Adaptation of PG-13 on Indonesian Adolescents

Procedure

Adaptation was based on the guidelines proposed by Beaton on self-report questionnaires (D. Beaton et al., 2007; D. E. Beaton et al., 2000). This guideline proposed six steps on the adaptation process: forward translation, synthesis I, back translation, expert committee/synthesis II, pre-testing and finalisation. Forward translation involved two Indonesians that are fluent in both English (TOEFL ITP 580) and Indonesian as their mother tongue. Both are familiar with translating documents and one is knowledgeable in psychology and counselling terms. After translation, a meeting (synthesis I) was held between the two forward translators and the researcher (also fluent in both languages; TOEFL ITP 630) to discuss the major differences between translations. This meeting produced a compiled translated PG-13 in Indonesian that is then back translated to English. Back translators are non-Indonesians, English as their mother tongue, but fluent in Indonesian and have experience living in Indonesia for at least 5 years (or more). The compiled Indonesian version and the two back translated versions are then considered and discussed in the expert committee meeting (synthesis II). Expert committee consisted of: a psychology lecturer who is well versed in psychological research and child clinical psychology, a practising clinical psychologist that has handled real cases of PGD and an English linguistic major. The expert committee criticised, reviewed the versions mentioned previously and produced a revised Indonesian PG-13. All discussion considered the characteristics of the test subjects (Indonesians ages 13–18 years old), ethical issues, research methodology, characteristics of the disorder in Indonesians in general and the reconstruction into online questionnaire for basic validation purposes.

Pre-testing was done on forty Indonesian adolescents by phone interviews. These adolescents must have an experience with the loss of an immediate family member(s) by sudden death. They were contacted through friends, families and acquaintances in organisations such as churches and NGOs whom are familiar with these adolescents and their situations. Once the adolescents agreed to be included as a participant, they were contacted personally and asked again for their consent by signing an informed consent. One of their parents/guardians is also asked to sign the informed consent. These informed consents were signed digitally or signed on paper, then sent to the researcher by post. After gaining consent, participants were given time to complete the PG-13 Indonesian Version on their own as much as they understood the questionnaire. A phone interview was scheduled with these participants. On the appointed time, participants were given the opportunity to ask clarification on the items that were unclear to them and complete the remaining parts of the questionnaire. Afterwards, the participants were interviewed on the comprehensibility and ease of use of the questionnaire. Participants were asked for inputs in order to clarify the questionnaire further more. Interview questions presented were based on Braun’s research, which are

Was there anything in the introduction, the definitions or the question that you did not understand or did not understand well? (Apakah ada istilah atau pertanyaan di pendahuluan yang tidak dipahami baik atau tidak dipahami sama sekali?)

The results of pre-testing were then considered for changes in the finalisation step, and then readjusted for the online questionnaire.

The PG-13 Indonesian Version was reconstructed into an online questionnaire for basic validation purposes, specifically the content validity. Referring to previous researches (Kaplow et al., 2018; Killikelly et al., 2018; Killikelly & Maercker, 2017), the best way to confirm convergent-content validity is to calculate the correlation between PG-13 Indonesian Version with depression and anxiety of the individual. Theoretically, individuals that experience PGD should also experience some intensity of depression and/or anxiety. To do so, Depression, Anxiety and Stress Scales (DASS) was reconstructed and presented alongside the PG-13 Indonesian Version. Basic content validity was also confirmed by Principal Component Analysis (PCA) as mentioned by Taherdoost & Group (2017).

Other than the content validity, the PG-13 Indonesian Version results were used to calculate the internal consistency of the questionnaire by calculating the Cronbach’s Alpha. The online questionnaires were made on a Google form and the link was passed to organisations (e.g. schools, orphanages and NGO) that are focused on providing the needs of children and adolescents. These organisations then passed the questionnaires to adolescents that are willing to complete the questionnaire and have lost an immediate family member(s) at the minimum 6 months prior completion of questionnaire. Completion of the PG-13 Indonesian Version took 5–10 minutes; interview for pre-testing took half an hour to an hour.

Participants

All participants of this research ranges from the age 13–18 years old. These adolescents must have an experience of losing an immediate family member (parents or sibling), due to sudden death. The cause of death includes accidents, natural disasters, medical issues and unnatural deaths (e.g. suicide, murder and manslaughter). Adolescents were recruited from all over Indonesia, though most participants are from the West Java area.

Adaptation

Adaptation process went according to the steps presented in the methods section. All meetings, synthesis I and expert committee was held online due to COVID-19 pandemic that restricts gathering of any sort. Equivalence of semantic (meaning of words and sentences), idiomatic (colloquialisms and idioms unique to the origin language), experiential (experiences in daily life related to the culture of target language) and conceptual (ideas, attitude and definitions related to the culture of target language) were discussed and agreed upon by translators and experts. One notable change made was the providence of multiple emotional experiences in Indonesian, to capture the variety of emotions that may be experienced. Pre-testing interviews were also done by phone. Participants did not find much difficulties in understanding most of the items. The only issues presented were the role of the blank space and the differences between item 14a (onset of experience) and 14b (duration of experience). This results in the two notable changes for the Indonesian Version: an explanation on the role of the blank space and an example for answering item 14.

Correlation of PG-13 Indonesian Version and DASS

Results of the Spearman correlation indicated that there was a significant positive association between PGD measured by PG-13 Indonesian Version and general distress measured by DASS (rs (129) = .38, p < .001). However, based on Akoglu (2018), this result is considered a weak towards moderate correlation.

Principal Component Analysis

Kaiser–Meyer–Olkin measure of Sampling Adequacy (KMO = 0.885) and Bartlett’s test (Χ²(55) = 556, p <.001) of sphericity indicates an adequate sample size and data to conduct PCA. PCA, with oblique rotation method and parallel analysis, extracted one significant factor (Figure 1) that accounts for 46.1% of the variance within the data. All item loadings ranged from .55 to .81; loadings are higher than the baseline .4 proposed by Taherdoost, et al. (2017) for an acceptable construct validity .

OPEN IN VIEWER

Figure 1.

Scree Plot of Principal Component Analysis on PG 13 Indonesian Version.

Reliability

Cronbach’s Alpha coefficient for PG-13 Indonesian Version is .88; a good and acceptable internal consistency.

Adaptation

The main highlights of adaptation are the semantic, idiomatic, experiential and conceptual equivalence of the translated forms compared to the original forms. Throughout the adaptation process, these aspects must be kept in mind while ensure the comprehensibility of the translated forms. The semantic quality of the translated form was the most difficult to maintain. It was found during the translation process, both by the forward translators and back translators, that there is a limited number of words in Indonesian to describe the variety and intensity of emotions, compared to English. Forward translators also struggled to translate the appropriate terms for yearning that is understandable for adolescents. In light of the items’ purpose and language constraints, the expert committee (including the forward translators) agreed that it is best to provide a series of ‘options’. These options are based on the words that are mentioned by those that has been found to be experienced and mentioned by cases of PGD in Indonesia. This allows a range of experiences to be captured by the item. In hopes that the completion of the assessment will be based on what is most appropriate to the respondents’ personal experience. Back translators had an issue in translating the Indonesian version to the appropriate English tense. However, this does not pose a threat to the core purposes of items in the Indonesian PG-13.

There were conceptual issues that were encountered during the pre-testing, rather than the translation and expert committee meeting. Parts of concern and presented in the expert committee meeting was the use of a blank space in the form, and the two questions regarding the time and duration of loss. Experts were apprehensive on whether the adolescents will be able to immediately recognise how to answer/respond the form appropriately, if there were no further instructions or examples. Experts decided that it was best to maintain the original form of the PG-13 for pre-testing. In case there should be any difficulties for the participants, experts recommended a rough idea of an example and further instruction regarding the blank spaces and questions of time. True to the concerns of the experts, a reasonable number of pre-testing participants struggled to give an appropriate response. Some participants initially thought the blank spaces were to be completed with the multiple-choice answer. Hence, an extra instruction was presented before the first item to explain the purpose of the blank space provided. (Bayangkan isi titik-titik di bawah ini dengan nama dari orang yang telah meninggal dan kamu rindukan. Kemudian pilih jawaban yang sesuai dengan pengalamanmu. Imagine that the blank space provided is the name of a person that has died and you miss. Then choose an answer according to your experience).

In regards to the items enquiring the time of loss and duration of experience, an example was provided to ensure the correct are obtained. Two different examples were provided and participants were asked to give their preference and reasoning behind their choice. The first example provided was focused on how to respond appropriately, though not related to the context of loss.

Semisalnya kamu mengikuti ekskul badminton. Kamu ikut ekskul ini dari kelas 1 SD dan terus mengikutinya hingga kamu kelas 3 SD. Dengan contoh ini, nomor 14a dijawab “sejak kelas 1 SD”; nomor 14b dijawab “selama 3 tahun.”

Most participants chose the second example as it clarifies the intention of the items, while maintaining the context of the item to the whole form. In terms of idiomatic and experiential, there are no issues in the process of adaptation.

Validation

The small correlations between DASS and PG-13 Indonesian Version results have a few plausible explanations. The first explanation lies on the theoretical framework that defines the symptoms that DASS measures. The underlying theory of DASS is similar to the tripartite model by Watson & Clark, though it differs in defining the coherency of the Stress syndrome. DASS assumes that depression, anxiety and stress are strong intercorrelated due to the common causes they share, such as genetic and environmental vulnerabilities and environmental triggers. By this framework, there are diagnostical symptoms that are included in the measurement by DASS. No matter how closely correlated the DASS Scale is to the symptom criteria for anxiety disorders (except Generalized Anxiety Disorder) and Mood Disorders, there is a chance that within the Indonesian population these ‘excluded’ symptom criteria is more definitive in determining diagnostical depression, anxiety and stress.

Another plausible explanation is the intensity of experience in the times of pandemic and the expressive method of such experiences in Indonesian adolescents. In regards to expressing emotional experiences, this research was conducted during the COVID-19 pandemic which has impacted the mood stability global-wise. Some surveys have found that there is a sudden influx of requests for mental health professional in dealing the increase of depressive and/or anxious tendencies within the population. Such experiences may have impacted the quality and frequency of emotional experiences the participants involved in this research. As adolescents in general are prone to over- or underreport their experiences in self-report questionnaires, there is a bigger chance for this tendency to happen due to the impact of their personal experiences during the pandemic.

It is also worth considering that within Asian countries, validation for three-factor scores are considerably lower compared to other cultures and countries. This is credited to interpretation of emotional experience and expression of such emotional experiences. In doing so, some research suggests a shorter version of DASS or considering DASS results as a measurement of general distress, instead of three-factor psychological issues. Similar conditions of those findings may have happened in this research as a further analysis was done on DASS and PGD, by Principal Component Analysis (PCA), and DASS was found to have three major components that contribute to the results. Although these three major components cannot be conclusively categorized into depression, anxiety and stress.

Despite the low correlation coefficient between DASS and PG-13 Indonesian Version, PCA proved that the items that adapted are measuring a common component that is assumed to be PGD. This component also differentiates itself from the components measured by DASS. It shows recognisable impact on the variability of data attained, though comparatively smaller than the components measured by DASS which supports the possibility of over- or under-report of emotional state by participants involved in the validation study.

This study has several notable limitations. One of the most notable limitations is this study is done on a convenient sample of adolescents, which impacts the characteristics of the sample. Although it includes adolescents all over Indonesia, most of the participants are from West Java. Further research that includes representatives from different cultures in Indonesia is highly recommended. Other limitations include the tendency over- or under-reporting of one’s condition and the inability of the assessment to make a clear-cut diagnosis, especially without a standardised norm.