Psychometric properties of the physical environment problems and barriers scale in older adults’ homes using an ergonomic approach

Abstract

BACKGROUND:

As older adults prefer to live in their homes, it is of paramount importance to examine the adaptation of the environment to the older adults’ capabilities.

OBJECTIVE:

The aim of this study was to develop and evaluate psychometric properties of a scale to measure physical environment problems and barriers in older adults’ homes using an ergonomic approach in Iran.

METHODS:

This mixed-method investigation was conducted in two stages in Yazd, Iran. The primary 71-item version of the questionnaire was developed according to qualitative findings and a thorough review of the literature. Then the psychometric characteristics, including face, content, construct validity, were assessed. Content validity was also assessed using CVI and CVR. Finally, its reliability and construct validity were confirmed by composite reliability (CR), Fornell-Larker matrix, and confirmatory factor analysis (CFA) using Smart PLS software version 3.

RESULTS:

The face validity of the developed scale was acceptable, and the mean scores of CVI and CVR were 0.78 and 0.84, respectively. The preliminary draft of the scale was categorized into seven dimensions. Factor validity and reliability were confirmed by acceptable factor loadings, and desirable realms of composite reliability (>0.7) average variance extracted (>0.5). The cross-loading method and the Fornell-Larker matrix were used to evaluate the divergent validity of the scale, and the results confirmed its acceptable fit.

CONCLUSIONS:

The findings indicated that reliability reached acceptable values, and different aspects of validity were almost confirmed. Accordingly, the questionnaire was to measure physical environment problems and barriers in older adults’ homes; however, it requires further validation for future use in other contexts.

Keywords

Psychometrics older adults barriers home environment ergonomic

1 Introduction

Older adults prefer to spend their time at home [1]. In this regard, their tendency to live in their homes with their families and communities is called “aging in place” [2 –4]. The home environment plays a vital role in their autonomy, health status, and independence [5]; however, the main problem is that their homes are not designed based on their needs and abilities, and older adult population cannot immediately meet their needs due to their disabilities and limitations [6]. In the aging process, various organs of the body degrade [7]. This degradation can affect the performance and work activities of the older adults [8 –10]. Studies have also revealed that more than half of older adults have difficulty doing daily routine activities such as bathing and dressing, and that the rate of disability increases in older age groups [11, 12]. Given that the existing houses are often not suitable for older adults and the disabled, some improvements may reduce their dependence on daily activities, reduce the risk of accidents, and enable them to spend more hours safely and independently at home [13].

Environment adaptation significantly affects a person’s quality of life (QOL), spatial attachment, and a sense of well-being and belonging [14]. Regarding older adults, adaptation is the structural modification of the physical environment to remove barriers so that older adults can access the facilities of their living environment without others’ assistance [15]. One of the main goals of ergonomics is environment adaptation for humans by designing the environment properly and paying attention to human capabilities [16, 17].

Appropriate scales can be used to investigate the problems and barriers of the physical environment for older adults. As such, the scales play a critical role in identifying barriers and planning to eliminate them. Various measurement scales have been developed to study the physical environment of houses, including Evaluation of Older People’s Living Environments (EVOLVE) by Lewis et al. in the United Kingdom [17, 18]. This scale examines the characteristics and physical design of older adults’ living environment, the suitability of the living environment, and its relationship with older adults’ QOL [19]. It addresses all living spaces, including entrance, hallway, hall, living room, kitchen, bathroom, and bedroom, as well as assistive technologies used by their residents. This scale can be used to design, evaluate, and relocate older adults’ living environment [17]. One of the disadvantages of this scale is that it contains a large number of items (about 1900 items in 60 pages), and older adults themselves do not complete it. In other words, it is an object-oriented method that includes a detailed checklist. Physical Fitness Questionnaire for Nursing Homes is another scale evaluating, which contains 12 subscales and 70 items [20]. This scale is also an expert-oriented checklist, in which the adaptation is examined without evaluating the disability of the elderly. In another study by Hrovatin et al. in Slovenia, a questionnaire was developed to assess the environment of older adults’ homes ergonomically [21], which disregarded the relationship between older adults’ limitations and living environment.

The existing scales for assessing the living environment of older adults are primarily checklists addressing cultural and environmental conditions. The way older adults interact with the barriers and constraints in the physical home environment is not usually explored; however, the issue is of paramount significance. Accordingly, in this study, a developed scale was used as a basis for identifying the problems and barriers of the physical environment of older adults’ homes regarding the type of limitations and disabilities of older adults. In other words, when the needs of an environment exceed an individual’s capabilities, an unadaptable situation arises. Therefore, the physical, mental, and cognitive limitations may challenge older adults to meet the environmental needs [22]. The problems of older adults in the home environment are rooted in their decreased abilities; hence, classifying barriers and problems of the physical home environment according to the limitations of older adults can be a new approach in scales to delve into the problems and barriers of the physical environment in older adults’ homes.

Given the significance of identifying barriers and problems in the physical environment of older adults’ homes and the lack of appropriate and valid scales in this field, a suitable and reliable scale is necessary. On the other hand, the adjustment and design of measurement scales should be in accordance with the native culture of the community under study. To this end, the present study aimed to develop and evaluate the psychometric properties of a scale to measure physical environment problems and barriers in older adults’ homes by adopting an ergonomic approach.

2 Materials and methods

2.1 Type of study

In the present study, an instrument was developed based on qualitative results and the literature review and was then employed to collect quantitative data. This study was conducted to develop and evaluate the psychometric properties of a scale to measure physical environment problems and barriers in older adults’ homes by adopting an ergonomic approach. The research setting of this study was the homes of older adults residing in Yazd, Iran, in 2020.

2.2 Inclusion and exclusion criteria

Inclusion criteria consisted of age above 60 and residence in Yazd. On the other hand, older adults living in nursing homes, including full- and part-time centers, and older adults suffering from a remarkable disability that made them fully dependent on others, were excluded from the present study.

2.3 Scale development

The present study was a mixed-methods study using both qualitative and quantitative phases. In the qualitative phases, semi-structured interviews were conducted to ask the older adults residing in their homes in Yazd about the barriers and problems of the physical environment of homes. This qualitative study adopted the conventional content analysis and encompassed 53 participants, including 36 61–94-year-old older adults and 17 24–58-year-old caregivers. Then, the questionnaire items were developed using the data extracted from interviews and the literature review. In the quantitative phase, the psychometric properties of the questionnaire were assessed. Then some items were merged, or some others were modified. The items formed the first draft of the questionnaire. In the next step, the psychometric evaluation of the scale was performed. Face validity, content validity, and construct validity were used to determine the scale’s validity. In this regard, face validity was first examined using both qualitative and quantitative methods.

2.3.1 Face validity

Face validity was investigated both qualitatively and quantitatively. From the qualitative perspective, face validity addresses whether the appearance of the instrument is associated with what it measures. To investigate qualitative face validity, 10 participants (older adults) meeting the inclusion criteria were asked to comment on the items’ level of difficulty, appropriateness, and ambiguity [23, 24].

To determine face validity quantitatively, the item impact was applied. The scale was submitted to 10 older adults, who were asked to select one of the options from a Likert scale to show the significance of each item. The item impact scores > 1.5 were considered appropriate [25]. The impact score of each item was calculated by the following equation [26]. $Impact Score = Frequency (%) \times Importance$ (1)

2.3.2 Content validity

Content validity addresses the relevance of an instrument’s content to the construct it measures [27]. Both quantitative and qualitative methods were used to assess content validity [28]. Qualitative content validity was assessed by experts on qualitative and instrument development, ergonomists and gerontologists [29]. In the qualitative phase, they evaluated wording, grammar, item allocation, and scaling of the questionnaire. In the quantitative phase, the content validity index (CVI) and the content validity ratio (CVR) were calculated. The clarity, simplicity, and relevancy of each item were assessed by CVI. To determine the content validity ratio, 10 experts were asked to review each phrase based on a 3-point Likert scale (namely necessary, useful but not necessary, and not necessary). In this regard, according to the Lawshe Table, phrases whose content validity ratio was≥0.62 were retained [30].

To determine the content validity index, the method proposed by Waltz and Basel and the opinions of 10 experts on a 4-point Likert scale were used to assess the relevance, simplicity, and clarity criteria for each phrase in the questionnaire. According to Waltz and Basel’s method, items with a score > 0.79, 0.70–0.79, and < 0.70 are appropriate, need to be modified and reviewed, and are unacceptable, respectively. After modifying and reviewing the items with CVI between 0.7 and 0.79, they were re-submitted to the experts, and their content validity index was evaluated once more. The items with scores < 0.7 were deleted. In the next step, the questionnaire was calculated based on the mean, ratio, and content validity index [31].

2.3.3 Construct validity

To perform factor analysis by using the PLS (partial least squares regression) method, sampling was performed on 200 eligible older adults. According to some sources, this number is sufficient to perform confirmatory factor analysis (CFA) [32, 33]. The sampling method was multi-stage, including selecting the municipal areas and older adults using a systematic random method. PLS test is based on variance and can be used to evaluate formative (real) items. The test needs to check the reliability and validity of research constructs and scales. The reliability of the test depends on the accuracy of the measurement and its stability, so it implies two different definitions, including the reliability (stability) and consistency of test scores over time. This means that if a test is performed several times on a respondent, the same score will be obtained in all cases. The second meaning of reliability refers to the similarity of items, indicating to what extent test questions are interrelated. To evaluate the reliability of the constructs, Fornell and Larcker [33, 34] suggested three criteria, including (1) reliability of each item, (2) composite reliability (CR) of each construct, and (3) average variance extracted (AVE).

Regarding the reliability of each item, the factor loading of≥0.6 for each item is defined in the CFA as a good construct indicator. Moreover, the factor loading of the items should be significant, at least at p = 0.01 [31, 32]. Dillon-Goldstein coefficient (pc) is used to evaluate the CR of each construct. Given that PLS uses the subjects’ factor scores for analysis, in contrast to multiple regression, it is necessary to consider the factor loading of each item in calculating the reliability index. However, Cronbach’s alpha coefficient assigns equal weights to the items and shows less reliability; hence, the CR coefficient was used. The acceptable values of CR should be≥0.7 [35, 36].

The third indicator of reliability is AVE. Fornell and Larcker recommended AVE≥0.5, indicating that the construct explains about≥50% of the variance of its indicators [34, 36]. Validity refers to whether the items measure the same concept as desired. To examine the validity or divergent validity of constructs, Chin [37] recommended two criteria. First, the construct items should have the maximum factor loading on their construct, indicating that they should have a small cross-sectional load on other constructs. Gefen and Straub suggested that the factor loading of each item on its construct should be at least 0.1 more than the factor loading of the same item on other constructs [38]. The second criterion is that the AVE of a construct must be greater than the correlation of that construct with other constructs. This indicates that the correlation of that construct with its markers is greater than its correlation with other constructs. According to Tannen et al., the goodness of fit (GOF) index in the PLS model acts as the overall fit index of the model and can be used to check the validity or quality of the PLS in general [39]. The GOF index consists of three absolute types to compare and contrast different groups. The value of this index can be obtained by dividing the absolute GOF by the maximum GOF of the tested model. The GOF of the external model measures the overall quality of the measurement model, and the GOF of the internal model measures the quality of the structural model. This index also acts as the fit index of the LISREL model and ranges between zero and one; values closer to one indicate the appropriate quality of the model [40].

Finally, the stability of composite reliability was used to determine the reliability of the scale. To determine the scale’s reliability, the retest was reviewed over a two-week interval, and the scores obtained in these two stages were compared using the intra-cluster correlation index. If this value is > 0.80, the stability rate is desirable. Sampling was performed on 200 qualified older adults; according to some references, this number is acceptable to perform CFA.

2.4 Ethical consideration

The ethics committee at Tabriz University of Medical Sciences, Iran, approved the study protocol (IR.TBZMED.REC.1398.517). In this study, the research objectives and the confidentiality of the participants’ information were observed. Moreover, the participants’ written consent was obtained before they participated in this study. The participants were informed based on the Declaration of Helsinki.

3 Results

The participants’ mean age (standard deviation) in the qualitative phase of the study was 71 years (1.8), and the mean age (standard deviation) of older adult caregivers was 48 years (2.3). Moreover, 63% of the subjects were male, and 37% were female, and 18% of older adult caregivers were male, and 82% were female. In the quantitative phase of the study (psychometric phase), 200 persons participated in this study, 184 of whom were male, and 216 were female, and the mean age (standard deviation) of older adults participating in this study phase was 74 years (1.4). The participants’ characteristics for this phase are shown in Table 1.

Table 1
Participants’ characteristics in the quantitative phase of the study

Characteristic Group n %

Gender Male 84 42

Female 116 58

Age 60–70 years 94 47

70–80 years 60 30

Over 80 years 46 23

Type of residential Single storey house 156 78

buildings Apartments 44 22

Characteristic	Group	n	%
Gender	Male	84	42
	Female	116	58
Age	60–70 years	94	47
	70–80 years	60	30
	Over 80 years	46	23
Type of residential	Single storey house	156	78
buildings	Apartments	44	22

In the qualitative phase of the study, the data obtained from the structured interviews with the older adults and the literature review led to the development of scale. The scale consisted of two parts. The first part included the general demographic information of each individual and the second part includes physical environment problems and barriers in the older adults’ homes. This part consists of 71 Likert-type questions, anchored in three points ranging from ‘0-no difficulty’ to ‘3-very’, which were classified into seven main dimensions (Table 2).

Table 2

The domain and item in the scale

Domain	Item no.	Total item
Vision	C1, C2, C3, C4, D1, D2	6
Hearing	C5, C6, C7, C8, D3	5
Cooling and heating	C9, C10, C11, C12, C13	5
Movement	C14, C15, C16, C17, C18, C19, C20, C21, D4, C22, C23,	20
	C24, C25, C26, C27, C28, C29, C30, C31, C32
Accessibility	C33, C34, C35, C36, C37, C38, C39, C40, C41,	20
	C42, C43, C44, C45, C46, C47, C48, C49, D5, D6, D7
Handgrip	C50, C51, C52, C53, D8	5
Accident	C54, C55, C56, C57, C58, C59, C60, C61, C62, C63	10

3.1 Qualitative face validity

In evaluating the qualitative face validity, after collecting information about the level of difficulty, the degree of appropriateness and ambiguity was determined by interviewing the participants, and the necessary modifications were made in the questionnaire items.

3.2 Quantitative face validity

In this phase, two items (D3, D8) did not get the minimum impact score (1.5) and were removed from the questionnaire.

3.3 Qualitative content validity

To evaluate the qualitative content validity, the questionnaire was provided to 10 experts in ergonomics and older adults, and finally, some modifications were made to the questionnaire.

3.4 Quantitative content validity

The results of CVR showed that six items (D1, D2, D4, D6, D7 and D8) of the scale scored < 0.62, which were removed from the scale. Using the Waltz and Basel method, the results of CVI showed that 51 items scored > 0.79, which were retained from the scale, and 12 items scored between 0.7 and 0.79, which were provided to experts after modifications and with a score > 0.79, were retained in the scale. In the next step, the CVR and CVI of the scale were calculated based on the mean value. Finally, 63 questions with a mean CVI of 0.78 and a mean CVR of 0.84 were obtained (Table 3).

Table 3
Content validity ratios and content validity index of the scale

Construct Item CVI CVR Acceptance or non-acceptance

Vision C1 1 0.82 Acceptance

D1 0.70 0.40 Non-acceptance

C2 0.90 0.80 Acceptance

C3 1.00 1.00 Acceptance

C4 0.87 0.80 Acceptance

D2 0.80 0.40 Non-acceptance

Hearing C5 0.87 1.00 Acceptance

C6 0.70 0.80 Acceptance

C7 1.00 1.00 Acceptance

C8 0.77 1.00 Acceptance

Cooling and C9 1.00 1.00 Acceptance

heating C10 1.00 1.00 Acceptance

C11 0.77 0.80 Acceptance

C12 0.80 0.80 Acceptance

C13 0.87 1.00 Acceptance

Movement C14 0.77 0.80 Acceptance

C15 1.00 0.80 Acceptance

C16 1.00 0.80 Acceptance

C17 0.97 1.00 Acceptance

C18 1.00 1.00 Acceptance

C19 1.00 1.00 Acceptance

C20 1.00 0.80 Acceptance

C21 0.90 1.00 Acceptance

D4 0.70 0.40 Non-acceptance

C22 0.80 0.80 Acceptance

C23 0.70 0.80 Acceptance

C24 0.70 0.82 Acceptance

C25 0.84 0.82 Acceptance

C26 1.00 1.00 Acceptance

C27 0.97 1.00 Acceptance

C28 1.00 1.00 Acceptance

C29 1.00 1.00 Acceptance

C30 0.97 1.00 Acceptance

C31 0.97 1.00 Acceptance

C32 0.97 0.82 Acceptance

Accessibility C33 0.70 0.80 Acceptance

C34 1.00 1.00 Acceptance

C35 0.93 1.00 Acceptance

C36 1.00 1.00 Acceptance

D5 0.70 0.40 Non-acceptance

C37 1.00 1.00 Acceptance

C38 0.70 0.80 Acceptance

C39 0.97 1.00 Acceptance

C40 1.00 0.80 Acceptance

C41 0.83 0.80 Acceptance

D6 0.80 0.40 Non-acceptance

C42 0.87 0.80 Acceptance

C43 1.00 1.00 Acceptance

C44 1.00 1.00 Acceptance

C45 0.93 1.00 Acceptance

C46 0.90 0.80 Acceptance

C47 0.83 0.80 Acceptance

D7 0.70 0.40 Non-acceptance

C48 0.87 0.82 Acceptance

C49 0.87 0.80 Acceptance

Handgrip C50 0.97 0.80 Acceptance

C51 0.97 1.00 Acceptance

C52 0.97 1.00 Acceptance

C53 0.97 0.80 Acceptance

Accident C54 0.97 1.00 Acceptance

C55 1.00 1.00 Acceptance

C56 0.97 1.00 Acceptance

C57 0.93 0.80 Acceptance

C58 0.93 1.00 Acceptance

C59 0.97 1.00 Acceptance

C60 1.00 0.80 Acceptance

C61 1.00 0.82 Acceptance

C42 1.00 1.00 Acceptance

C63 1.00 1.00 Acceptance

Construct	Item	CVI	CVR	Acceptance or non-acceptance
Vision	C1	1	0.82	Acceptance
	D1	0.70	0.40	Non-acceptance
	C2	0.90	0.80	Acceptance
	C3	1.00	1.00	Acceptance
	C4	0.87	0.80	Acceptance
	D2	0.80	0.40	Non-acceptance
Hearing	C5	0.87	1.00	Acceptance
	C6	0.70	0.80	Acceptance
	C7	1.00	1.00	Acceptance
	C8	0.77	1.00	Acceptance
Cooling and	C9	1.00	1.00	Acceptance
heating	C10	1.00	1.00	Acceptance
	C11	0.77	0.80	Acceptance
	C12	0.80	0.80	Acceptance
	C13	0.87	1.00	Acceptance
Movement	C14	0.77	0.80	Acceptance
	C15	1.00	0.80	Acceptance
	C16	1.00	0.80	Acceptance
	C17	0.97	1.00	Acceptance
	C18	1.00	1.00	Acceptance
	C19	1.00	1.00	Acceptance
	C20	1.00	0.80	Acceptance
	C21	0.90	1.00	Acceptance
	D4	0.70	0.40	Non-acceptance
	C22	0.80	0.80	Acceptance
	C23	0.70	0.80	Acceptance
	C24	0.70	0.82	Acceptance
	C25	0.84	0.82	Acceptance
	C26	1.00	1.00	Acceptance
	C27	0.97	1.00	Acceptance
	C28	1.00	1.00	Acceptance
	C29	1.00	1.00	Acceptance
	C30	0.97	1.00	Acceptance
	C31	0.97	1.00	Acceptance
	C32	0.97	0.82	Acceptance
Accessibility	C33	0.70	0.80	Acceptance
	C34	1.00	1.00	Acceptance
	C35	0.93	1.00	Acceptance
	C36	1.00	1.00	Acceptance
	D5	0.70	0.40	Non-acceptance
	C37	1.00	1.00	Acceptance
	C38	0.70	0.80	Acceptance
	C39	0.97	1.00	Acceptance
	C40	1.00	0.80	Acceptance
	C41	0.83	0.80	Acceptance
	D6	0.80	0.40	Non-acceptance
	C42	0.87	0.80	Acceptance
	C43	1.00	1.00	Acceptance
	C44	1.00	1.00	Acceptance
	C45	0.93	1.00	Acceptance
	C46	0.90	0.80	Acceptance
	C47	0.83	0.80	Acceptance
	D7	0.70	0.40	Non-acceptance
	C48	0.87	0.82	Acceptance
	C49	0.87	0.80	Acceptance
Handgrip	C50	0.97	0.80	Acceptance
	C51	0.97	1.00	Acceptance
	C52	0.97	1.00	Acceptance
	C53	0.97	0.80	Acceptance
Accident	C54	0.97	1.00	Acceptance
	C55	1.00	1.00	Acceptance
	C56	0.97	1.00	Acceptance
	C57	0.93	0.80	Acceptance
	C58	0.93	1.00	Acceptance
	C59	0.97	1.00	Acceptance
	C60	1.00	0.80	Acceptance
	C61	1.00	0.82	Acceptance
	C42	1.00	1.00	Acceptance
	C63	1.00	1.00	Acceptance

3.5 Validity and reliability of the construct

Factor loading coefficients and CR were used to evaluate the reliability of the constructs. The evaluation of the preliminary 63-item scale showed that 25 items were removed from the model due to factor loadings < 0.6, and other items (44 items) with factor loadings > 0.6 were included in the scale. After removing the 25 items, factor loadings for the retained 44 items were recalculated, as shown in Table 4. Figure 1 shows the research model with standardized path coefficients.

Table 4
Construct Validity and Reliability

Construct Item Factor loading AVE CR

Barriers and problems related to vision C1 0.897 0.788 0.918

C2 0.877

C3 0.889

Barriers and problems related to hearing C5 0.864 0.754 0.925

C6 0.882

C7 0.882

C8 0.845

Barriers and problems related to cooling and heating C9 0.894 0.785 0.916

C10 0.890

C13 0.874

Barriers and problems related to movement C15 0.866 0.755 0.979

C16 0.866

C17 0.853

C18 0.853

C19 0.882

C20 0.844

C21 0.878

C22 0.861

C23 0.878

C25 0.869

C26 0.880

C27 0.876

C28 0.865

C30 0.880

C31 0.881

Barriers and problems related to accessibility C34 0.859 0.743 0.959

C36 0856

C37 0.917

C39 0.859

C40 0.843

C43 0.855

C44 0.839

C49 0.866

Barriers and problems related to handgrip C50 0.898 0.798 0.922

C52 0.882

C53 0.900

Barriers and problems related to accidents C54 0.875 0.753 0.961

C55 0.847

C57 0.875

C58 0.907

C59 0.831

C60 0.884

C61 0.846

C63 0.876

Construct	Item	Factor loading	AVE	CR
Barriers and problems related to vision	C1	0.897	0.788	0.918
	C2	0.877
	C3	0.889
Barriers and problems related to hearing	C5	0.864	0.754	0.925
	C6	0.882
	C7	0.882
	C8	0.845
Barriers and problems related to cooling and heating	C9	0.894	0.785	0.916
	C10	0.890
	C13	0.874
Barriers and problems related to movement	C15	0.866	0.755	0.979
	C16	0.866
	C17	0.853
	C18	0.853
	C19	0.882
	C20	0.844
	C21	0.878
	C22	0.861
	C23	0.878
	C25	0.869
	C26	0.880
	C27	0.876
	C28	0.865
	C30	0.880
	C31	0.881
Barriers and problems related to accessibility	C34	0.859	0.743	0.959
	C36	0856
	C37	0.917
	C39	0.859
	C40	0.843
	C43	0.855
	C44	0.839
	C49	0.866
Barriers and problems related to handgrip	C50	0.898	0.798	0.922
	C52	0.882
	C53	0.900
Barriers and problems related to accidents	C54	0.875	0.753	0.961
	C55	0.847
	C57	0.875
	C58	0.907
	C59	0.831
	C60	0.884
	C61	0.846
	C63	0.876

Fig. 1

Research model with standardized path coefficients (evaluation of the measurement model after modification).

The results of the second round of factor analysis showed that all the items had factor loadings > 0.6, and thus these 44 items were approved, and the model had a proper fit (p < 0.01). The results in Table 1 are significant at the 99% confidence level (p < 0.01). CR was used to investigate the scale’s internal correlation. The CR coefficients for the studied constructs were > 0.7, confirming their acceptable reliability. Moreover, the AVE index value for all the constructs was > 0.5 and within an acceptable level, indicating the convergent validity of the model. In general, the measurement model explains the convergent reliability and validity of the questionnaire (Table 5).

Table 5

Summary of reliability analysis

Construct	Cronbach’s α	rho_A	CR
Vision	0.866	0.867	0.918
Hearing	0.891	0.893	0.925
Cooling and heating	0.863	0.863	0.916
Movement	0.977	0.977	0.925
Accessibility	0.950	0.951	0.959
Handgrip	0.873	0.874	0.922
Accident	0.953	0.955	0.961

To evaluate divergent validity, a matrix should be created based on Fornell and Larker so that the values of the original diameter are the square root of the AVE coefficients of each construct, and the smaller values of the original diameter are the coefficients of correlation between each construct with other constructs. Table 6 reveals that the square root of the AVE coefficients is greater than the correlation coefficients between the constructs. These results indicate that the measurement model has acceptable divergent validity.

Table 6

Factor correlations and the square root of the AVE (divergent validity of constructs)

Vision	Hearing	Cooling and	Movement	Accessibility	Handgrip	Accident
heating
Vision	0.888
Hearing	0.257	0.868
Cooling and heating	0.310	0.357	0.886
Movement	0.361	0.367	0.353	0.869
Accessibility	0.218	0.333	0.335	0.245	0.862
Handgrip	0.772	0.259	0.243	0.292	0.217	0.893
Accident	0.285	0.315	0.287	0.377	0.280	0.288	0.868

The results of the divergent validity test were also calculated by using the cross-loading method, as represented in Table 7. It should be noted that the factor loading of each item on its construct in this test should be at least 0.1 above the factor loading of the same item on other constructs.

Table 7

Results of divergent validity study through cross-factor loading method

Item	Vision	Accessibility	Movement	Cooling and heating	Hearing	Accident	Handgrip
C1	0.897	0.194	0.258	0.314	0.248	0.279	0.705
C2	0.877	0.160	0.337	0.274	0.183	0.238	0.778
C3	0.889	0.223	0.364	0.241	0.251	0.244	0.741
C5	0.275	0.276	0.326	0.303	0.864	0.234	0.279
C6	0.256	0.344	0.304	0.354	0.882	0.316	0.271
C7	0.180	0.338	0.324	0.337	0.882	0.233	0.164
C8	0.176	0.190	0.322	0.240	0.845	0.312	0.181
C9	0.294	0.335	0.305	0.894	0.289	0.258	0.222
C10	0.273	0.318	0.284	0.890	0.304	0.283	0.225
C13	0.257	0.237	0.349	0.874	0.356	0.222	0.197
C15	0.339	0.229	0.866	0.313	0.362	0.361	0.279
C16	0.301	0.198	0.865	0.216	0.259	0.321	0.255
C17	0.278	0.201	0.853	0.353	0.333	0.330	0.249
C18	0.346	0.205	0.853	0.294	0.276	0.316	0.256
C19	0.339	0.249	0.882	0.306	0.297	0.294	0.284
C20	0.294	0.208	0.844	0.299	0.318	0.329	0.229
C21	0.287	0.263	0.878	0.319	0.289	0.319	0.225
C22	0.304	0.188	0.861	0.313	0.323	0.262	0.207
C23	0.273	0.199	0.878	0.323	0.346	0.410	0.223
C25	0.338	0.186	0.869	0.341	0.311	0.317	0.292
C26	0.291	0.200	0.880	0.298	0.358	0.363	0.261
C27	0.350	0.226	0.876	0.344	0.355	0.391	0.283
C28	0.303	0.117	0.865	0.256	0.265	0.270	0.250
C30	0.362	0.262	0.880	0.310	0.344	0.319	0.276
C31	0.300	0.251	0.881	0.306	0.332	0.317	0.229
C34	0.167	0.859	0.178	0.335	0.190	0.246	0.148
C36	0.208	0.856	0.208	0.325	0.319	0.242	0.210
C37	0.209	0.917	0.202	0.279	0.281	0.251	0.221
C39	0.160	0.859	0.201	0.329	0.316	0.289	0.134
C40	0.217	0.843	0.240	0.312	0.263	0.253	0.211
C43	0.194	0.855	0.248	0.271	0.271	0.192	0.203
C44	0.230	0.839	0.178	0.238	0.356	0.202	0.216
C49	0.113	0.866	0.222	0.221	0.296	0.254	0.148
C50	0.725	0.210	0.276	0.184	0.195	0.236	0.898
C52	0.747	0.189	0.223	0.187	0.234	0.293	0.882
C53	0.746	0.182	0.281	0.276	0.266	0.246	0.900
C54	0.260	0.235	0.331	0.241	0.325	0.875	0.275
C55	0.242	0.252	0.328	0.251	0.315	0.847	0.248
C57	0.237	0.202	0.284	0.237	0.205	0.975	0.240
C58	0.251	0.264	0.351	0.275	0.259	0.907	0.231
C59	0.205	0.231	0.281	0.234	0.244	0.831	0.216
C60	0.266	0.250	0.399	0.298	0.310	0.884	0.259
C61	0.237	0.215	0.303	0.190	0.269	0.846	0.227
C63	0.274	0.284	0.323	0.257	0.253	0.876	0.300

Table 7 indicates that the divergent validity was confirmed by the cross-loading method for 44 items for each construct. The results in Table 7 are significant at the 99% confidence level (p < 0.01).

The goodness of fit (GOF) considers both measurement and structural models and is used as a criterion in measuring the model’s overall performance. The index is calculated according to the following formula. Wetzels et al. (2009) proposed three values of 0.01, 0.25, and 0.36 for weak, medium, and strong GOF, respectively. According to Table 8 and the following formula, its mean communality was 0.579, and its mean R² was 0.388. According to the following formula, the GOF value for the study model was 0.473, indicating a strong and acceptable GOF [41].

Table 8

Structural model fit criteria and GOF

Construct	R²	Communality	GOF
Handgrip	0.344	0.517	0.473
Vision	0.332	0.548
Accessibility	0.300	0.510
Movement	0.707	0.674
Cooling and heating	0.310	0.633
Hearing	0.296	0.531
Accident	0.429	0.642

$\begin{matrix} GOF = \sqrt{\bar{Communality} \times \bar{R^{2}}} \\ = \sqrt{0.579 \times 0.388} = 0.473 \end{matrix}$ (2)

4 Discussion

One of the ways to evaluate the degree of adaptation between the home environment and the characteristics of older adults is to examine the barriers and problems reported by older adults in the home environment, which reflect a lack of adaptation. In this regard, the measurement scales for barriers and problems in older adults’ homes can play a crucial role, and the present study addressed the gap in the literature, i.e. the lack of proper assessment scales in this field.

One of the significant challenges of measurement scales is determining their psychometric properties. Studies have revealed that some of the developed scales are used without checking their validity and reliability. However, the quality of a scale depends on its validity and reliability [39]. Accordingly, the present study aimed to evaluate the psychometric properties of a scale to measure physical environment problems and barriers in older adults’ homes using an ergonomic approach.

In this study, decreased abilities in older adults and their vital role in posing problems in the physical environment of older adults’ homes [17] were considered in the development of the scale. A comprehensive scale was also developed to measure the barriers and problems in the physical environment of older adults’ homes. In this regard, the scale validity was examined by assessing face validity, content validity, and construct validity.

In the present study, the face validity of the questionnaire was determined, and the impact score of each question was calculated according to the opinions of the target group. Questions with an impact score of < 1.5 were eliminated or modified [42]. Lawshe’s method was used to evaluate the content validity of the questionnaire quantitatively. To this end, CVR and CVI coefficients were determined. The results showed that the developed scale had acceptable validity to measure the physical environment barriers and problems in older adults’ homes.

The evaluation of the preliminary 63-item scale showed that the scale had a proper fit by removing one item from the construct of barriers and problems related to vision, two items from the cooling and heating construct, four items from the movement construct, 9 items from the access construct, one item from the hand-grip construct, and two items from the accident construct, as their factor loading values were < 0.6 [32]. Given that the scale items in this study were selected by both the interviews with older adults and a review of literature, the opinions of older adults as end users can be more important [43].

To evaluate the scale, internal Consistency Reliability (CR) was used, the value of which was > 0.7 for the studied constructs, confirming the acceptable reliability of the studied model. Moreover, the AVE index value for all the constructs was > 0.5, indicating the convergent validity of the model. In general, the measurement model explains the convergent reliability and validity of the questionnaire.

To evaluate the divergent validity of the questionnaire, cross-loading and Fornell-Larker methods were used, and the results showed an acceptable fit, confirming the necessary coordination between items and constructs. The GOF of the model was also 0.473, representing that the questionnaire structure generally fits well with the data.

According to the findings of this study, the scale is efficient in measuring the physical environment problems and barriers in older adults’ homes by using an ergonomic approach. It can be used for ergonomic assessments of older adults’ homes with appropriate psychometric properties.

Although the reliability and validity of the measurement scale for older adults living in Yazd were confirmed, and the multidimensional measurement scale of physical barriers and problems was approved by an ergonomic approach, other researchers should address the limitations of this study. The study limitations were as follows: small sample size, inclusion of older adults’ homes in Yazd and not the other cities, cross-sectional nature of the data, older adults’ tendency to underestimate the problems [44], and older adults forgetting some problems.

Accordingly, it is recommended to conduct studies on a larger statistical population from other geographical areas with different cultures and lifestyles to determine the role of culture, lifestyle and the type of architecture in older adults’ problems. On the other hand, given that the present study focused on urban houses, it is recommended to examine rural homes in future studies.

Given the tendency of older adults to underestimate access problems and environmental risk factors in their homes [44] and their lack of understanding of how environmental barriers affect their daily activities [45], it is recommended to solve this problem to some extent by expanding the statistical community or by including their family and caregivers in the evaluation. It should be noted that individuals with disabilities report more home adaptation needs than older adults [43, 44]. Given the limitations of this study and the need to address such limitations in future studies, new items should be added to the developed scale so that its results can be used to develop interventions and improve the physical environment in older adults’ homes.

Home designers usually ignore a decrease in human abilities caused by aging, and this issue, in most cases, causes a lack of physical environment adaptation for older adults. By focusing on home design based on older adults’ abilities, it is possible to make an environment adaptable to older adults’ conditions and help them maximize their capabilities [11]. On the other hand, it is recommended to consider inclusive design principles in home designs from the very beginning to make them suitable for older adults. This is because previous studies have indicated that older adults are usually reluctant or even opposed to making recommendations and taking corrective action in their homes and usually postpone repairs at least until their limitations are aggravated [46].

5 Conclusion

The present study examined the validity and reliability of the scale to measure physical environment barriers and problems in older adults’ homes using an ergonomic approach, and the findings revealed the acceptable psychometric properties of the scale. Accordingly, this scale can be used as a reliable and valid tool in future studies. Since measuring and evaluating the validity of a scale is a continuous process, and the desired validity is achieved by conducting several studies [47], future studies should examine the validity and the factor structure of this scale.

Footnotes

Acknowledgments

This study was extracted from a doctoral dissertation on ergonomics approved at the Tabriz University of Medical Sciences and financially supported by the Vice-Chancellor for Research. The authors thank the university authorities and older adults and their caregivers for their cooperation in this study.

References

Mast

, Burmester

, Kruger

, Fatikow

, Arbeiter

, Graf

, et al. User-centered design of a dynamic-autonomy remote interaction concept for manipulation-capable robots to assist elderly people in the home, Journal of Human-Robot Interaction 2012;1(1):96–118.

Healthy places terminology USA: Centers for Disease Control and Prevention; 2010 [Available from: https://www.cdc.gov/healthyplaces/terminology.htm.

Jaschinski

, Allouch

, Peters

, Cachucho

, van Dijk

. Acceptance of Technologies for Aging in Place: A Conceptual Model, Journal of Medical Internet Research 2021;23(3):e22613.

Wagner

. Aging in Place with Age-Related Cognitive Changes: The Impact of Caregiving Support and Finances, Societies 2021;11(2):31.

Iwarsson

, Wahl

H-W

, Nygren

, Oswald

, Sixsmith

, et al. Importance of the home environment for healthy aging: conceptual and methodological background of the European ENABLE-AGE Project, The Gerontologist 2007;47(1):78–84.

Burton

, Mitchell

. Inclusive urban design: Streets for life: Routledge; 2006.

Patti

, Bianco

, Karsten

, Montalto

, Battaglia

, Bellafiore

, et al. The effects of physical trainingwithout equipment on pain perception and balance in the elderly: A randomized controlled trial, Work. 2017;57(1):23–30.

da Silva

, Cavalcante Neto

. Quality of life and functional capability of elderly Brazilian women, Work 2019;62(1):97–106.

Richert-Kaźmierska

, Stankiewicz

. Work–life balance: Does age matter? Work 2016;55(3):679–88.

10.

Barros

, Carnide

, Cunha

, Santos

, Silva

. Will I be able to do my work at 60? An analysis of working conditions that hinder active ageing, Work 2015;51(3):579–90.

11.

Nasiri

, Foroughan

, Rashedi

, Makarem

, Jafari Mourjan

. Compliance to Universal Design Criteria in Nursing Homes of Tehran, Iranian Journal of Ageing 2016;11(2):340–7.

12.

Nazari

, Rashedi

, Mohammadi

, Yousefi

. Relationship between cognitive status and activities of daily living among the elderly of nursing homes, Journal of Kermanshah University of Medical Sciences (J Kermanshah Univ Med Sci) 2015;18(12):744–6.

13.

Beck

, Enguidanos

, Rahman

, Nabors

, Tracton-Bishop

, Olshevski

, et al. Provision of Universal Design Educational Material Following a Homeowner’s Insurance Claim, Journal of Housing For the Elderly 2016;30(3):237–52.

14.

Friesen

, Brémault-Phillips

, Rudrum

, Rogers

. Environmental Design That Supports Healthy Aging:Evaluating a New Supportive Living Facility, Journal of Housing For the Elderly 2016;30(1):18–34.

15.

Goli

, Khodadad

. Pathology of the laws of architectural rights and urban planning of the elderly in Iran. The third Annual conference for research in architecture, urban planning and urban management; Tehran, Iran2017.

16.

Tauke

, Smith

, Maisel

, Basnak

, Steinfeld

. Inclusive Design: Implementation and Evaluation: Routledge; 2017.

17.

Bentley

, Green

, Tappin

, Haslam

. State of science: the future of work–ergonomics and human factors contributions to the field. Ergonomics. 2020:1-13.

18.

Nakhodaeezadeh

, Jafarabadi

, Allahverdipour

, Matlabi

, Dehkordi

. Home Environment and Its Relation with Quality of Life of Older People. Journal of Housing For the Elderly. 2017:1-14.

19.

Lewis

, Torrington

, Barnes

, Darton

, Holder

, McKee

, et al. EVOLVE: a tool for evaluating the design of older people’s housing, Housing, Care and Support 2010;13(3):36–41.

20.

Barnes

, Torrington

, Darton

, Holder

, Lewis

, McKee

, et al. Does the design of extra-care housing meet the needs of the residents? A focus group study, Ageing & Society 2012;32(7):1193–214.

21.

Hrovatin

, Pisot

, Plevnik

. Healthy ageing at home: Ergonomics adaptations of interior design and self-assesed quality of life of · older adults of the municipality of Ljubljana, Annales Kinesiologiae 2017;7(2):139–52.

22.

Zarghami

, Sharghi

, Olfat

. Effective environmental factors in increasing the quality of life of residents ofnursing homes Case Study: Seniaran homes in the province of Tehran, Iranian Architecture Quarterly. 2015;7:111–26.

23.

Houser

. Precision, reliability, and validity: Essential elements of measurement in nursing research, Journal for Specialists in Pediatric Nursing 2008;13(4):297.

24.

Taghizadeh

, Ebadi

, Montazeri

, Shahvari

, Tavousi

, Bagherzadeh

. Psychometric properties of health related measures, Part Translation, development, and content and face validity. Payesh (Health Monitor) 2017;16(3):343–57.

25.

Broder

, McGrath

, Cisneros

. Questionnaire development: face validity and item impact testing of the Child Oral Health Impact Profile, Community dentistry and oral epidemiology 2007;35:8–19.

26.

Waltz

, Strickland

, Lenz

Measurement in nursing and health research: Springer publishing company; 2010.

27.

Cook

, Beckman

. Current concepts in validity and reliability for psychometric instruments: theory and application, The American journal of medicine 2006;119(2):166.e7–e16.

28.

Taherdoost

. Validity and reliability of the research instrument; how to test the validation of a questionnaire/survey in a research. How to test the validation of a questionnaire/survey in a research (August 10, 2016). 2016.

29.

Almanasreh

, Moles

, Chen

. Evaluation of methods used for estimating content validity, Research in Social and Administrative Pharmacy 2019;15(2):214–21.

30.

Lawshe

. A quantitative approach to content validity, Personnel Psychology 1975;28(4):563–75.

31.

Waltz

, Bausell

. Nursing research: Design, statistics, and computer analysis: FA Davis Co; 1981.

32.

Grove

, Gray

, Faan

. Understanding Nursing Research: First South Asia Edition. E-Book: Building an Evidence- Based Practice: Elsevier India. 2019.

33.

Grove

, Gray

, Understanding Nursing Research E-Book: Building an Evidence-Based Practice: Elsevier Health Sciences; 2018.

34.

Fornell

, Larcker

. Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research. 1981:39-50.

35.

Hair

, Risher

, Sarstedt

, Ringle

. When to use and how to report the results of PLS-SEM. European business 2019.

36.

Hair

Jr , Hult

GTM

, Ringle

, Sarstedt

Aprimer on partial least squares structural equation modeling (PLS-SEM): Sage publications; 2016.

37.

Chin

. The partial least squares approach to structural equation modeling, Modern Methods for Business Research 1998;295(2):295–336.

38.

Gefen

, Straub

. A practical guide to factorial validity using PLS-Graph: Tutorial and annotated example, Communications of the Association for Information systems 2005;16(1):5.

39.

Tenenhaus

, Vinzi

, Chatelin

Y-M

, Lauro

. PLS path modeling, Computational Statistics & Data Analysis 2005;48(1):159–205.

40.

Sánchez-Franco

, Martínez-López

, Martín-Velicia

. Exploring the impact of individualism anduncertainty avoidance in Web-based electronic learning: An empiricalanalysis in European higher education, Computers & Education 2009;52(3):588–98.

41.

Wetzels

, Odekerken-Schröder

, Van Oppen

. Using PLS path modeling for assessing hierarchical construct models: Guidelines and empirical illustration. MIS quarterly. 2009:177-95.

42.

Mousazadeh

, Rakhshan

, Mohammadi

. Investigation of content and face validity and reliability of sociocultural attitude towards appearance questionnaire-3 (SATAQ-3) among female adolescents, Iranian Journal of Psychiatry 2017;12(1):15.

43.

Tosi

. Ergonomics and Design. Design for Ergonomics: Springer; 2020. pp. 3–29.

44.

Esposito Vinzi

, Chin

, Henseler

, Wang

. Handbook of partial least squares: Concepts, methods and applications: Heidelberg, Dordrecht, London, New York: Springer; 2010.

45.

Kimberlin

, Winterstein

. Validity and reliability of measurement instruments used in research, American Journal of Health-System Pharmacy 2008;65(23):2276–84.

46.

Steinfeld

, Shea

. Enabling home environments, Technology and Disability 1993;2(4):69–79.

47.

Fox

. Rehabilitation in practice: Environmental modifications in the homes of elderly Canadians with disabilities, Disability and Rehabilitation 1995;17(1):43–9.