OPINION and ORDER
EATON, Senior Judge:Before the court are the United States Department of Commerce’s (“Commerce” or the “Department”) final results, dated October 30, 2013, following remand of its antidumping review of Floor-Standing, Metalr-Top Ironing Tables and Certain Parts Thereof from the People’s Republic of China, 74 Fed.Reg. 11,085 (Dep’t of Commerce Mar. 16, 2009) (final results of antidumping duty administrative review), and accompanying Issues and Decision Memorandum (“Issues & Dec. Mem.”) (collectively, the “Final Results”) made pursuant to the court’s order in Since Hardware (Guangzhou) Co. v. United States, 37 CIT -, -, Slip Op. 13-71, 2013 WL 3116513 (2013) (“Since Hardware III ”). See Final Results of Redetermination Pursuant to Ct. Remand (ECF Dkt. No. 169) (“Third Remand Results”).
On this third remand,1 Commerce was instructed to support with substantial evi*1272dence its claim that the import data from the United States Customs and Border Protection Agency (the “Customs Data”) corroborated the 157.68-percent rate for plaintiff Since Hardware (Guangzhou) Co., Ltd.’s (“plaintiff’ or “Since Hardware”) subject merchandise. See Since Hardware III, 37 CIT at —, Slip Op. 13-71, at 16-17; Ex. 1 to Draft Second Results of Redetermination Pursuant to Ct. Remand (ECF Dkt. No. 179) (“Customs Data”).
In the Third Remand Results, Commerce determined that its selected rate of 157.68 percent was corroborated, to the extent practicable, by information from independent sources. See Third Remand Results at 7. The Department then continued to assign the 157.68-percent rate to Since Hardware’s merchandise. See Third Remand Results at 7. Since Hardware objects to Commerce’s determination, alleging that the Department failed to corroborate the assigned rate. See PL Since Hardware (Guangzhou) Co., Ltd.’s Objections to the Department of Commerce’s Redetermination Pursuant to Third Remand 6 (ECF Dkt. No. 174) (“PL’s Br.”). Defendant-intervenor, Home Products International, Inc., urges the court to sustain Commerce’s determination. See Resp. of Home Products International, Inc., Def.int., to PL’s Objections to the Department of Commerce’s Third Remand Determination 5 (ECF Dkt. No. 80). For the following reasons, the court finds that the Third Remand Results are not supported by substantial evidence and remands this matter to Commerce.
STANDARD OF REVIEW
“The court shall hold unlawful any determination, finding, or conclusion found ... to be unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B)(i). “The results of a re-determination pursuant to court remand are also reviewed for compliance with the court’s remand order.” Yantai Xinke Steel Structure Co. v. United States, 38 CIT -, -, Slip Op. 14-38, at 4, 2014 WL 1387529 (2014) (quoting Xinjiamei Furniture (Zhangzhou) Co. v. United States, 38 CIT -, -, 968 F.Supp.2d 1255, 1259 (2014)) (internal quotation marks omitted).
LEGAL FRAMEWORK
Pursuant to 19 U.S.C. § 1677e(a)(2)(C), the Department, when a party “significantly impedes a proceeding,” must use “ ‘facts otherwise available ... to fill in the gaps when Commerce has received less than the full and complete facts needed to make a determination’ from the respondents.” See 19 U.S.C. § 1677e(a)(2)(C); Foshan Shunde Yongjian Housewares & Hardware Co. v. United States, 35 CIT -, -, Slip Op. 11-123, at 7, 2011 WL 4829947 (2011) (alteration in original) (quoting Gerber Food (Yunnan) Co. v. United States, 29 CIT 753, 767, 387 F.Supp.2d 1270, 1283 (2005)) (internal quotation marks omitted). When Commerce has made the further finding that a party has failed to cooperate to the best of its ability, it “may use an inference that is adverse to the interests of that party in selecting from among the facts otherwise *1273available,” to determine that party’s anti-dumping duty rate. 19 U.S.C. § 1677e(b). The use of this inference by Commerce is referred to as adverse facts available (“AFA”).
When determining a rate based on an adverse inference, Commerce may use appropriate “secondary information.” See Uruguay Round Agreements Act, Statement of Administrative Action, H.R. Doc. No. 103-316, at 870 (1994), reprinted in 1994 U.S.C.C.A.N. 4040, 4199 (“SAA”) (“Secondary information is information derived from the petition that gave rise to the investigation or review, the final determination concerning the subject merchandise, or any previous review under section 751 [ (19 U.S.C. § 1675) ] concerning the subject merchandise.”).
Although the Department is authorized to use an adverse inference with respect to the facts used to determine rates for uncooperative parties, Congress has expressly established limitations on the Department when it relies on secondary information to assign a rate. See 19 U.S.C. § 1677e(c) (‘When the [Department] ... relies on secondary information rather than on information obtained in the course of an investigation or review, [Commerce] ... shall, to the extent practicable, corroborate that information from independent sources that are reasonably at their disposal.”). This corroboration requirement intentionally restricts the Department’s ability to select questionable secondary information when it assigns a rate based on adverse inferences. See SAA, H.R. Doc. No. 103-316, at 870, reprinted in 1994 U.S.C.C.A.N. at 4199; see also Hubscher Ribbon Corp. v. United States, 38 CIT -, -, 979 F.Supp.2d 1360, 1365 (2014) (“In practice ‘corroboration’ involves confirming that secondary information has ‘probative value,’ by examining its ‘reliability and relevance.’ ” (quoting 19 C.F.R. § 351.308(d); Mittal Steel Galati S.A. v. United States, 31 CIT 730, 734, 491 F.Supp.2d 1273, 1278 (2007))).
In order to satisfy the relevance requirement, the information used to corroborate a rate must bear some relationship to the respondent. That is, where a rate based on adverse inferences is derived from secondary information, Commerce must support the rate by demonstrating that the information “has some grounding in [the] commercial reality” of the respondent during the period of review (“POR”). See Gallant Ocean (Thai.) Co. v. United States, 602 F.3d 1319, 1324 (Fed. Cir.2010). Therefore, “the assignment of a rate resulting from an adverse inference [must be] based on secondary information [that is] corroborated by evidence showing that the rate is ‘reliable and relevant to the particular respondent.’” Foshan Shunde Yongjian Housewares & Hardware Co. v. United States, 38 CIT -, -, 991 F.Supp.2d 1322, 1328 (2014) (emphasis added) (quoting Foshan Shunde Yongjian Housewares & Hardware Co. v. United States, 37 CIT -, -, Slip Op. 13-47, at 8, 2013 WL 3116253 (2013)); see also F.lli De Ceceo di Filippo Fara S. Martino S.p.A. v. United States, 216 F.3d 1027, 1032 (Fed.Cir.2000) (“It is clear from Congress’s imposition of the corroboration requirement in 19 U.S.C. § 1677e(c) that it intended for an adverse facts available rate to be a reasonably accurate estimate of the respondent’s actual rate, albeit with some built-in increase intended as a deterrent to non-compliance. Congress could not have intended for Commerce’s discretion to include the ability to select unreasonably high rates with no relationship to the respondent’s actual dumping margin.”). In addition, unlike the remainder of the record, it is the responsibility of the Department to develop the corroborating information. See Foshan Shunde, 38 CIT *1274at -, 991 F.Supp.2d at 1829 (“It is important to note that Congress placed the obligation to corroborate secondary information using independent sources on the Department, not on the interested parties who are normally responsible for generating the administrative record.”).
DISCUSSION
1. Background
This case involves Since Hardware’s challenge of the Department’s final results of the third administrative review of the antidumping duty order on floor-standing, metal-top ironing tables and certain parts thereof from the People’s Republic of China (“PRC”) (the “Order”) for the POR, August 1, 2006 through July 81, 2007. See Final Results, 74 Fed.Reg. at 11,085. In the Final Results, Commerce found the use of facts available appropriate because “Since Hardware ... provided unreliable and incomplete documentation in support of its claimed purchases of market economy inputs,” and thereby significantly impeded Commerce’s investigation. See Issues & Dec. Mem. at cmt. 1 (citing 19 U.S.C. § 1677e(a)(2)(C)). As a result of Since Hardware’s failure to cooperate, Commerce further applied AFA in selecting the facts relating to Since Hardware’s independence from the PRC government as well as to those facts surrounding Since Hardware’s cost and country of origin information. See Issues & Dec. Mem. at cmt. 1. This led the Department to determine that Since Hardware could not establish its independence from the PRC government, and as a result, the Department assigned Since Hardware the PRC-wide antidumping duty rate of 157.68 percent. See Issues & Dec. Mem. at cmt. 1.
The court questioned the use of an adverse inference with respect to the information relating to Since Hardware’s entitlement to a separate rate, and remanded the question twice. See Since Hardware (Guangzhou) Co. v. United States, 34 CIT -, -, Slip Op. 10-108, at 22-23, 2010 WL 3982277 (2010) (“Since Hardware I ”); Since Hardware (Guangzhou) Co. v. United States, 35 CIT -, -, Slip Op. 11-146, at 29-30, 2011 WL 5926137 (2011). The Department issued its Second Remand Results on November 29, 2011. Final Results of Redetermination Pursuant to Ct. Remand (ECF Dkt. No. 133) (“Second Remand Results”). In the Second Remand Results, Commerce found that Since Hardware was “entitled to a separate rate.” See Second Remand Results at 5. Commerce then determined a separate rate of 157.68 percent for Since Hardware, based on the rate of a cooperative respondent in the investigation. See Second Remand Results at 7-8. Commerce supported its selection of that rate, in part, by explaining the importance of incentivizing future cooperation. See Second Remand Results at 9.
Commerce also offered several reasons why the rate was “both reliable and relevant,” and therefore corroborated.2 See Second Remand Results at 7. First, the Department asserted that the rate was reliable because it was calculated for a respondent during the investigation. Second Remand Results at 7-8. To corroborate the relevance of its selected rate to Since Hardware, Commerce used the Customs Data derived from rates at which *1275imports of the ironing tables, that entered into the United States during the POR from non-party producers and exporters, were eventually liquidated. See Second Remand Results at 8. In doing so, it argued that, because other companies were able to import and presumably sell ironing tables subject to the 157.68-percent rate, this meant that the rate was representative of the commercial reality of companies in the same business as plaintiff. See Second Remand Results at 8-9. In other words, Commerce contends that the liquidation rate for other exporters’ entries of ironing tables, made during the POR, is indicative of their commercial reality, and thus that of Since Hardware, during that time period.
Commerce also expressly declined to reopen the record of the proceeding to gather more information and calculate a rate specific to Since Hardware. See Second Remand Results at 13. Plaintiff challenged Commerce’s assigned rate of 157.68 percent, arguing that the rate did not reflect plaintiffs commercial reality during the POR. Second Remand Results at 11.
In Since Hardware III, the court rejected plaintiffs claim that Commerce was required to reopen the record, holding that “nothing in th[e] court’s order directed that the Department must reopen the record.” Since Hardware III, 37 CIT at -, Slip Op. 13-71, at 8 (citations omitted). Further, the court found that the rate of 157.68 percent was reliable, but found that Commerce did not demonstrate the relevance of that rate to plaintiffs commercial reality. See id. at -, Slip Op. 13-71, at 11. Accordingly, the court remanded the Second Remand Results. Id. at -, Slip Op. 13-71, at 16.
II. Analysis
A. Commerce May Not Presume that the Highest Prior Margin Is Relevant
In this remand, Commerce endeavors to demonstrate that the assigned rate is relevant to plaintiffs commercial reality in several ways. First, it reasserts its position that the court should permit it to apply a modified version of the Rhone Poulenc presumption. See Third Remand Results at 10 (“[E]ven before looking to the corroborative evidence, the 157.68 percent rate is relevant to Since Hardware by virtue of its choice not to cooperate to the best of its ability.” (citing SAA, H.R. Doc. No. 103-316, at 870, reprinted in 1994 U.S.C.C.A.N. at 4199)). By using this presumption, Commerce attempts to assume that the highest previously calculated rate in the history of the antidumping duty order is automatically relevant to a party that has failed to cooperate with Commerce’s investigation. “In other words, Since Hardware’s actions demonstrate that it preferred to avoid providing to the Department the necessary information for calculating its true margin. Instead of providing that information, Since Hardware knowingly chose to risk incurring an AFA rate of 157.68 percent.” Third Remand Results at 10. Thus, the Department initially seeks to corroborate an assigned rate by means of a presumption rather than by using direct evidence.
The Rhone Poulenc presumption, which was first found in the Federal Circuit’s case, Rhone Poulenc v. United States, confirmed Commerce’s use of the “best information available”3 to determine a rate in those instances where a respondent had *1276refused to answer the questionnaires or where the answers were highly deficient. See Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1190 (Fed.Cir.1990). Put differently, “the case stands for the proposition that a respondent can be assumed to make a rational decision to either respond or not respond to Commerce’s questionnaires, based on which choice will result in the lower rate.” Tianjin Mach. Imp. & Exp. Corp. v. United States, 35 CIT-, -, 752 F.Supp.2d 1336, 1347 (2011).
This Court, however, has continually rejected the expansion of Rhone Poulenc-type presumptions beyond what has been permitted by the Federal Circuit. See, e.g., Tianjin Mach., 35 CIT at -, 752 F.Supp.2d at 1348. This restriction of the inference reflects the Federal Circuit’s reasoning that the use of a Rhone Pou-lenc-type presumption must be based on the “common sense inference” found in the case itself. See Rhone Poulenc, 899 F.2d at 1190 (“It reflects a common sense inference that the highest prior margin is the most probative evidence of current margins because, if it were not so, the importer, knowing of the [“best information available”] rule, would have produced current information showing the margin to be less.” (internal quotation marks omitted)).
The presumption, though, is particularly ill-suited as a substitute for corroboration. As this Court noted in Tianjin Machinery, Rhone Poulenc was decided before the Uruguay Round Agreements Act, which, for the first time, required corroboration of AFA rates based on secondary information. Tianjin Mach., 35 CIT at -, 752 F.Supp.2d at 1347-48 (“Congress directed Commerce to make additional findings in AFA cases.” (citing 19 U.S.C. § 1677e(c))); see also Foshan Shunde, 38 CIT at —— n. 6, 991 F.Supp.2d at 1328 n. 6 (noting that the corroboration requirement of 19 U.S.C. § 1677e(c) was added as part of the Uruguay Round Agreements Act (citing SAA, H.R. Doc. No. 103-316, at 869-70, reprinted in 1994 U.S.C.C.A.N. at 4198-99)). Thus, while Rhone Poulenc may provide a rationale for assigning a particular rate when a respondent has failed to adequately answer Commerce’s questionnaires, it says nothing about whether that rate represents a respondent’s commercial reality.4 First, because corroboration was not a requirement when Rhone Poulenc was decided, the case’s application to the corroboration of a rate is beyond the scope of the Federal Circuit’s holding. Second, “common sense” does not support the idea that a respondent who has been found to be uncooperative can be presumed to know that there is a “rule” that the assignment of a rate previously calculated for another respondent in an earlier segment of the *1277proceeding is self-corroborating. This is because there is no such rule and it is difficult to see how there could be one. That is, the manner in which a respondent decides to behave with respect to answering questionnaires simply tells Commerce nothing about the market conditions in which the respondent operates. Thus, this Court has held that, when a respondent fails to cooperate when answering Commerce’s questionnaires, Commerce cannot “dispense with [the] corroboration requirement by employing the Rhone Poulenc presumption.” Tianjin Mach., 35 CIT at -, 752 F.Supp.2d at 1348. Indeed, the court has found unlawful the use of the presumption in Since Hardware III, and finds no reason to revisit its conclusion. See Since Hardware III, 37 CIT at - n. 4, Slip Op. 13-71, at 11 n. 4.
Thus, the Department’s reliance on the Rhone Poulenc presumption does nothing to satisfy its duty of corroboration. Therefore, the Department may not rely on this presumption to corroborate the 157.68-percent rate.
B. Commerce Has Not Supported with Substantial Evidence its Conclusions Relating to Liquidation Rates
In addition, the Department has not supported with substantial evidence its conclusions drawn from the Customs Data. The Customs Data placed on the record by Commerce comprises information for imports of ironing tables into the United States during the POR. The summary spreadsheet, from which the Customs Data is derived, is made up of numerous columns, including a column that lists the antidumping duty order to which the merchandise is subject, the liquidation rate, the value of the shipment subject to the antidumping duty order, the quantity of merchandise liquidated at that particular rate, and a description of the primary items invoiced in that shipment. See Customs Data at Tab 1. The Department relies on this data to corroborate the 157.68-percent rate5 by arguing that the “AD_DUTY” column in the Customs Data “shows that the entries at issue contained subject merchandise that was liquidated at the 157.68 percent rate.”6 See Third Remand Results at 18. As Commerce explains, the values in this column are equal to the “AD_VALUE” column multiplied by the “AD_RATE” column. See Third Remand Results at 18; Customs Data at 25. From this, Commerce concludes that the values in the “AD_DUTY_BOND” column represent both the amount of duty deposit and the amount at which the merchandise *1278was liquidated.7 See Third Remand Results at 11-12 (“The Customs [D]ata available in this review consist of a summary spreadsheet that lists, inter alia, ... the antidumping duty liquidation rate (variable AD_Duty).... However, regarding the tariff classification for all of the entries that were liquidated at the 157.68 percent rate, we note that the variable ‘CA-SENBR’ lists the ironing tables case number of ‘A-570-888-000,’ and that the column ‘AD__Duty’ specifically includes the antidumping duty that was collected on the shipment.” (quoting Customs Data at 4, 18)). Assuming the “AD_DUTY_BOND” column represents both the amount of the deposit and the liquidated amount, the Department posits that the “AD_RATE” column would represent both the cash-deposit rate and the ultimate liquidation rate. See Third Remand Results at 12; Customs Data at 18, 25. In other words, Commerce infers from the Customs Data that importers actually paid cash deposits at the 157.68-percent rate during the POR and that the merchandise was eventually liquidated at that rate.
Since Hardware disputes that the Customs Data shows that the importers actually paid cash deposits at the 157.68-per-cent rate. For instance, Since Hardware points to the “AD_DEPOSIT_VALUE” column, which does not reflect that 157.68-percent deposits were made on most of the entries.8 See Pl.’s Br. 8; Customs Data at 25. More, importantly, plaintiff asserts that the deposit rate, not the ultimate liquidation rate, is what could be representative of the commercial reality of exporters during the POR. PL’s Br. 8. Plaintiff contends that, if the cash-deposit rate for the relied upon entries during the POR was not 157.68 percent, then doubt is cast on whether the rate of 157.68 percent is representative of plaintiffs commercial reality during the POR. See PL’s Br. 9.
The parties have drawn conflicting conclusions from the Customs Data, ■ and it appears that none of them know with certainty whether a cash deposit was actually paid on any of the relevant entries, or, if paid, the amount of the cash-deposit rate assessed.9 During oral argument, both Commerce and plaintiff conceded that they do not know what the “AD_DEPOS-IT_VALUE” column represents. When pressed on the issue, counsel for Commerce admitted that she did not know whether a cash deposit was actually paid. See Confidential Oral Arg. at 6:05-6:55, Since Hardware (Guangzhou) Co. v. United States, Court No. 09-00123 (June 10, 2014).
Because Commerce has chosen to rely on the Customs Data as its primary source for corroborating the secondary information, the Department has the burden of explaining how this data serves to corroborate the assigned rate. See Foshan Shunde, 38 CIT at-, 991 F.Supp.2d at 1329 (“It is important to note that Congress placed the obligation to corroborate secondary information using independent sources on the Department, not on the *1279interested parties who are normally responsible for generating the administrative record.”). By statute, each “determination by Commerce must include ‘an explanation of the basis for its determination.’ ” See NMB Sing. Ltd. v. United States, 557 F.3d 1316, 1320 (Fed.Cir.2009) (quoting 19 U.S.C. 1677f(i)(3)(A)). Thus, where, as here, “the Department has reached important conclusions that are not fully explained with reference to record evidence [and relevant statutes or regulations], remand is appropriate for Commerce to ‘explain its rationale ... such that a court may follow and review its line of analysis, its reasonable assumptions, and other relevant considerations.’ ” See Since Hardware III, 37 CIT at -, Slip Op. 13-71, at 15 (alteration in original) (quoting Olearon Corp. v. United States, 35 CIT -, -, Slip Op. 11-142, at 27-28, 2011 WL 5909576 (2011)) (internal quotation marks omitted).
The Federal Circuit in Gallant Ocean held that, “[although Commerce has discretion in choosing from a list of secondary information to support its adverse inferences, Commerce must select secondary information that has some grounding in commercial reality.” Gallant Ocean, 602 F.3d at 1324. Thus, in order to have probative value, Commerce must at least prove that its interpretation of the data is more than mere guesswork. Accordingly, should the Department continue to rely on the Customs Data to corroborate the rate assigned to plaintiff, it must (1) provide a full explanation of what the data represents and (2) supply substantial evidence tending to demonstrate the amounts deposited during the POR. Based on the evidence now before the court, however, the Department has failed to show that substantial evidence supports its conclusion that the Customs Data corroborates the 157.68-percent rate.
C. The Ultimate Liquidation Rate Is Not Indicative of Commercial Reality During the POR
The Department contends that the 157.68-percent rate is reflective of plaintiffs commercial reality because, according to Commerce, at liquidation, importers actually paid duties assessed at this rate on entries made during the POR.10 See Third Remand Results at 19. Thus, Commerce believes that the “data do establish that some importers paid antidumping duties of 157.68 percent” and, as a result, the rate is relevant to plaintiff because “other companies were able to participate in the U.S. Market while their imports were liquidated at a rate of 157.68 percent of the value of the entered merchandise.” See Third Remand Results at 12,13.
Plaintiff argues that “[t]he antidumping rate [at which] the entries of subject merchandise may ultimately be liquidated does not establish the commercial reality for a respondent during a review period.” PL’s Br. 8. That is, according to plaintiff, it is the cash-deposit rate, which is assessed during the relevant period, and not the ultimate liquidation rate, which is assessed after the POR, that could be representative of an exporter’s commercial reality during the POR. See PL’s Br. 8. Thus, for plaintiff, commercial reality, at least for other exporters, is demonstrated by the amount paid when the merchandise entered the United States, not the rate at which the merchandise was ultimately liquidated months or years later. Put anoth*1280er way, plaintiff argues that, the fact that the entries were ultimately liquidated at an antidumping duty rate of 157.68 percent does not offer any insight into plaintiffs commercial reality during the POR.
Plaintiff is correct that, regardless of whether the liquidation rate of 157.68 percent may have been eventually paid by importers, that alone is not indicative of plaintiffs commercial reality during the POR. As noted, the Department conceded at oral argument that it did not know whether the Customs Data contained information as to the cash-deposit rate, which would have been the rate at which Since Hardware or other exporters of subject merchandise would have willingly made sales into the United States. The subsequently imposed liquidation rates of entries imported during the POR does not necessarily show that exporters were able to export subject merchandise at a 157.68-percent rate.
This is because, during the POR, only the cash-deposit rates were known to the market participants. Therefore, when making a decision to import the ironing tables, an importer (who normally pays the duties) would only know how much was being paid as a cash deposit and, based on that amount, reach a judgment as to whether it could make a profit. The ultimate liquidation rate of merchandise is not determined until after the completion of the review of the antidumping duty order. Because liquidation rates are only known after the review is completed, neither a respondent nor any potential importer would have had that information at the time of its decision to participate in the U.S. market. See Foshan Shunde, 38 CIT at -, 991 F.Supp.2d at 1333 (“At the time of importation only the cash-deposit rates were known. The ultimate liquidation rate of merchandise subject to a review of an antidumping duty order is unknown until after the completion of the review itself. Thus, for Commerce to assert that [the respondent] or other exporters ‘chose’ to participate in the U.S. market knowing that its products were subject to a 157.68 percent rate simply assumes too much.”).
Thus, despite the Department’s arguments to the contrary, any liquidation rate that differs from the cash-deposit rate of imports subject to the Order sheds no light on whether exporters willingly participated in the U.S. market while being aware that they were likely to be assessed antidump-ing duties of 157.68 percent. As a result, these liquidation rates do not serve to corroborate the 157.68-percent rate because they do not constitute substantial evidence of the commercial reality of plaintiff or other importers during the POR.
D. Commerce Failed to Support with Substantial Evidence its Reliance on the Small Quantity of the Entries
On remand, Commerce was directed to “either identify record evidence indicating that the Customs Data represents a relevant quantity of exports of the subject merchandise or reopen the record to place such additional evidence thereon.” Since Hardware III, 37 CIT at -, Slip Op. 13-71, at 17. Thus, Commerce was required to explain why its reliance on a small number of entries of limited value supplied it with the substantial evidence necessary to support its determination that the 157.68-percent rate was sufficiently corroborated. Id. at -, Slip Op. 13-71, at 16 (“[T]he Department shall explain why the Customs Data represents a sufficiently large number of entries to demonstrate the relevance of the selected rate or ... otherwise corroborate its selected rate in a manner supported by substantial evidence and in accordance with law.”). *1281When Commerce relies on minimal data, courts have held that the Department must establish why that data is sufficient to be considered relevant. For example, in Gallant Ocean, the Federal Circuit held that, “[b]ecause Commerce did not identify any relationship between the small number of unusually high dumping transactions with [the plaintiffs] actual rate, those transactions cannot corroborate the ... rate.” Gallant Ocean, 602 F.3d at 1324. In this case, as in Gallant Ocean, if the Department is relying upon a small amount of data, from a small number of entries,11 which substantially deviates from the rest of the data as a whole, it must explain why that minority of data merits reliance.
The Department argues that, “when corroborating a margin selected based on adverse inferences, Commerce is not required to establish that the selected AFA rate is paid by most, all[,] or a specific number of the importers____Commerce is required to establish only that the secondary information has probative value.” Defi’s Resp. to Comments on the Remand Redetermination 14 (ECF Dkt. No. 183) (“Def.’s Br”). If by this, Commerce means “some probative value” or “a little probative value,” then it misstates its responsibility. Commerce’s responsibility is to corroborate its use of secondary information with substantial evidence. See Consol. Edison Co. of N.Y. v. Nat'l Labor Relations Bd., 305 U.S. 197, 229, 59 S.Ct. 206, 83 L.Ed. 126 (1938) (“Substantial evidence is more than a mere scintilla. It means such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” (citations omitted)). A small amount of data can be insufficiently probative of plaintiffs commercial reality. This does not mean that Commerce is required to provide a certain number of entries for that information to have probative value, but it is required to explain why the limited data in this case is sufficient to substantiate the Department’s reliance thereon. See Yangzhou Bestpak Gifts & Crafts Co. v. United States, 716 F.3d 1370, 1380 (Fed.Cir.2013) (“What could have been'a coincidental correlation of the three data points is not enough to be substantial supporting evidence of economic reality.”); Gallant Ocean, 602 F.3d at 1324 (“Because Commerce did not identify any relationship between the small number of unusually high dumping transactions with Gallant’s actual rate, those transactions cannot corroborate the adjusted petition rate.”). The Department has not done so here. Should the Department continue to rely upon the Customs Data on remand, it must provide a sufficient explanation as to why its small quantity amounts to substantial evidence.
E. Commerce Did Not Adequately Explain the Relevance of the Subject Merchandise that Entered at Lower Rates
On remand, the court also directed. Commerce to explain “the significance, if any, of the subject merchandise being entered at rates below the selected rate.” Since Hardware III, 37 CIT at -, Slip Op. 13-71, at 16. In other words, the Department was required to explain why the large quantity of entries that purportedly entered the United States at lower rates than 157.68 percent did not also provide evidence of the commercial reality of *1282many market participants and thus give some indication of plaintiffs commercial reality.
The Department argues that the entries liquidated at lower rates do not reflect plaintiffs commercial reality because they were assigned rates rather than calculated rates.12 Third Remand Results at 22 (explaining that the selected, lower “rate was not a calculated rate,” and that, “[r]ather than a calculated rate, this [selected, lower] rate represents the separate rate assigned to [other respondents] during the [less-than-fair-value investigation], and is the weighted-average rate for Since Hardware and” the 157.68-percent rate assigned to a cooperating party). Further, according to Commerce, Since Hardware’s filing of unreliable and incomplete “information has precluded the Department from determining Since Hardware’s actual margin or from considering whether Since Hardware’s commercial reality differed from the commercial reality of those companies detailed in the Customs [D]ata.” Third Remand Results at 22. That is, for Commerce, Since Hardware’s filing of unreliable and incomplete documentation “rendered the calculation of a ‘true’ margin for Since Hardware impossible. Lacking information to determine Since Hardware’s own commercial reality, Commerce [thus] used a calculated rate from the [less-than-fair-value] investigation as it [was] representative of the market for the subject merchandise.” Third Remand Results at 22-23. Moreover, the Department insists that, because the lower, selected “rate is not a calculated rate and, because the rate must serve as a deterrent, selecting a rate that is indicative of commercial reality is not the only ... consideration when selecting an AFA rate.” Def.’s Br. 17 (citations omitted).
Plaintiff contends that the Department was unwarranted in concluding that the lower rates were irrelevant to Since Hardware merely because they were assigned rates, rather than calculated rates. See Pl.’s Br. 15 (“The Department’s very conclusion in this proceeding is that the 157.68 percent rate is reflective of the commercial reality during the instant review because importers paid the rate and exporters participated in the market while their merchandise is assessed at the [157.68] percent rate. As directly provided in the Customs Data, the [separate rate] is also a rate paid by importers and a rate by which exporters participated in the market. The Department has not provided explanation for why its conclusion should somehow be different for a separate rate ... as opposed to a calculated rate. This is particularly the case here where the [separate] rate is based, in part, on Since Hardware’s rate. This Court should find the Department’s redetermination in this regard unsupported, and remand to the Department to consider the additional lower rates in the Customs Data.” (footnote omitted)).
The Department defends its finding, claiming that the rates at which subject merchandise were liquidated, which are lower than 157.68 percent, were not appropriate as an AFA rate. Def.’s Br. 17, 18 (“[A]s explained by Commerce, the [rate ' urged by plaintiff, which is] ... the weighted-average rate of the two respon*1283dents in the less than fair value investigation, ... is not a calculated rate and, because the rate must serve as a deterrent, selecting a rate that is indicative of commercial reality is not the only ... consideration when selecting an AFA rate____ Thus, while [the rate sought by plaintiff] was the rate assigned to separate rate respondents in the less than fair value investigation, [that] rate would provide an uncooperative respondent, like Since Hardware, with a lower margin than the highest previously published rate which it had the opportunity to rebut and does not provide proper inducement for cooperation with Commerce’s proceedings.” (citations omitted)).
The Department has not sufficiently explained why the lower rates are not relevant for purposes of corroboration. While it may be that the Department has provided some explanation for why it did not use the separate rates as the basis for assigning a rate to Since Hardware, it has not explained why the rates lack probative value for purposes of corroboration. Instead, Commerce asserts that the entries were assigned rates, rather than calculated rates, but does not explain why this difference makes the lower rates irrelevant. If the relevance of the 157.68-percent rate is, as the Department argues, established by the fact that market participants are paying the rate, why does merchandise entered at an assigned rate also give some indication of commercial reality? That is, what is important here is the entered rate—the rate at which importers are willing to participate in the U.S. market—i.e., whether they believe that, despite paying the rate, they can make a profit. The Department, however, has not sufficiently explained the significance of the subject merchandise entered at lower rates, as directed on remand. See Since Hardware III, 37 CIT at -, Slip Op. 13-71, at 16. The Department was required to explain why the assigned rate represented Since Hardware’s commercial reality. See id. at -, Slip Op. 13-71, at 17. Should the Department continue to rely on the Customs Data on remand, it must fully explain why merchandise entered at rates less than 157.68 percent is not relevant to this inquiry.
F. The Department Has Not Satisfied its Burden of Corroboration
Commerce maintains that it has satisfied its burden to corroborate the rate to the extent practicable. Third Remand Results at 10 (“The question of relevance, then, is whether the 157.68 percent rate has any basis in reality that corresponds with the POR. As we explain below, our examination of the evidence here demonstrates that the limited independent information on the record of this proceeding supports the selected rate and, therefore, that rate is corroborated to the extent practicable.”). The court is unpersuaded. Merely stating that the Department has not identified other sources that support its assessment is an inadequate explanation to support the conclusion that the Customs Data is the only information relevant to Since Hardware. “[T]he Department is not excused from its obligation to corroborate secondary information used to draw an adverse inference simply because the interested parties have placed no corroboratiye information on the record.” Foshan Shunde, 38 CIT at -, 991 F.Supp.2d at 1329 (citing Wash. Int’l Ins. Co. v. United States, 34 CIT -, - n. 3, Slip Op. 10-16, at 5 n. 3, 2010 WL 446048 (2010)). If, on remand, Commerce continues to maintain that there are no other independent sources reasonably at its disposal that could corroborate the assigned rate, Commerce “must describe the steps that it has taken so that a reviewing [c]ourt can determine if the Department’s *1284finding that corroboration was not practicable is supported by substantial evidence and in accordance with law.” Foshan Shunde, 38 CIT at -, 991 F.Supp.2d at 1334, (citing Toyota Motor Sales, U.S.A., Inc. v. United States, 22 CIT 643, 651, 15 F.Supp.2d 872, 877-79 (1998)). Thus, given the Department’s scant explanation in the Third Remand Results of the steps it took to corroborate the assigned rate of 157.68 percent, the court must conclude that Commerce did not satisfy its burden.
G. Plaintiffs Claim that the Rate Is Inconsistent with the Remedial Nature of the Statute Is Waived
Plaintiff claims that “the Department's] decisions do not provide U.S. importers full due process of law because the Department[’s] investigations and reviews are not conducted pursuant to the Administrative Proceduref ] Act.” Pl.’s Br. 16. Plaintiff appears to assert that the rate is inconsistent with the remedial nature of the statute. See Pl.’s Br. 16. Commerce responds that the Federal Circuit has already rejected a similar argument. Def.’s Br. 19 (citing KYD, Inc. v. United States, 607 F.3d 760, 768 (Fed.Cir.2010)). In any event, because plaintiff raises this argument here for the first time in its papers, after three remands, it has waived this argument. See Yantai 38 CIT at -, Slip Op. 14-38, at 18 (“Moreover, plaintiff waived its claim that Commerce used the wrong type of steel for the steel input by failing to raise it before the court prior to the remand. Because plaintiff did not raise this issue until after remand, the court’s instructions necessarily did not direct the Department to reconsider its selection of the input itself.”).
CONCLUSION and ORDER
For the foregoing reasons, it is hereby
ORDERED that the Third Remand Results are remanded; it is further
ORDERED that, on remand, Commerce shall issue a redetermination that complies in all respects with this Opinion and Order, is based on determinations that are supported by substantial record evidence, and is in all respects in accordance with law; it is further
ORDERED that Commerce support the rate assigned to Since Hardware by demonstrating that the information has some grounding in the commercial reality of plaintiff during the POR; it is further
ORDERED that Commerce no longer rely on the presumption that, because the 157.68 percent rate had already been calculated in a prior segment of the proceeding at the time Since Hardware took the risk of providing unreliable, incomplete, and unusable data, it could have anticipated the assignment of that rate, to corroborate the assigned rate to Since Hardware; it is further
ORDERED that, should the Department continue to rely upon the Customs Data, the Department shall clarify all apparent inconsistencies in the data and conclusively establish the cash-deposit rate for the relevant entries; it is further
ORDERED that the Department shall explain with specificity why the cash-deposit rates for other market participants during the POR tend to corroborate the selected rate; it is further
ORDERED that, should the Department continue to rely upon the Customs Data, the Department shall explain with specificity why the Customs Data represents a sufficient quantity of exports of the subject merchandise to be relevant to Since Hardware; it is further
ORDERED that, should the Department continue to rely upon the Customs Data, the Department shall explain with specificity the significance, if any, of.the subject merchandise being entered at rates below the selected rate; it is further
*1285ORDERED that the Department may reopen the record to solicit any information it finds to be necessary to make its determination; it is further
ORDERED that, if Commerce reopens the record, it may seek clarification and further information from the U.S. Customs and Border Protection Agency regarding the Customs Data and what the data represents; it is further
ORDERED that Commerce may select a new antidumping duty rate for Since Hardware and corroborate that rate; and it is further
ORDERED that the remand results shall be due on June 18, 2015; comments to the remand results shall be due thirty (30) days following filing of the remand results; and replies to such comments shall be due fifteen (15) days following filing of the comments.
. The parties to this 2009 case have, no doubt, reached the point at which they wonder how much longer it can last, a point identified by a recent law review article:
The continual remands caused partly by the [Federal Circuit's] refusal to accept, or discouragement of, appeals following disposi-tive remands by the [United States Court of International Trade ("CIT”) ], together with lack of clear recognition of the authority of the CIT to issue a substantive reversal, can result in decisions that have no effect upon the parties and thus are purely advisory opinions. Under current practice, the CIT’s decisions cease to be advisory and have some effect only when the administrative agency agrees to accept the CIT’s direction, under protest, often after several *1272remands, and the case proceeds to a final conclusive result. This is not a rational system for resolving cases in our constitutional system. This Article suggested various practical remedies, but the one principle that is clear is that Article III courts reviewing agency action for lawfulness have the power to reverse the decision of the reviewed agency.
Jane Restani & Ira Bloom, The Nippon Quagmire: Article III Courts and Finality of United States Court of International Trade Decisions, 39 Brook. J. Int’l L. 1005, 1025 (2014).
. In Since Hardware III, the court found that the rate was reliable. See Since Hardware III, 37 CIT at -, Slip Op. 13-71, at 11 ("Plaintiff has neither placed evidence on the record to challenge the reliability of the selected rate, nor pointed to any court’s holding declaring the selected rate to be unreliable. Thus, Commerce has sufficiently demonstrated reliability.”).
. The Rhone Poulenc presumption allowed Commerce to assign a previously established rate to a respondent that did not supply current information from which a rate could be calculated. Thus, the Department could assign a prior margin as the "best information available,” where a party had refused to respond to the Department's questionnaires. *1276See Rhone Poulenc, Inc. v. United States, 899 F.2d 1185, 1190 (Fed.Cir.1990). The "best information available” standard was superseded by the "facts available” and "adverse facts available” provisions following the Uruguay Round Agreements Act in 1994. See SAA, H.R. Doc. No. 103-316, at 869-70, reprinted in 1994 U.S.C.C.A.N. at 4198-99; see also Shandong Mach. Imp. & Exp. Co. v. United States, 34 CIT -, - n. 3, Slip Op. 10-88, at 3 n. 3, 2010 WL 3199830 (2010) (citing SAA, H.R. Doc. No. 103-316, at 868, reprinted in 1994 U.S.C.C.A.N. at 4198; Shandong Huarong Mach. Co. v. United States, 30 CIT 1269, 1282 n. 9, 435 F.Supp.2d 1261, 1274 n. 9 (2006)).
. The recent case of Dongtai Peak Honey Indus. Co. v. United States does not conflict with this conclusion. See Dongtai Peak Honey Indus. Co. v. United States, 777 F.3d 1343 (Fed. Cir.2015). In that case, the Federal Circuit cites KYD, Inc. v. United States for the proposition that a rate determined in a prior segment is reliable. See id. at 1355-56 (citing KYD, Inc. v. United States, 607 F.3d 760, 766-67 (Fed.Cir.2010)). The Court, however, did not authorize the use of the presumption as a substitute for actual evidence of a respondent’s commercial reality.
. The Customs Data has thirty-two columns: "CASENBR,” "CNTRYOR,” "CNTRYEX," "ORGNAME,” “EXPNAME,” “ENTRYNO,” “ENTRYDTE,” “CREDTE,” "LIQDTE,” "EXPDTE,” "HTS,” "DESCRIPTION,” "TYPE,” "TYPEDESC,” "UOM,” "MFRNBR,” "MFRNAME,” "IMPORTER,” "CONSIGNEE,” "IMPNAME,” "AD_DUTY_BOND,” "AD-DEPOSIT-VALUE,” "AD-QTY,” "AD-VALUE,” "AD-RATE,” "DISTPOR,” "DPNAME,” "PORTLAD,” "PLADCNTRY,” "PLAD-NAME,” “YEAR,” and "MONTH.” Commerce relies on the "AD-DUTY-BOND” column (column 21). See Customs Data at Tab 1. Since Hardware relies on the "AD-DEPOSIT-VALUE” column (column 22). See Customs Data at 25. The court has an obligation to review the data as a whole, and other columns appear to be relevant to whether the Customs Data supports Commerce's Determination. The court refers to the "AD_QTY” column (column 23) and the "AD-VALUE” column (column 24). See Customs Data at 25.
. The "AD-DUTY” column, to which Commerce refers, appears to be an excerpt from the Customs Data containing the same data as the “AD_DUTY_BOND” column in the full exhibit. See Customs Data at 18.
. "Liquidation means the final computation or ascertainment of duties on entries for consumption or drawback entries.” 19 C.F.R. § 159.1; see also Shinyei Corp. of Am. v. United States, 524 F.3d 1274, 1276 (Fed.Cir.2008).
. The "AD_DEPOSIT_VALUE” column reflects [[Confidential Data Deleted ]] as the amount deposited and paid for all but [[Confidential Data Deleted]]. See Customs Data at 37. Plaintiff argues that this indicates that a cash deposit was actually paid on only [[Confidential Data Deleted]], and that the cash-deposit rate for the other entries must have been [[Confidential Data Deleted]]. See Pl.’s Br. 9.
.There were [[Confidential Data Deleted]] relevant entries relied upon by Commerce. Third Remand Results at 12.
. According to the Department, [[Confidential Data Deleted ]] of ironing tables came into the United States during the POR. See Third Remand Results at 18. The Department claims that all of these entries were ultimately liquidated at the 157.68-percent rate. See Third Remand Results at 18.
. Commerce relied upon a total of [[Confidential Data Deleted]] entries made during the POR, that were liquidated at the 157.68-percent rate, upon which [[Confidential Data Deleted ]] different importers paid antidump-ing duties, to determine that this rate was “reflective of the commercial reality during the instant review period for these [[Confidential Data Deleted]] importers.” See Third Remand Results at 12-13.
. The "separate rate” ([[Confidential Data Deleted ]]), which plaintiff urges as the rate that should have been assigned to Since Hardware in this review as AFA, was the highest rate at which entries of subject merchandise were liquidated during the POR, and is the weighted-average of the rates calculated for Since Hardware and Shunde Yongjian Housewares Co., Ltd., which were the two respondents in the less-than-fair-value investigation. This was also the rate assigned to other separate rate companies that were not individually investigated.