Accounting for variation in designing greenhouse experiments with special reference to greenhouses containing plants on conveyor systems

Table 1 Results of hypothesis tests from the mixed model analyses for all response variables

	Day 21		Day 51
	Total area		Total area		Fresh weight		Height		Density index
	P ^b	Action^c	P ^b	Action^c	P ^b	Action^c	P ^b	Action^c	P ^b	Action^c
Random model change ^a
Add heterogeneous Locations / Tactics variances^d	0.023	Retain	0.021	Retain	0.061	Retain	0.017	Retain	0.161	Retain
Add columns / Positions ar1, differing for Locations / Tactics	0.416	Omit	0.810	Omit	0.771	Omit	0.015	Omit	0.589	Omit
Drop Columns∧Locations / Positions∧Tactics deviations	0.347	Omit	0.131	Omit	0.468	Omit	0.245	Omit	0.469	Omit
Drop spl(Columns)∧Locations/ spl(Positions)∧Tactics	0.005	Retain	0.262	Retain	0.045	Retain	0.073	Retain	0.464	Retain
Drop position deviations			DNF	Omit	DNF	Omit	0.120	Omit	DNF	Omit
Drop spl(Position)			DNF	Omit	DNF	Omit	DNF	Omit	DNF	Omit
Check heterogeneous Locations / Tactics variances^d	0.038	Retain	0.015	Retain	0.025	Retain	0.007	Retain	0.392	Omit
Check ar1 on Columns / Positions, differing for Locations / Tactics	0.395	Omit	0.839	Omit	0.622	Omit	0.224	Omit	0.467	Omit
Fixed model testing ^a
lin(Columns)∧Locations / lin(Positions)∧Tactics	<0.001	NA	0.026	Retain	0.036	NA	0.005	NA	0.012	Retain
lin(Positions)			<0.001	NA	<0.001	NA	0.710	NA	<0.001	NA
Locations∧Rows / Tactics∧Lanes	0.210	ns	0.405	ns	0.311	ns	0.311	ns	0.772	ns
Locations / Tactics	0.099	NA	<0.001	NA	<0.001	NA	0.007	NA	<0.001	NA

^aThe italicized term before the slash is for Day 21 and the one after the slash is for Day 51.
^bDNF indicates that the parameters in the changed model “Did Not Fit”, either because of a lack of convergence or because the estimates for one or more parameters was zero; on occasion terms with a P-value between 0.05 and 0.10 are retained on the grounds that there is some indication that the terms are required.
^cNA indicates that testing was not applicable because, for this term, (i) its factors are a subset of those for a significant fixed term or (ii) it has the same factors as a higher-order term that is significant; ns indicates not significant.
^dA significance level of 0.25 is used for this hypothesis test in order to decrease the chance of a Type I error and to err on the side of allowing for heterogeneous variances if there is some evidence for them.

ISSN: 1746-4811