Here we have modeled the data using a curve instead of a line. Scatterplot with linear and curvilinear regressions compare the lung cancer rate predicted by the two models when the cigarette consumption is 3200 cigarettes a year per person. The linear model predicts [ select ] deaths per 100,000 due to lung cancer 30 years later. The curved model predicts [ select ] deaths per 100,000 due to lung cancer 30 years later. Which model gives the more reliable prediction? [ select ]