Blackwell Reference Online

Extract

Time‐series data are defined as information gathered for specific variables which are observed over a period of time. For example, a firm might have quarterly pricing and sales data available for 10 years which could be used to estimate the demand for its product. But a manager could also estimate demand if the firm had pricing and sales data available from 40 branches located across the country. The data in this second example are cross‐section data and are obtained by gathering information from different locations at the same point in time. It is also possible to combine cross‐section and time‐series data (a pooled sample) to examine how demand varied across locations and over time. Each data source raises different estimation problems for analysts and requires different econometric methods. Since economic theory specifies relationships between variables holding all else constant (the ceteris paribus assumption), researchers must be careful to ensure that the linear regression is properly specified. For example, when using time‐series data, changes in inflation are likely, so variables must be adjusted to reflect real values. This problem is not likely to be significant in cross‐section analysis . See also consumer price index ; real income ; real prices ( 2000 ). Modern Industrial Organization , 3rd edn . Reading, MA : Addison‐Wesley . ( 2002 ). Econometric ... log in or subscribe to read full text