Research design and sampling
The present study is a descriptive-correlational study in which structural equation modeling was used to investigate the mediating effect of various factors. The study was conducted on ICU nurses from July 2022 to July 2023. The research population consisted of nurses working in ICUs of the teaching hospitals affiliated with Urmia University of Medical Sciences, Urmia, northwestern Iran (Motahari, Imam Khomeini, Seyed al-Shohada, and Taleghani Hospitals).
The following formula was used to calculate the minimum required sample size. A minimum correlation coefficient of 0.15 between the research variables was assumed to determine the larger sample size with a confidence level of 95% and a test power of 80% [
35]. The calculation required a minimum sample size of
n = 350. Considering a potential 20% sample attrition rate, 420 eligible nurses were recruited for the study.
After determining the sample size, nurses were selected using a proportional quota sampling method based on the number of nurses in each ICU. Initially, the total population of ICU nurses (n = 503) was identified, and then the desired sample number (n = 420) was divided by this total.
$$\:\frac{\text{n}}{\text{N}}=\:\frac{420}{503}=0.835$$
The resulting fraction (0.835) was then multiplied by the number of nurses in each category to determine how many from each category should be selected. For example, to determine the sample size for Imam Khomeini Hospital, the sample size was calculated to be
n = 184 (220 × 0.835), considering the number of ICU nurses there (
n = 220). The sample sizes for other hospitals were calculated using a similar method (Fig.
1). The sampling process lasted eight months, from October 2022 to June 2023.
$$w = \frac{1}{2}\ln \,\frac{{1 + r}}{{1 - r}}$$
$$\:n=\frac{{\left({Z}_{1-\frac{\alpha\:}{2}}+{Z}_{1-\beta\:}\right)}^{2}}{{W}^{2}}+3\:\Rightarrow\:\:n=\frac{{\left(1.96+0.84\right)}^{2}}{({0.15)}^{2}}+3\:\Rightarrow\:n=350$$
Data were gathered using a demographic questionnaire, the Varjus Professional Autonomy Scale (V-PAS), the Porter Organizational Commitment Questionnaire (P-OCQ), the Minnesota Satisfaction Questionnaire (MSQ), and the Paterson Job Performance Questionnaire (P-JPQ).
I.
The demographic questionnaire covered personal details such as marital status, age, gender, employment status, work experience, educational level, department type, and hospital.
II.
Varjus Professional Autonomy Scale (V-PAS).
This scale consists of 18 items categorized into three bases: knowledge, action, and value. Responses are measured using a 6-point Likert scale, ranging from “strongly disagree = 1” to “strongly agree = 6.” The total score can range from 18 to 108, which indicates three levels of professional autonomy: low (18-47.99), moderate (48-77.99), and high (78–108). This scale has been used in Finland by Varjus et al. (2003) to assess the autonomy of ICU nurses and has been validated with Cronbach’s alpha of 0.56 for knowledge, 0.62 for action, and 0.76 for values [
36]. In Iran, this scale was applied in a study by Yeganeh et al. (2019), validated by faculty members, and confirmed reliable with a Cronbach’s alpha of 0.99. A preliminary study was conducted on 30 participants using a test-retest method over two weeks, confirming its reliability with an alpha of 0.8 [
37].
III.
Porter Organizational Commitment Questionnaire (P-OCQ).
This questionnaire consists of 15 items developed by Mowday, Steers, and Porter. It utilizes a 7-point Likert scale, ranging from “strongly disagree = 1” to “strongly agree = 7.” The questionnaire includes three subscales: willingness to exert effort, desire to maintain membership in the organization and acceptance of organizational values. Total scores indicate levels of organizational commitment: high commitment (70–105), moderate commitment (35–69), and low commitment (below 35). Questions 3, 7, 9, 11, 12, and 15 require reverse scoring. The validity of this questionnaire for the general organizational commitment, emotional commitment, continuous commitment, and excitement commitment, respectively, is 0.88, 0.77, 0.61, and 0.79. Its reliability through Cronbach’s alpha is 0.91 [
38]. Another study in Australia reported a Cronbach’s alpha of 0.89 for this questionnaire [
39]. Seyyedmoharrami et al. (2019) confirmed its validity and reliability in Iran, with a Cronbach’s alpha of 0.80 [
40]. In recent research conducted in Iran, Ebrahimi et al. reported that all coefficient alphas were above 0.70. The Cronbach’s alpha coefficient for organizational commitment was 0.858 [
41].
IV.
Minnesota Satisfaction Questionnaire (MSQ).
The MSQ has been developed to assess job satisfaction levels. This questionnaire comprises 19 items and is divided into six subscales, including compensation system (3 items), job type (4 items), advancement opportunities (3 items), organizational climate (2 items), leadership style (4 items), and physical conditions (3 items). Each item is rated on a 5-point Likert scale ranging from “strongly disagree = 1” to “strongly agree = 5”. The total score for each respondent can range from 0 to 45. Total scores indicate levels of job satisfaction: high job satisfaction (above 57), moderate job satisfaction (39–57), and low job satisfaction (19–38) [
42]. A study validated the reliability of the MSQ, showing a Cronbach’s alpha of 0.89 [
43]. The reliability of the MSQ in an African context has been confirmed, with Cronbach’s alpha values of 0.79 for the intrinsic subscale, 0.82 for the extrinsic subscale, and 0.86 for the overall scale, indicating strong internal consistency [
44]. Heydari et al. confirmed the validity of this questionnaire and obtained its reliability to be 0.86 in the Iranian population [
45]. Otaghi et al. reported an internal consistency reliability measured with a Cronbach’s alpha of 0.87 [
46].
V.
Paterson Job Performance Questionnaire (P-JPQ).
The P-JPQ is a 15-item questionnaire developed by Paterson to assess employees’ job performance. This questionnaire utilizes a 4-point Likert scale with points assigned as follows: “rarely = 0”, “sometimes = 1”, “often = 2”, and “always = 3”. The total score for each respondent can range from 0 to 45. Total scores indicate levels of job performance: strong job performance (above 22), moderate job performance (15–22), and poor job performance (0–15) [
47]. The reliability of this questionnaire was evaluated by Shokrkon et al. (2008), who found a Cronbach’s alpha of 0.85 and confirmed its validity through a significant correlation with a self-assessment job performance questionnaire (
p < 0.05) [
48]. Alinejad et al. (2011) reported a Cronbach’s alpha of 0.80 in a pilot study [
49]. Additionally, ten faculty members validated all the questionnaires, and their reliability was confirmed through a split-half test conducted with 60 participants in a pilot study, resulting in an alpha coefficient greater than 0.88.
Procedure
The researcher first obtained approval from the Research Council of Urmia University of Medical Sciences and then received ethical clearance from the Regional Ethics Committee. Subsequently, the researcher introduced himself to the relevant units. The leading researcher obtained lists of nurses from nursing offices and hospital administrators to identify eligible participants. The questionnaires were distributed to participants anonymously, ensuring that ethical considerations were respected. Visits were made to hospitals throughout the week, covering morning, afternoon, and night shifts to ensure accessibility for all nurses. The study’s objectives were explained to potential participants individually, and they then provided written informed consent. Questionnaires were given to eligible nurses, with flexible completion times to accommodate their workload and response speed. The nurses completed the questionnaires independently and collected the completed forms. To ensure consistency, the questionnaires were administered in person, with the researcher present to assist the nurses as they completed them and to answer any questions. Nurses were encouraged to complete the questionnaires at their own pace and anonymously in a relaxed setting to boost response rates. They were also given extra time to finish the questionnaires. The researcher followed up with the participants to ensure the nurses completed the questionnaires. We assured the confidentiality and anonymity of the data provided by the participants.