Assume that you want to analyze the effect of crime on neigh

Assume that you want to analyze the effect of crime on neighborhood housing prices. So you collect the following data in 100 different neighborhoods: median housing price (P), number of crimes per 1,000 residents (C), quality of the school in the neighborhood (S), median square-footage of houses in the neighborhood (SF).

Then, you run the regression: P = B0 + B1 C + B2 S + B3 SF + E.

Discuss why the following regression may not provide a good cause-effect relationship.

Solution

P = B0 + B1 C + B2 S + B3 SF + E.

is arrived from taking data from 100 households in the neighbourhood

The regression may not provide a good cause-effect relationship due to one or more of the following reasons:

1) The housing price depends on the transportation facilities to other places in the city also which was not considered in this

2) Sample size 100 is very small compared to the area size. Hence may not represent the true population

3) Nearby supermarkets, police stations and hospitals are immediate requirements for household persons. Hence they give preference to these points while selecting the houses and hence have effect on price. This was not at all considered in the regression equation

4) Though quality of school in the neighbourhood is the main criterion, colleges also play a vital role in this

5) The development of the area in the near future for example, laying down roads or highways nearby or construction of industries increasing employment also is important to be considered

6) Hygiene and drainage facilities, water availability, pollution free are also other important factors to be considered.

Due to reasons stated as above the regression equation is not sufficient to estimate the price.