Nursing Research Today is a free monthly online journal that collates and summarizes the latest research about Nursing, including details on health care, hospital staff, professional care.

Reliability of reasons for early termination of breastfeeding: application of a bivariate probability model with sample selection to data from surveys in Malaysia in 1976-77 and 1988-89.

Kuate-Defo B, DaVanzo J

Demography and Preventive Medicine-Epidemiology, University of Montreal, Canada. barthelemy.kuate.defo@umontreal.ca

Although extensively collected, data on people's reasons for their behaviour provided retrospectively have been met with some scepticism on the grounds that they may be subject to biases and errors that jeopardize their usefulness. This study investigates, for a sample of 1,327 births, the reliability with which women in Peninsular Malaysia recalled, at intervals 12 years apart, reasons for not initiating or for stopping breastfeeding less than 3 months after a birth. Overall, we find low to moderate reliability of recall. Levels of reliability are relatively high for some reasons (the child died and no or insufficient milk) but low for some others (child ill, breastfeeding inconvenient). Results from selection models show that reliability does not vary with the length of time since the child's birth but is inversely related to socio-economic status (proxied by education and employment). Social status, social norms, and health-related factors appear to be significant influences on women's consistency of reporting.

Published 8 February 2006 in Popul Stud (Camb), 60(1): 83-98.
Full-text of this article is available online (may require subscription).

© 2005-2008 Nursing Research Today. All Rights Reserved.