Measuring Occupational Identity in Hospitalized Older Adults: A Psychometric Examination of the Occupational Identity Questionnaire Open Access

Occupational identity refers to the concept that clients define who they are and who they aspire to be as occupational beings (Kielhofner, 2008). They form this identity by engaging in meaningful activities (Kielhofner, 2008). Previous research has identified three dimensions of occupational identity: construction, discrepancy, and disruption. These dimensions highlight the profound connections between occupation and identity, as well as the roles of belonging and meaning in shaping this concept (Hansson et al., 2022). These findings emphasize that occupational identity is influenced not only by individual actions but also by relationships and the broader context in which individuals engage in occupations. Occupational therapy services are tasked with promoting the acquisition and maintenance of occupational identity for individuals who are dealing with or at risk for illness, injury, disease, disorder, condition, impairment, disability, activity limitation, or participation restriction (American Occupational Therapy Association, 2020). To support their clients’ engagement in activities, occupational therapy practitioners must understand their clients’ occupational identities, thereby aiding in the acquisition and maintenance of these identities.

As people age, they often experience a decline in physical, cognitive, and psychological functions, which makes daily activities more challenging and reduces quality of life (Christensen et al., 2009). A study by Johansson et al. (2023) involving 20 frail older adults revealed that, although these individuals desired to remain physically active and independent in their daily lives, illness and functional limitations often hindered this goal. Christiansen (1999) suggested that the loss of former competencies as a result of aging, disability, or disease may lead to an identity crisis. Consequently, older adults are often confronted with such a crisis.

Howie et al. (2004) argued that occupational identity plays a crucial role in enabling older adults to choose meaningful occupations throughout their lives. Ballmer and Gantschnig (2024) further noted that, for older adults with chronic conditions, autonomy and agency are central to interpreting, navigating, and overcoming daily life challenges that are closely tied to their sense of identity. Thus, although occupational identity can be threatened by aging, its importance increases later in life.

The process of adjustment after a chronic disease, sudden illness, or injury has been shown to be central to reestablishing occupational identity and reacquiring purpose and meaning (Walder & Molineux, 2017). Occupational therapy practitioners possess the skills and knowledge to support older adults in remaining active in their daily lives by engaging them in meaningful activities aligned with their beliefs, values, and needs (Hinojosa & Kramer, 1997). It is essential that occupational therapy practitioners understand older adults’ occupational identity to help them overcome life challenges and remain active, which necessitates their having a deeper understanding of that occupational identity.

As the global population of older adults increases, the need for support for this demographic is also expected to increase. Therefore, the Occupational Identity Questionnaire (OIQ) was developed as a specialized assessment tool to evaluate occupational identity among older adults. Its design is rooted in the theoretical framework of the Model of Human Occupation (Shikata, Notoh, et al., 2021). The OIQ consists of 14 items across three factors: Sense of Past Self, Sense of Present Self and Future Expectations, and Satisfaction With the Current Circumstances (see Table A.1 in the Supplemental Material, available online with this article at https://research.aota.org/ajot). This self-reported assessment was designed for community-dwelling elderly individuals requiring care. Although reports on use of the OIQ are limited, it has been used in both intervention and survey studies in Japan (Shikata, Shinohara, et al., 2021; Shikata et al., 2024). One assessment method similar to the OIQ is the Canadian Occupational Performance Measure, which aims to identify occupations that are significant to clients and track changes in their personal perceptions of occupational performance (Law et al., 2019). Conversely, the OIQ focuses on uncovering clients’ perceptions of occupations as shaped by their personal history of occupational participation, thereby aiming to identify occupations that are meaningful to them.

With the increase in the elderly population, the number of older adults using health care services is increasing as well (Conneely et al., 2023). Hospitalization challenges elderly patients’ ability to engage in activities of daily living and other significant occupations (Andreasen et al., 2015; Gill et al., 2004). The experience of not being able to do and be what they used to may disrupt their occupational identity (Hansson et al., 2022). Therefore, it is crucial that occupational therapy practitioners understand the occupational identity of these individuals and intervene in their lives after hospitalization, potentially contributing to the maintenance or acquisition of occupational identity. However, the reliability and validity of the OIQ have only been verified for community-dwelling older adults, and its use with hospitalized older adults has not been explored.

Thus, this study aimed to investigate the OIQ’s reliability and validity with hospitalized older adults to determine whether it can be effectively used to support their transition to home. This would enable the use of the OIQ with hospitalized or institutionalized older adults, facilitating support for occupational identity, with a focus on the transition to community living.

The inclusion criteria were as follows: hospitalized (in acute care hospitals, community-based integrated care wards, or convalescent rehabilitation wards) or residing in a geriatric health services facility, age 65 yr or older, and able to respond to the questionnaire. The exclusion criterion was a diagnosis of dementia or cognitive dysfunction made by a physician. If older adults understood the study content and were able to express willingness to participate, they were considered to have the cognitive ability to complete the questionnaire. Although geriatric health service facilities are not medical institutions, they were included as target facilities because they are considered intermediate facilities between medical institutions and home that enable patients to return home (Ministry of Health, Labour and Welfare, 2023).

The sample size was determined on the basis of the COnsensus-Based Standards for the Selection of health Measurement INstruments (COSMIN) guidelines, a comprehensive methodological guide for evaluating the methodological quality of patient-reported outcome measures (Mokkink et al., 2018). To meet the “adequate” criteria for both factor and Rasch analyses as outlined by COSMIN, a sample size ranging from 100 to 199 participants was required.

The participants were recruited from hospitals and nursing homes in Japan, including acute care hospitals and rehabilitation wards. Occupational therapists affiliated with these facilities were recruited as the research collaborators. Recruitment was conducted via an announcement on the Japan Society for Occupational Behavior website and through referrals by the researchers’ professional networks. Before study initiation, the purpose and methods were explained to all participants, and they were confirmed to meet the inclusion criteria and not the exclusion criterion. If participants expressed willingness to participate, written consent was obtained before the study commenced. The study was approved by the Ethics Review Committee of Tokoha University (No. 2023-012H).

The OIQ consists of 14 items across three factors, rated on a 4-point scale (1 = not at all, 2 = no, 3 = yes, 4 = very much so), with higher scores indicating a stronger perception of occupational identity. Scores can range from 14 to 56 points. In addition to the OIQ, basic demographic and clinical information—including age, sex, level of care, and FIM™ scores—was collected from the participants.

When a scale includes at least 20 items, item response theory (IRT) is considered superior to classical test theory (CTT; Jabrayilov et al., 2016). Although the OIQ consists of 14 items, which suggests use of the CTT, the drawbacks of the CTT include item and participant dependency (Doucette & Wolf, 2009). Cross-validating a scale with different theories provides more comprehensive measurement information and allows for a deeper investigation into its effectiveness and appropriateness (Gao & Liu, 2024). Therefore, we used both the CTT and the IRT to examine the OIQ’s psychometric properties. Validity was examined with confirmatory factor analysis (CFA) within the CTT framework and principal-components analysis and goodness-of-fit indices within the IRT framework. Reliability was assessed through internal consistency under the CTT, whereas the person separation index, person separation reliability coefficient, item separation index, and item separation reliability were calculated under the IRT.

CFA was conducted to verify the fit of the previously established three-factor structure. Structural equation modeling was performed using robust weighted least squares estimation. The fit was evaluated using the comparative fit index (CFI), standardized root-mean-square residual (SRMR), and root-mean-square error of approximation (RMSEA). The fit criteria were as follows: CFI > 0.90, SRMR ≤ 0.08, and RMSEA < 0.08 (Hair et al., 2014). CFA was performed using R (Version 4.4.0).

Internal consistency was evaluated using Cronbach’s α, which assesses the relationships between multiple items on a scale to determine whether the items measure the same construct. A value of ≥.7 is considered acceptable reliability (Terwee et al., 2007). Cronbach’s α was calculated using IBM SPSS Statistics (Version 25).

We implemented the Rasch rating scale model (RSM), for which it is crucial to confirm unidimensionality. This was achieved through a principal-components analysis of the residuals using the Rasch model. Unidimensionality was deemed acceptable if the explained raw variance exceeded 30%, the eigenvalue of the first unexplained contrast was <3.0, and the essential unidimensionality was >50% (Linacre & Wright, 2000).

The model fit was assessed using the infit mean square (Infit MnSq) and standardized z scores (Zstd). Items or participants were flagged as misfits if Infit MnSq exceeded 1.40 or if |Zstd| was more than 2.00 (Bond & Fox, 2007).

Reliability was evaluated using the person separation index and person separation reliability coefficient, with the latter being analogous to Cronbach’s α. A person separation Index of ≥2.0 and reliability coefficient of ≥.8 were considered satisfactory. Additionally, the item separation index and item separation reliability were calculated to determine whether the sample was sufficient to verify the item difficulty hierarchy. An item separation index of ≥3.0 and item separation reliability of ≥.9 were deemed adequate. Rasch analysis was performed using Winsteps (Version 5.8.2.0).

The participating facilities were located in six cities across Japan, including Tokyo and Shizuoka. In total, 126 participants (48 men, 78 women) with no missing data were enrolled. Table 1 presents participant characteristics (Tables A.2 and A.3 in the Supplemental Material present the definitions of types of facilities and level of care required, and Table A.4 presents the OIQ results). The participant’s minimum score on the cognitive items of the FIM was 15 points, indicating problems with comprehension and expression. The occupational therapist assisted participants in understanding the questionnaire and expressing their answers.

CFA was conducted using a three-factor model. Figure 1 illustrates the CFA results, which show that the model fit the data well (χ²/df = 1.16, CFI = .949, SRMR = .092, RMSEA = .057). However, the standardized coefficients for Items 6 and 12 were low (.32 and .38, respectively).

Cronbach’s α for the entire OIQ was .822, which exceeded the recommended threshold of .70. The α values for Factors 1, 2, and 3 were .788, .736, and .666, respectively.

The OIQ was analyzed using the RSM; Table 2 presents the results. None of the OIQ items were misfitted; however, 12 participants (9.52%) were identified as misfitted. Descriptive statistics for the fitted and misfitted participants were calculated, revealing that misfitted participants had a higher sense of past self and lower FIM scores (Table A.5). A principal-components analysis of the residuals revealed that the explained variance was 39%, the eigenvalue of the first contrast was 2.7, and the essential unidimensionality reached 50%. Item difficulty ranged from −1.28 to 1.55. The most difficult item (i.e., the item that, compared with other items, received lower scores, indicating it had the highest level of difficulty) was Item 12, “I would like to challenge myself to do what I do not do now,” followed by Item 14, “I think that I live as I expected I would,” and Item 7, “I have fun in my life.” The easiest item (i.e., the item for which participants assigned a higher score, indicating less perceived difficulty) was Item 10, “I want to live happily in the future,” followed by Item 2, “Previously, I lived with a well mind and body,” and Item 11, “From now on as well, I want to do what I can do.”

The person separation index was 2.25, with a stratum of 3.33 and a reliability coefficient of .83. The item separation index was 4.94, with a stratum of 6.92 and a reliability coefficient of .96.

The OIQ was initially developed for community-dwelling elderly individuals, and this study aimed to verify its reliability and validity with hospitalized older adults. The study confirmed that the OIQ can be used with hospitalized older adults.

This study assessed the structural validity of the OIQ using CFA with a three-factor model. The CFA results indicate that the model demonstrated adequate fit, meeting all required fit criteria. Although these values were slightly lower than those reported in previous studies with community-dwelling older adults (CFI = .95, SRMR = .07, RMSEA = .034), structural validity for hospitalized older adults was confirmed. Standardized coefficients were low for Items 6 and 12, which may be attributed to the fact that the participants were hospitalized at the onset of their illness or injury and were engaged in rehabilitation for the transition to living at home. In particular, health and recovery status resulting from their illness or injury may have affected the participants’ perception of their own abilities and their willingness to take on future challenges (Mjøsund et al., 2022; van Seben et al., 2019), resulting in inconsistent responses. In this study, participants’ cognition was based on the internationally used FIM (Manee et al., 2020), but information on awareness should also be obtained and examined by using assessments that evaluate self-perception, such as the OIQ. Internal consistency, as measured by Cronbach’s α, exceeded the recommended threshold of .70 for all factors except the third, for which the α was slightly lower. Given that Cronbach’s α depends on the number of items (Sijtsma, 2009), the lower reliability of the third factor may be attributed to its limited number of items (two), suggesting that adding more items could improve the scale’s reliability.

In the RSM analysis, only 12 participants (9.52%) were misfitted, indicating that consistent response patterns were obtained from the older adults included in this study (Kielhofner et al., 2009). The descriptive statistics indicated that the lower FIM scores observed for the misfitted participants could have contributed to their higher perception of underperformance (Item 6), desire to accomplish what they were capable of (Item 11), and a tendency toward lower satisfaction with their current lives (Item 14). No items were identified as misfitted, suggesting that the OIQ items functioned appropriately, and there were no problems at the item level.

Through the RSM, specific items representing varying difficulty levels for hospitalized older adults were identified within the OIQ. The items with high difficulty levels were consistent with those identified in previous studies involving community-dwelling older adults requiring assistance (Shikata, Notoh, et al., 2021). Items 10, 2, and 11 were identified as having the lowest difficulty levels, listed in ascending order. Conversely, a previous study highlighted Items 4, 2, and 11 as the easiest, revealing partially different results. The participants in the current study required medical services for injuries and illnesses and included individuals in acute care hospitals and rehabilitation wards. Because these participants were in the early stages of recovery, their perceptions of prehospitalization as occupational beings likely influenced their desire to return to their former lives. This may explain why the items related to the desire for future independence and enjoyment of life were the easiest to endorse. These findings underscore occupational therapy practitioners’ importance in restoring older adults’ independence and enjoyment of life according to their desires.

Moreover, the person separation index and person separation reliability coefficient met the recommended values, indicating that the OIQ demonstrated high reliability. Low item reliability is often attributed to an insufficient sample size, which can hinder the establishment of a stable item difficulty hierarchy. In this study, the sample size was adequate to support the development of a reliable and reproducible item difficulty hierarchy (Linacre & Wright, 2000). The strata suggest that the OIQ is sufficiently sensitive to distinguish between individuals with high, intermediate, and low levels of self-awareness of occupational identity. The item separation index and reliability coefficient met the required values, further supporting the structural validity of the OIQ. Consequently, the OIQ captures a broad range of item difficulties concerning occupational identity among hospitalized older adults and provides occupational therapy practitioners with valuable information for tailoring interventions according to individual client needs.

This study has several limitations that warrant acknowledgment. The study participants represented a wide range of individuals requiring rehabilitation services because of injury or illness, encompassing settings from acute care hospitals to geriatric health facilities, as they transitioned to home life. However, because the sample was drawn from a specific region of Japan, the generalizability of the findings to other cultural contexts is limited. Moreover, although the study’s sample size is considered adequate, some recommend a minimum sample size of 3 to 20 times the number of variables (Mundfrom et al., 2005). Given these considerations, future research should focus on collecting more data across diverse populations and environments to further validate the OIQ and enhance its applicability. Moreover, although the structural validity and internal consistency of the OIQ were confirmed in this study, other psychometric properties, such as test–retest reliability and predictive validity, remain unexamined and should be addressed in future studies. Longitudinal research could offer valuable insights into how occupational identity evolves over time and the long-term effects of occupational therapy interventions on maintaining or enhancing this identity.

The confirmation of the OIQ’s reliability and validity with hospitalized elderly patients has several significant implications for occupational therapy practice:

• ▪ Although its test–retest reliability has not yet been investigated, the OIQ provides a means to quantitatively assess clients’ perceptions of occupational identity.

• ▪ In addition to its use with community-dwelling older adults, it can be used to measure occupational identity status among older adults who are hospitalized and undergoing rehabilitation aimed at returning to community living transitioning to home.

• ▪ The information obtained through the OIQ can be used to design client-centered occupational therapy programs that align with clients’ values and aspirations, especially in preparation for their return to community living.

The OIQ has been demonstrated to be a reliable and valid tool for assessing the occupational identity of hospitalized older adults. Such individuals are likely to encounter significant threats to their occupational identities. Occupational therapists can use the OIQ to gain a deeper understanding of older adults’ occupational identities and provide the necessary guidance to support their desired posthospitalization life outcomes.

We express our gratitude to the clients and occupational therapists who contributed to this study. This study was financially supported by a grant from the Japan Society for the Promotion of Science KAKENHI (JP23K12672). The authors declare no conflict of interest regarding the research, authorship, or publication of this study.