Measurement refers to a procedure which identifies the value of a quantitative variable through its numerical relationship to another value (Michell, 1990). For example, we want to know the body temperature of a particular person. Suen (1990) says, “The science of developing educational and psychological tests and measurement procedures has become highly sophisticated and has developed into such a large body of knowledge that it is considered a scientific discipline of enquiry in its own right. This discipline is referred to as psychometrics”. Presently, psychological researchers are intended to conduct cross-cultural researches and they must have reliable and valid instruments conducting those researches.
Psychology is a growing field in Bangladesh both for research and practice. A number of psychological researchers of Bangladesh give emphasis on quantitative research which requires standardized instruments to measure behaviors and mental processes. However, the reasonable way to get an instrument can be cross-cultural adaptation whereas development of a new instrument is relatively tricky and time consuming. Meanwhile, many researchers adapted numerous instruments or psychometric tools to measure a range of behaviors and mental processes of human being.
The adaptation of psychometric tools differs in the process, study design, sampling technique, sample size, reliability and validity. Therefore, the objective of the present study is to systematically review the adaptation of psychometric tools in Bangladesh, more specifically the adaptation procedure and research methodology. Table 1 shows the review.
Sousa & Rojjanasrirat (2010) provided a user-friendly guideline for translation, adaptation and validation of instruments or scales for use in cross-cultural health care research. The guideline includes seven steps for the adaptation of psychometric tools. The steps are (1) Forward translation, (2) Comparison of the two translated versions (synthesis I), (3) Blind back-translation, (4) Comparison of two back-translated versions (synthesis II), (5) Pilot testing of the pre-final version, (6) Preliminary psychometric testing of the pre-final version, (7) Full psychometric testing of the pre-final version.
The Review of the Adaptation of Psychometric Tools in Bangladesh
1. Adaptation of pervasive developmental disorder assessment scale (Huq & Das, 2007)
i.Test-retest reliability of 3 subscales: 1. social interaction(.90) 2. language & speech( .88) 3. imaginative play(.92).
2. Adaptation of inventory of interpersonal problems (Hossain, Yasmin, & Uddin, 2008)
i.Split-half(.90) ii.Cronback Alpha(.88) iii.Parallel form(.074-.651)
3. Adaptation of inferiority questionnaire (Uzzaman, Islam, & Haque, 2011)
i.Split-half(.74) ii.Cronbach alpha(.79)
4. Preliminery evidences for psychometric properties of the Bangla parental power-prestige questionnaire (Ahmed, Uddin, & Islam, 2011)
5. Adaptation of questionnaire on stress in cancer patient revised (Jerin, Khatun, &Ahmed 2013)
i.Test-retest(.83) ii.Cronbach alpha(.967)
6. Bangladeshi adaptation of Warnick-Ebinburgh mental wellbeing scale (Rahman & Imran, 2013)
i.Test-retest(.72) ii.Cronbach alpha(.77) iii.Splithalf(.87)
Method of Review
The aim of this systematic review is to investigate the quality and utility of psychometric tools adapted in Bangladesh. For conducting this review, the psychological journals published in different public universities were taken for literature review. Then the studies on adaptation of psychometric tools were picked up initially. Finally six original papers on adaptation of psychometric tools which met the criteria of current systematic review were selected from these journals. Among them four papers were selected from ‘The Dhaka University Journal of Psychology’ and two other papers were selected from ‘Jagannath University Journal of Psychology’.
The papers were basically selected based on the following criteria
Has the study been done in the context of Bangladesh?
Has it been done on the adaptation of psychometric tools for use in Bangladesh?
Whether it is an original study or review?
Whether it is published?
Results of Review
Information about the Reviewed Studies
All the selected studies represent the adaptation of psychometric tools for use in Bangladesh. The pervasive developmental disorder assessment scale presented in first study (shown in table 1) contains 40 items and 3 subscales. Similarly, the inventory of interpersonal problems contains 64 items and 8 subscales. The inferiority questionnaire and the Bangla parental power-prestige questionnaire contain 20 and 10 items respectively. The questionnaire on stress in cancer patient contains 23 items and 5 homogenous subscales shown in fifth study of table 1. Finally the Warnick-Ebinburgh mental wellbeing scale is composed of 10 items. The sample size of the reviewed studies ranges between 50 and 232. Only the first study shown in review table followed random sampling technique. On the contrary, other five studies followed convenience and purposive sampling technique. The design of all six studies was cross-sectional survey. The reliability ranges between .72 and .96. In case of validity most of the studies followed content and construct validity.
Adaptation Process Followed by the Studies
Researchers of three studies (such as study number 1, 3, and 5 shown in table 1) among six mentioned that they followed the guidance provided by International Test Commission (ITC). However, the researchers of other studies did not mention that which guideline they followed during adaptation.
The first step of adaptation process is forward translation by two different translators. Two translators should have different background for example one translator must have the knowledge of health terminology whereas the second translator must possess cultural and linguistic knowledge (Sousa & Rojjanasrirat, 2010). However, no researcher among six mentioned the background of the translators in their papers. In fact, most of the researchers did not even clarify the different phases of adaptation process although the reliabilities of the adapted instruments are relatively high. Moreover, the researchers of all the above studies except study number 6 (shown in review table) did not make contact with the original developers for their consent and proof reading which is considered as an important phase of adaptation of psychometric tools.
On the other hand, the blind back translation step needs two translators who are native speaker of the original language from which the instrument was translated in first step. In addition the translators must be chosen form distinct background like the translators of forward translation (Sousa & Rojjanasrirat, 2010). It is paradoxical that none of the reviewed studies followed the blind back translation phase strictly.
However, the other steps mentioned in introduction part were reasonably followed by the above researchers. That is why the reliabilities of their adapted instruments are relatively satisfactory.
Sample size and sampling technique
Whether the result of a study can be generalized or not, depends on the sample size as well as the sampling technique by which the sample is drawn. Sousa & Rojjanasrirat (2010) suggested using 300 to 500 subjects for the adaptation of psychometric tools. Contrary to that all the researcher of current reviewed studies used less than 300 subjects in their research. Moreover, five out of six studies did not use random sampling. That is only available and interested people were the subjects of these studies. Consequently, Passable generalization may not be possible by these studies.
Most of the psychological researchers in Bangladesh show interest in quantitative research. Reliable and valid instruments are needed for this type of research. In fact, lots of such instruments or questionnaires have been adapted at different times in Bangladesh. However, after this systematic review it is difficult to say that the researchers could follow standardized adaptation procedure and research methodology to adapt these instruments. Standardized adaptation procedure and research methodology are important because cultural diversity among countries is a gigantic issue of discussion in research world. If the adaptation process of a psychometric tool becomes flawed it will not be able to measure a sample of behavior accurately. Ultimately the field of psychology including mental health practice may suffer as it is sometimes very much dependent on the psychological tests and measurement. The further research on adaptation of psychometric tools in Bangladesh should follow the standardized procedure and research methodology strictly.
Ahmed, F., Uddin, M. K., & Islam, M. J. (2011). Preliminery evidences for psychometric properties of the bangla parental power-prestige questionnaire. Jagannath University Journal of Psychology, 1, 97-106.
Hossain, M. S., Yeasmin, M., & Uddin, M. K. (2008), Adaptation of inventory of interpersonal problems. The Dhaka University Journal of Psychology, 33, 1-10.
Haque, S. & Das, A. (2007), Adaptation of pervasive developmental disorder assessment scale. The Dhaka University Journal of Psychology, 31, 11-22.
Jerin, M. I., Khatun, M. N., & Ahmed, A. (2013), Adaptation on questionnaire on stress in cancer patient revised. The Dhaka University Journal of Psychology, 37, 23-30.
Michell, J. (1990), An Introduction to the Logic of Psychological Measurement. Hillsdale, New Jersey: Lawrence Erlbaum.
Rahman, S. T. & Imran, M. H. (2013), Bangladeshi adaptation of Warwick-Ebinburgh mental well-being scale. The Dhaka University Journal of Psychology, 37, 49-60.
Sousa, V. S. & Rojjanssrirat, W. (2010), Translation, adaptation and validation of instruments or scales for use in cross-cultural health care research: a clear and user-friendly guideline. Journal of Evaluation in Clinical Practice, 17, 268-274.
Suen, H.K. (1990), Principles of Test Theories. Hillsdale, New Jersey: Lawrence Erlbaum.
Uzzaman, M. A., Islam, A. S. M. A., & Haque, M. A. (2011), Adaptation of inferiority questionnaire. Jagannath University Journal of Psychology, 1, 35-42.