081Q700081Q800081Q900081Q1000081Q1100081Q1200081Q1300081Q1400081Q1500261Q1601340.875① Structural validity: the structural validity of the questionnaire was examined in this study using EFA. The KMO value was used to compare the simple correlation and partial correlation coefficients between the items and took a value between 0 and 1. When the KMO value is between 0.6 and 0.5, it is not suitable for EFA; below 0.5, it should be abandoned [25]. Bartlett’s sphericity test was used to test whether the correlation coefficients between the entries were significant; if P<0.05, it indicated that it was suitable for EFA.This study conducted EFA, and the results showed that the KMO was 0.968, > 0.9; The χ2 value of Bartlett’s Sphericity Test was 3594.914, with a degree of freedom of 120, and P < 0.001, suitable for EFA [25].② Common factor variance: view the common degree after factor analysis; the extracted value is proportional to the degree expressed by the entry’s common factor. The larger the extracted value, the greater the entry’s dependence on the factor; otherwise, it is smaller. A value greater than 0.5 means better, 0.4 to 0.5 means average, and less than 0.4 should be excluded [37]. The results of this study showed that the common degree of all entries was greater than 0.4. Accordingly, these 16 entries were retained. (4)Validity analysis: The CVI was measured using the expert evaluation method, which contains the following two levels: the entry-level and the overall level of the questionnaire. The results of this study showed that the Chinese version of the MHFAKQ was scored by a panel of eight experts, with I-CVI of 0.875–1.000 and S-CVI of 0.992, suggesting that the content validity of the Chinese version of the MHFAKQ was good (see Table 6).Table 6Results of content validity analysisItemsNumber of entry evaluation scoresI-CVIS-CVI1 score2 scores3 scores4 scoresQ1000810.992Q200081Q300081Q400081Q500081Q600081Q700081Q800081Q900081Q1000081Q1100081Q1200081Q1300081Q1400081Q1500261Q1601340.875① Structural validity: the structural validity of the questionnaire was examined in this study using EFA. The KMO value was used to compare the simple correlation and partial correlation coefficients between the items and took a value between 0 and 1. When the KMO value is between 0.6 and 0.5, it is not suitable for EFA; below 0.5, it should be abandoned [25]. Bartlett’s sphericity test was used to test whether the correlation coefficients between the entries were significant; if P<0.05, it indicated that it was suitable for EFA.This study conducted EFA, and the results showed that the KMO was 0.968, > 0.9; The χ2 value of Bartlett’s Sphericity Test was 3594.914, with a degree of freedom of 120, and P < 0.001, suitable for EFA [25].② Common factor variance: view the common degree after factor analysis; the extracted value is proportional to the degree expressed by the entry’s common factor. The larger the extracted value, the greater the entry’s dependence on the factor; otherwise, it is smaller. A value greater than 0.5 means better, 0.4 to 0.5 means average, and less than 0.4 should be excluded [37]. The results of this study showed that the common degree of all entries was greater than 0.4. Accordingly, these 16 entries were retained. Validity analysis: The CVI was measured using the expert evaluation method, which contains the following two levels: the entry-level and the overall level of the questionnaire. The results of this study showed that the Chinese version of the MHFAKQ was scored by a panel of eight experts, with I-CVI of 0.875–1.000 and S-CVI of 0.992, suggesting that the content validity of the Chinese version of the MHFAKQ was good (see Table 6).Table 6Results of content validity analysisItemsNumber of entry evaluation scoresI-CVIS-CVI1 score2 scores3 scores4 scoresQ1000810.992Q200081Q300081Q400081Q500081Q600081Q700081Q800081Q900081Q1000081Q1100081Q1200081Q1300081Q1400081Q1500261Q1601340.875① Structural validity: the structural validity of the questionnaire was examined in this study using EFA. The KMO value was used to compare the simple correlation and partial correlation coefficients between the items and took a value between 0 and 1. When the KMO value is between 0.6 and 0.5, it is not suitable for EFA; below 0.5, it should be abandoned [25]. Bartlett’s sphericity test was used to test whether the correlation coefficients between the entries were significant; if P<0.05, it indicated that it was suitable for EFA.This study conducted EFA, and the results showed that the KMO was 0.968, > 0.9; The χ2 value of Bartlett’s Sphericity Test was 3594.914, with a degree of freedom of 120, and P < 0.001, suitable for EFA [25].② Common factor variance: view the common degree after factor analysis; the extracted value is proportional to the degree expressed by the entry’s common factor. The larger the extracted value, the greater the entry’s dependence on the factor; otherwise, it is smaller. A value greater than 0.5 means better, 0.4 to 0.5 means average, and less than 0.4 should be excluded [37]. The results of this study showed that the common degree of all entries was greater than 0.4. Accordingly, these 16 entries were retained. Table 6Results of content validity analysisItemsNumber of entry evaluation scoresI-CVIS-CVI1 score2 scores3 scores4 scoresQ1000810.992Q200081Q300081Q400081Q500081Q600081Q700081Q800081Q900081Q1000081Q1100081Q1200081Q1300081Q1400081Q1500261Q1601340.875 Results of content validity analysis ① Structural validity: the structural validity of the questionnaire was examined in this study using EFA. The KMO value was used to compare the simple correlation and partial correlation coefficients between the items and took a value between 0 and 1. When the KMO value is between 0.6 and 0.5, it is not suitable for EFA; below 0.5, it should be abandoned [25]. Bartlett’s sphericity test was used to test whether the correlation coefficients between the entries were significant; if P<0.05, it indicated that it was suitable for EFA.This study conducted EFA, and the results showed that the KMO was 0.968, > 0.9; The χ2 value of Bartlett’s Sphericity Test was 3594.914, with a degree of freedom of 120, and P < 0.001, suitable for EFA [25]. ① Structural validity: the structural validity of the questionnaire was examined in this study using EFA. The KMO value was used to compare the simple correlation and partial correlation coefficients between the items and took a value between 0 and 1. When the KMO value is between 0.6 and 0.5, it is not suitable for EFA; below 0.5, it should be abandoned [25]. Bartlett’s sphericity test was used to test whether the correlation coefficients between the entries were significant; if P<0.05, it indicated that it was suitable for EFA. This study conducted EFA, and the results showed that the KMO was 0.968, > 0.9; The χ2 value of Bartlett’s Sphericity Test was 3594.914, with a degree of freedom of 120, and P < 0.001, suitable for EFA [25]. ② Common factor variance: view the common degree after factor analysis; the extracted value is proportional to the degree expressed by the entry’s common factor. The larger the extracted value, the greater the entry’s dependence on the factor; otherwise, it is smaller. A value greater than 0.5 means better, 0.4 to 0.5 means average, and less than 0.4 should be excluded [37]. The results of this study showed that the common degree of all entries was greater than 0.4. Accordingly, these 16 entries were retained. ② Common factor variance: view the common degree after factor analysis; the extracted value is proportional to the degree expressed by the entry’s common factor. The larger the extracted value, the greater the entry’s dependence on the factor; otherwise, it is smaller. A value greater than 0.5 means better, 0.4 to 0.5 means average, and less than 0.4 should be excluded [37]. The results of this study showed that the common degree of all entries was greater than 0.4. Accordingly, these 16 entries were retained. Total Variance Explained: the factors with eigenvalues greater than 1 were extracted according to the maximum variance method, and the 16 topics were extracted through principal component analysis. The results of the study showed that a total of 1 common factor with an eigenvalue of 8.367 was extracted, with a variance contribution rate of 52.291% (> 40%), which is more satisfactory. Through plot (Fig. 2), it can be seen that only the eigenvalue of the first 1 factor has a vertical coordinate above 1 and is on a steep slope, and the slope of the 2nd factor begins to flatten. Accordingly, the first 1 factor is chosen as the main factor. Fig. 2Scree plot Scree plot The factor loading matrix reflects the degree of correlation between the original variables and each factor. To accurately interpret the naming of each principal factor, orthogonal rotation of the factor loading matrix was implemented using the method of maximum variance. For the selection of measurement entries, the size of the factor loading value was used as the criterion for retention and exclusion, the factor loading matrix was established, factor loading less than 0.5 were excluded, and the topics that were in the same column in the final factor arrangement were categorized as the same. The results of the study show that 16 items belong to the same component, and the absolute value of the factor loading coefficients corresponding to all items is greater than 0.5. This finding indicates a correspondence existing between the item and this component. The factor loading matrix reflects the degree of correlation between the original variable and each factor. Table 6 of the study results shows that 16 entries belong to the same component, and the absolute value of the factor loading coefficients corresponding to all entries is greater than 0.5, indicating that there is a correspondence between the item and this component. (see Table 7) Table 7Factor loading tableItemsComponentItemsComponentQ10.783Q90.74Q20.73Q100.73Q30.638Q110.728Q40.726Q120.752Q50.76Q130.706Q60.707Q140.701Q70.664Q150.731Q80.753Q160.709 Table 7Factor loading tableItemsComponentItemsComponentQ10.783Q90.74Q20.73Q100.73Q30.638Q110.728Q40.726Q120.752Q50.76Q130.706Q60.707Q140.701Q70.664Q150.731Q80.753Q160.709 Factor loading table Discussion Importance and scientific validity of the mental health first aid knowledge questionnaire (MHFAKQ) Mental health problems are common among general hospital patients, especially those with depressive symptoms of suicidal ideation [38]. In addition, nurses often lack training and education on recognizing, communicating, and responding to such problems [39]. MHFA is a training for mental health problems designed to help laypeople assess mental health crises and provide initial help. Nurses, as the healthcare workers with the longest and closest contact with patients, need to improve non-mental health nurses’ MHFA knowledge and skills. However, a specialized tool for assessing MHFA knowledge is lacking in the country. This study translated the “Mental Health First Aid Knowledge” questionnaire developed by Nicola J Reavley scholars into Chinese through translation, back-translation, cultural debugging, and pilot study. Cronbach’s α coefficient, test-retest reliability, and EFA were used to assess the reliability and validity of the questionnaire. The results showed that the Chinese version of the MHFAKQ showed good reliability and validity. Therefore, the questionnaire can be used to assess the knowledge of MHFA of non-mental health nurses in China for subsequent research. The Chinese version of the MHFAKQ is reliable Item analysis showed that the entries were sufficiently homogeneous and well discriminated, which was the basis for retaining or deleting the entries. The correlation coefficient values between the scores of the Chinese version of each entry and the total score of the questionnaire ranged from 0.646 to 0.779 (P < 0.001), and all > 0.4, indicating that the homogeneity of the entries and the scale was good and that no entries were deleted. Discriminant analysis of the entries revealed a significant difference between the high grouping (top 27%) and the low grouping (bottom 27%), with a CR value of > 3 (P < 0.001) for both, suggesting a reasonable degree of discrimination between the entries [25, 40]. Reliability analysis, also known as reliability analysis, is a test of stability, consistency, and dependability of measurement results, often using Cronbach’s α coefficient to test-retest reliability. A Cronbach’s α coefficient of > 0.7 is acceptable, and > 0.8 is considered good. The Cronbach’s α coefficient of the Chinese version of the MHFAKQ in this study is 0.939, which suggests good reliability. The coefficient of the test-retest reliability within 2 weeks is 0.822, which is slightly higher than the retest coefficient of the original version 0.61 [11], showing acceptable stability [25]. A test-retest reliability ICC value of ≥ 0.8 indicates that the questionnaire has good test-retest reliability [29]. The reason for the higher retest reliability than the original version may be due to cultural differences; the perception and expression of mental health problems in Chinese culture may have made the nurses more familiar with the content of the questionnaire, and thus more stable in their responses; the process of translation and cultural adaptation may have optimized the questionnaire presentation so that it is more in line with the understanding habits of the Chinese nurses, which in turn improves the test-retest reliability. Content validity refers to the degree of agreement between the scale entries and the measurement content [25]. When the number of experts involved in the evaluation is ≤ 5, the I-CVI should be 1. When the number of experts is ≥ 6, the I-CVI should be ≥ 0.78 and the S-CVI ≥ 0.90, which indicates that the questionnaire has good content validity [25, 41]. The results of this study, show that the Chinese version of the MHFAKQ was scored by an expert panel of eight people, with an I-CVI of 0.875 to 1.000 and an S-CVI of 0.992, indicating that the content validity of the Chinese version of the MHFAKQ was good. Structural validity is an index to test the stability of the questionnaire structure. When the factor structure of an equivalence scale is tested in different cultural contexts, differences usually occur [25]. The current study extracted a total of 1 main factor with eigenvalues of 8.367 using EFA, with a variance contribution rate of 52.291%. Moreover, the factor loading of each entry were > 0.4, which had good construct validity. In conclusion, the Chinese version of the MHFAKQ showed acceptable reliability and validity among non-mental health nurses. Therefore, the instrument can be used to assess the level of mental health first aid knowledge among non-mental health nurses. In addition, further validation of the Chinese version of the MHFAKQ requires a more representative and larger sample. Furthermore, validation factor analysis should be conducted in future studies. The practical value of the MHFAKQ in non-mental health nurses The item analysis showed that although the correlation between the scores of each item and the total score of the questionnaire was good and the difference between the high and low score groups was significant, potential weak points could still be identified. For example, the overall scores generally do not reach the ideal level, MHFAK score of clinical nurses was (8.06 ± 5.71), it may indicate that the nurses’ mastery of the knowledge is poor, which may affect the nurses’ ability to accurately judge the patient’s problems in the clinic. Special training content should be designed to address low-scoring knowledge points. If mental disorder symptom recognition scores are low, trainers can add clinically common mental disorder cases to the design content, such as showing videos of actual performance of patients with depression and anxiety disorders. Carry out situational simulations so that nurses can more intuitively feel the characteristics of different symptoms in role-playing. Organizing group discussions for nurses to analyze relevant cases encountered in the clinic and deepen the memory of disease symptoms. The training method can choose a mixed mode of online and offline training. For theoretically strong contents, such as the learning of mental health knowledge, nurses can learn repeatedly through online courses, making full use of their fragmented time. For the practical operation content, different scenarios are set up for nurses to role-play, so as to exercise the ability of nurses to transform theory into practice and thus improve their clinical response ability. Strengths and limitations This research is the first study of non-mental health nurses on mental health first aid training in the country, and the findings reveal an essential public health issue. This study also has its limitations. First, subjects were recruited through a convenience sample from 1 hospital in Xinjiang Province, resulting in an under representative sample. Future studies should use stratified random sampling to obtain more representative samples from different regions and different levels of medical institutions to further validate the current findings. Second, this study only assessed reliability and structural validity as well asEFA; however, it did not assess criterion validity. Such studies could be conducted in the future. Finally, the small number of male participants limited the study of the instrument’s applicability among male nurses, and the male sample size could be increased in future studies to explore the effect of gender differences on the instrument’s validity. Conclusion The Chinese version of the MHFAKQ showed acceptable reliability and validity among non-mental health nurses in China. The Chinese version of MHFAKQ can be used to assess the level of MHFA knowledge among Chinese non-mental health nurses. Further research is warranted to test the instrument in a broader range of participants, particularly those who have more opportunities to contact with people with mental health problems studies. Supplementary Information Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2.