Skip to content

why does my inter-annual SD = 0?

5 messages · Scott Bennett, Douglas Bates, Ben Bolker +1 more

#
Can you tell us how many levels there are of region, site, zone and
year?  If the data are collected in only a few years it is difficult
to estimate a variance component and the ML and REML  criteria drive
the estimates to zero.  You may be overmodeling the data but it is
difficult to say without access to the data or at least a feeling for
how many observations and in what configuration they are.
On Mon, Oct 11, 2010 at 10:32 AM, Scott Bennett <sbennett at ceab.csic.es> wrote:
#
Hi all, thanks for the quick replies! I have attached the data, if it helps for you to look at it directly. The end product that I would like to achieve is an estimate of the variance associated with each factor (depth, surveyor, year, water body, site and zone), to then model the probability of misclassifying the health status of a water body, based on the variability associated with each of the respective factors (An uncertainty analysis of our seagrass index). The index we are using POMI_14 (Posidonia oceanica multivariate index, Romero et al. 2008) is comprised of 14 metrics relates to the health and status of /P. oceanica./ There are only 4 years of data. The 2005, 2006 and 2008 data is from annual sampling of  30 seagrass meadows (sites) sampled at a single depth along the Catalan coast, Spain. The 2002 data is of a subset of those sites, but includes replication between 2 discrete depths (5m and 15m) and among three discrete zones nested within each site. The sites are nested within 'water-bodies'. A water body represents an area of coastal water (15 - 50 km in length) which has been classified based on its exposure to water quality pressures. The surveyor factor, is only from the 2008 series, where we calculated POMI based on two separate surveyors. Needless to say the design is unbalanced.  

In short the data looks like this:  

'data.frame':   231 obs. of  8 variables: 
 $ year    : Factor w/ 4 levels "2002","2005",..: 4 4 4 4 4 4 4 4 4 4 ...
 $ WB      : Factor w/ 17 levels "1","2","3","4",..: 1 2 3 3 3 4 5 6 7 8 ...
 $ Site    : Factor w/ 30 levels "Balis    ","Cadaques ",..: 22 19 27 7 9 2 16 21 20 15 ...
 $ Zone    : Factor w/ 3 levels "a","b","c":..: 1 1 1 1 1 1 1 1 1 1 ...
 $ Depth   : Factor w/ 2 levels "p","s": 1 1 1 1 1 1 1 1 1 1 ...
 $ surveyor: Factor w/ 2 levels "1","2": 1 1 1 1 1 1 1 1 1 1 ...
 $ POMI_14 : num  0.781 0.633 0.717 0.936 0.86 ...
 $ POMI_9  : num  0.803 0.67 0.745 0.942 0.873 ... 

I hope this makes things clearer. Any help will be greatly appreciated.  

Kind regards  

Scott Bennett
-------------- next part --------------
year,WB,Site,Zone,Depth,surveyor,POMI_14,POMI_9
2008,1,Rovellad ,a,p,1,0.781107622,0.80279966
2008,2,Port_Selva ,a,p,1,0.633441988,0.669767557
2008,3,Tamariua ,a,p,1,0.716925282,0.744977731
2008,3,Culip    ,a,p,1,0.935657559,0.942033837
2008,3,Jugadora ,a,p,1,0.859538869,0.873458441
2008,4,Cadaques ,a,p,1,0.684976423,0.716194975
2008,5,Montjoi  ,a,p,1,0.749745797,0.774545763
2008,6,Roses_2 ,a,p,1,0.705540388,0.73472107
2008,7,Roses_1  ,a,p,1,0.406455662,0.465275371
2008,10,Montgo ,a,p,1,0.468556701,0.521222254
2008,10,Medas    ,a,p,1,0.675423785,0.707588995
2008,14,Sa_Tuna  ,a,p,1,0.560703188,0.604237106
2008,14,Llafranc ,a,p,1,0.575450184,0.617522688
2008,14,Palamos  ,a,p,1,0.567427391,0.610294947
2008,14,St_Feliu ,a,p,1,0.630364509,0.666995053
2008,14,Tossa    ,a,p,1,0.696983928,0.727012548
2008,14,Canyelle ,a,p,1,0.666440392,0.699495849
2008,14,Fenals   ,a,p,1,0.884870258,0.896279512
2008,16,Balis    ,a,p,1,0.56050751,0.60406082
2008,16,Mataro   ,a,p,1,0.566466791,0.609429541
2008,23,Sitges   ,a,p,1,0.414958755,0.472935815
2008,24,Vilanova ,a,p,1,0.346702757,0.411443925
2008,25,Comarug ,a,p,1,0.303724782,0.372725028
2008,25,Torredem ,a,p,1,0.290578787,0.36088179
2008,29,Salou    ,a,p,1,0.268360454,0.340865274
2008,30,Montroig ,a,p,1,0.399909348,0.459377791
2008,31,L_Hospit ,a,p,1,0.517707424,0.565502184
2008,31,Calafat  ,a,p,1,0.620118104,0.657764058
2008,32,L_Ametll ,a,p,1,0.519628625,0.567232995
2008,32,Cap_Roig ,a,p,1,0.301643481,0.370849983
2006,1,Rovellad ,a,p,1,0.830061339,0.631092868
2006,3,Tamariua ,a,p,1,0.644537158,0.59184892
2006,3,Culip    ,a,p,1,0.696764496,0.647210919
2006,3,Jugadora ,a,p,1,0.699754821,0.69911104
2006,5,Montjoi  ,a,p,1,0.531083947,0.565331223
2006,7,Roses_1  ,a,p,1,0.338586935,0.416484533
2006,10,Montgo ,a,p,1,0.618630194,0.518771081
2006,10,Medas    ,a,p,1,0.599916547,0.744704099
2006,14,Sa_Tuna  ,a,p,1,0.438411429,0.579487076
2006,14,Llafranc ,a,p,1,0.297618103,0.422441203
2006,14,Palamos  ,a,p,1,0.513128216,0.494888336
2006,14,St_Feliu ,a,p,1,0.578708665,0.717093178
2006,14,Tossa    ,a,p,1,0.556797987,0.661815274
2006,14,Canyelle ,a,p,1,0.504377726,0.600790934
2006,14,Fenals   ,a,p,1,0.65746308,0.797038708
2006,16,Balis    ,a,p,1,0.398365877,0.540523441
2006,16,Mataro   ,a,p,1,0.293938763,0.546353971
2006,23,Sitges   ,a,p,1,0.419821804,0.295669234
2006,24,Vilanova ,a,p,1,0.473647656,0.447921739
2006,25,Comarug ,a,p,1,0.392950219,0.288899653
2006,25,Torredem ,a,p,1,0.488461479,0.284022191
2006,29,Salou    ,a,p,1,0.47425399,0.284442662
2006,30,Montroig ,a,p,1,0.448925798,0.459134295
2006,31,L_Hospit ,a,p,1,0.672318243,0.579893531
2006,31,Calafat  ,a,p,1,0.613972453,0.549185141
2006,32,L_Ametll ,a,p,1,0.740379146,0.63193381
2006,32,Cap_Roig ,a,p,1,0.377695427,0.425496625
2005,1,Rovellad ,a,p,1,0.613089417,0.6010795
2005,3,Tamariua ,a,p,1,0.627720981,0.664700276
2005,3,Culip    ,a,p,1,0.654396909,0.707133319
2005,3,Jugadora ,a,p,1,0.757197896,0.812208357
2005,5,Montjoi  ,a,p,1,0.675401462,0.724559462
2005,7,Roses_1  ,a,p,1,0.525139248,0.542237478
2005,10,Montgo ,a,p,1,0.5626171,0.557706059
2005,10,Medas    ,a,p,1,0.716986675,0.775244974
2005,14,Sa_Tuna  ,a,p,1,0.662041572,0.690167778
2005,14,Llafranc ,a,p,1,0.577818725,0.66366392
2005,14,Palamos  ,a,p,1,0.506707278,0.570545365
2005,14,St_Feliu ,a,p,1,0.683543101,0.779121715
2005,14,Tossa    ,a,p,1,0.651795093,0.688517285
2005,14,Canyelle ,a,p,1,0.525928563,0.574115038
2005,14,Fenals   ,a,p,1,0.693760347,0.759027913
2005,16,Balis    ,a,p,1,0.536876657,0.578529149
2005,16,Mataro   ,a,p,1,0.521017269,0.545864726
2005,23,Sitges   ,a,p,1,0.288373944,0.208780166
2005,24,Vilanova ,a,p,1,0.379861414,0.380757776
2005,25,Comarug ,a,p,1,0.254243373,0.163065328
2005,25,Torredem ,a,p,1,0.245195483,0.203003438
2005,29,Salou    ,a,p,1,0.245663225,0.143969499
2005,30,Montroig ,a,p,1,0.414678981,0.400678852
2005,31,L_Hospit ,a,p,1,0.518064646,0.487310582
2005,31,Calafat  ,a,p,1,0.537417484,0.548781878
2005,32,L_Ametll ,a,p,1,0.546363055,0.531374927
2005,32,Cap_Roig ,a,p,1,0.396583201,0.327903593
2008,1,Rovellad ,a,p,2,0.835797869,0.644942527
2008,2,Port_Selva ,a,p,2,0.791972534,0.653946258
2008,3,Tamariua ,a,p,2,0.662135843,0.685428269
2008,3,Culip    ,a,p,2,0.720143502,0.734483079
2008,3,Jugadora ,a,p,2,0.495271707,0.441691062
2008,4,Cadaques ,a,p,2,0.535878572,0.517213736
2008,5,Montjoi  ,a,p,2,0.782317124,0.548866508
2008,6,Roses_2 ,a,p,2,0.567130818,0.54638272
2008,7,Roses_1  ,a,p,2,0.153828129,0.35502239
2008,10,Montgo ,a,p,2,0.633124494,0.436583773
2008,10,Medas    ,a,p,2,0.562227915,0.713541642
2008,14,Sa_Tuna  ,a,p,2,0.657488613,0.546460338
2008,14,Llafranc ,a,p,2,0.520237409,0.547158904
2008,14,Palamos  ,a,p,2,0.525982528,0.44949947
2008,14,St_Feliu ,a,p,2,0.596548236,0.60346327
2008,14,Tossa    ,a,p,2,0.675325554,0.635845654
2008,14,Canyelle ,a,p,2,0.786167256,0.640580374
2008,14,Fenals   ,a,p,2,0.780692849,0.816354936
2008,16,Balis    ,a,p,2,0.393468792,0.767377744
2008,16,Mataro   ,a,p,2,0.458078826,0.685723218
2008,23,Sitges   ,a,p,2,0.583298366,0.354789535
2008,24,Vilanova ,a,p,2,0.425879086,0.458021967
2008,25,Comarug ,a,p,2,0.768751423,0.32299705
2008,25,Torredem ,a,p,2,0.614370138,0.29794184
2008,29,Salou    ,a,p,2,0.509138199,0.209037759
2008,30,Montroig ,a,p,2,0.403996498,0.293455498
2008,31,L_Hospit ,a,p,2,0.803718446,0.620585882
2008,31,Calafat  ,a,p,2,0.680318695,0.527645645
2008,32,L_Ametll ,a,p,2,0.591013671,0.439766127
2008,32,Cap_Roig ,a,p,2,0.490880149,0.328476907
2002,3,Jugadora ,a,s,1,0.852607907,0.787071029
2002,3,Jugadora ,b,s,1,0.798730237,0.736424513
2002,3,Jugadora ,c,s,1,0.757097491,0.675645884
2002,3,Jugadora ,a,p,1,0.645718219,0.612211263
2002,3,Jugadora ,b,p,1,0.659123934,0.626193861
2002,3,Jugadora ,c,p,1,0.593213678,0.562871663
2002,5,Montjoi  ,a,s,1,0.694542203,0.675182139
2002,5,Montjoi  ,b,s,1,0.754818378,0.611438355
2002,5,Montjoi  ,c,s,1,0.736443912,0.70729293
2002,5,Montjoi  ,a,p,1,0.529950591,0.500069421
2002,5,Montjoi  ,b,p,1,0.578123401,0.485370127
2002,5,Montjoi  ,c,p,1,0.555587076,0.492537087
2002,14,Fenals   ,a,s,1,0.691045676,0.63625571
2002,14,Fenals   ,b,s,1,0.751378474,0.615541791
2002,14,Fenals   ,c,s,1,0.777184831,0.645390071
2002,14,Fenals   ,a,p,1,0.554638625,0.54375977
2002,14,Fenals   ,b,p,1,0.468867526,0.438602201
2002,14,Fenals   ,c,p,1,0.533701927,0.552093118
2002,30,Montroig ,a,s,1,0.634379276,0.648467648
2002,30,Montroig ,b,s,1,0.690196317,0.695347997
2002,30,Montroig ,c,s,1,0.744229703,0.736817994
2002,30,Montroig ,a,p,1,0.60022089,0.457278455
2002,30,Montroig ,b,p,1,0.440371532,0.45543753
2002,30,Montroig ,c,p,1,0.510061436,0.381589738
2002,23,Sitges   ,a,p,1,0.394930827,0.344377759
2002,23,Sitges   ,b,p,1,0.436082269,0.332727936
2002,23,Sitges   ,c,p,1,0.269961811,0.282910539
2002,25,Comarug ,a,p,1,0.394520304,0.351614983
2002,25,Comarug ,b,p,1,0.375438039,0.260383801
2002,25,Comarug ,c,p,1,0.303681362,0.240555212
2002,16,Mataro   ,a,p,1,0.530247867,0.429819162
2002,16,Mataro   ,b,p,1,0.495197654,0.475842286
2002,16,Mataro   ,c,p,1,0.496528316,0.467607309
2002,10,Montgo ,a,p,1,0.641315708,0.541384837
2002,10,Montgo ,b,p,1,0.58235604,0.540696246
2002,10,Montgo ,c,p,1,0.578491456,0.477121097
2008,1,Rovellad ,a,p,1,0.624182987,0.671245913
2008,3,Tamariua ,a,p,1,0.704355158,0.72756147
2008,3,Culip    ,a,p,1,0.80911293,0.782600491
2008,3,Jugadora ,a,p,1,0.610176765,0.614042332
2008,5,Montjoi  ,a,p,1,0.506873409,0.518672196
2008,7,Roses_1  ,a,p,1,0.533132584,0.354350654
2008,10,Montgo ,a,p,1,0.401936325,0.459082049
2008,10,Medas    ,a,p,1,0.717146046,0.711780534
2008,14,Sa_Tuna  ,a,p,1,0.555387136,0.586032558
2008,14,Llafranc ,a,p,1,0.606630382,0.562351903
2008,14,Palamos  ,a,p,1,0.452283014,0.429277725
2008,14,St_Feliu ,a,p,1,0.636435941,0.621054026
2008,14,Tossa    ,a,p,1,0.667536527,0.673521476
2008,14,Canyelle ,a,p,1,0.573776528,0.637519481
2008,14,Fenals   ,a,p,1,0.853203639,0.858952101
2008,16,Balis    ,a,p,1,0.913492156,0.832514871
2008,16,Mataro   ,a,p,1,0.826884694,0.747560523
2008,23,Sitges   ,a,p,1,0.245477066,0.320827728
2008,24,Vilanova ,a,p,1,0.45419567,0.463707177
2008,25,Comarug ,a,p,1,0.127709248,0.311373966
2008,25,Torredem ,a,p,1,0.157156183,0.272041879
2008,29,Salou    ,a,p,1,0.128287029,0.146090398
2008,30,Montroig ,a,p,1,0.352565888,0.27387343
2008,31,L_Hospit ,a,p,1,0.533232202,0.463263165
2008,31,Calafat  ,a,p,1,0.516596078,0.534305128
2008,32,L_Ametll ,a,p,1,0.48930088,0.434772377
2008,32,Cap_Roig ,a,p,1,0.370736122,0.344452881
2006,1,Rovellad ,a,p,1,0.830061339,0.631092868
2006,3,Tamariua ,a,p,1,0.644537158,0.59184892
2006,3,Culip    ,a,p,1,0.696764496,0.647210919
2006,3,Jugadora ,a,p,1,0.699754821,0.69911104
2006,5,Montjoi  ,a,p,1,0.531083947,0.565331223
2006,7,Roses_1  ,a,p,1,0.338586935,0.416484533
2006,10,Montgo ,a,p,1,0.618630194,0.518771081
2006,10,Medas    ,a,p,1,0.599916547,0.744704099
2006,14,Sa_Tuna  ,a,p,1,0.438411429,0.579487076
2006,14,Llafranc ,a,p,1,0.297618103,0.422441203
2006,14,Palamos  ,a,p,1,0.513128216,0.494888336
2006,14,St_Feliu ,a,p,1,0.578708665,0.717093178
2006,14,Tossa    ,a,p,1,0.556797987,0.661815274
2006,14,Canyelle ,a,p,1,0.504377726,0.600790934
2006,14,Fenals   ,a,p,1,0.65746308,0.797038708
2006,16,Balis    ,a,p,1,0.398365877,0.540523441
2006,16,Mataro   ,a,p,1,0.293938763,0.546353971
2006,23,Sitges   ,a,p,1,0.419821804,0.295669234
2006,24,Vilanova ,a,p,1,0.473647656,0.447921739
2006,25,Comarug ,a,p,1,0.392950219,0.288899653
2006,25,Torredem ,a,p,1,0.488461479,0.284022191
2006,29,Salou    ,a,p,1,0.47425399,0.284442662
2006,30,Montroig ,a,p,1,0.448925798,0.459134295
2006,31,L_Hospit ,a,p,1,0.672318243,0.579893531
2006,31,Calafat  ,a,p,1,0.613972453,0.549185141
2006,32,L_Ametll ,a,p,1,0.740379146,0.63193381
2006,32,Cap_Roig ,a,p,1,0.377695427,0.425496625
2005,1,Rovellad ,a,p,1,0.613089417,0.6010795
2005,3,Tamariua ,a,p,1,0.627720981,0.664700276
2005,3,Culip    ,a,p,1,0.654396909,0.707133319
2005,3,Jugadora ,a,p,1,0.757197896,0.812208357
2005,5,Montjoi  ,a,p,1,0.675401462,0.724559462
2005,7,Roses_1  ,a,p,1,0.525139248,0.542237478
2005,10,Montgo ,a,p,1,0.5626171,0.557706059
2005,10,Medas    ,a,p,1,0.716986675,0.775244974
2005,14,Sa_Tuna  ,a,p,1,0.662041572,0.690167778
2005,14,Llafranc ,a,p,1,0.577818725,0.66366392
2005,14,Palamos  ,a,p,1,0.506707278,0.570545365
2005,14,St_Feliu ,a,p,1,0.683543101,0.779121715
2005,14,Tossa    ,a,p,1,0.651795093,0.688517285
2005,14,Canyelle ,a,p,1,0.525928563,0.574115038
2005,14,Fenals   ,a,p,1,0.693760347,0.759027913
2005,16,Balis    ,a,p,1,0.536876657,0.578529149
2005,16,Mataro   ,a,p,1,0.521017269,0.545864726
2005,23,Sitges   ,a,p,1,0.288373944,0.208780166
2005,24,Vilanova ,a,p,1,0.379861414,0.380757776
2005,25,Comarug ,a,p,1,0.254243373,0.163065328
2005,25,Torredem ,a,p,1,0.245195483,0.203003438
2005,29,Salou    ,a,p,1,0.245663225,0.143969499
2005,30,Montroig ,a,p,1,0.414678981,0.400678852
2005,31,L_Hospit ,a,p,1,0.518064646,0.487310582
2005,31,Calafat  ,a,p,1,0.537417484,0.548781878
2005,32,L_Ametll ,a,p,1,0.546363055,0.531374927
2005,32,Cap_Roig ,a,p,1,0.396583201,0.327903593
#
Dear Scott,

Thanks for sharing your data. That makes things a lot easier.

IMHO you have several problems with your model. First you have only 4
years. Estimating a variance based on only 4 levels will give you very
unreliable estimates. Therefore it is better to treat them as a fixed
effect.
Another possible problem might be the structure of your nested random
effects. I noticed that many levels of WB have only one level of site,
and many levels of site have only one levels of zone. In such a case
there is competition between the nested effects. I'm not sure how lmer
handles that. If it were just a few cases I guess it will not be a big
issue. But your data has quite a lot of this things. A possible solution
is to use (1|WB:site:zone) as random effect.

HTH,

Thierry

------------------------------------------------------------------------
----
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek
team Biometrie & Kwaliteitszorg
Gaverstraat 4
9500 Geraardsbergen
Belgium

Research Institute for Nature and Forest
team Biometrics & Quality Assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium

tel. + 32 54/436 185
Thierry.Onkelinx at inbo.be
www.inbo.be

To call in the statistician after the experiment is done may be no more
than asking him to perform a post-mortem examination: he may be able to
say what the experiment died of.
~ Sir Ronald Aylmer Fisher

The plural of anecdote is not data.
~ Roger Brinner

The combination of some data and an aching desire for an answer does not
ensure that a reasonable answer can be extracted from a given body of
data.
~ John Tukey