3.11 Comparison of a variance component estimate with its standard error

As before (Section 3.4), each variance component in the output of the REML statement is accompanied by an SE: for example, the variation among breeding lines is estimated to be c03-math-0134 ± 6242. This SE provides an indication of the precision with which the variance component is estimated. It may also provide a tentative indication of whether any variation is really accounted for by the term under consideration: that is, whether the null hypothesis

can be rejected. If H0 is true, the ratio

3.29 equation

is expected to have a value of about ±1, on average. In the present case,

equation

This is considerably larger than the expected value, suggesting that H0 is false.

If c03-math-0138 were an observation of a normally distributed variable, this ratio could be made the basis for a formal significance test. z would then be approximately an observation of a standard normal variable: that is, of a variable Z such that

3.30

(For an account of the normal ...

Get Introduction to Mixed Modelling: Beyond Regression and Analysis of Variance, 2nd Edition now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.