TY - JOUR AU - González, María Isabel AU - Chen Mok, Mario PY - 2004/01/01 Y2 - 2024/03/28 TI - El numero deseado de hijos en Costa Rica: 1993-1999 JF - Población y Salud en Mesoamérica JA - PSM VL - 1 IS - 2 SE - Scientific articles DO - 10.15517/psm.v1i2.13930 UR - https://revistas.ucr.ac.cr/index.php/psm/article/view/13930 SP - AB - <span style="color: #000000; font-family: 'Times New Roman'; font-size: medium; font-style: normal; font-variant: normal; font-weight: normal; letter-spacing: normal; line-height: normal; orphans: auto; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: auto; word-spacing: 0px; -webkit-text-stroke-width: 0px; display: inline !important; float: none;">This study examines the ideal number of children in Costa Rica using the 1999 National Reproductive Health Survey. The survey interviewed a total of 1030 women between 18 and 44 years of age from a sample of 50 census tracks around the country. We study the evolution of the ideal number of children since the last national survey of this kind conducted in 1993. The ideal number of children maintains the downward tendency observed since 1964 with a decrease from 3.4 to 2.7 between 1993 and 1999. This decrease came out to be statistically significant (p&lt;0.001) based on a t-test for independent samples. However, given that most of the census tracks were part of the sample of the census tracks used in the previous sample of 1993, an alternative paired test using the average ideal number of children per census track was performed as well. This test led to similar results, but with a smaller p-value. With the purpose of identifying a set of variables of easy measurement for the prediction of the ideal number of children, we fitted multivariable models based on ordinary least squares and Poisson. The variables analyzed were: age, occupation, education, number of children, and religion. Both models had problems related to the base distribution, but did not seem to have major problems regarding the rest of the assumptions of the models. The ordinary least squares model led to a higher predictive probability (R<sup>2</sup> =.173). The only significant variable in both models was the number of children.</span> ER -