Abstract

PIP: Principal components factor analyses with orthogonal rotation were performed on Importance, Sponge Agreement, Pill Agreement, and Diaphragm Agreement scales on the Contraceptive Attributes Questionnaire (CAQ) to ascertain reliability and validity. Specifically, the intent was to measure whether the scales included more than 1 dimension. Internal consistency of the factors was also measured with Cronbach's alpha for each of the factors in each of the Importance and Agreement factors. Data were obtained from a US national telephone survey of 1187 sexually active women 15-44 years. 385 were currently using the contraceptive sponge, 407 women who had formerly used the sponge, and 396 who had never used the sponge and were using other forms of contraception. 91.9% were white, 5.8% were black, and 1.4% were of Hispanic origin. 75.9% were employed outside the household. 61.4% were currently married, 25.2% were never married, and 13.4% were widowed or divorced. The instrument was designed to measure the subjective importance of specific attributed that contraceptive methods may have (Importance) and the perceived characteristics of specific contraceptive methods (Agreement). Importance is measured by the respondent's rating on a 5-point scale of how important each of 18 attributes was in selecting a birth control method. Agreement was determined by asking how much one agrees that the 18 attributes describes each of 3 methods: the pill, the sponge, and the diaphragm (14 attributes). The model of contraceptive decision making assumes that choice of a specific method is a function of value, importance, and likelihood. The attributes were based on 2 other sources, and pretesting reduces 63 items down to 18. The results yielded ratings of agreement for all 3 methods on convenience and absence of health problems and side effects attributes. For vaginal methods only, the attribute of greatest importance was convenience. The conclusion is that the subscales have reasonable internal consistency. In the comparison of women who used different methods, there were differences by user status on the ratings of the Importance and Agreement attributes for all 3 methods. Sponge users emphasized obtaining contraceptives without consulting medical staff. Pill users were concerned with the reputation of the method and convenience of use. Diaphragm users related to the reputation and familiarity with the method. Limitations were built into the sample which was not randomly selected, predominately white, and may reflect more innovative users. Items also were all stated in the positive. The CAQ would be useful is determining the characteristics of methods of importance to her and her perceptions of methods. Future research might examine different populations and compare positive and negative items.