<p>pool 针对“窄而长”的截面少、时期较长的data，侧重时间序列分析；panel则是相反，重视截面分析。</p><p>另外，在模型设定过程中，变截距模型有固定效应和随机效应两种，固定效应适合于数据是一个整体的全部状况，比如全国各省经济分析过程；而随机效应适合数据来源于随机抽样得到的数据。</p><p>高铁梅，计量经济分析方法与建模，EVIEWS  中针对POOL进行详细讲解。p330-367</p><p>Generally speaking, we distinguish between the two by noting that pooled time-series,<br/>cross-section data refer to data with relatively few cross-sections, where variables are held<br/>in cross-section specific individual series, while panel data correspond to data with large<br/>numbers of cross-sections, with variables held in single series in stacked form.</p>

[此贴子已经被作者于2008-4-16 23:49:31编辑过]