I have field measurements of spectral reflectance obtained for a research to map different vegatation assemblages. I have been able to use the median test and, Kruskal-wallis H test with Tukey post-hoc in SPSS to determine the vegetation assemblages with significant differences between them. However, I need to know the number of vegetetion pairs with significant differences between them per wavelength of the captured spectral reflectances. The ASD spectral radiometer I used recorded spectral measurements in 1 nm intervals over 350 to 2500 (i.e., 2151). I require the number of of statistical significant differences for each one of these.
I have spent over 3 weeks reviewing papers and books that describe the non-parameteric techniques and this specific approach, but none of them have really provided the details on how they did it or how it should be done.
It could be that I am not just interpering the results properly. I am not from a statistical background but need to apply statistics in this particular research.
NB. After reading a few of the posts here while in search for the answer I observed that including a sample of the data I have may make the problem much more understandable. Below therefore is a sample of the data for 3 vegetation assemblages. There are 21 of these vegetations assemblages and 2151 wavelength bands. With Kruskal-Wallis F test, 21 vegetations results in 210 pair combinations. My problem as explained above is how many of these are significantly different at each of the wavelength bands; 350, 351, 352, ..... Please note that the number of wavelengths per vegetation may be different following removal of noise in the measured spectra. (in the data below wv = wavelength band; Veg(i)_Refl = reflectance for vegetaion assemblage number (i) )
wv Veg1_Refl. wv Veg2_Refl. wv Veg3_Refl.
350 0.008174487 350 0.013188644 351 0.028514129
351 0.007655592 351 0.013734324 352 0.049531318
352 0.009765395 352 0.012374247 353 0.018516597
353 0.010755514 353 0.011760141 356 0.014577343
354 0.011269412 354 0.012087280 357 0.050192549
355 0.013393519 355 0.012631961 358 0.054180159
356 0.014837899 356 0.014939956 359 0.038724873
357 0.013845443 357 0.015511826 360 0.046552907
358 0.011693407 358 0.014272921 361 0.036904282
359 0.011682005 359 0.013951152 362 0.006326519
360 0.011172930 360 0.013224544 363 0.011451388
361 0.010111401 361 0.012961764 364 0.007719737
362 0.009413968 362 0.013361601 365 0.006463090
363 0.010810984 363 0.013334547 366 0.013768701
364 0.012677799 364 0.012788702 368 0.004883761
365 0.012832758 365 0.012118639 369 0.039288373
366 0.010936107 366 0.011785503 370 0.027853873
367 0.012028101 367 0.012143695 371 0.024068664
368 0.012188279 368 0.011838059 372 0.026245151
