1
$\begingroup$

I am currently bothered on how to statistically analyze my simple data as shown below.

Date             Spatial Features
          1      2      3      4      5      6 
----------------------------------------------
1988 0.7784 0.8310 0.8726 0.9307 0.9189 0.8814
1998 0.8956 0.9036 0.9368 0.9575 0.9283 0.9045
2009 0.9255 0.9479 0.9703 0.9797 0.9434 0.9555

Note: These values are Relative Entropy values. For my purpose, Im using Relative Entropy to measure compactness or dispersion. Entropy value ranges from 0 to 1, where 0 means compact, while 1 means dispersed. The entropy values were calculated for every spatial feature in 3-time period (1988, 1998, & 2009). As shown in the table, there are six spatial features being considered numbered 1, 2, 3, 4, 5, & 6.

My goal is to prove whether there is a significant difference among these values of compactness/dispersion based on spatial features. Or whether compactness/dispersion differs significantly with spatial features.

Does the time (year) also need to be taken into account in the analysis?

Thank you very much and I would really appreciate any help or advice you could offer me on how to attain my goal, particularly on the appropriate stat test or method to be used.

$\endgroup$

closed as unclear what you're asking by kjetil b halvorsen, mdewey, Jan Kukacka, gung Aug 13 '18 at 17:04

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 1
    $\begingroup$ In this regard, what would 'statistically significant' mean? Without any knowledge on the distribution (i.e. the variance on the measurements), there is no clear answer: it is both alright to say the values 1,2,3,4,5,6 are close to each other as to say that they are different. At best you can use some arbitrary criterion like outliership as it is used in boxplots, but you may need to provide more info to get a proper answer. $\endgroup$ – Nick Sabbe Jun 21 '11 at 13:31
  • $\begingroup$ Hi Nick,I have updated the data. I realized that only one observation per factor does not allow further test analysis. Thanks and I would really appreciate any help. $\endgroup$ – Ronald Jun 21 '11 at 15:21
  • $\begingroup$ How much data do you have? You should use a Friedman test if this is all the data. $\endgroup$ – fgregg Jun 21 '11 at 18:50
  • $\begingroup$ Hi fgregg, yeah those are only my data. Thanks.It seems to me that Friedman test only determines if there is a significant difference of the mean entropy values for the 3 measurement periods. It does not take into account the differences among the spatial variables. For example, is there a significant difference between Spatial features 1 & 2; 1 & 3; 2 & 3; and so on. The test I am looking for is the one that will take both time and spatial features into account. I might have misunderstood the concept so please correct me if I am wrong. Thank you very much. $\endgroup$ – Ronald Jun 21 '11 at 19:23
  • 1
    $\begingroup$ I'm voting to close this question as off-topic because it is unclear, and abandoned by OP (not seen since 2011) $\endgroup$ – kjetil b halvorsen Aug 9 '18 at 12:24