Validity and reliability of the Persian version of the Ageism Scale for Dental Students in Ahvaz Jundishapur University of Medical Sciences

Introduction

Population aging is one of the most impressive human achievements indicating sustainable development in human societies resulting from improved literacy, access to medical technologies, and reduced mortality. According to the World Health Organization (WHO), the world’s elderly population has risen from 9.2% in 1990 to 11.7% in 2025 and is predicted to increase significantly to 21.1% (2 billion people) by 2050 (Mohammadi-Shahboulaghi et al. 2018). According to the statistical models, the elderly population in Iran is estimated to increase to 10.5% in 2025 and 21.7% in 2050 (Ardane et al. 2018). Therefore, the phenomenon of aging should be considered as a major challenge of the future to be addressed.
Along with the worldwide increase of the elderly population, the phenomenon of ageism has been proposed. The WHO defines ageism as “stereotyping and discrimination against individuals or groups solely based on their age” (WHO 2020). In this regard, Butler believes that age discrimination not only is as important as racism but also can be considered as a major problem in the coming years because according to his theory, “not all of us turn black or white, but we all grow old” (Butler 1969). Unlike other forms of discrimination, such as sexism and racism, ageism against the elderly is socially acceptable, fully institutionalized, unrecognizable, and unchallenged (Officer et al. 2016). Ageism is prevalent in government agencies, families, health care systems, labor market, and media in various forms, such as negligence, early retirement, restrictions on receiving social services, stereotypes, and media misconceptions about the elderly. As a consequence of these factors, the seniors’ psychological and physical health is affected by social isolation (Sum et al. 2016). Negative attitudes and stereotypes about the senior usually represent them as sick, incapacitated, with a crooked and inappropriate appearance, affected with dementia and mental illness, useless to society, isolated, low-income, and depressed (Palmore 2014). Although older people are the largest group of recipients of health care services, negative attitudes toward them have been observed among the health care providers reflected in their treatment choices and decisions for the elderly (Ben-Harush et al. 2017).
This type of attitude in nurses and medical students has led to shorter and shallower conversations with less humor in dealing with the elderly (Eymard and Douglas 2012). A longitudinal study showed that the elderly who perceived age discrimination by the medical staff had a lower quality of life than others (Jackson et al. 2019). A 6-year study also found that the risk of heart disease, arthritis, diabetes, and depression was significantly higher in older people who experienced age discrimination (Jackson et al. 2019).
Ageism leads to the misconception that providing care to the elderly is not necessary and desirable. These negative and incorrect attitudes towards the elderly are deemed serious obstacles to the formulation of appropriate and effective policies for this age group (De Visschere et al. 2009).
Studies in different parts of the world indicate that dentists are no exception to ageism. They usually have little knowledge of physiological, pathological, and psychological changes in old age so that their lack of knowledge and information has decreased their willingness to work with seniors (Tahani et al. 2019).
Given the growing proportion of the elderly population in Iran and the resulted increasing burden of their referrals to receive dental services in the coming years, identifying the medical staff attitudes towards ageism against seniors is of paramount importance. The findings can pave the way for conducting educational interventions to provide appropriate and equitable health care services (Koch-Filho et al. 2017).
Numerous questionnaires have been developed and administered to assess and determine ageism as well as age discrimination attitudes and stereotypes against the elderly: Children’s Attitudes toward Older People (Mehri et al. 2020), the Aging Attitude Scale (Robinson and Howatson-Jones 2014), the Fraboni Scale of Ageism (Fraboni et al. 1990), the Perspectives on Caring for Older Patients Scale (Lucchetti et al. 2018), and the Maxwell-Sullivan Attitudes Scale (Burbank et al. 2018).
Recently, an age discrimination scale was designed for dental students, known as the ASDS by American and European geriatricians (Rucker et al. 2018). The validity and reliability of this scale have been confirmed by three studies in Brazil, the United States, and Greece (Kossioni et al. 2019; Rucker et al. 2020; Rucker et al. 2019). This scale contains 27 items related to geriatric dentistry, which should be answered based on a six-point Likert scale, ranging from strongly disagree (zero points) to strongly agree (6 points) (Rucker et al. 2018). The most important questions of this scale are related to the time-consuming work with the elderly, complexities of taking the patient’s history considering their comorbidities, rejection of the treatment plans by seniors, high costs of providing dental care at home, and inadequacy of spending such high costs with regard to the remaining lifespan of the elderly (Rucker et al. 2018).
To the best of our knowledge, the ASDS has not been validated in Persian. So, due to the increasing number of elderly persons in Iran and the increasing burden of their referrals to receive dental services, the present study aimed to evaluate the validity and reliability of the Persian version of the ASDS among the dental students of Ahvaz University of Medical Sciences.

Material and methods

The statistical population of this study consisted of all clinical dental students in Ahvaz University of Medical Sciences in 2020 selected using the census method. The sample size was estimated as 5-20 respondents per item of the questionnaire based on the structural equation modeling approach. In this regard, many researchers consider that an average of 10 people would be sufficient per item. Since the number of items in the main tool is 27 items, the sample size was estimated as 270. The selected participants included all dental students who entered university in 2020. Inclusion criteria were studying in the dental school of Ahvaz University of Medical Sciences, being willing to participate in the study, and completing the informed consent form. Exclusion criteria included unwillingness to continue participating in the study or incomplete responses to the questionnaires. The research tools included demographic questionnaires and the ASDS designed by Ryan Rocker in 2018. The ASDS contains 27 items related to geriatric dentistry. All items should be answered on a six-point Likert scale ranging from strongly disagree (zero points) to strongly agree (6 points).
After obtaining formal permission from the scale developer, the ASDS was translated into Persian. Later, two professional English translators were asked to translate the Persian version back into English. The obtained translated versions were compared and revised by the researchers and the final version was approved.
In order to determine face validity of the translated scale, the impact score was calculated for readability of each question. To this end, a 5-point Likert scale was used for each item: very strong (5 points), strong (4 points), moderate (3 points), poor (2 points), and very poor (1 point). Later, 10 dental students were asked to read and examine the translated items and determine their validity and clarity based on the developed Likert scale. Face validity of the Persian version of the questionnaire was calculated using the item impact method (impact score = frequency (%) × importance). Impact scores greater than 1.5 were considered appropriate.
To determine the content validity index (CVI), a panel of experts, including several experts and professors in the field of geriatrics, was asked to review the scale meticulously and write their revisions in detail with regard to the relevance, simplicity, and clarity of the translated items. As a result, the revisions recommended by experts were evaluated and finalized by the research team and the required changes were applied in the questionnaire. The content validity index was calculated using the CVI formula (ratio of the number of evaluators who rated an item with a score of 3 or 4 divided by the total number of evaluators). Scores higher than 0.79 indicate appropriateness of the questionnaire, a score within the range of 0.70 to 0.79 shows that the item needs revision, and scores lower than 0.70 are unacceptable.
Given that ensuring sufficiency of the sample size is necessary for using the factor analysis method and determining construct validity of the translated scale, the sample adequacy index developed by Kaiser, Meyer and Olkin (KMO) was applied, which should be higher than 0.7. Exploratory factor analysis was also performed with varimax rotation. Moreover, eigenvalues of greater than one and pebble curves were used to extract the factors. Considering that the minimum factor load was 0.5, items with a factor load of less than 0.5 were removed. Finally, factors that played the most important role in explaining variance of the data were extracted as effective factors. Construct validity of the questionnaire was re-evaluated and confirmatory factor analysis was performed by collecting opinions of the study participants via Smart PLS software version 3.
In order to observe ethical considerations, the participants were ensured about confidentiality and anonymity of their information. To this end, all questionnaires were coded. The Ethics Code of the present study is IR.AJUMS REC.1399.415.

Results

The mean age of participants was 24.74 years with a standard deviation of 6.01 years. Of all participants, 174 (64.7%) were women; 88 (32.7%) and 78 (29%) students entered university in 2015 and 2016, respectively; 38 (51.3%) of the participants lived with their parents; 58 (21.6%) stated that they had elderly parents; and 246 (91.4%) had passed the geriatrics course.
The findings corroborated appropriate face (impact factor = higher than 1.5) and content (impact factor = higher than 0.79) validity of the translated questionnaire (Table 1).
Reliability of the translated questionnaire was confirmed using internal consistency (Cron-bach’s α = 0.712) and ICC of 0.646 (Table 2).
Prior to application of the factor analysis method, sufficiency of the sample size was corroborated for exploratory factor analysis by calculating the sample adequacy index (Table 3).
According to Table 3, the KMO statistic value is 0.70, showing that the data are suitable for exploratory factor analysis. Furthermore, results of the Bartlett sphericity test are also significant, confirming a significant correlation between the variables.
Exploratory factor analysis was performed by varimax rotation. The eigenvalue of greater than one and the pebble curves were used to extract the factors. The minimum factor load was considered as 0.5 and items with a factor load of less than 0.5 were removed. As a result, five factors were extracted, which explained 59.55% of the total variance.
Table 4 represents the contribution of variables in the factors after rotation. Each variable was attributed to its factor based on high correlation coefficients.
In order to evaluate construct validity of the questionnaire, all participants were required to confirm 15 questions of the questionnaire related to five factors of the confirmatory factor analysis. As a result, 13 questions were confirmed and two questions (questions 5 and 27) were removed from the items due to their lowest factor loadings (Fig. 1).

Discussion

Based on the findings, the Persian version of the ASDS has desirable levels of reliability and validity. The CVI and content validity ratio (CVR) values were obtained at an acceptable level. Followed by exploratory factor analysis, five factors were identified, which explained 59.55% of the total variance. The ICC of 0.64 and Cronbach’s α of 0.74 confirmed appropriate reliability of this scale. These results are consistent with other studies conducted on this scale by Ryan Rucker and Anastassia E. Kossioni in the United States, Brazil, and Greece (Kossioni et al. 2019; Rucker et al. 2020; Rucker et al. 2019). In these studies, one to five factors were determined that predicted 51% to 63% of the total variance. In this regard, the highest consistency in the findings was associated with the study by Ryan Rucker in the United States (Kossioni et al. 2019). Their estimated Cronbach’s α coefficient was completely consistent with our findings. Considering that the appropriate value for Cronbach’s α is 0.7 (Cronbach 1951), for combined reliability it is 0.7 (Nunnally 1978), and for the average variance extracted it is 0.5 (Fornell and Larcker 1981), our findings confirmed appropriate reliability and validity of the translated questionnaire concerning these variables.
Based on the findings, the method proposed by Fornell and Larcker (1981) shows that the average variance extracted root of the latent variables is greater than the correlation value between them. Therefore, it can be stated that the model constructs (hidden variables) have more interaction with their indices than with other constructs; in other words, the divergent validity of the model is appropriate.

Conclusions

Based on the findings, the Persian version of the ASDS has good levels of validity and reliability. So, it can be included in the health evaluation checklist of the Iranian seniors as an auxiliary tool in determining age discrimination against them.

Acknowledgements

This study was a part of a dissertation by Ms. Golnaz Hosseinipur in the School of Dentistry. This study was supported by Social Determinants of Health Research Center in Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran. The authors would like to express their gratitude to all participants for their cooperation in this study.

Statement of ethics

This study was part of a dissertation in the School of Dentistry with the ethics committee number IR.AJUMS.REC.1399.415 and was implemented with the support of Social Determinant of Health Research Center (Reference No. SDH-9920), Deputy of Research in Ahwaz Jundishapur University of Medical Sciences.

Disclosure

The authors declare no conflict of interest.

References