For any novel psychometric instrument to be applied in a different culture (UAE) from the setting in which it was originally developed (Canada), it needs to be appropriately adapted. Cultural adaptation of an existing instrument has many benefits over creating an entirely new tool, such as reducing cost and time spent developing it. Psychometric instruments, such as SSTICS, can not only stimulate cross-cultural research but also be employed in many fields such as public health, primary health care, psychiatry, neurology, and other disciplines. In practice, cultural adaptation is done while translating the tool from its original language and it precedes its psychometric validation through the assessment of different aspects of reliability and validity using well-established scientific measures. In this process, the self-reporting scales are potentially susceptible to significant alterations resulting from a series of influences, including social attractiveness, suppression, habits, and response style. Thus, describing and analyzing the cultural adaptation and psychometric validation process may serve as a guide for future similar undertakings (Table 5).

Cross-cultural validation of an assessment instrument is a complex process that requires a significant investment in time and financial resources (Bullinger et al., 1993). Even though a 2015 review of cross-cultural adaptation guidelines for questionnaires did not find a consensus among the 31 guidelines investigated in this study, the authors reported that similar results were achieved using most of them as long as the adaptation and validation processes were properly and separately considered (Epstein et al., 2015). Before venturing into this process for SSTICS, it is vital to ensure that there is no equivalent instrument in the Arabic language, or an equivalent instrument translated and validated. Otherwise, the researcher must make an exhaustive review of the instruments relevant to his study to select the one that has passed the validation stages, is the most rigorous, and is in its original language (Haccoun, 1987; Flaherty et al., 1988; Guillemin et al., 1993).

In general, the cross-cultural validation of an instrument involves three main steps: (1) the translation and verification of its equivalence; (2) empirical verification of the validity of the translated version; (3) adaptation of scores to cultural context and development standards (Epstein et al., 2015). Each of them also includes steps necessary for creating a valid version, and several options are available to the researcher with their advantages and disadvantages. The semantic aspect of a questionnaire is the first step to consider before going any further with the validation. It is often a neglected step. The difference between dialects needs to be considered (Table 6).

There are different types of translation methods. The translation process must ensure that an instrument retains inferential equivalence, that is, it is possible to produce the same inferences from the translated version as with the original instrument. The correspondence between words (semantic equivalence) is difficult to achieve from one culture to another, considering the vocabulary, dialect, and grammar specific to each language.

This can be the case between the UAE and Canada. Some translated expressions have no meaning in another culture, and expressions specific to the target culture, and retaining the meaning of the items must be found (equivalence of expression). Certain situations that are evoked in the culture of the original instrument may not correspond to the reality in another culture. These items should be replaced by other situations appropriate to the target culture while preserving the objective and the meaning aimed by these items (experiential equivalence) (Guillemin et al., 1993).

Finally, the same exercise should be applied to some concepts, which, when translated, do not allow the exact representation of the target culture (conceptual equivalence). Translating items into different contexts is a delicate task that raises many questions. Only translating an instrument into the local language does not guarantee validity and accuracy in a different context.

Capitulo et al. (2001) identified the decentering of the translation process as an extension of back-translation (Capitulo et al., 2001). Once the final version has been developed, after the back-translation, whoever owns the copyright to the original instrument and the translation team/individual will agree on the semantic equivalence between the “original’ and the translation instrument by modifying the original instrument for universally better-understood language (Capitulo et al., 2001).

Hilton and Skrutkowski (2002) describe symmetric and asymmetric categories of translation. Symmetric translation requires that the original and the translation instruments be faithful to meaning and familiarity. Decentering refers to the translation process in which the source and target of language versions are seen as equally important and subject to change (Hilton and Skrutkowski, 2002).

In the asymmetric translation category, the original language is kept without modification, thus the translated version may present a literal translation of words and may lack conceptual equivalence. One of the reasons that an instrument developer or organization resists making changes is to avoid compromising the validity and instrument reliability. The translation process is considered incomplete until the instrument is pretested with members of the target culture.

Pretesting an instrument primarily serves two purposes: it checks the quality of the translation and the practicalities of test administration (Hilton and Skrutkowski, 2002).

Several objectives are targeted when translating the tool, not only the meaning but also the equivalence of content, technique, and concept. The ‘subject’s belonging to a different ethnic group influences their results in the various clinical tests’. A great deal of documentation exists revealing the difficulties caused by using assessment instruments in a different cultural context other than the one they were created for. Flaherty et al. (1988) show that to ensure the correct use of a diagnostic instrument in a new cultural context (Flaherty et al., 1988), it is essential to verify the equivalence of these five dimensions:

Content equivalence corresponds to the content of each instrument item that is relevant in the cultures where its use is conceived. This step includes the verification of the items by a team of specialists from each culture. They decide on the relevance or not of each item. At best, the author of the tool is contacted.

Semantic equivalence corresponds to the fact that the meaning of each item remains the same after translating the tool into the target language. Back-translation is considered the best method to ensure semantic equivalence, translating the scale into the desired language and then reverse translation by another translator into the original language.

A high-importance concept, technical equivalence should be directly related to the assessment method and not interfere with results obtained from one culture to another, for example, when the paper-and-pencil method is not necessarily familiar or a male interviewer interviewing women in certain contexts, which can skew the data.

Translation is the first step; the adaptation process is the second step. The term “adaptation” has a different meaning from “translation”; hence, adaptation contains all modifications of the translated document related to cultural, dialectal, linguistic, and contextual aspects (Beaton, 2010; Gjersing, 2010; Paula, 2014). Adaptation ensures that the resulting tool achieves the required uniformity and offers contextual, dialectal, experiential, and hypothetical equivalence, as backed by the expert committee review and the creators of the scale (Bravo et al., 1991).

In this study, our general objective was to present the different steps necessary at the very beginning of the process of cultural adaptation and validation of the SSTICS. In the UAE, the questionnaire’s clientele is from several countries and with various ethnicities making the study complex. We started the back-translations with the Emirates Arabic language, which is the predominant language in the United Arab Emirates, and then we added a comparison of back-translations from 9 other languages for the scale to compare the differences between all back-translations. All the versions will be added to the first Arabic back-translation since UAE has residents from different countries and nationalities (Table 7).

Specifically, we aimed to (1) translate the SSTICS into UAE Arabic; (2) explain the cultural adjustment needed to make the scale applicable and adjusted to suit the UAE; (3) present and post the scale and make it available to clinicians, readers, and researchers who are planning to validate it in the UAE. (4) We tested the feasibility of it in the UAE with a pilot (N = 13) and an exploratory study (N = 23). (5) Additionally, we added comparisons for the scale back-translations in different 9 languages, that can be used and validated in those countries, as well as for patients with those nationalities living in UAE such as Egyptian Arabic, Pakistan/Urdu, Hindi, Marathi, Lithuanian, Serbian, German, Romanian, Sinhala, and Russian (Table 8).

The initial translation aims to develop the first version of the tool in the language used in the UAE. A native speaker of both languages Arabic and English did the initial translation. The first step in developing a subjective scale to investigate cognitive deficits in the Emirate population was performed through four main stages. The first stage was the examination of the two versions of (SSTICS), the initial French language version that was translated to Tunisian Arabic (SACCS) and the English language version (SSTICS). The second stage was the translation (by a single researcher) of the English SSTICS into a second translated Arabic version in the Emirates Arabic language, named the Subjective Scale to Investigate Cognition in Emirates (SSTICE).

The third stage involved translating the two Arabic SSTICS versions (SACCS and SSTICE) back into the English language. Two independent researchers processed each scale version, each blind to the other “investigator’s work.” After completing this stage, we obtained four different back-translations (two back-translations for each scale), performed by four independent researchers. In the fourth stage, one reviewer compared the four translations and chose the version most like the original English-language SSTICS, so that adaptation and validation could be performed, and the final version of SSTICE-GCC was presented (Figure 1). The author of the scale (ES) reviewed all the versions.

During the harmonization stage, meetings were held between the different people involved in translation to examine the divergent items and concepts and ensure inter-translation validity. During these meetings, participants made decisions regarding concepts identified as potentially problematic or confusing in the instructions’ meaning and understanding. When necessary, the author of the tool has provided clarifications on the meaning of the concepts so that the Arabic version is as faithful as possible to the original. Decisions made by meeting participants were noted, and a draft of the tool was developed.

We provided the final version of the scale to (N = 13) residents in psychiatry to check the visibility and preference, and to ensure the understanding of each item by the participants. For each word questioned by participants or not understood, the final version was reviewed and adapted by the two translators by consensus. The step of reviewing the results of the application and finalization allows us to make the last necessary changes in the preliminary tool. The instructions for the various items were well understood by clinicians and were easily administered; few reformulations were necessary (Table 9).

The Arabic language is known for its complexity and richness of vocabulary. Its singularity appears from its designation. If one mentions Arabic without an adjective, some will understand that it is classical Arabic, which is also called literary, literal, or Koranic. Others perhaps, less numerous, would consider it dialectal Arabic, i.e., one of the forms of Arabic spoken in the Maghreb like Tunisia or Saharan Africa (Mauritania) or the East (Egyptian, Sudanese, Chadians, Saudis, Jordanians, Palestinians, Lebanese, Syrians, Iraqis, speak of Dathina, Hadramawt, Yemeni, and Omani).

Diglossia that prevails in the Arab world, namely the existence of a literary variant reserved for written use and the dialectal variants that are spoken in the various Arab countries, might be very old. The Koranic Arabic was not a spoken language whose dialects would be emitted historically but a standard of prestige in which the prophet received and wanted to transmit the revelation of Islam. In addition, literary Arabic is currently used in all Arabic literature. It can be used on the radio and television, in official speeches, in scientific communication, and in any other formal circumstance (Figure 2).

What nevertheless makes it possible to say that literary Arabic is not a spoken language in the ordinary sense of the term is, in fact, simple: it is not the mother tongue of people; the inhabitants of Arab countries speak and transmit to their children only the dialect of the country where they live. We know that Arabic dialects can be quite different, even within the same geographic and cultural settings, and more so between two sets: an Emirati and a Moroccan may have trouble communicating, each speaking his/her dialect. In this case, using literary Arabic may be considered. However, because of the substantial heterogeneity of the levels of schooling and education, not every Arabic speaker knows literary Arabic well enough to use it in their oral communication. In addition, the use of literary Arabic in oral communication appears to most Arabic speakers as being artificial, the dialects alone having a reality as instruments of conversation (Table 10).

The initial translation aims to develop the first version of the tool in the language used in the UAE. A native speaker in both languages Arabic and English performed the initial translation. The process of adopting the SSTICS to produce the SSTIC-E was performed through four main stages:

Finally, in the harmonization stage, meetings were held between the different people involved in translation to examine the divergent items and concepts and ensure inter-translation validity. During these meetings, participants made decisions regarding concepts identified as potentially problematic or confusing in the meaning and understanding of the instructions. When necessary, the author of the tool has provided clarifications on the meaning of the concepts so that the Arabic version is as accurate as possible to the original. Decisions made by meeting participants were noted, and a draft of the tool was developed.

We provided the final version of the scale to 13 residents in the Al Ain Hospital psychiatry clinic, the main goal was to check the translated version in terms of wording adaptability, visibility, and preferred language (Arabic or English) of using the SSTIC-E, as well as to ensure the understanding of each item of the scale by the participants. No obstacles were documented by participants, and all their comments were taken into consideration for a better and clearer version of the final Arabic version of the scale. The final version was reviewed and adapted by the two translators by consensus.

To enhance the importance of linguistic and textual differences in a different type of cross-translation and cross-adaptation based on each country’s language, context, and culture, we received different versions of back-translation from different countries. All the unpublished translations are displayed in the supplementary section. These versions of the scales are Pakistani/Urdu, Hindi, Marathi, Lithuanian, Serbian, German, Romanian, Sinhala, and Russian.

When we compared the back-translations, we found slight differences in using alternative synonymous words without changing the general meanings of the sentences or the questions. Moreover, we found some of the back-translations had slightly different grammar, others added more words or used completely different vocabulary (for example, in some back-translations, they used “you face difficulty” and in other back-translations “you struggle,” also, some used “house maintenance” and some used “home repairs”).

Some of the back-translations we found typical as the original scale, those highlighted in light orange (original scale BT, and BT-Version 1, 3) and others also had typical back translations but differed slightly from the original scale, highlighted in light green (BT-Version 4, 5). We classified differences into two types: grammarian, and semantic and contextual.

Because most of the back-translations come from countries where English is not the first language, we have noticed some mistakes in grammar and punctuation. These variations may affect the context meanings because of their misplaced positions in the sentence or using the wrong verb or noun sometimes.

Context and culture: The socio-political context would also have influenced the formulation of the question. For example, “Who is the first minister of Canada?” (Q #10) is a question related to explicit semantic memory. In some countries, the election of a first minister or president of the republic is every 4 or 5 years. In other countries, there is no election, and the leader is for life; this is the case with monarchical models. In the use of the scale in France, for example, we will instead ask who the president of the Republic is. In the Emirates, the question would be “What is the name of the leader of the country?”

In addition, the sociological context is not the same relative to cooking, laundry, etc. For example, Q.17 tests the executive tasks, but in the UAE culture, this question is not significant since Emirati citizens do not do laundry themselves, instead, they rely on house helpers or automatic laundries (see Supplementary Table S1). Considering that the SSTICS was developed from a construct of cognition, the adequacy between the question on executive functions and real life is crucial if one wants a minimum of ecological validity (Supplementary Table S1)

In this process, we had to choose between different methods such as the traditional translation. It simply consists of the translation by a bilingual researcher or professional translator of the original instrument. This method, used alone, is not recommended because it introduces significant bias, especially in the interpretation of the researcher or the translator. This difficulty can be overcome by performing several parallel translations by different bilingual translators or researchers, but the following methods are recommended. Therefore, the team was composed of several psychiatrists from different cultures within the Arab world.

The method of translation by a committee of experts involves participation in the translation of several bilingual people who know the field for which the instrument is intended, which also limits the bias of a single researcher. This committee can consider a first translated version or participate in the development of a first version. Under ideal conditions, if the author (ES) of the original version participates in it, this makes it possible to clarify certain ambiguities that the translation process generates. We were able at this stage to document certain differences in dialect and semantic usage. For instance:

Comments: “some minor comments on the translation.”

Some repeated words, some literal translations, and redirecting for the correct meanings of the words in Arabic. Here are some examples: all the texts written in Arabic are accompanied by an interlinear gloss, to help the reader follow the relationship between the source text and its translation, and the structure of the original language:

التعليمات: سترى على الورقة التي أمامك، قائمة من العبارات التي تصف بعض مشاكل الذاكرة أو التركيز والتي قد يواجهها كل منكم في أنشطته اليومية.جميع األسئلة عليك تقدير تكرار مثل هذه االضطرابات التي الحظتها مؤخراً في سلوكك، من خالل اإلجابة على.

Back-translation implies that when the first version translated from the instrument has been produced, it is re-translated by a second person in its original language. The gap between the original version and the re-translated version identifies problematic items. This method can be even more sophisticated by performing two reverse translations in parallel, involving four people. This method can be considered ideal and was used. However, researchers who have used it find it challenging to obtain a perfect equivalence between the re-translated and original versions. For instance: Q# 20 Do you have difficulty finding words, making sentences, or understanding the meaning of words (may need to specify whether you mean written or spoken words)? or pronouncing them, or naming objects?

In addition, a recent study by Cella and colleagues showed that the SSTICS could be used as a screening measure and be completed by service users independently (Cella et al., 2020). However, item complexity may limit self-administration, which can affect item interpretation (Eremenco et al., 2005).

As a first opportunity, a population of our area and hospital agreed to participate in our questionnaire when they had a follow-up during the COVID-19 pandemic. This gave us the opportunity to test the use of the SSTIC-E (UAE Arabic) with a sample of 23 subjects. The mean age of our sample was 30.18% (18–62) 62% were female. The mean SSTICS total score was 16.5 (SD:16.9), in accordance with scores found in previous studies using the SSTICS in similar samples (Mancini et al., 2002). Working memory 2.13 (SD 1.83); Explicit memory 6.20 (SD 6.02); Attention 4.65 (SD 4.12); Language 1.13 (SD 1.06); Praxis 0.19 (SD 0.46). No correlations were found with age since this sample size was small. The Ethical Board at UAEU accepted the project and every subject gave consent to participate.

Second, and although this article is not focused on carrying out a clinical validation study at this stage but rather on describing the processes to be respected beforehand in the cultural and linguistic validation of the construct, we summarize here the feasibility results obtained from a population in the Emirates: The Arabic version of the SSTIC-E was administered to a total of 210 participants, including 126 patients and 84 healthy control participants (Al Mugaddam, 2023). The healthy group has a lower mean score of 22.55 (SD = 12.04), with a range of scores from 1 to 63. On the other hand, the patient group has a higher mean score of 34.06 (SD = 15.19), with a range of scores from 1 to 73 (p < 0.001). Additionally, the scores are broken down into five cognitive domains: Memory, Attention, Executive Function, Language, and Praxia. In the Memory domain, the control group has a mean score of 13.50 (SD = 7.21), with a range of scores from 1 to 36, while the patient group has a mean score of 17.58 (SD = 8.22), with a range of scores from 0 to 38; (p = 0.0005). In the Attention domain, the control group has a mean score of 5.81 (SD = 3.65), with a range of scores from 0 to 16, while the patient group has a mean score of 9.68 (SD = 4.64), with a range of scores from 0 to 20; (p < 0.0001). In the Executive Function domain, the control group has a mean score of 3.01 (SD = 2.52), with a range of scores from 0 to 12, while the patient group has a mean score of 4.76 (SD = 3.29), with a range of scores from 0 to 12; (p = 0.0002). In the Language domain, the control group has a mean score of 0.90 (SD = 9.6), with a range of scores from 0 to 4, while the patient group has a mean score of 1.45 (SD = 1.25), with a range of scores from 0 to 4; (p = 0.0786). In the Praxia domain, the control group has a mean score of 0.24 (SD = 0.63), with a range of scores from 0 to 3, while the patient group has a mean score of 0.76 (SD = 1.15), with a range of scores from 0 to 4; (p = 0.0009). To explore comparison, we compared the proportions between the control and patient groups for the SSTIC-E Scores, performing independent sample t-tests for each domain of the SSTIC-E score. In all domains except for Language, the mean scores of the patient groups were significantly higher than those of the control group, indicating worse performance in cognitive functioning. The effect sizes (Cohen’s d) were moderate to large for all domains except for Language, where the effect size was small. Overall, the patient group has higher scores in all domains, indicating that they have more cognitive impairment than the control group. Comparing the study results with the literature, the total mean score on the SSTICS for Stip et al.’s (2003) study population was 25.94 (SD = ±9.72) (Stip et al., 2003), which is similar to the study findings by Raffard et al. (2020) in France, where patients scored 25.56 (SD = ±9.10), but lower than the score of our patient participants, 34.06 (SD = ±15.19). However, the control group in the latter study scored similar results (22.55, SD = ±7.87) (Raffard et al., 2020) to the control group of our study (22.55, SD = ±12.04).

Most of the studies had SSTICS mean scores relatively close to our study results. For example, Stratta et al. (2020) found a mean score of 23.34 (SD = ±14.91) when validating the SSTICS in an Italian patient population (Stratta et al., 2020), while Lecardeur et al. (2009) reported a mean score of 24.73 (SD = ±9.56) for patients with schizophrenia who had preserved awareness of their cognitive deficits, as measured by the SSTICS (Lecardeur et al., 2009). In contrast, Haddad et al. (2021) found that subjective cognitive complaints were associated with poorer cognitive performance in Lebanese patients with schizophrenia, with patients scoring 25.15 (SD = 16.67; min = 0, max = 76; median = 23.50) and controls scoring 9.15 (SD = 7.63; min = 0, max = 37; median = 7.00) (Haddad et al., 2021). The lowest mean score for SSTICS among patients was reported by Baliga et al. (2020) in India, with a mean score of 16.22 (SD = ± 10.8) (Baliga et al., 2022). Similarly, Shin et al. (2016) in Korea reported a mean SSTICS score of 18.9 (SD = ±14.3) for patients, Shin et al. (2016) and Chuang et al. (2019) in Taiwan found a mean SSTICS score of 16.22 (SD = ± 10.8) for their study population of patient participants (Chuang et al., 2019).

The intended application of translated versions is multiple depending on the languages and countries where the scale may be useful. In the Emirates, it gives the possibility of choosing a version, depending on the origin of the person tested, their mother tongue, or the language they use at home, so that the complaint best reflects their cognitive reality. The diversity of the authors of the article reflects this concern and may reflect the motivation of a healthcare professional to use the scale with a variety of people. Additionally, the scale can be used by professionals who propose a cognitive remediation plan. For example, neurocognitive disorders, which affect approximately 80% of people with schizophrenia, are present from the first psychotic episode. They must be objectified as soon as possible. A neuropsychological evaluation is considered when a cognitive complaint is formulated. Such a complaint may initially be raised by the SSTIC-E. In this case, a neuropsychological assessment and an assessment of social cognition are implemented. In the end, the results must be returned to the patient, so that they can take ownership of them. The preserved processes on which they can rely on to recover are put into perspective with those that are altered so that they can manage them as best as possible. The application will also eventually be broadened in transdiagnostic approaches. In certain geographic or economic situations, there are no means to carry out a complete neuropsychological assessment. The scale will still provide a signal of cognitive difficulty. Finally, the scale can be used in clinical or epidemiological research.