• http://mpra.ub.uni-muenchen.de/4823/1/MPRA_paper_4823.pdf
• http://www.stata.com/statalist/archive/2013-09/msg00917.html
• http://www.bauer.uh.edu/povel/documents/cleary_povel_raith_jfqa_forthc.pdf
• http://ftp.iza.org/dp3075.pdf
• http://www.dartmouth.edu/~blnchflr/papers/scnudgbajo.pdf
  

If "sets of data" denotes "variables", then you want some sort of test of interaction.  What is appropriate depends on the sort of scaling and on what is conventional in your area.

If "sets of data" denotes separate data collection samples, then whatever you mean by "range" is more of a question, and your question is not well-defined.

建议先做一下散点图，看看图示情况。
根据图示找拐点，选择区间，再分区间进行检验。