WITHDRAWN: A Measure of College Student's Privacy Management about Condom Use Measure

Generally, adolescents are uncomfortable broaching sexual health topics, and confidentiality is especially important to teens when they seek out sexual health care. In light of thick boundaries surrounding young people information about condom use, understanding privacy management is of utmost importance in developing sexual health interventions. The theory of Communication Privacy Management has been studied extensively in health contexts, though most prolifically via qualitative methods. This study seeks to extend understanding of CPM in the context of HIV prevention through development of a privacy management scale assessing how young people manage their private information about condom use. Items were created from belief statements extracted from the natural language of interviews and focus groups exploring college students’s privacy management of condom use information. Each of the 79 items required a 5-point Likert-type scale response, ranging “never true,” “rarely true,” “sometimes true,” “often true,” to “always true.” The questionnaire contained three subscales with domains corresponding to the theory of CPM: boundary linkages (BL), boundary ownership (BO), and boundary permeability (BP). Two-hundred and thirty nine participants from a participant pool at a large Northeastern university responded to the online questionnaire. The three sub-scales all returned good reliability. The descriptive statistics were slightly skewed, with skewness values all less than .5, and slightly kurtotic, with absolute values all less than one. Within the three-factor solution, most of the items associated within each factor in the matrix also corresponded to the a priori theoretical assignment of items, and then were compared to existing quantitative explorations of CPM. To confirm the underlying structure of the data, the fit was tested for four models. The first model contained all of the items from the Exploratory Factor Analysis with a single latent factor. The second model contained three uncorrelated factors. The third model contained three correlated factors, with one of the observed variables for each latent variables scaled to one. The fourth model was a hierarchical model, containing three first-order factors and one second-order factor, with the second-order factor scaled to one, and one of the observed variables for each latent variables scaled to one. The global fit improved progressively from model one to model four. Ultimately, the best fit was found for three first-order factors, one second-order factor. Researchers interested in the theoretical development of CPM can learn that there are in fact, likely three factors in this management process, with a second order privacy management factor, confirming the claims of the theory.