Thumbnail Image

The Probability Distribution Framework for estimating the Prevalence of Undernourishment

Exploding the Myth of the Bivariate Distribution

Year of publication 2007

Product type Document

Edition 1

Synopsis (short abstract) In his pioneering study carried in the early 1960’s, Sukhatme had formulated the estimate of the prevalence of undernourishment in a population within a bivariate distribution framework where dietary energy consumption (DEC) and dietary energy requirement (DER) are considered as random variables. The evaluation of the formula required the specification of the joint distribution of DEC and DER. In the absence of data on the joint distribution Sukhatme had, as an approximation, formulate d the estimate within a univariate distribution framework involving the distribution of DEC and a cut-off point reflecting the lower limit of the distribution of DER. FAO’s methodology for estimating the prevalence of undernourishment has been traditionally based on this univariate distribution framework. However, since this approach appeared to ignore the risk of undernourishment at DEC levels overlapping the range of variation of requirement, it has been criticised as yielding an und erestimate of the magnitude of the problem of undernourishment. In view of this some analysts have attempted to apply the bivariate distribution framework by modeling the joint distribution of intake and requirement. Others have applied the univariate distribution framework but used the average DER requirement rather than the lower limit of the distribution of DER as the cut-off point. All these attempts have led to very high estimates of the prevalence of undernourishment. In further studies undertaken in the 1970’s Sukhatme has attempted to justify the univariate distribution framework that he proposed earlier by postulating the theory of intra-individual variation in energy requirement which implies that an individual cannot be considered to be undernourished or overnourished as long as his or her DEC is within the range of variation of DER.

Language English

Also available in:
No results found.

Related items

Showing items related by metadata.

  • Thumbnail Image
    Document
    The FAO parametric versus the IFPRI non-parametric approach to estimating the Prevalence of Undernourishment:
    Issues Relating to the Use of Household Level Data from National Household Surveys
    2007
    Also available in:
    No results found.
    A non-parametric approach suggested by researchers from the International Food Policy Research Institute (IFPRI) for measuring food deprivation (undernourishment) is not an improvement to the current FAO parametric approach. This is mainly due to flaws arising from the use of an inappropriate methodological framework and the reliance on single household data from national household surveys (NHS) that are subject to undesirable sources of variation. FAO’s parametric approach is still th e only choice for estimating the prevalence of undernourishment for the purpose of monitoring hunger reduction at country, regional and global levels. The FAO approach estimates the average food consumption parameter from national food balances such as those from the FBS compiled and prepared by FAO on yearly basis. The FBS is the only data source for global monitoring. The parameter on inequality in food access is derived from NHS data, which are collected less frequently. For estim ating the prevalence of undernourishment at subnational levels and identifying population groups at high risk of food insecurity, countries are applying the FAO method to derive both the average and the inequality parameters from the NHS data.
  • Thumbnail Image
    Book (series)
    Methodological issues in the estimation of the prevalence of undernourishment based on dietary energy consumption data: A review and clarification 2014
    Also available in:
    No results found.
    Sukhatme had in the early 1960’s originally formulated the estimate of the proportion of undernourished in a population (PU) within a bivariate distribution framework where dietary energy consumption (DEC) and dietary energy requirement (DER) are considered as random variables. However, in the absence of data on DEC and DER of individuals expressed in the form of bivariate distribution, Sukhatme had suggested a formula that considers the part of the distribution of DEC below a cut-off point repr esenting the lower limit of the distribution of DER as an estimate of PU. However, this univariate approach has been criticised as yielding an underestimate of the magnitude of the prevalence undernourishment in a population. In response to this critic, Sukhatme has attempted to justify the approach by invoking the theory of intra-individual changes in DER. As this theory has led to a controversy rather than a clarification of the univariate approach, doubts regarding its validity still prevail. Following a review of these developments including the concept of DER, this article shows that the formulation of PU within the bivariate distribution framework is inappropriate. Subsequently, the relevance of the univariate approach is clarified. Finally, the article addresses certain issues relating to practical estimation of the prevalence measures based on household rather than individual data pertaining to DEC.
  • Thumbnail Image
    Book (series)
    Methodological note on new estimates of the prevalence of undernourishment in China 2020
    Also available in:
    No results found.
    This paper presents new estimates of the extent of food consumption inequality in mainland China and discusses their implications for the estimated prevalence of undernourishment (PoU). The new food consumption inequality estimates are based on the joint analysis of food consumption and food expenditure data obtained from two separate household surveys, covering the period from 2011 to 2017. The results reveal much less inequality in dietary energy consumption than previously assumed and imply a substantial downward revision of the estimated series of the PoU for China, which becomes more in line with other assessments of food insecurity and with other development indicators.
Show more

Users also downloaded

Showing related downloaded files

No results found.