The short-form of the revised junior Eysenck personality questionnaire: A Bengali edition.

BACKGROUND: Personality measuring instrument plays an important role in many fields of human civilization and therefore, present study was aimed to find such an instrument for Bengali speaking juniors.
MATERIALS AND METHODS: Bengali translation of the short-form of the revised junior Eysenck personality questionnaire developed by Corulla was administered on a sample of 226 Bengali speaking students (99 boys and 127 girls) studying in class seven and eight taken from two urban and two rural schools. Internal consistency of each item under a subscale was calculated; internal consistency of each of the four subscales of the translated questionnaire was calculated; test-retest reliability was found with an interval of 3 months and inter-correlations between different subscales were found.
CONCLUSION: The findings provided satisfactory psychometric properties of the extraversion, neuroticism, psychoticism, and lie scale.

Since the development of the Eysenck's scale to measure personality dimensions among adults and juniors, increasing literatures on it indicate its importance. In the series of Eysenck's scale to measure personality the first one was the Maudsley medical questionnaire containing 40 items.[1] As the Eysenck's dimensional model of personality had been refined and clarified the length of the instrument was increased. The Maudsley personality inventory contained 48 items,[2] the Eysenck personality inventory contained 57 items,[3] the Eysenck personality questionnaire contained 90 items,[4] and the revised Eysenck personality questionnaire contained 100 items.[5]

Like the adult scale, the scale to measure personality among juniors was increased in length. The early junior Maudsley personality inventory contained 44 items,[6] the new junior Maudsley inventory contained 62 items,[78] the junior Eysenck personality questionnaire contained 81 items,[4] and the junior revised Eysenck personality questionnaire (JEPQR) contained 89 items.[9]

In practical situation sometimes it becomes difficult to work with a long measuring tool. Therefore alongside the full-length questionnaires some shorter even abbreviated questionnaires were developed. Eysenck, Eysenck and Barrett devised a short-form of the revised Eysenck personality questionnaire (EPQR-S) for adults containing 12 items in each dimension of extraversion, neuroticism, psychoticism, and lie scale.[5] Corulla developed the short-form of the revised junior Eysenck personality questionnaire (JEPQR-S) containing 48 items.[9] Francis developed an abbreviated-form of the junior Eysenck personality questionnaire (JEPQR-A) containing 24 items from the JEPQR-S.[10] He recorded alpha coefficients for different dimensions of the JEPQR-S; extraversion 0.7657, neuroticism 0.7895, psychoticism 0.7124 and lie scale 0.6810 but for the reduced 24 items JEPQR-A, the respective alpha coefficients were reduced to 0.6644 for extraversion, 0.7027 for neuroticism, 0.6112 for psychoticism and 0.5747 for lie scale and this was because of the shorter length of the questionnaire.

The Eysenck personality questionnaires were originally developed in England and then extended to other English-speaking areas with different culture. Cross-culture studies among some English-speaking countries were made by Francis, Brown, and Philipchalk.[11] More recently different forms of Eysenck personality questionnaire were translated in different languages and consistency of the translated questionnaires were examined. Francis, Lewis, and Ziebertz translated the EPQR-S into German and applied over a sample of 331 students.[12] Reported alpha coefficients for the four scales were extraversion - 0.8523, neuroticism - 0.8104, psychoticism - 0.4181, and lie scale - 0.6379. Tiwari, Singh, and Singh translated the same questionnaire (EPQR-S) into Hindi and applied over a sample of 202 Hindi speaking students studying in the Banaras Hindu University.[13] Reported alpha reliabilities for the four subscales were extraversion - 0.766, neuroticism - 0.772, psychoticism - 0.238, and lie scale - 0.624.

Because of the importance of the personality measuring tools in research work and educational purpose, present study was aimed to examine the psychometric properties of the Bengali translation of the JEPQR-S developed by Corulla[9] for Bengali speaking school students.

MATERIALS AND METHODS

Sample

A total of 226 Bengali medium students (male 99 and female 127), studying in class seven and eight, were taken from two urban and two rural schools of West Bengal.

Tool

The JEPQR-S

JEPQR-S developed by Corulla is 48 item self-reported questionnaire having four dimensions extraversion, neuroticism, psychoticism, and lie scale each of which has 12 items.[9] Each question has a binary response and scored either 1 or 0. The maximum possible score of a dimension is 12 and the minimum is 0.

Procedure

The items of the questionnaire were translated into Bengali by two bilingual Bengali personnel. Where there were inconsistencies, both the translators were consulted. Also back translation of the questionnaire by an expert was made and where there were any inconsistencies, the translators were consulted to get best possible solutions.

Internal consistency of each items of each of the four subscales of the translated JEPQR-S was completed. The internal consistency of the four subscales of the translated questionnaire was also calculated. Test–retest reliability was found with an interval of 3 months. Inter-correlations between different subscales were also found.

Statistical analysis

To find internal consistency of items under a subscale, item rest of test correlations were found. The internal consistency of the subscales of the translated questionnaire was found by calculating Chronbach's alpha. The statistical analysis was made with the help of Statistical Package for the Social Science (SPSS) statistics 17.0.

RESULTS

Table 1 presents that the item rest of test correlations vary for the extraversion between 0.082 and 0.417, for neuroticism between 0.127 and 0.552, for psychoticism between 0.085 and 0.504, and for the lie scale between 0.051 and 0.363.

Table 1

Item rest of test correlations

Alpha reliability found for the four subscales were 0.620 for extraversion, 0.663 for neuroticism, 0.604 for psychoticism, and 0.620 for lie scale.

The test–retest reliabilities were found 0.659 for extraversion, 0.652 for neuroticism, 0.631 for psychoticism, and 0.621 for lie scale.

Table 2 presents the correlation coefficients between the four subscales.

Table 2

Inter-correlations between different subscales

DISCUSSION

The present study was aimed to examine to what extent the Bengali translation of the JEPQR-S is functional in Bengali cultured population. The results [Table 1] of the first stage of data analysis indicate that most of the items within a subscale have significant correlations with rest of the items. The alpha coefficients of all the four subscales recorded in this study do not cross 0.7, the level recommended by Kline,[14] but they are very close to it and satisfactory. Alpha coefficients for extraversion, neuroticism and lie scale recorded in this study are less well relative to main stream studies, perhaps, due to cultural difference, as a measuring tool like personality test is always culturally loaded. Significant observation of this study is that the alpha coefficient of 0.604 for psychoticism is impressive against the historic difficulties encountered by the subscale. Francis also found a better alpha coefficient of 0.7124 for psychoticism.[10]

The test–retest reliabilities found in this study are satisfactory and the results are very similar to the findings of Eysenck and Eysenck which were for 13 years boys and girls with an interval of 6 months; extraversion 0.67 and 0.76, neuroticism 0.72 and 0.74, psychoticism 0.63 and 0.66, and lie scale 0.65 and 0.66.[4]

Results relating to the main stream questionnaires provide that the dimensions of personality are not all orthogonal.[410] The inter correlations [Table 2] between the dimensions of the translated JEPQR-S of the present study also show that they are not all orthogonal. There is a significant (0.05 level) negative correlation between extraversion and psychoticism. Neuroticism and lie scale also showed negative correlation significant at the 0.05 level. On the other hand, extraversion and lie scale also neuroticism and psychoticism both showed positive correlations significant at the 0.05 level. However, there is a significant negative correlation at the 0.01 level between psychoticism and lie scale while extraversion is independent of neuroticism.

On the basis of the present study, it may be concluded that the Bengali version of the JEPQR-S may be accepted as a functional tool for research work and educational purpose. It is further recommended to administer the Bengali version of JEPQR-S on a larger sample for determining the psychometric properties more accurately.