Simple_Regression教程.pptVIP

ANOVA Table Source? df SS MS F Regression k SSR = b1(SSXY) Residual (Error) n-k-1 SSE = SSYY -SSR Total n-1 SST = SSYY ? ? Problem 4.31 The Conference Board produces a Consumer Confidence Index (CCI) that reflects people’s feelings about general business conditions, employment opportunities, and their own income prospects. Some researchers may feel CCI is a function of median household income. Shown here are the CCIs for 9 years and the median household income for the same 9 years. Determine the equation of the regression line. Compute the standard error of the estimate. Compute the value of r2. Does median household income appear to be a good predictor of CCI? The Data Output Analyzing the model:Standard Error of the Estimate This is the variation we expect to see around any predicted value from our model. n is the sample size k is the number of independent variables. (1 for linear regression). Analyzing the model:Standard deviation of the regression slope This is the variation in the regression slope based upon the data we have taken. If we took a different set of data, we would expect a different slope. A large SSxx represents a wide range of x-values. Is the linear relation meaningful? We can use hypothesis testing to evaluate how meaningful the relation is: H0: No relation exists, b1=0. Ha: A relation exists, b1≠0. The appropriate statistic is the t-statistic. The appropriate degrees of freedom is n-2. Try it out Let’s look at whether 4.5 describes a meaningful relation. Let’s look at whether 4.31 describes a meaningful relation. Let’s look at whether the housing cost data describes a meaningful relation. Confidence in Our Prediction We can also define the standard deviation for any prediction. We can use this standard deviation to create a confidence interval about our prediction. Please Ignore Everything I Am Telling You John Maynard Keynes A “sound banker” is one who, “when he is ruined, is ruined in a conventional and orthodox way.” Not all future eve