Memorandum Oeinion and Order
DiCarlo, Chief Judge:Plaintiffs in this consolidated action, Usinor Sacilor and its subsidiaries, Sollac and GTS (collectively “Usinor”), move for Judgment Upon an Agency Record pursuant to USCIT R. 56.2, challenging certain aspects of the final determinations of sales at less than fair value by the Department of Commerce in Certain Hot-Rolled Carbon Steel Flat Products, Certain Cold-Rolled Carbon Steel Flat Products, Certain Corrosion-Resistant Carbon Steel Flat Products, and Certain Cut-to-Length Carbon Steel Plate From France, 58 Fed. Reg. 37,125 (Dep’t Comm. 1993), amended by 58 Fed. Reg. 44,169 (Dep’t Comm. 1993). The court has jurisdiction pursuant to 19 U.S.C. § 1516a(a)(2) (1988) and 28 U.S.C. § 1581(c) (1988).
Background
On June 30,1992, the United States steel industry filed antidumping duty petitions seeking over 40 investigations of various types of steel *1156products from 20 countries, including France. For the French investigations, Commerce named Usinor Sacilor as a mandatory respondent. Usinor Sacilor is a French holding company owning virtually every steel company in France, and its subsidiaries, Sollac and GTS, produced virtually all of the merchandise investigated.
The investigations covered four separate classes of merchandise: hot-rolled carbon steel flat products (hot-rolled steel), cold-rolled carbon steel flat products (cold-rolled steel), corrosion-resistant carbon steel flat products (corrosion-resistant steel), and cut-to-length carbon steel plate (steel plate). The period of investigation (POI) was from January 1, 1992 through June 30, 1992.
In the final determinations, Commerce found dumping margins for all four classes of merchandise sold by Usinor, based on the differences between Foreign Market Value (FMV) and United States Price (USP). The amended final determinations resulted in the weighted-average dumping margins of 80.56% ad valorem for hot-rolled steel, 78.68% for cold-rolled steel, 39.40% for corrosion-resistant steel, and 52.76% for steel plate. Amended Determination, 58 Fed. Reg. at 44,169.
The issues presented in this action include: (1) whether Commerce properly used the arm’s length test in determining if home market sales from Usinor to its related steel service centers should be included in the calculation of FMV; (2) whether it was proper for Commerce to reject Usinor’s data for home market downstream sales from its related secondary steel centers to their unrelated customers, and if so, whether Commerce’s choice of BIA was reasonable; (3) whether it was proper for Commerce to reject Usinor’s revised B-2 product concordance; and (4) whether Commerce properly determined that Usinor miscoded hot-rolled steel product grade E-24.
Discussion
This court must uphold Commerce’s final determination in an anti-dumping investigation unless that determination is “unsupported by substantial evidence on the record, or otherwise not in accordance with law.” 19 U.S.C. § 1516a(b)(l)(B) (1988). Substantial evidence has been defined as “such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Universal Camera Corp. v. NLRB, 340 U.S. 474, 477 (1951) (quoting Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229 (1938)). It “is something less than the weight of the evidence, and the possibility of drawing two inconsistent conclusions from the evidence does not prevent an administrative agency’s finding from being supported by substantial evidence.” Consolo v. Federal Maritime Comm’n, 383 U.S. 607, 620 (1966) (citations omitted).
1. The Arm’s Length Test:
In calculating FMV Commerce normally uses prices for home market sales, as defined in 19 U.S.C. § 1677b(a) (1988), but excludes sales to related parties — because such sales are subject to manipulation— unless the sales were made at arm’s length, i.e., that the prices are “com*1157parable” to the prices of sales made to unrelated parties. See 19 C.F.R. § 353.45(a) (1994). The arm’s length test Commerce used in all the flat-rolled steel investigations is set out in Appendix II. A., Issues Common to All Antidumping Investigations of Flat-Rolled Steel Products, to Notice of Preliminary Determination of Sales at Less Than Fair Value: Certain Cold-Rolled Carbon Steel Flat Products From Argentina, 58 Fed. Reg. 7066,7069 (Dep’t Comm. 1993), and Appendix II. A., Issues Common to All Antidumping Investigations of Flat-Rolled Steel Products, to Notice of Final Determination of Sales at Less Than Fair Value: Certain Cold-Rolled Carbon Steel Flat Products From Argentina, 58 Fed. Reg. 37,062, 37,077 (Dep’t Comm. 1993). Pursuant to this test, for each related customer, Commerce compared, on a product-by-product basis, the weighted average price of total sales from the respondent to the related customer with the weighted average price of total sales from the respondent directly to unrelated customers. If the customer-specific price ratio of related-party versus unrelated-party sales was greater than or equal to 99.5%, all sales to that related customer were determined to be at arm’s length. Certain Cold-Rolled Carbon Steel Flat Products From Argentina, 58 Fed. Reg. at 7069. Conversely, if the price ratio of related-party versus unrelated-party sales was less than 99.5%, all sales to that related customer were considered not at arm’s length because, “on average, that customer was paying less than unrelated customers for the same merchandise.” Id. Commerce excluded from its calculation of foreign market value “all sales to any related customer” that were determined not to be at arm’s length. Id.
Responding to the comments from the parties, Commerce modified the arm’s length test to take into account the effect on prices that different levels of trade may have, as prices to end-users are generally higher than prices to distributors. 58 Fed. Reg. at 37,077. Commerce, however, did not adopt Usinor’s proposal that the arm’s length test should incorporate the quantity of sales in its calculation. Usinor contended that price differences may result when related party sales made in certain quantities are compared to unrelated party sales made in different quantities. Commerce explained:
We agree that Usinor has highlighted a method for fine-tuning our arm’s-length test * * *. Adopting Usinor’s proposal, however, would require us to identify quantity ranges within which quantities are “comparable.” Given the failure of Usinor or other parties to provide factual information on this subject, this would be a highly subjective task that neither petitioners nor respondents could comment on at this point in our investigations. Moreover, * * * we believe that consideration of level of trade could address Usinor’s concern to the extent that different quantities are associated with different levels of trade.
Id.
Usinor’s home market sales to its related secondary steel centers ranged from [ ] percent to [ ] percent of its total home market sales *1158during the POI for the different classes or kinds of merchandise. (Conf. Doc. 11, Letter from Counsel for Usinor to Commerce, Sept. 1,1992, at 4.) Applying the arm’s length test, Commerce found that many of these sales were not at arm’s length.
Usinor contends that Commerce’s arm’s length test is flawed because it did not take into account differences in quantities of sales and a statistical variance. To demonstrate the methodological flaws of Commerce’s test, Usinor engages in a statistical analysis of hypothetical situations. According to Usinor, Commerce should have used Usinor’s methodology, prepared by a statistician in a study of related and unrelated sales in the home market. .
Under the applicable statute and regulations, Commerce has considerable discretion in deciding whether to include related party sales for the calculation of foreign market value The statute merely states that the prices of sales to a related party “may be used in determining foreign market value.” 19 U.S.C. § 1677b(a)(3). According to Commerce’s implementing regulation:
If a producer or reseller sold such or similar merchandise to a person related as described in [19 U.S.C. § 1677(13)], the Secretary ordinarily will calculate foreign market value based on that sale only if satisfied that the price is comparable to the price at which the producer or reseller sold such or similar merchandise to a person not related to the seller.
19 C.F.R. § 353.45(a) (emphasis added).
The court “must accord substantial weight to an agency’s interpretation of a statute it administers. Though a court may reject an agency interpretation that contravenes clearly discernible legislative intent, its role when that intent is not contravened is to determine whether the agency’s interpretation is ‘sufficiently reasonable.’” American Lamb Co. v. United States, 4 Fed. Cir. (T) 47, 54, 785 F.2d 994, 1001 (1986) (citations omitted). Thus, the court will uphold Commerce’s arm’s length test unless the test is shown to be unreasonable.
Usinor has failed to demonstrate that application of Commerce’s test resulted in actual distortion of price comparability. Usinor is unable to point to any record evidence showing that its products were sold at different prices for different quantities, or that Commerce actually compared Usinor’s sales to related parties in larger quantities and at lower prices, with its sales to unrelated parties in smaller quantities and at higher prices. Although Usinor had quantity discount price lists for various products, the record shows that the actual sales prices were determined by Usinor’s negotiation with individual customers, and not based solely on quantity levels. See (Conf. Doc. 12, Usinor’s Questionnaire Response to Section A, at 6) (noting “[t]here is no policy of pricing by customer category nor is there any standard policy governing the use of discounts, allowances, rebates, or other price reductions. Also, terms of sale vary by customer.”); (Conf. Doc. 24, Usinor’s Questionnaire Response to Sections B, C, D, and E, at B-18); (Conf. Doc. 36, Usinor’s *1159Corrected Supplemental Questionnaire Response, at 13.) Given the lack of evidence showing any distortion of price comparability, the court finds the application of Commerce arm’s length test reasonable.
2. Rejection of Downstream Sales Data in Calculating FMV:
During the investigations, Commerce requested Usinor to report all of its home market sales for calculation of foreign market value, including sales through its related secondary steel centers to unrelated customers in France. (Pub. Doc. 50, Antidumping Duty Questionnaire, at B-4, Q. 1.) Usinor holds majority interests in secondary [ ] steel centers and minority interests in [ ] others. (Conf. Doc. 24, Usinor’s Questionnaire Response to Sections B, C, D, and E, at B-3 n.2)
Usinor initially advised Commerce that it was not possible to report accurate information on downstream sales. This was due to a number of factors, including the large number of sales involved, the lack of a centralized data base for the secondary steel centers, the differences in record maintenance, and the time constraints. Usinor also emphasized the difficulties involved in tracing the country of origin of the merchandise sold by the secondary steel centers. Accordingto Usinor, sales by the secondary steel centers to unrelated customers included sales of non-French merchandise which the centers were unable to segregate from sales of French merchandise.
Further, according to Usinor, under French law it could not compel the secondary steel centers in which it held a minority interest to provide the data requested. In any case, Usinor claimed, the downstream sales did not have identical or most similar matches to its U.S. sales.1 Due to these factors, Commerce and Usinor reached an agreement permitting Usinor to provide limited reporting.
Pursuant to the agreement, Commerce allowed Usinor to provide sales information regarding limited product characteristics for all home market sales of the three largest resellers [ ]. Under this arrangement, Commerce would receive sales information for [ ] of related sales of hot-rolled products, [ ] of cold-rolled products, and [ ] of corrosion-resistant steel. Usinor was also required to report total volume and value of merchandise for the other majority-owned centers not reporting information, to list invoices for all sales made during the period of investigation, and to provide limited information for randomly selected invoices from those lists. Finally, Usinor had to demonstrate it did not have control over the centers in which it held a minority interest. This limited reporting arrangement, however, failed to address the inability of the secondary steel centers to identify the source of the steel they processed and/or resold.
*1160Commerce expressly conditioned limited reporting of downstream sales on the lack of “identical or * * * most similar match[s] to the U.S. products sold in all cases * * (Pub. Doc. 87, Letter from Commerce to Counsel for Usinor, Oct. 14, 1992, at 1.) Commerce warned, however, that if, upon review, Commerce determines that any of the downstream sales “would have been selected as the most appropriate match to the U.S. product, the Department may use best information available for its determination.” Id. at 1-2. Commerce required all information to be submitted by December 2,1992. (Pub. Doc. 132, Letter from Commerce to Counsel for Usinor, Nov. 25,1992, at 1.)
On November 24, 1992, Usinor submitted the bulk of the required invoices, and on December 2, 1992, provided the remainder of the invoices and reported its home market sales by its related secondary steel centers per Commerce’s instructions. Usinor discovered, however, upon reviewing its submission, that there were some product matches for the three largest centers for home-market sales with products sold in the United States.
Commerce established December 21, 1992, as the deadline for submission of information to remedy deficiencies in the questionnaire responses. On that date Usinor reported sales data for [ ] of the majority-owned secondary steel centers. The information supplied covered the vast majority of the sales by the cen- [ ters [ [ (Pis.’ Conf. Mem. Supp [ only constituted [ [ [ ] of hot-rolled products; ] of cold-rolled products; and ] of corrosion-resistant products. J. Agey. R. at 13). The remaining ] of the majority-owned centers, which ] of hot-rolled sales, ] of cold-rolled sales, and ] of corrosion-resistant sales, did not have any product matches. Id. at 14. Usinor did not report data from these centers.
Commerce determined that it would not use any of the information placed on the record for Usinor’s downstream sales and instead would resort to BIA. The basis for this action was Usinor’s failure to meet the preconditions of the agreement. Moreover, because Commerce did not use the downstream sales information, Commerce never reached the question of whether Usinor had operational control over the secondary steel centers in which it held a minority interest. Final Determinations, 58 Fed. Reg. at 37,129.
(a) Commerce’s Resort to BIA:
Usinor contends that Commerce should have used Usinor’s downstream sales data, because such data covered nearly all of the downstream sales in the home market and was timely placed on the record. Usinor claims that Commerce was arbitrary in insisting upon Usinor’s submission of downstream sales data and then rejectingit in its entirety. *1161According to Usinor, it had timely supplied almost all of the data originally requested, the errors had been timely corrected, and the data was verifiable. The reason for Commerce’s rejection of its data, Usinor claims, was to punish Usinor for its misrepresentation that downstream sales did not contain appropriate matches to U.S. sales.
Whether Commerce may resort to BIA for downstream sales does not depend upon whether Usinor breached a condition for submitting the requested data. Commerce shall use BIA “whenever a party or any other person refuses or is unable to produce information requested in a timely manner and in the form required, or otherwise significantly impedes an investigation * * 19 U.S.C. § 1677e(c) (1988).
The best information available is not necessarily the most accurate information; rather, it is information that has become usable due to a respondent’s failure to provide accurate information. Asociacion Colombiana de Exportadores de Flores v. United States, 13 CIT 13, 16, 704 F. Supp. 1114, 1117 (1989) aff’d 8 Fed. Cir. (T) 97, 901 F.2d 1089, cert. denied, 498 U.S. 848 (1990).
Usinor was unable to provide accurate information. Commerce requested Usinor to report its home market sales to unrelated customers. Most of the downstream sales reported by Usinor, however, did not identify which sales involved Usinor’s products, due to the inability of most of the secondary steel centers to match the supplier of the materials processed by the centers to the products sold to their customers. (Conf. Doc. 44, Usinor’s Questionnaire Response to Sections A, B, and C, at 5-6.) As Usinor itself recognized, the downstream sales data it had provided was “unreliable” for this reason. Final Determinations, 58 Fed. Reg. at 37,128. Given Usinor’s inability to provide the information requested, Commerce was reasonable in resorting to BIA.
(b) Selection of the Highest Non-Aberrant Margin as BIA:
Usinor next contests Commerce’s choice of the highest non-aberrant margin as BIA. Commerce based its decision to reject Usinor’s data on Usinor’s failure to meet the preconditions for limited reporting. Commerce contended:
[t]he fact that Usinor never ‘guaranteed’ that downstream sales would never be matched to U.S. sales is irrelevant. The letter to Usi-nor from the Department granting Usinor permission to submit less than full reporting of sales data expressly laid out the terms and conditions of such permission * * *. Usinor accepted these conditions when it made the decision to provide limited data for these sales.”
Final Determinations, 58 Fed. Reg. at 37,129. Commerce further argued that Usinor’s submissions were deficient due to Usinor’s failure to report downstream sales from its minority-owned secondary steel centers. (Def.’s Br. at 20).
The court finds Commerce’s reasoning unpersuasive. Usinor substantially met the requirements of the original and modified questionn*1162aire requests. Usinor supplied more data than was required under the limited reporting arrangement and provided well over 99% of the data demanded by the original questionnaire. Rejection of Usinor’s data solely on the basis that Usinor failed to meet a precondition of its agreement with Commerce would be unreasonable. The question, therefore, is whether Commerce may use adverse BIA on the sole basis of Usinor’s inability to trace the source of the steel processed by its secondary steel centers.
Commerce may not choose a BIA methodology that rejects “low margin information in favor of high margin information that was demonstrably less probative of current conditions. ” Rhone Poulenc, Inc. v. United States, 8 Fed. Cir. (T) 61, 67, 899 F.2d 1185, 1190 (1990). Commerce’s selection of a severely adverse BIA is “improper when only a minor adjustment in the data is involved or there is an inadvertent gap in the record, but otherwise the respondent has substantially complied with the questionnaire, National Steel Corp. v. United States, 18 CIT 1126, Slip Op. 94-194 (December 13, 1994) (citations omitted), or when the missing data is beyond the control of the respondent, see Allied-Signal Aerospace Co. v. United States, 11 Fed. Cir. (T)_,_, 996 F.2d 1185, 1190 (1993) (finding respondent’s inability to provide necessary data did not constitute failure to cooperate mandating more adverse BIA); Holmes Prods. Corp. v. United States, 16 CIT 628, 629-31, 795 F. Supp. 1205, 1206-07 (1992) (requiring Commerce to use respondent’s data as respondent had substantially complied with Commerce’s requests and could not control conduct of uncooperative affiliate).
The deficiencies in Usinor’s data were a result of factors outside Usi-nor’s control. Usinor’s subsidiaries did not maintain the sourcing data. Therefore, any tracing would have been done manually. (Conf. Doc. 36, Usinor’s Corrected Supplemental Questionnaire Response, at 25-26.) Due to the time limitations and the large number of invoices involved (180,000), this would have been unreasonable. Given Usinor’s inability to provide the data requested, selection of the highest non-aberrant margin is improper.
Further, that Usinor failed to report sales by the secondary steel centers in which it maintained a minority interest is not necessarily prejudicial, as Commerce declined to investigate whether Usinor lacked operational control. Commerce may resort to BIA, however, if Usinor inaccurately represented that it has no such control.
To resolve this issue, the court directs Commerce to determine whether Usinor had operational control over the secondary steel centers in which it held a minority interest. If Commerce finds that Usinor did not have operational control, Commerce is directed to select the weighted average calculated margin as BIA. If Commerce finds Usinor maintained operational control, Commerce may reapply the highest non-aberrant margin as BIA in a manner consistent with this court’s decision in National Steel Corp. v. United States, 18 CIT 1126, Slip Op. 94-194.
*11633. Rejection of Usinor’s Correction of Product Concordance:
After the publication of the preliminary determinations and before verification, Usinor informed Commerce that it might have misunderstood Commerce’s product matching instruction and placed level of trade before physical similarity in constructing the product concordance. Usinor explained that it did not realize the problem until it reviewed Commerce’s preliminary determination in Certain Hot-Rolled Carbon Steel Flat Products, Certain Cold-Rolled Carbon Steel Flat Products, and Certain Corrosion-Resistant Carbon Steel Flat Products From Japan, 58 Fed. Reg. 7103 (Dep’t Comm. 1993) (prelim, determ.), in which Commerce stated that one of the respondents, Nippon Steel, had incorrectly “matched sales of less similar models at the same or the most comparable level of trade rather than matching sales of more similar products at other levels of trade.” Id. at 7105. Usinor requested Commerce to either confirm its product matching hierarchy or to accept Usinor’s submission of a revised product concordance.
Commerce rejected Usinor’s revised product concordance as unsolicited and untimely submitted information. See 19 C.F.R. § 353.31(b)(2) (1994). In its final determinations, Commerce acknowledged that Usi-nor’s error could have been caused by confusion over the “General Instructions” on Section A of the Questionnaire. Final Determinations, 58 Fed. Reg. at 37,130.
However, at the point when we discovered Usinor’s error, we could not allow the correction of the errors nor can we now simply use the existing information in our margin analysis since we know that this information is incorrect. Because of the arguable ambiguity in Section A of the questionnaire and the limited number of U.S. sales involved, we have applied the weighted-average margin calculated for all other sales on a class or kind basis to sales of each U.S. product that were incorrectly matched to a home market product * * *.
Id. (emphasis added).
Commerce further explained that, because of “the large volume of cases” and “the limited time available to conduct these investigations,” it had decided not to accept corrections to, nor to significantly revise on its own, the product concordances submitted prior to verification. Id.
Usinor contends that Commerce should have either accepted Usi-nor’s revision, or modified the product concordance on its own, as the error was caused by Commerce’s misleading instructions. Usinor further claims that the correction was not new information, but clarification of the existing information on the record. Furthermore, Usinor claims Commerce could have saved more time and effort by accepting Usinor’s revised concordance before verification than by resorting to the use of weighted-average margins as best information available for the sales affected by the errors.
During an antidumping proceeding, Commerce has discretion to set time limits for the submission of a questionnaire response or other factual information requested. See 19 C.F.R. § 353.31(b). Because Com*1164merce must complete the investigation within strict statutory time limits, it is imperative that the requested information be submitted “within a period that allows Commerce sufficient time for adequate analysis and comment while still meeting statutory deadlines. ” Ansaldo Componenti, S.p.A. v. United States, 10 CIT 28, 37, 628 F. Supp. 198, 205 (1986). Thus, in general, the court has affirmed Commerce’s rejection of corrections submitted after the deadline. See e.g., Mantex, Inc. v. United States, 17 CIT 1385, 1409, 841 F. Supp. 1290, 1310 (1993); NSK Ltd. v. United States, 16 CIT 745, 749-50, 798 F. Supp. 721, 725 (1992), aff’d, 11 Fed. Cir. (T)_, 996 F.2d 1236 (1993); Sugiyama Chain Co. v. United States, 16 CIT 526, 533, 797 F. Supp. 989, 996 (1992).
The court, however, also has required Commerce to accept late submission of corrections where the error was “so obvious or egregious” that the failure to correct it constituted an abuse of discretion and undermined the interests of justice, Tehnoimportexport v. United States, 15 CIT 250, 260, 766 F. Supp. 1169, 1178 (1991), or where the errors were clerical and the limited burden of correction was “far outweighed by the preference for accuracy in final dumping determinations.” Koyo Seiko Co. v. United States, 14 CIT 680, 683, 746 F. Supp. 1108, 1111 (1990). See also Serampore Indus. Pvt., Ltd. v. United States Dep’t of Comm., 12 CIT 825, 834, 696 F. Supp. 665, 673 (1988). Furthermore, the court has held that it was arbitrary for Commerce to reject a respondent’s late submission where the untimeliness was of Commerce’s own making. Bowe-Passat v. United States, 17 CIT 335, 343, Slip. Op. 93-68 (May 7, 1993) (requiring Commerce to accept party’s late submission clarifying deficiencies that Commerce had been aware of but failed to disclose until too late).
Essentially, when deciding whether Commerce abused its discretion in rejecting the information submitted after the deadline, the court considers the interests of accuracy and fairness, and the burden imposed upon the agency by accepting the late submission.
In this case, the court finds the interests of accuracy and fairness outweigh the burden imposed upon the agency. The court recognizes that this investigation was among the large number of steel investigations that engaged Commerce in an unusually intensive and demanding period of work. This circumstance made it more difficult for Commerce to accept corrections submitted after the deadline. In addition, Commerce used weighted-average margins as the best information for the sales affected by the errors, which minimized the adverse effect that any inaccuracy resulted from the use of BIA might have on Usinor’s final margins.
There are, however, other significant facts the court must consider. First of all, the court finds the relevant section of the questionnaire, when read by itself, could be misleading as to the priority of the product matching hierarchy. See (Pub. Doc. 50, Antidumping Duty Questionnaire, at A-4, Q. 4C.) Where, as here, an ambiguity in the questionnaire *1165causes a party to make errors in its response, Commerce’s obligation to remedy the situation it caused is greater than it would otherwise be.
Here, Usinor submitted its revised product concordance before verification. At the time Usinor approached Commerce with the revision, on February 17, 1993, Commerce was still seeking “clarifications” of the information previously submitted, although it specifically excluded “new product concordances” from the scope of clarification. (Pub. Doc. 190, Letter from Commerce to Counsel for Usinor, Feb. 10,1993, at 1.) These corrections merely sought to change the format of the data, and not what constituted the data itself. Finally, given the limited number of U.S. sales involved in the revision, see (Conf. Doc. 60, Letter from Counsel for Usinor to Commerce, Feb. 17,1993), the court does not agree that acceptance of the revision would have placed an undue burden upon Commerce.
Having considered the interests of accuracy, fairness and administrative economy, the court holds that Commerce abused its discretion in rejecting Usinor’s revised product concordance. Accordingly, the court remands the issue to Commerce with the instruction that Commerce accept the revised product concordance.
4. Miscoding:
Commerce found at verification that Usinor incorrectly assigned product codes for an entire grade of steel in both the U.S. and home markets. Final Determinations, 58 Fed. Reg. at 37,131. Usinor reported the product, grade E-24 hot-rolled steel, as having yield strength of one category, whereas Commerce found the product should be coded as another category having a different yield strength. Consequently, Commerce applied partial BIA to all sales involving that product. The BIA margin assigned was the higher of the average margin in the petition or the highest non-aberrant calculated margin in the investigation. Id. Commerce applied this adverse BIA margin because the errors were not limited in nature. Id.
Usinor challenges Commerce’s use of BIA, claiming that it did not miscode the product. Usinor contends that although, according to the industry standard printed on the product sheet, the product should have one category of yield strength, the actual product made by Usinor had the yield strength of another category. Usinor claims it coded the product according to its actual yield strength, and that Commerce was silent in its instructions as to whether the actual yield strength or the industry standard should be used. In addition, Usinor points out that Commerce verified Usinor’s coding of the product by its actual yield strength.
(a) Whether Usinor is barred from raising the arguments:
Defendant and defendant-intervenor contend that Usinor is barred from raising these arguments before the court because it did not raise them in a timely fashion below, and thus failed to exhaust administrative remedies as required by law.
*1166The court “shall, where appropriate, require the exhaustion of administrative remedies.” 28 U.S.C. § 2637(d) (1988). “A reviewing court usurps the agency’s function when it sets aside the administrative determination upon a ground not theretofore presented and deprives the [agency] of dn opportunity to consider the matter, make its ruling, and state the reasons for its action.” Unemployment Compensation Comm’n v. Aragon, 329 U.S. 143, 155 (1946).
The record shows Usinor argued, at the administrative level, that Commerce should not apply “punitive BIA” to sales involving “arguably incorrectly coded products.” (Conf. Doc. 81, Resp’t’s Rebuttal Br. before ITA, at 24.) Usinor stated that its product experts decided to code grade E-24 steel as having the yield strength of that category because “most of the products actually produced fill [sic] within [that category] range.” Id. at 25.
Apparently, Usinor argued at the administrative level that Commerce should not apply a punitive BIA rate because Usinor had substantially complied with Commerce’s instructions. Before the court, Usinor argues that Commerce should not apply BIA at all because Usinor reported the product according to its actual yield strength, which was not inconsistent with Commerce’s instructions.
While Usinor appears to have shifted its arguments, the ground upon which Usinor bases its arguments remains the same, that is, Usinor’s compliance with Commerce’s instructions. Because the question of whether Commerce properly resorted to BIA depends on whether Usi-nor complied with Commerce’s information request, see 19 U.S.C. § 1677e(c), and because Usinor raised the issue of compliance during the administrative proceeding, the court concludes that it would not be appropriate to bar Usinor from raising its arguments before the court.
(b) Whether Usinor complied with Commerce’s instruction:
Usinor argues that it did not miscode the product according to its actual yield strength, because Commerce did not specify whether Usi-nor should report the product by the industry standard or by the product’s actual yield strength. The record shows that Commerce’s questionnaire required Usinor to classify its hot-rolled products according to certain ranges of minimum yield strength, but did not indicate whether the classification should be by the actual quality of the product. (Pub. Doc. 50, Antidumping Duty Questionnaire, App. v. at 5.)
It is unclear whether Commerce, when drafting the questionnaire instructions, should have anticipated that the actual quality of the product may differ from the industry standard. The fact remains, however, that Commerce did not require Usinor to report the product according to the industry standard. Under the circumstances, it was not unreasonable for Usinor to report the product by its actual yield strength. Usi-nor’s reporting of the product according to its actual yield strength was verified by Commerce. (Conf. Doc. 102, Verification Report, at 7.) Moreover, Usinor classified the product consistently for both the U.S. market *1167and home market, and there is no suggestion that Usinor acted in bad faith.
“Before [Commerce] may find any non-compliance on the part of the parties to the proceeding, there must be a clear and adequate communication requesting the information* * Daewoo Elecs. Co. v. United States, 13 CIT 253, 266, 712 F. Supp. 931, 945 (1989), rev’d in part on other grounds, 11 Fed. Cir. (T)_, 6 F.3d 1511, cert. denied, International Union of Elec., Elec., Technical, Salaried, & Mach. Workers v. United States, 114 S. Ct. 2672 (1994). As Commerce did not specify the standard by which Usinor should classify the product, the court finds there was a lack of clear communication requesting the information.
Accordingly, the court remands the issue to Commerce for redeter-mination. Commerce may either provide Usinor with an opportunity to reclassify the product according to the industry standard, or accept the classification based on actual yield strength. If Commerce chooses the latter, it should make a further finding as to whether Usinor correctly reported all of its grade E-24 steel by the actual yield strength.
Conclusions
Commerce’s final determinations are affirmed in part, and remanded in part. On remand, Commerce shall (1) determine whether Usinor had operational control over the secondary steel centers in which it maintained a minority interest and recalculate the appropriate BIA margin as directed herein after making such determination; (2) accept Usinor’s revised product concordance submitted on February 17, 1993; and (3) redetermine the issue of miscoding grade E-24 steel, in accordance with the specific instruction set forth in this opinion.
The remand results shall be filed with the Court within [60] days from the date of this opinion. Any party contesting the remand results shall file comments within 30 days of the remand results. Commerce may file its response to any comments within 15 days of the filing of the comments.
Usinor’s representation did not extend to cut-to length carbon steel plate since the U.S. sales of these products all had identical matches of French merchandise sold to unrelated purchasers in France. (Conf. Doc. 24, Usinor’s Questionnaire Response to Sections B, C, D, and £ at B-4.)