In Exercise 2.205, we used a random sample of 10 countries to use the size of a country to predict the percent of the population living in rural areas. We now see how results change if we add the United States (Land Area: 9147.4, Rural: 18.3%) to the sample.

(a) Use technology to find the new regression line using the 11 data points.

(b) The slope of the regression line using the original 10 points in Exercise 2.205 is about 0.08. Compare the slope with US added to the slope without US. Does adding US have a strong effect on the slope? Why or why not?

(c) Predict the percent rural for US with the new regression line. Is this prediction better than the prediction given in Example 2.205 (which was 752%)? 

Exercise 2.205

Two variables in the dataset AllCountries are the size of the country (in 1000 sq km) and the percent of the population living in rural areas. We are interested in using the size of the country (LandArea) to predict the percent rural (Rural). The values of these variables for a random sample of ten countries is shown, with the 3-letter country codes, in Table 2.37, and is also available in TenCountries. Figure 2.74 shows a scatterplot of the data.

Table 2.37

Figure 2.74

Transcribed Image Text:

Country Land Area Rural ROU 229.9 45.8 LBN 10.2 13 SRB 88.3 BHS SVN 20.1 UZB 425.4 63.2 TUN ARM 28.5 MKD PRK 155.4 33.5 120.4 10.0 25.2 16.3 36.1 48.0 51.4 33.1 37.3 60 40 20 100 200 300 400 Land Area Rural