This section describes the materials and methods used in the study. It discusses the methods of constructing the wealth indices, the sampling procedure, the selection of durable assets, and the analytical techniques. Methods of constructing wealth indices In the present study, three methods of constructing asset-based indices have been used. These are the unweighted asset index, the weighted asset index using the inverse of proportion, and the weighted asset index using Principal Component Analysis (PCA). For convenience, the indices are named index 1, index 2, and index 3 respectively. Among others, Filmer and Pritchett (2001), McKenzie (2003), Prakongsai (2006), Vyas and Kumaranayake (2006), Howe et al. (2008), Zerin (2009), Rahman et al. (2013) have developed and applied several methods for constructing asset-based indices to measure the level of asset ownership of households and to classify their socioeconomic positions. In light of those studies, the aforementioned methods of constructing asset indices are described. Unweighted asset index: This is the easiest method, which involves simply summing up the number of assets possessed by a household. This method is often called the index using equal weight. The unweighted asset index is widely used in the literature. Suppose there are n households selected as the sample for the study. Let there be k types of assets considered in the study that a household may possess. Sampling procedure: In this study, a two-stage sampling design has been used, where the Primary Sampling Units (PSUs) are designated as villages, while the Secondary Sampling Units (SSUs) comprise households within the village. In the first stage, 20 PSUs (villages) have been selected from the list of all villages in Bangladesh following the standard systematic probability proportional to size (PPS) sampling method. A village wise household list was used from the Population and Housing Census 2011 (BBS, 2012) to apply the PPS method. While conducting PPS, all villages of the whole of Bangladesh were arranged geographically from the starting of the north to the end of the south, irrespective of the division, district, or other administrative units or any other criteria. It is noted that this study intends to provide the results, in general, in the overall areas of rural Bangladesh. The study was conducted for three months, started on 1st September 2013 and ending on 30th November 2013. The data were collected using face-to-face interviews. To collect primary data, a well-designed questionnaire was prepared for this study. Taking in mind the difficulties in collecting monetary indicators, data on household income and expenditure were collected technically and carefully. Data on household income were collected in the following ways: * Income comes out from different occupations of main earning members of the households; Income from different occupations of secondary earning members in the households if any; * Income from many other sources excluding the above two ones. It may include agriculture (crops/vegetables/trees/fruits), poultry, livestock, fishing, fishery, irregular business or handicraft, rent of house/shop, labor, scholarship, donation, allowance, private tuition, Vulnerable Group Development (VGD)/Vulnerable Group Feeding (VGF) card, etc. Analytical techniques: For the analytical purpose of the wealth indices as proxy measures, different analytical techniques are cited in the literature, such as Pearson's product-moment correlation, scatter plots, Spearman's rank correlation, use of Kappa statistic, classification in different quantiles (e. g. tercile, quintile), graphical representation, etc. For example, Ucar (2015) and Prakongsai (2006) conducted Pearson's correlations between the asset index and household income and household expenditure, and Howe et al. (2008) used classification in quintiles and a graphical representation, and Kappa statistic, while Filmer and Pritchett (2001) performed rank correlation and classification in terciles. Zerin (2009) and Bhuiya et al. (2007) used the scatter plot along with other methods.