This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.View Full Document
Unformatted text preview: Chapter 1 Simple regression analysis Overview This chapter introduces the least squares criterion of goodness of fit and demonstrates, first through examples and then in the general case, how it may be used to develop expressions for the coefficients that quantify the relationship when a dependent variable is assumed to be determined by one explanatory variable. The chapter continues by showing how the coefficients should be interpreted when the variables are measured in natural units, and it concludes by introducing R 2 , a second criterion of goodness of fit, and showing how it is related to the least squares criterion and the correlation between the fitted and actual values of the dependent variable. Further material Definition of cross-sectional, time series, and panel data Data sets may be of three types: cross-sectional data, time series data, and panel data. Cross-sectional data sets are generated at one moment in time and the observations generally relate to households, individuals, enterprises, or geographical areas. It is usually very desirable that the sample should be drawn from a well-defined population using a statistically respectable sampling scheme, so that one may generalize from the results. For example, the Consumer Expenditure Survey data set, whose observations relate to households, employs a sampling scheme that makes it nationally representative of the United States. Time series data sets consist of repeated observations on a set of variables over an interval of time. Generally the interval between the observations is fixed, often being a year, a quarter, or a month, but in some cases, such as analysis using stock market prices, the frequency may be much greater. An example is the Demand Functions data set, used for examples and exercises in Chapters 11–13 of the text. It consists of 45 annual observations over the period 1959–2003 on aggregate expenditure on different categories of consumer expenditure, the prices of the categories, and aggregate disposable income. Panel data sets have both cross-sectional and time series dimensions, being repeated observations over an interval of time on the same cross-section sample. The National Longitudinal Survey of Youth is an example. The 6,000 males and females in the core sample were first interviewed in 1979. From then until 1994 they were reinterviewed every year, and since 1994 every two years. The EAEF data sets derived from it and used as examples and exercises in the first part of the text are effectively cross-sectional data sets because the earnings data relate to only one year, 2002, and the other characteristics, such as educational attainment, are mostly fixed. Learning outcomes After working through the corresponding chapter in the text, studying the corresponding slideshows, and doing the starred exercises in the text and the additional exercises in this guide, you should be able to explain what is meant by: the least squares criterion of goodness of fit • • •...
View Full Document
- Spring '10