Journal of Pakistan Psychiatric Society

Abstract

and self-report measures. As far as adolescents are concerned, the utility of self-report measures are significantly appropriate because the experiences and other related aspects which are entirely subjective in nature can be better assessed through self-report measures. Among the scales specifically designed for the measurement of attachment of adolescents, IPPA stands out it simultaneously measures attachment of adolescents with parents and peers. Greenberg and Colleagues developed the very first version of the instrument using a two dimensional format (i.e., cognitive and behavioral) based on the theory of Bowlby. But due to the high correlation of these dimensions of attachment with self-esteem and life satisfaction, this two dimensional instrument did not appeared to be a suitable measure of the construct. Keeping in view this phenomenal problem, the revision of the scale was done by Armsden and Greenberg and they converted it into a multidimensional measure with two parallel versions. One version to measure attachment of peers and the other measure attachment of parents. The parent version included 28 items and the peer version included 25 items. The items included in the revised version are combined to produce a total score depicting attachment and scores based on three dimensions of attachment as well. Armsden and Greenberg described these sub scales/dimensions of IPPA as 1) Trust, which is primarily concerned with the needs and desires of adolescents to be recognized and respected by their parents and peers. 2) Communication, stating the adolescents' perception for their parents and peers in terms of emotional support and sensitive response. It is also concerned with quality and extent of communication between parents and adolescents. 3) Alienation, is mainly concerned with the adolescents' feelings of separation, unacceptability, annoyance, and isolation with parents and peers. The IPPA demonstrated sound psychometrics and good reliability indices. The Cronbach's alpha reliability for the subscales ranged from 0.72 to 0.92. Test re-test reliability was reported as 0.93 for parent version and 0.86 for the peer version. But the inter correlations among latent dimensions (Trust, Communication, and Alienation) was high, having the r value ranging from .73 to .78 for parent version and .45 to .78 for the peer version. After the development of the first version, another version of (IPPA) was also introduced. Armsden and Greenberg split the parental version into two forms i.e., father and mother. They revised the inventory as the researches on attachment were consistently arguing about the different pattern of attachment toward father and mother. The three final versions were modified using 25 items only. Responses were presented as choices on a 5-point Likert-type scale. In addition, Armsden and Greenberg introduced classification regarding scores on (IPPA) in terms of secure and insecure attachment. In 2005 Gullone and Robinson revised the IPPA again. They revised the instrument with the objective to validate IPPA for youth age ranging from 9 years to 15 years by simplifying some items. This revision was applied to the original version of IPPA with two forms. Authors named that revised inventory as the inventory for Parent and Peer Attachment-Revised for children (IPPA-R). A number of strengths have been recognized in scholarly literature regarding the IPPA. IPPA is known for its good reliabilities on all three subscales. 2 9 10 1 9 9 11 12 11 13 14 Further the strengths such as strong conceptual background; psychometric properties, quickness, and self-scored procedure proved it a particularly handy, convenient and a useful tool for adolescents' attachment research. In recent years, the factorial structure of IPPA is criticized and a number of studies have been conducted to validate the factor structure of IPPA-R. These studies reported different number of factors for IPPA. Some studies reported a three factor model as the best presentation of the attachment construct in both parents and peers while others categorized it into a two factor model combining the items of trust and communication. With aim to validate IPPA-R in Pakistani context; present study was designed with two main objectives. First, to translate the latest version of IPPA-R into Urdu, and second to validate the dimensional structure of IPPA-R by comparing three models suggested in the literature using confirmatory factor analyses: the one-factor model (overall global score on attachment), the two-factor model, and the three-factor model (alienation, communication, and trust) of revised and simplified version of IPPA-R on adolescents. The present study was extended to assess the psychometric evaluation of best fitted model to data to present the validated measure of the construct. 15-18 Participants Total number of 706 volunteers participated in the study. Among the participants, 396 were females and 303 were males. Most of these adolescents belong to intermediate level of education and were recruited from different private and government colleges. This study used the latest revised version of the instrument i.e., IPPA-R. The IPPA-R was revised to simplify the wording of some items (e.g., item14 of Parent form and item 25 of the peer form). There was no change in the subscales and numbering of items. Scoring was adopted from the original IPPA-R. All items were scored as 1 = Never True, 2 = Sometimes True, 3 = Always True. Instruments 13 SUBJECTS AND METHODS Procedure The first step of the study was to translate the above described instrument into Urdu. In this regard, formal permission was taken from the authors of IPPA-R. By adopting the forward and back translation method, the inventory was translated into national language of Pakistan (Urdu). The forward translation of the inventory was done by three bilingual experts. Bilinguals were selected on the base of their higher education in the resource language (English) and target language (Urdu) as their first language. So it is expected that they will be able to understand items correctly and translate them keeping their meaning and context intact. Next step was to select the best translated items, and for that purpose, a committee consisting of three experts was approached. Same selection criterion was adopted again for the selection of members of the committee (as higher education in English and Urdu as their first language). The objective of the committee was to choose the best 19