Therefore, the model will attribute the effect of the missing variables to the estimates of the included variables, which compromises precision of the latter estimates. [...] The second objective is realized by comparing estimates of the ITPTE from the ‘traditional’ and the ‘spatial’ regression model that accounts for the distribution of domestic goods prices. [...] The aspect of the spatial weights matrix will be introduced in the next section to capture the closeness of the districts to each other, i.e. [...] In Figure 1-3, the darker the colour the higher the average goods price, the map’s key shows the district average prices in US$. [...] AIC, BIC and LM criteria are used for the purpose; the lower the AIC and BIC values and the greater the LM value, the better is the model.
Authors
- Pages
- 59
- Published in
- South Africa