[R-sig-ME] P-values for balanced mixed model with nesting

[Prew, Paul]

My apologies for sending this message twice, my first attempt was interpreted as an HTML document and "scrubbed", which I guess means removed.  MS Outlook says the current format is plain text.  If it also causes problems, I would appreciate advice on how to post to this list.

Hello, I am working with data from a balanced designed experiment and I'm unsure of how to model it to get the answers I'm seeking.  Two Detergents (current vs. experimental) are being compared against 6 different types of fabric Soils (motor oil, lipstick, etc).  Thus Soil and Detergent have fixed effects.  Each Detergent was tested on 3 different Loads of laundry. A single laundry Load consists of 3 "Backers"  where a Backer is a fabric sample that's been partitioned into a grid of 6 areas.  Each area contains one of the 6 soils.  All experiments were conducted in a lab using the same washing machine.  Soil Removal is the response measurement.
Whole plots:  Loads    --- 3 nested within each Detergent
Whole plot factor:  Detergent   --- 2 levels
Split Plot:  Backer   --- 3 nested within each Load
Split Plot factor:  Soil   --- 6 levels
Here's my question:  I have somehow gotten the impression that for balanced data, the p-values from a mixed effects analysis can be trusted.  However, the lmer function doesn't output p-values regardless of the case.  I'd like to do multiple comparisons to find the Soil:Detergent combinations that stand out as statistically significant.  Does anyone have advice for accomplishing this? 
Ultimately, my purpose is to provide a method to this chemist who is studying detergents, because he wants do the analysis himself.  Admirable, wouldn't you agree?    I want to avoid a discussion of 'p-values aren't that important, just make a qualitative comparison using the t-values'.  That would be a mixed message considering our statistics group's attempts to get the scientists to do less eyeballing and make decisions more objectively, i.e. consider statistical significance.
I asked a similar question a few months ago, and got a reasonable answer that the blocking factors such as Load and Backer could be modeled as fixed effects.  However, if I do that for this nested case, the Soil effects are conditional on the Load and the Backer.  I apologize if I'm not providing enough information.  Please let me know if I could add anything.  Thank you for taking the time to consider my request.

Regards, Paul

====================================================
> str(SoilOut)
'data.frame': 108 obs. of 6 variables:
$ Soil : Factor w/ 6 levels "DSB.P/C","EMPA.101",..: 1 1 1 2 2 2 3 3 3 4 ...
$ Detergent : Factor w/ 2 levels "EXP2","Xtra": 2 2 2 2 2 2 2 2 2 2 ...
$ X.SoilRemoval: num 76 76 76.5 41.2 34.8 ...
$ Load : Factor w/ 6 levels "1","2","3","4",..: 1 1 1 1 1 1 1 1 1 1 ...
$ Backer : Factor w/ 18 levels "1","2","3","4",..: 1 1 1 1 1 1 2 2 2 2 ... 
==============================================
m2<-lmer(X.SoilRemoval~Detergent*Soil + (1 | Load/Backer), SoilOut) 
> summary (m2)
Fixed effects:
                                   Estimate Std. Error t value
(Intercept)                         74.1804     1.7712   41.88
Detergent[T.Xtra]                   -3.2044     2.5049   -1.28
Soil[T.EMPA.101]                   -35.1601     2.0543  -17.12
Soil[T.EMPA.104]                   -26.2447     2.5049  -10.48
Soil[T.EMPA.106]                   -35.4014     2.5049  -14.13
Soil[T.LIPSTICK]                   -12.1608     2.5049   -4.85
Soil[T.MAKE-UP]                    -38.1436     2.5049  -15.23
Detergent[T.Xtra]:Soil[T.EMPA.101]   0.8651     2.9053    0.30
Detergent[T.Xtra]:Soil[T.EMPA.104]  -4.0126     3.5425   -1.13
Detergent[T.Xtra]:Soil[T.EMPA.106]   0.6405     3.5425    0.18
Detergent[T.Xtra]:Soil[T.LIPSTICK]  -0.9912     3.5425   -0.28
Detergent[T.Xtra]:Soil[T.MAKE-UP]    2.9547     3.5425    0.83

Paul Prew   ▪  Statistician
651-795-5942   ▪   fax 651-204-7504 
Ecolab Research Center   ▪  Mail Stop ESC-F4412-A 
655 Lone Oak Drive   ▪   Eagan, MN 55121-1560 


CONFIDENTIALITY NOTICE: 
This e-mail communication and any attachments may contain proprietary and privileged information for the use of the designated recipients named above. 
Any unauthorized review, use, disclosure or distribution is prohibited. 
If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message.