High-risk adolescents for early pregnancy are very likely to have a higher rate of functional illiteracy than adolescents at lower risk and may be unable to complete self-administered questionnaires. Personal interviews may result in response biases that inaccurately portray the level of sexual activity among very high-risk adolescents. The goal of this project is to develop and evaluate a computer-based interactive instrument with an audio component for collecting sexual behavior and pregnancy information from adolescents that can be used for evaluation of pregnancy prevention programs and research with high-risk, low-literacy adolescents. Functionally illiterate adolescents will be able to complete this instrument independently, assuring their sense of anonymity.
The specific aims of the project are: (1) to develop an interactive computer-based program with an audio component in order to address the issues associated with anonymity and reporting sexual behavior and pregnancy for functionally illiterate adolescents; (2) compare the effects of different methods (i.e. self-administered paper-and-pencil questionnaire, personal interview, and the computer-based instrument) for collecting sexual behavior information on data quality, subjects' reactions to the method of data collection, and response bias; (3) examine the interaction of literacy level and method of data collection on data quality, reactions to assessment, and response bias, (4) examine the relationship of individual differences in impression management, trait anxiety, and perceived social norms to degree of response bias for each of the different methods of data collection. The evaluation of this instrument will involve comparing the responses of adolescents to different methods of collecting this information under different expectations of anonymity for their answers. The design of the project is a 2 x 3 factorial experiment. The first factor is the level of expected anonymity, included to determine the direction adolescents distort their answers as a result of social desirability or impression management. The second factor is the type of method used to collect the information -- (1) self-administered paper-and-pencil questionnaire, (2) personal interview, or (3) computer instrument. Individual differences in impression management, anxiety, and reading level will be examined as possible covariates. Gender, age, and ethnic group differences will also be examined. Analysis of covariance will be the primary data analytic technique.
