1. “Was Commerce’s decision to modify Marmen’s reported product-specific raw material costs arbitrary and unsupported by substantial evidence, given that Commerce failed to apply its longstanding practice and relied on assumptions and factual findings belied by the record evidence?” 2. “Was Commerce’s decision to reject a minor correction to one line of a cost reconciliation worksheet unsupported by substantial evidence, given that Commerce’s decision was illogical and contrary to the record evidence?” 3. “Was Commerce’s decision to apply the average-to-transaction comparison method based on a finding of ‘significant’ price differences unreasonable and unsupported by substantial evidence, given that Commerce blindly relied on the results of a Cohen’s d test that falsely attributed significance to price differences of less than one percent?”
“We conclude that it was unreasonable to rely on Cohen’s d test to determine whether prices differ significantly when the underlying data is not normally distributed, equally variable, and equally and sufficiently numerous. Because there is no dispute that Marmen’s data does not satisfy these assumptions, Cohen’s d test cannot be used here to determine “a pattern of export prices (or constructed export prices) for comparable merchandise that differ significantly among purchasers, regions, or periods of time.” § 1677f-1(d)(1)(B). We therefore vacate Commerce’s calculated dumping margin based on the unreasonable use of Cohen’s d test to justify the A-to-T methodology. On remand, Commerce may re-perform a differential pricing analysis, and that analysis may not rely on Cohen’s d test for data sets like those here.”
