Investigating the household standard of living is an important subject with respect to targeting problem in poverty alleviation programs. By doing so, at first, we can identify the poor, indirectly and then, we can targetize the assistants to them. Most of the studies on standard of living, based on survey data, are confronted with the sample selection error (bias). Sample selection bias is not the only feature of standard of living studies, but also, it occurs when the observations have self-selectivity property, which in turn might result in bias and inconsistency with estimators. The main purpose of the paper is the introduction of a general methodology of correcting selectivity bias, with application to the regression of standard of living in rural areas of Iran.