How to interpret this result from lbqtest?

6 次查看（过去 30 天）
silicon
silicon 2013-10-7
编辑： Shashank Prasanna 2013-10-7
The time series 'residual' has 61 data points. Can I claim that no auto-correlations are detected?
[h,pValue] = lbqtest(residual,'Lags',[5,10,15,20])
h =
0 0 0 0
pValue =
0.6959 0.8321 0.2296 0.1376

请先登录，再回答此问题。

回答（1 个）

Shashank Prasanna
Shashank Prasanna 2013-10-7
编辑：Shashank Prasanna 2013-10-7
Those p-values are quite large compared to default significance of 0.05 If the p-value was less than 0.05, h will be 1 and it is 'statistically significant' that there is some autocorrelation in the residuals. For p-value > 0.05 and h = 0, you cannot be that confident.
h = 0 means that you cannot confidently reject 'no autocorrelation' in the residuals.
h = 1 means you can confidently reject 'no autocorrelation'
You can change set your own significance using alpha
http://www.mathworks.com/help/econ/lbqtest.html

类别

在 Help Center 和 File Exchange 中查找有关 Descriptive Statistics 的更多信息

标签

产品

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by