The main objective of this paper is to explore the relationship between poverty variables and eight sociodemographic
correlates like location, gender, age, household size, marital status, occupation, land
ownership and house ownership in one of the poorest regions of Bangladesh. Poverty is defined and
measured by ten indicators, which incorporate multi-dimensionality of poverty - food, income, assets,
consumption, capability and well-being. Data for this study come from an UNDP-assisted survey of two
Upazilas of Bangladesh, Gaibandha Sadar and Tanore in Gaibandha and Rajshahi districts respectively,
carried during April-May 2002 by Bangladesh Rural Development Board under Community Empowerment
Project-2. A total of 5,180 heads of household, 3,158 in Gaibandha Sadar and 2,022 in Tanore, are surveyed
through a multi-stage stratified sampling. Chi-square measure of association and Pearson’s correlation is
used to ascertain the degree and direction of relationship between poverty variables and factors of poverty. It
is found that the incidence of rural poverty ranges between 46 per cent to 67 per cent and that income,
capacity and well-being poverty is greater than food poverty. Furthermore, it is found that land ownership
and occupation are the crucial correlates of poverty followed by marital status, age, geographic location and
gender. Most of them are significantly related to multi-dimensionality of poverty at the significance level of
α=. 01.