Abstract
This study assesses the Professional Belonging Scale for Occupational Therapy Students, a self-assessment that was created by the authors to measure students’ sense of belonging in occupational therapy education programs.
Belonging refers to the dynamic perception of feeling deeply connected, included, valued, and respected through support and reciprocity in the context of experiences and relationships with people, communities, systems, cultures, and places (Allen et al., 2021; Hammell, 2014; Hitch et al., 2014). Researchers have identified that belonging is not only an individual experience but also a communal and contextual one that may help mitigate loneliness for individuals, groups, and communities (Allen et al., 2021; Lim et al., 2021; Pascale, 2018). Belonging is crucial for students transitioning into college and professions (Gray et al., 2020). Students who feel they belong during health care education have demonstrated increased well-being, motivation, competence, professional identity, retention, degree completion, and preparation for careers (Berg-Poppe et al., 2017; Gray et al., 2020; Samz et al., 2021; Soerensen et al., 2023; Strayhorn, 2018). However, according to the 2020 National Survey of Student Engagement, students rated their belonging an average of 40 out of 60. Students identifying with historically minoritized racial and ethnic groups, the LGBTQ+ community, nonbinary gender identities, disabilities, and first-generation status have reported lower ratings and unique belonging needs (National Survey of Student Engagement, n.d.; Taff & Clifton, 2022; Tyminski et al., 2023). Therefore, higher education institutions must understand the belonging needs of all students, especially those from diverse backgrounds, to ensure support.
Measuring Belonging in Occupational Therapy Education
The American Occupational Therapy Association (AOTA) advocates for a diverse, equitable, and inclusive workforce and calls for the professional community to recognize belonging as a right with implications for sustainability (AOTA, 2020; Taff & Blash, 2017). Academic programs must demonstrate how diversity, equity, and inclusion are integral to occupational therapy education, including admissions, curriculum, and preparation for practice (Brown et al., 2021). To ensure inclusive education, it is essential to understand contextual supports and barriers to belonging. Higher education institutions have assessed student belonging with various tools, including the Sense of Social Fit Scale (Walton & Cohen (2007) and the University Belonging Questionnaire (Slaten et al., 2018). Discipline-specific scales, such as the Belongingness Scale–Clinical Placement Experience (Levett-Jones et al., 2009) and the Midwifery Belonging Scale (Baskaya et al., 2020), have measured students’ belonging within their fields.
No published tools measure belonging specifically among students in occupational therapy education programs. Existing scales are not likely adequate to assess belonging within occupational therapy education because the experience of belonging is context-dependent (Allen et al., 2021). This study aimed to (1) create the Professional Belonging Scale for Occupational Therapy Students (PBSOTS) and (2) assess its validity and internal consistency reliability for measuring belonging among occupational therapy and occupational therapy assistant students. In this article, we describe the process of scale development and initial psychometric study.
Method
This study was conducted in three phases from January 1, 2022, to May 8, 2023. Following scale development, we conducted a preliminary psychometric study. The institutional review boards of two universities in the northeastern United States approved this study (IRBNet ID 1844717-2 and 1980697-2).
Phase 1: Scale Development
We used a literature review, cognitive interviews, and a student focus group to develop the PBSOTS. First, we appraised literature on belonging, belonging scales, and scale development guidelines to identify themes related to the content domain of belonging and to create items. Next, we conducted cognitive interviews with three occupational therapy community members whose experience, research, or service demonstrated expertise in belonging, diversity, equity, inclusion, and assessment. To improve the scale and provide a comprehensive assessment, we asked experts to share their thoughts aloud as they reviewed each survey item for relevance to belonging and recommended edits during cognitive interviews (Willis & Artino, 2013). After member-checking and sharing the revised scale with the experts, we presented the PBSOTS to a convenience sample of 10 occupational therapy students from our institution, consistent with focus group size recommendations from Krueger and Casey (2008). The students discussed item clarity, readability, and impressions of the PBSOTS related to what they would expect in a belonging questionnaire.
Phase 2: Content Validity
We invited a convenience sample of occupational therapy experts in belonging, diversity, equity, inclusion, or assessment to participate in an anonymous QualtricsXM survey using the content validity index (CVI; Polit et al., 2007). Recruitment included the experts from Phase 1 plus 13 additional experts, aiming for 8 to 12 experts. We selected experts based on service on national committees focusing on diversity, equity, and inclusion and related research publications. After consenting, experts rated the relevance of each scale item to the construct of belonging using a 4-point Likert scale (1 = not relevant to 4 = highly relevant). The survey included an optional comment box for each item. We calculated item–CVI (I-CVI) by dividing the total 3 or 4 ratings by the number of experts and the scale–CVI average (S-CVI) values using Microsoft Excel (Version 16.53), considering an I-CVI of .78 or higher and S-CVI of .80 or higher as evidence of good content validity (Polit et al., 2007). We retained, revised, or removed items based on CVI ratings, expert comments, and literature.
Phase 3: Exploratory Factor Analysis to Help Determine Internal Consistency
Phase 3 included occupational therapy and occupational therapy assistant students from Accreditation Council for Occupational Therapy Education (ACOTE®)–accredited programs who were older than 18 yr and could read, write, and understand English. Students at our institution were excluded because of the risk of bias. We emailed the directors of all ACOTE-accredited programs a recruitment flyer to share with students. Students who wished to participate followed a link to a QualtricsXM survey that included demographic and PBSOTS items. The response rate could not be calculated because of the anonymity of participants and the inability to know how many program directors shared the study with students at their institution. Students were required to consent before beginning the survey. We transferred the data from QualtricsXM to IBM SPSS (Version 28.0). We calculated Cronbach’s α coefficient to assess internal consistency and conducted an exploratory factor analysis after verifying sampling adequacy with the Kaiser–Meyer–Okin (KMO) measure.
Results
Figure 1 shows a summary of the results for each study phase. In Phase 1, we developed a 24-item PBSOTS based on a literature review, expert recommendations, and student input. The tool included directions and definitions for belonging, occupational therapy program, and occupational therapy community. Each scale item was followed by a 5-point Likert scale response option (1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = strongly agree).

Summary of study phases and results of the Professional Belonging Scale for Occupational Therapy Students
In Phase 2, 11 experts rated the 24 items using the CVI survey. See Table 1 for the items and I-CVI values. Items 4, 12, and 13 were removed from the scale because of low I-CVI values (<.78). The S-CVI for the remaining 21 items was .84. Three items with low I-CVI values were retained during this phase: Item 14 was kept on the basis of literature linking self-efficacy and belonging, and Items 3 and 7 were included after expert feedback prompted revisions to their wording. Three additional revisions were made. Item 1 was reworded from “occupational therapist” to “member of the occupational therapy community” to include students pursuing the occupational therapy assistant degree. Item 11 was changed from “an occupational therapy program” to “my occupational therapy program” to increase specificity. We changed the wording of Item 16 from positive to negative to prevent response bias. Before Phase 3 recruitment, following expert feedback, we added Item 25 to distinguish perspectives of feeling valued by others from perspectives of being valuable members of their programs, represented by Item 9.
Phase 2 PBSOTS Items and I-CVI Values
Note. Items with item content validity index (I-CVI) values at or above the cutoff score of .78 are shown in bold type. OT = occupational therapy; PBSOTS = Professional Belonging Scale for Occupational Therapy Students.
During Phase 3, 162 students completed the PBSOTS. See Table 2 for participants’ demographic information. We initially conducted an exploratory factor analysis with data from 22 PBSOTS items (determinant value = 0.000006645) before excluding data from Items 3 and 14 (determinant value = 0.00002012) to meet a determinant value threshold greater than 0.00001 (Field, 2013). See Table 3 for the results of the principal axis factor analysis conducted on the 20 items with a varimax rotation, suppressing factor loading < .30. Bartlett’s test of sphericity was p < .05. The KMO test verified sampling adequacy (KMO = 0.91), with all KMO values for individual items greater than 0.50. Four factors had eigenvalues greater than Kaiser’s criteria of 1 and explained 61.13% of the variance. Cronbach α values for item clustering included the following: Factor 1, Cronbach α = .90; Factor 2, Cronbach α = .78; Factor 3, Cronbach α = .57; and Factor 4, Cronbach α = .48. The α value for the 20-item scale was .92.
Phase 3 Student Demographic Information (N = 161)
Note. A total of 162 participants completed all PBSOTS items, but 1 participant did not answer demographic questions.
Participants were asked to check all that apply; percentages were calculated for 162 participants.
Results From an Exploratory Factor Analysis of the PBSOTS
Note. In this table, 20 items are grouped by the factors on which they have primary loadings and arranged in the order in which the factors were identified, with the primary loadings shown in boldface. OT = occupational therapy; PBSOTS = Professional Belonging Scale for Occupational Therapy Students.
Discussion
This study introduced the PBSOTS to measure belonging among occupational therapy and occupational therapy assistant students. To our knowledge, this is the first assessment of belonging designed for occupational therapy education. Literature and expert interviews were instrumental in shaping the construct of belonging and developing the items. We assessed item and scale validity with a sufficient sample of 11 experts (Polit et al., 2007). We further enhanced the validity and internal consistency reliability of the scale by excluding two items from analysis in Phase 3. The revised PBSOTS demonstrated good content validity with S-CVI greater than .8 (Polit et al., 2007) and a high α value (.92) when tested with a sample of 162 students with demographics similar to the overall population of students in accredited programs (AOTA, 2023). Cronbach’s α can be affected by the number of items and dimensionality of the construct, and high α values (>.90) have been interpreted differently across studies, for example, as strong interrelatedness among test items or redundancy in test items (Tavakol & Dennick, 2011).
We investigated possible dimensions of belonging through a factor analysis and found four potential factors. Although the PBSOTS is brief, it is comparable with other established assessments with multiple subscales. Statistics, including the KMO measure and Bartlett’s test, indicated that the sample size was adequate and there was a strong relationship between variables (Shrestha, 2021), which suggested the appropriateness of an exploratory factor analysis to identify aspects of belonging measured in the study’s sample. Instability was found in the fourth factor because it had fewer than three items (Velicer & Fava, 1998), which emphasizes the challenge of capturing all aspects of belonging within a fixed number of scale items. We refrained from prematurely naming or eliminating factors and instead examined findings in light of existing literature to enhance the understanding of student belonging.
Despite some consensus on the structure of belonging scales having three or four factors, there are discrepancies in how factors have been interpreted in other studies. For example, Slaten et al. (2018) developed the 24-item University Belonging Questionnaire (α = .93), including three subscales (University Affiliation, α = .92; University Support and Acceptance, α = .85; and Faculty and Staff Relations, α = .88). Walton and Cohen’s (2007) 17-item, single-factor Sense of Social Fit Scale was validated by Maghsoodi et al. (2023) using a multidimensional model with high α values for the scale (.90) and four subscales (Identification With the University, Social Match, Social Acceptance, and Cultural Capital). The 30-item Midwifery Belonging Scale, piloted with 300 midwives, demonstrated a scale α value of .91 and four factors (Emotional Belonging, α = .88; Performance of Professional Roles and Responsibilities, α = .85; Professional Development and Utilization of Opportunities, α = .91; and Limits of Duty and Authority in the Profession, α = .81; Baskaya et al., 2020). The 34-item Belonging Scale–Clinical Placement Experience, when tested with 362 nursing students, showed a high α value for the scale (.92) and three subscales (Esteem, α = .9; Connectedness, α = .82; and Efficacy, α = .8; Levett-Jones et al., 2009). A comparison of the PBSOTS with these assessments suggests that the PBSOTS requires refinement before a factor structure can be claimed, which underscores the need for more rigorous and consistent methodologies to develop and validate these scales. Such an approach could lead to a more universally accepted understanding of what constitutes belonging for students and address the ongoing challenges and complexities in measuring a multilayered construct.
Limitations
This study had several limitations. First, although the CVI survey allowed experts to comment on the relevance of items to belonging, we could not verify an accurate interpretation of these comments. It would have been beneficial to rerun the expert survey after scale revisions before administering it to students. Second, keeping items that did not reach the desired I-CVI between Phases 2 and 3 created a gap in methodology. For this reason, we reran the exploratory factor analysis using data from 20 items. Last, we note the challenge of representing the universal domain of belonging. Belonging is historically not well-defined, and there is little agreement on interpreting factors across belonging assessments. Although this study measured the scale’s content validity, which is important in construct validation, we recognize there was considerable variability in defining the domain of belonging. We also acknowledge that the 20 PBSOTS items, while relevant, may not represent all dimensions of belonging. This study highlights methodological challenges in measuring and defining belonging, emphasizing the need for precise validation methods and a broader consensus on defining student belonging in educational research.
Future Research Directions
We propose further refinement and research of the PBSOTS, including exploring new theoretical frameworks to gain a more comprehensive understanding of belonging as a dynamic system and conducting expert interviews to identify a comprehensive scale relevant to the domain of belonging in occupational therapy education. The CVI should be recalculated upon revisions before administering the scale to a student sample to examine internal consistency reliability and explore factors. Eventually, scoring guidelines should be established, and utility studies should be conducted. It is our hope that scholars within the professional community will consider this study and collaborate to understand, define, and prioritize belonging in occupational therapy education. These ongoing efforts are essential in developing a valid and reliable assessment that accurately represents the broad, multidimensional construct of belonging and serves as a foundation for meaningful discussion and action toward creating a professional community that fosters a sense of belonging for every member.
Implications for Occupational Therapy Education
The following points offer a rationale for continued development and research of the PBSOTS: ▪ Belonging is emphasized as a vital construct within occupational therapy education. ▪ Faculty and administrators who want to improve occupational therapy and occupational therapy assistant student outcomes may be interested in a tool that offers dependable results related to students’ sense of belonging. ▪ To our knowledge, the PBSOTS is the first assessment designed to measure belonging for students in occupational therapy education programs. ▪ The PBSOTS may inform academic teams about student belonging if the tool is refined and researched.
Conclusion
Student belonging is important to academic success and career transitions. Students who feel connected and valued within their school community are more likely to be engaged, motivated, and resilient when faced with academic challenges. To support the study of belonging in occupational therapy education, we developed a 20-item self-assessment questionnaire called PBSOTS. This tool aims to provide educators and administrators with empirical data on students’ subjective experiences in occupational therapy programs. Establishing the reliability and validity of the PBSOTS is crucial to building trust in the scale’s ability to provide meaningful insight into student belonging. Achieving a broader consensus on belonging is crucial for standardizing measurements and interpretations, allowing for more effective comparisons across various educational settings. The initial content validity and internal consistency reliability of the PBSOTS indicate that further refinement is required, and this research may lead to a resource that can help academic teams evaluate students’ belonging.
Footnotes
Acknowledgments
We acknowledge Rochelle Mendonca, PhD, OTR/L, Assistant Professor of Rehabilitation and Regenerative Medicine and Director of Post-Professional Programs in Occupational Therapy, Columbia University Medical Center, for contributing to the study design and supporting our pursuit to learn about belonging assessment.
