The researchers adopted the Brislin method in translating the questionnaire. According to Brislin (1970), the most common and highly recommended procedure for translating is translation and back-translation. Steps for translating an instrument are translating from the source to the target language, blindly translating back from the translated version to the source language, and comparing the two versions in the original language [14]
a. Translating from the source to the target language
b. Blindly translating back from the target to the source
Back-translation is translating from the newly translated version back into the original source language without looking at the original [14]. The second translator back-translated from the first translator’s Indonesian version into English. The forward-translator and backward-translator did not discuss translation with each other during this stage to avoid the back-translator being influenced by exposure to the original English version of the questionnaire.
c. Comparing the two versions in the original language
The equivalence of the source and target language versions of the instrument must then be evaluated. If the two English versions of the questionnaire are identical, the translated version of the questionnaire (in this case, the Indonesian version) may be considered as equivalent to the original one [14]. Therefore, the researcher conducted a panel discussion to compare if the original English version to the back-translated version were alike, as well as the final version in the Indonesian language.
During this process, there were several distinct opinions between the two translators and the researcher considered terms and sentence structure of the original and translated versions of Indonesian and English. Differences were found in items 3, 4, and 5. First, on item 3, the forward translator used the word “very” in Indonesian translation. Whereas, in the original English version, there is no “very” term on item 3. Therefore, the researcher agreed to remove the word “very” in both Indonesian and back-translated versions.
Then, on item 4: the back-translator wrote the word “stiff” rather than “restrain” as written in the original English version. It because the forward translator also translated “restrain” as “stiff” in Indonesian. Due to the two English version of item 4 was unidentical, the researcher agreed to revise the Indonesian version in order to make it alike.
Lastly, on item 5, the back translator put “aggrieved” rather than the original, “harmed.” The definition of “harmed” is “physically injured”, which is different with “aggrieved”. After a discussion, we decided to change the translated version into a term that is more identical with “harmed” which seemed more suitable. In the end, the researcher and the translators reached agreement on all items.
The process of translation and back translation of Pregnancy-related Anxiety Scale is described in Fig. 1.
Content Validity Index of the PrAS
To test for validity, the researcher invited six experts to evaluate the questionnaire. These six experts include: 2 maternity nurses, 3 lecturers of maternity nursing, and 1 obgyn doctor. The minimal number of content validity experts should be three [15,16]. The content validity index (CVI) in this study that was rated by six experts ranged from 0.83 to 1 and had average of 0.984. The scale’s content validity (S-CVI) average for an instrument should be at least 0.90. In addition, according to Polit and colleagues, an item-level content validity (I-CVI) that is more than 0.50 is acceptable. Specifically, they noted that for six experts, an item with an I-CVI 0.67 is “fair,” while 0.86 and 1.00 are “excellent,” according to the kappa evaluation criteria [16]. These results indicate that the PrAS is valid and reliable for use in Indonesia.