Reliability and Validity of the Persian Version of the Professional Self Identity Questionnaire (PSIQ)

Background: Professional identity is a vital component of medical education that affects the future practice of medical students. There is a notice of the lack of a reliable and valid Persian scale to measure the students' readiness for professional identity. Objective: This present study aims to determine whether or not the PSIQ can be adapted for Persian healthcare professions. Methods: A cross-sectional study was performed among medical students (years of 4-7) in Shiraz medical school. A total of 175 students completed the Persian version of PSIQ using convenience sampling. Categorical con�rmatory factor analysis (CCFA) and Cronbach's alpha coe�cient were used to investigate the questionnaire's validity and reliability. Moreover, R.3.6.2 software, by using the lavaan and semPlot packages in it, was used for data analysis and the path diagram. Results: The indices of the goodness of �t of the model were used for the professional self-identity, which including root mean square error of approximation (RMSEA), Tuker-Lewise index (TLI), and comparative �t index (CFI) ( RMSEA=0.055, CFI=0.996, and TLI=0.994). The Cronbach's alpha coe�cients for these three factors of "profession-specic tasks," "generic attributes," and "inter-personal tasks" subscales were 0.762, 0.622, and 0.747, respectively (p-value<0.05). The internal consistency of the whole questionnaire measured by Cronbach's alpha was 0.873. The set of these �t statistics show that the hypothesized three-factor model �ts the sample data. Therefore, all nine items were signi�cantly loaded on the three domains (Profession speci�c tasks, Generic attributes, Interpersonal tasks). Conclusion: The Persian translated version of the PSIQ may be an appropriate, valid and reliable tool for assessing medical students' health professional self-identity attitudes in Iran.


Introduction:
Physician training involves far more than acquiring basic sciences, pathophysiology, diagnosis, and management of diseases.Medical students should learn some skills, such as interpersonal communication, empathy, self-care, and professional identity (1).Professional identity as a concept has had an extensive existence in medicine (2).Professional identity can be de ned as a subjective state that a person belongs to a particular professional group (3,4).In medicine, professional identity means how a doctor thinks of himself or herself as a doctor (5).Cruess RL and et al. de ned professional identity as a symbol achieved in stages over time in which the characteristics, norms, and values of the medical profession are internalized.This internalization will result in thinking, acting, and sensation like a physician (6).
Acquiring professional identity has a subjective nature under the in uence of some factors such as social, demographic, and personality (3).According to Cruess et al. study, identity formation includes three domains.The rst domain is the individual identity (personal characteristics, Self-esteem, and life experiences).The second domain is the relational identity (friends, family members, mentors, and colleagues).The third domain is the collective identity (social groups to which an individual belongs or wishes to join) (7).Crossley and Vivekananda-Schmidt show that the professional identity is under the in uence of three factors of 'profession-speci c tasks,' 'generic attributes,' and 'inter-personal tasks' (3).Pratt et al. study show socialization, identity work, and career/role transitions are different aspects of identity construction (8).According to Tagawa's study, professional identity formation evaluates self-control as a professional, awareness of being a medical doctor, re ection as a medical practitioner, execution of social responsibility, and external and internal self-harmonization.
However, medical learners are different from in the experiences and development process of professional identity (9).These ndings show that different experiences on the development of professional identity.Generally, professional identity is affected by the learning environment, existing personal characters, role models and mentors, clinical/non-clinical experiences, the attitude of peers, health care professionals (7), life experience, socialization in academic and clinical settings, and technological and societal factors (5).
Professional identity is a vital component of medical education (3,10).Professional identity affects the future practice of medical students in the workplace.The potential challenges associated with health settings (10) and delays in acquiring this critical competency are barriers for successful movement from medical student to physician (3).The absence of professional identity clarity has greatly impacted a practitioner's con dence in encouraging their professional opinions (10).Therefore, one of the vital missions of the medical education curriculum is to consider the goal for students to develop

Professional identity formation (8).
There are numerous studies on the professional identity of physicians (8,(11)(12)(13)(14).Along with this growth in professional identity, there is a growing concern over the lack of a reliable and valid Persian scale to measure professional identity.One of the questionnaires for measuring this ability is the professional self-identity questionnaire (PSIQ).The purpose of this study is to determine whether or not the PSIQ can be adapted for Persian healthcare professions.Using the Persian validated version of this questionnaire is a way to recognize the curricular features that lead to the development of professional selfidentity in Persian language medical students.

Materials And Methods: Study design and participants:
This study is a quantitative study using a cross-sectional design done on medical students (year of 4-7) in Shiraz Medical School in 2019.The sample size number was determined by an expert statistician based on sample size requirements for structural equation models (15).The PSIQ includes nine items, and ten respondents determined the minimum required sample size for each item (90 students).Due to drop-out probabilities, we increased the sample size to 200 medical students.
The number of medical students in each year was determined and based on the strati ed sampling method.One hundred seventy-ve medical students (year of 4-7) with a mean age of 23.52 ± 1.54 completed the questionnaire (response rate 87.5%).The percentage of withdrawn subjects was 12.5% (25 students).

Data Gathering Tool:
There are validated methods that have been developed for assessing professional identity; one of the tools for evaluating professional identity is the "Professional Self Identity Questionnaire "(PSIQ) (1) that was used in the present study.It was developed in 2009 by Crossley and Vivekananda-Schmidt (5) for examining the curricular features that contribute to the development of a professional identity (1).This questionnaire includes a 9-item that indicates a three-domain.The latent constructs proposed by these three factors are: "profession-speci c tasks" (conducting assessments, using records, dealing with emergencies and teaching), "generic attributes" (cultural awareness, ethical awareness, and re ection), and "interpersonal tasks" (teamwork and communication) (5).Present questionnaire was used in pharmacy(16), medicine (4,17), dentistry(4), nursing(17) professions.Procedure Professional Self Identity Questionnaire (PSIQ) was rst translated into Persian by two independent translators whose rst language was Persian and reached an agreement.The translated version of the PSIQ was reviewed for inconsistencies using the back-translation method by two independent translators who were native English speakers.Back translators were blinded to the original version of the PSIQ.That means they had no access to the original English version.
All translations were reviewed by an expert committee consists of English and medical education experts to reach on consensus on discrepancies.After determining the inconsistencies and resolving them, we distributed the nal translated version of the PSIQ to nal-year medical students.The purpose of the PSIQ was thoroughly described to the students, and we explained that we would use the study results anonymously only for research purposes.All items were scored on a six-point Likert response scale (from 1 as the lowest to six as the highest).The raw subscale scores were transformed into values ranging from 3 to 18, such that higher scores indicated better task performance.

Statistical analyses
Ordered-categorical con rmatory factor analysis (CCFA), initially designed for the Likert response scale, was used to evaluate the construct validity of the questionnaire.By using CCFA, we could measure how well our data can be made using extracted domains.In the present study, we investigated whether or not the hypothesized three-factor model ts the data well.Three criteria, including root mean square error of approximation (RMSEA), Tuker-Lewise index (TLI), and comparative t index (CFI), were used to assess the goodness of t of the model.Values of CFI more than 0.9 can support a good t model, and above 0.95 represents an excellent t model.Moreover, TLI > 0.90 is acceptable, and more than 0.95 is a good t model (18); and RMSEA < 0.08 can support acceptable and good model t, and it under 0.05 can show an excellent model t (19).
To assess the internal reliability of the consistency of the PSIQ, we calculated Cronbach's alpha for each of those subscales.
The coe cient alpha equal to or greater than 0.70 was considered satisfactory reliability of the scale score (20,21).
The statistical data analysis was used by the diagonal weighted least square (DWLS) estimation procedure in the lavaan package in R.3.6.2 software was used to perform the CCFA.Moreover, the path diagram was plotted by semPlot package in this statistical software.

Results:
A total of One hundred seventy-ve participants answered the questionnaires (male: 77, female: 98).The mean age of participants was 23.52 ± 1.54 years.
The results of the CCFA indicated that all values of CFI and TLI were greater than 0.95, and the RMSEA index was less than 0.08, which supported the excellent and satisfactory t of the three-factor CCFA model (Table 1).Moreover, the path diagram showed that all of the items (questions) should have been in profession-speci c tasks, generic attributes, and inter-personal tasks subscales were properly loaded (higher than 0.5) on these dimensions.(Fig. 1).
The result of the reliability test for the three dimensions of the questionnaire is shown in Table2.Cronbach's alpha was used to evaluate the internal consistency of the items with each subscale.The results revealed that the Cronbach's alpha coe cients for the three factors of "profession-speci c tasks," "generic attributes," and "inter-personal tasks" subscales were 0.762, 0.622, and 0.747, respectively (p-value < 0.05) (Table 2).The internal consistency of the whole questionnaire also measured by Cronbach's alpha was 0.873.We used the ANOVA with Tukey post-hoc and T independent tests to compare the students' subscale scores and educational years and gender (Table 3).In Table 3, according to the Tukey test, the "profession-speci c tasks" and "generic and attributes" subscales were statistically signi cant between students at the four and six, four and seven, ve and seven, and also at the six and seven educational years.Moreover, interpersonal tasks were statistically signi cant between students at all educational years except for students who were at the ve and six as well as six and seven education years.In addition, the generic attributes subscale was statistically signi cant across male and female medical students. Discussion: Professional identity is derived from the individual role that people take on in their work (22). it affects perceptions of the medical student's role in medicine (5).Professional identity formation is affected by students' emotional and spiritual maturity (22,23), leading to the difference between being a physician and medical practice (22).It is essential to assess the professional identity status of health professional students by a valid and reliable tool.The present study aimed to investigate whether the 3-factor model from the original version of the PSIQ could be applied to the Persian translation of the scale.
The ndings are consistent with Crossley and Vivekananda-Schmidt study (3).The internal consistency of the Persian (0.873) and original versions were approximately similar (0.93).The result of the present study con rmed that PSIQ includes three factors of "profession-speci c tasks," "generic attributes," and "inter-personal tasks" (3).The internal consistency of the three factors was similar(0.762,0.747, and 0.622).
The value of RMSEA suggests that the 3-factor model ts the population covariance matrix.It is a perfect and valuable t index because of its sensitivity to the number of assessed parameters in the model.
In our study, the CFI, estimated to examine the goodness-of-t a statistical model, is more than the threshold (0.996), suggesting that the 3-factor model ts an independent model.TLI was calculated to measure the difference between the χ2 value of the 3-factor model to the χ2 value of the null model.The values of the TLI are indicative of a good-tting model (24)(25)(26).
The values mentioned above showed that the 3-factor model ts the sample data.There are three subscales for each factor in the questionnaire.
Inter-personal tasks include teamwork and communication (3).Inter-professional teamwork and communication are prerequisites for professional self-identity (27); a Positive attitude toward teamwork is to the positive self-identity of health professions (28).As a treatment team member, developing a professional identity will lead to student help together inpatient care (5).
Profession-speci c tasks include conducting assessments, using records, dealing with emergencies, and teaching(3).
Re ection on medical records is essential for protecting professional image (29).Con dence in dealing with emergencies and education about one's profession appears to be critical determinants of students' attitudes to quali cation(3).Also, a relevant approach for us is based on the understanding that the more profound impact of initial teacher education touches upon the professional identity of the student-teacher.Van Huizen et al. ( 2005) specify that a teacher education's overall goal is best conceived as professional identity development (30).In addition, professional identity is created during all sorts of activities, e.g., bedside and ambulatory teaching, communication skills teaching, small group teaching, ward rounds, etc., and during exchanges in informal settings (31).
Generic attributes include cultural awareness, ethical awareness, and re ection(3).The effectiveness of re ection is reinforced when facilitated by a role model or mentor or carried out as a group activity (7).Re ection is a critical component of professional identity formation and development (32).Also, Family members, friends, home environments, and other outside in uences provide a basis for the professional identity formation process.They can either support the commitment essential to be a medical professional or inhibit it.These factors in the external environment have become increasingly essential as recent generations of students and physicians try to readjust the balance between commitment to medicine, personal well-being, and lifestyle.(7).The strength of this study is the high response rate from a convenience sample, and several signi cant limitations should be considered.First, the present study was done in a single medical school, so generalizability is limited.Second, the study sample only comprises of medical students, and they may not be a good representative of other allied health science students.Third, response bias may occur due to the nature of self-reports.
These study limitations recommend that further data collection is essential for using this questionnaire in other health professions.Other study designs such as using randomization sampling for increasing generalizability and representativeness are recommended.It is recommended that upcoming studies focus more on the factors affecting the formation of the professional identity of health professional learners.
In conclusion, this study revealed that the Persian translated version of the PSIQ is a valid and reliable tool for assessing the health professional self-identity of medical students in Iran.

Figure 1 Path
Figure 1

Table 1
The results of the CCFA for the hypothesized CFA models.

Table 2
Cronbach's alpha, Mean, Standard Deviation, maximum and minimum of the subscales of the PSIQ

Table 3
Comparing subscale scores across educational years and gender groups