Development and Psychometric Evaluation of the Prenatal Attachment Scale for Thai Pregnant Adolescents (PAS-Thai)

Abstract

The purposes of this study were to develop the Prenatal Attachment Scale for Thai pregnant adolescents (PAS-Thai) and to examine the psychometric properties of the developed PAS-Thai. Based on DeVellis (2017)'s scale development guideline, the PAS-Thai was developed in two phases with eight steps. The first phase was the PAS-Thai development. The second phase was the psychometric evaluation of the PAS-Thai. In the first phase, the PAS-Thai was developed based on extensive literature review and qualitative approaches using individual in-depth interview and focus group discussion with 13 Thai pregnant adolescents, generating three domains and 56 items: 1) cognitive attachment domain (18 items), 2) affective attachment domain (17 items), and 3) behavioral attachment domain (21 items). In the second phase, the PAS-Thai was reviewed for content validity by five experts. One item was excluded because of redundancy. With 55 items, the content validity of the PAS-Thai was satisfactory both at item level and scale level (I-CVI = .80-1.00, S-CVI/UA = .89, S-CVI/Ave = .98.). The PAS-Thai was tested for reliability with 30 pregnant adolescents, yielding a Cronbach's alpha coefficient of .94 for the total scale, .86 for cognitive attachment domain, .90 for affective attachment domain, and .92 for behavioral attachment domain. The questionnaires were distributed to 575 samples in five regions of Thailand, 354 cases were available without missing data for further analysis. Exploratory factor analysis (EFA) using PAF extraction with varimax orthogonal rotation method produced 35 items within three factors namely: having good wishes for the baby, preparing to care for the baby, and feeling of connectedness to the baby. The eigenvalues for the three factors were 6.82, 6.03, and 3.32, respectively. The percentage of total variance explained was 46.19, and the percentage of variance explained for each factor was 19.48, 17.23, and 9.48, respectively. The scale had a good internal consistency reliability with Cronbach's alpha coefficient of .94 for the total scale, .91 for "Having good wishes for the baby" domain, .90 for "Preparing to care for the baby" domain, and .89 for "Feeling of connectedness to the baby" domain. The contrasted group approach result revealed significantly different scores between the group of pregnant adolescents who wanted to and were ready to have a baby and the group of pregnant adolescents who did not want and were not ready to have a baby (t = 9.05, p < .001). Thus, the 35-item PAS-Thai possessed acceptable validity and reliability for use in measuring cognitive, behavioral, and affective aspects of prenatal attachment of Thai pregnant adolescents.