The work done by Mr. Anik is indeed a praiseworthy work. But I have some questions regarding the modeling and the practice done in the paper.

1. In the equation, HHS Adult equivalent household size calculated as HHS 0 (Mage<2yrs) + 0.5(M2≤age<10) + 1 (M10>age) where M is the number of household members in the given age range. My question is what is the justification for this equation? Why a person from age 2 to 10 will be given the weight of .5 in the equation? Is it based on Consumption? If yes then why we will treat the consumption of a 25 year\'s adult equal with a 65 years old adult?

2. There is no control for Gender in the model

3. How we have calculated the income quartile? What are the cutoff points?

4. What are the post estimations/diagnostics used to test that the regression is correct? I seriously think that the regression is infected with omitted variable which will create a biased and inconsistent estimates of the coefficients. Also there is multi-collinearity problem. One way to deal with this problem could be the Instrument variable approach.

5. Why the income used as a non-logarithmic form in the regression?

6. There is no control of seasonality, Market condition and inflation.

Hopefully by putting more effort we can answer these questions and can proceed forward with this very interesting research.