SPSS 和普通最小二乘法
我正在做回归,并且使用SPSS/PASW。但它似乎不支持普通最小二乘法,它只有偏最小二乘法和两阶段最小二乘法。关于做什么有什么建议吗?
I am doing regression and I am using SPSS/PASW. But it doesn't seem to support Ordinary Least Squares, it only has Partial least Squares and 2-stages Least Squares. Any suggestions about what to do?
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(3)
此链接提到了 SPSS 加权最小二乘法。我认为如果你让所有权重等于 1.0,你就得到了所谓的“普通”最小二乘法。
This link mentions SPSS weighted least squares. I think if you make all the weights equal to 1.0 you've got what you're calling "ordinary" least squares.
我同意巴里的观点 - OLS 在 SPSS/PASW 中是“标准” - 最小二乘法用于标准线性回归,而在 PASW 中,如果您选择“分析>回归>线性”,这将为您提供所谓的 OLS。
这是取自SPSS/PASW的帮助文档——它并没有直接说标准线性回归下的OLS,而是通过这个文档推断OLS...
I agree with Barry - OLS is 'standard' in SPSS/PASW - the least squares method is used in standard linear regressions and in PASW if you select "Analyze>Regression>Linear" that will give you what you are calling OLS.
This is taken from SPSS/PASW's help documents - it does not directly say OLS under standard linear regression, but infers OLS via this document...
SPSS 应默认为 OLS,除非您正在做一些事情来使其切换;我认为问题在于默认值是假设的,并且没有明确提及。
SPSS should default to OLS unless you are doing something to make it switch; I think that the problem is that the default is assumed, and not explicitly mentioned.