OPINION

TSOUCALAS, Judge:

Plaintiffs, Tehnoimportexport and Peer Bearing Corp. (“Tehnoimportexport”), challenge the final determinations of the Department of Commerce, International Trade Administration (“Commerce” or “ITA”) in Final Determinations of Sales at Less Than Fair Value; Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From Romania, 54 Fed.Reg. 19,109 (1989). Specifically, plaintiffs contest the decision to use surrogate pricing data from Portugal to determine the foreign market value (“FMV”) of plaintiffs’ imported antifriction bearings (“AFBs”) from Romania, and they also challenge the failure of the ITA to correct certain ministerial errors.

Background

On March 31, 1988, The Torrington Company (“Torrington”) filed an antidumping petition which alleged, inter alia, that imports of AFBs from Romania were being sold in the United States at less than fair value and were injuring, or threatening to injure, an American industry. General Administrative Record (“GAR”) (Pub.) Doc. 1. The ITA initiated an investigation of Romanian AFBs on April 27, 1988. 53 Fed.Reg. 15,077 (1988). Because Romania has a non-market, or state-controlled, economy, the ITA was obliged to use data from a market economy country to determine the foreign market value of Romanian AFBs. 19 U.S.C. § 1677b(c) (1988); 19 C.F.R. § 353.8 (1988). 1

The petition proposed Italy and Portugal as surrogate countries. GAR Doc. 1 at 95. Tehnoimportexport, on the other hand, suggested that Turkey, Yugoslavia or Indonesia should be considered. Romanian Record (“RR”) (Pub.) Doc. 75. Later, Tehnoimportexport wrote that Mexico or, in the alternative, Yugoslavia, would be an appropriate substitute for Romania. RR (Pub.) Doc. 162. In a memorandum dated August 8, 1988, the ITA listed five potential surrogate countries: Portugal, Brazil, Mexico, Korea and Yugoslavia. RR (Pub.) Doc. 94 at 1. The ITA stated that these countries were selected “because we determined that their economies were comparable to that of Romania.” Id. The factors considered in making these choices included gross per capita income and the distribution of labor between the agricultural and non-agricultural sectors; attached statistics support this conclusion.

Once the ITA had selected five possible surrogates for Romania, it sent questionnaires to companies and embassies in those countries requesting information regarding prices and the values of the factors of production. These questionnaires were sent out between August and September 1988. See e.g., RR (Pub.) Docs. 101, 103, 126, 127, 138 and 147. The ITA indicated that, “no affirmative responses [were] received” from any of the companies contacted by the ITA. Preliminary Determinations of Sales at Less Than Fair Value: Antifriction Bearings (Other Than Tapered Roller Bearings) and Parts Thereof From Romania, 53 Fed.Reg. 45,324, 45,-326 (1988). Therefore, the ITA “used the factors of production valued in a comparable economy as the basis of foreign market value.” Id. at 45,326-327. This methodology is expressly authorized by statute and regulation. 19 U.S.C. § 1677b(c); 19 C.F.R. § 353.8(c).

Furthermore, since no responses to its questionnaires were received by the ITA, the agency was authorized by Section 776 of the Tariff Act of 1930, as amended, 19 U.S.C. § 1677e(c) (1988), to use the best information otherwise available in its valuations of the factors of production. Where possible, “best information available was *1173 obtained from publicly available sources in Yugoslavia,” the country selected as the most comparable to Romania. 53 Fed.Reg. at 45,327. However, given the dearth of information gathered to that point, the ITA made clear that it would attempt to obtain more surrogate pricing data prior to the final determinations. Id. at 45,326.

The ITA’s preliminary determinations were released on October 27, 1988, and were published on November 9, 1988, in the Federal Register. In them, the ITA chose to use certain factors of production from Yugoslavia as the basis for calculating the FMV of Tehnoimportexport’s AFBs. The ITA reasoned that Yugoslavia’s level of economic development most closely approximated Romania’s. 53 Fed.Reg. at 45,327. However, since public information was not available for all factors of production from Yugoslavia, the ITA used data from other countries, including Mexico, for the valuation of the balance of the factors of production for purposes of the preliminary determinations.

The final determinations were issued on March 24, 1989 and published in the Federal Register on May 3, 1989. Final Determinations, 54 Fed.Reg. 19,109. The ITA stated that since it was unable to obtain adequate pricing information from Yugoslavia, it opted to change surrogates and use Portugal as the surrogate country for purposes of calculating the factors of production in the final determinations. 54 Fed.Reg. at 19,110. Portugal was chosen because its “level of economic development also closely approximates that of Romania” and it had been listed in the preliminary determinations as one of five possible surrogate countries. Id. Portuguese data was used for: Steel used to manufacture bearing cages; steel scrap; brass used in the manufacture of brass cages; overhead; and labor. Id. at 19,110-111. Because Portugal does not produce bearing quality steel, the ITA used “information obtained from the World Material Study — Europe, as provided by Torrington and verified by the Department” to value steel used to produce inner and outer rings, balls, and other components. Id. at 19,110.

Plaintiffs contend that the ITA should not have switched surrogates in midstream. They claim that Yugoslavia was the country most economically comparable to Romania and thus should have been used for the final determinations as well as the preliminary.

Discussion

A determination by the Department of Commerce will be affirmed unless that determination is not supported by substantial evidence or is otherwise not in accordance with law. 19 U.S.C. § 1516a(b)(l)(B) (1988). Substantial evidence is relevant evidence that “a reasonable mind might accept as adequate to support a conclusion.” Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229, 59 S.Ct. 206, 216, 83 L.Ed. 126 (1938); Alhambra Foundry Co. v. United States, 12 CIT 343, 345, 685 F.Supp. 1252, 1255 (1988). Under this standard, Commerce is granted considerable deference in its interpretation of its statutory authority and the methodology employed in the administration of the anti-dumping law. American Lamb Co. v. United States, 785 F.2d 994, 1001 (Fed.Cir. 1986). See also Ceramica Regiomontana, S.A. v. United States, 10 CIT 399, 404, 636 F.Supp. 961, 965-66 (1986), aff'd, 810 F.2d 1137 (Fed.Cir.1987).

Commerce’s determination “will not be overturned merely because the plaintiff ‘is able to produce evidence ... in support of its own contentions and in opposition to the evidence supporting the agency’s determination.’ ” The Torrington Co. v. United States, 14 CIT —, —, 745 F.Supp. 718, 723 (1990), quoting Hercules, Inc. v. United States, 11 CIT 710, 755, 673 F.Supp. 454, 490 (1987). The evidence introduced by plaintiffs must be “enough to convince the Court that a reasonable mind would not have found ITA’s evidence sufficient to support its conclusion.” Torrington, 14 CIT at —, 745 F.Supp. at 723.

Plaintiffs first contest the ITA’s decision to use Portugal as the surrogate in the final determinations. Second, Tehnoimportexport faults the ITA for calculating overhead costs using unverified data from a *1174 Portuguese company that was related to a company under investigation. Similarly, plaintiffs claim that the ITA violated its own precedent and policy by calculating packing costs based on confidential data from a Swedish firm whose prior written permission was not obtained by the ITA and which firm was related to a firm under investigation. Finally, plaintiffs contest the ITA’s decision not to make corrections regarding the weights of rivets.

I. Use of Surrogate Pricing Data

When in the course of an anti-dumping investigation the ITA determines that the economy of the country from which the subject merchandise is exported is state-controlled, the agency is authorized to determine the FMV of the merchandise based on the “the best available information” regarding the values of the factors of production used in manufacturing the merchandise in a “market economy country or countries considered to be appropriate by the administering authority.” 19 U.S.C. § 1677b(c)(l). If the ITA determines that adequate information is not available to determine FMV on this basis, then the agency may value FMV based on the price that merchandise which is “comparable to the merchandise under investigation,” and “produced in one or more market economy countries that are at a level of economic development comparable to that of the non-market economy country,” is sold in other countries, including the United States. 19 U.S.C. § 1677b(c)(2).

Commerce has interpreted this rather convoluted passage to mean that, if the economy of the country from which the imported merchandise originates is state-controlled, then FMV shall be determined based on the prices “at which similar merchandise produced in a non-state-controlled-economy country or countries is sold either: (i) For consumption in the home market of that country or countries, or (ii) to other countries, including the United States.” 19 C.F.R. § 353.8(a)(1). Alternatively, the ITA may determine FMV based on the constructed value of such or similar merchandise in a non-state-controlled-economy country. 19 C.F.R. § 353.8(a)(2). There is an express preference in the regulation for prices over constructed value.

The reason for this provision is so that the agency can acquire an accurate reading of the actual costs of a company operating in a state-controlled-economy. The non-state-controlled, or surrogate, country, however, should have an economy which is at a comparable level of economic development as the subject economy and should be a significant producer of comparable merchandise. 19 U.S.C. § 1677b(c)(4); 19 C.F.R. § 353.8(b); see The Timken Co. v. United States, 12 CIT 955, 958-59, 699 F.Supp. 300, 303 (1988), after remand, 13 CIT —, —, 714 F.Supp. 535 (1989), aff'd, 894 F.2d 385 (Fed.Cir.1990); Chemical Prods. Corp. v. United States, 10 CIT 626, 630, 645 F.Supp. 289, 293 (1986). Of course, no two countries have identically developed economies, but to the extent possible, Commerce is required to compare similarly developed economies; for example, the United Kingdom would not make a suitable surrogate for Albania. Commerce’s regulations further state that this comparability should be based on factors such as per capita gross national product and infrastructure development, particularly within that industry. 19 C.F.R. § 353.8(b)(1).

That Romania has a state-controlled-economy is not in issue. Tehnoimportexport’s complaint is that the ITA abused its discretion by choosing Portugal as the surrogate, instead of Yugoslavia. Since Yugoslavia had been selected as the most comparable country in the preliminary determinations, and there allegedly was more publicly available information from Yugoslavia than from Portugal and the ITA did not give notice that it intended to switch surrogates, plaintiffs aver that Commerce’s actions were improper. Memorandum of Points and Authorities in Support of Plaintiffs’ Motion for Judgment on the Agency Record (“Plaintiffs’ Memorandum”) at 2.

The fact that the ITA chose Yugoslavia as the most comparable country to *1175 Romania for the preliminary determinations, then used Portugal for the final determinations, is not evidence, in itself, to support the contention that the ITA abused its discretion. The reason for having both preliminary and final determinations is so the ITA can make corrections and adjustments to its preliminary findings and reach a more accurate conclusion in the final determinations. Some changes are to be expected in any final determination, and there was no obligation on the part of the ITA to notify the parties beforehand that there would be a different surrogate used in the final determinations than in the preliminary.

Nonetheless, the ITA did explain in the preliminary determinations that additional factfinding as to the surrogate country issue would be necessary prior to the final determinations. 53 Fed.Reg. at 45,326. Indeed, plaintiffs themselves proffered data on plausible proxies for Romania well after the preliminary determinations were issued. See RR (Pub.) Docs. 233, 249, 250 and 260. This included information on the value of the price of steel in Portugal and Korea, the cost of labor and electricity in Yugoslavia and the cost of gas and water in Mexico. Id. Hence, there exists ample evidence that Tehnoimportexport was aware that a country other than Yugoslavia might be selected as the surrogate in the final determinations, and that Portugal was among the finalists.

The issue facing the Court is whether the ITA’s choice of Portugal as a surrogate for purposes of the final determinations was supported by substantial evidence and was otherwise in accordance with law. The Court will not impose its choice of which economy is more comparable to Romania’s, provided the choice made by Commerce is sufficiently reasonable and supported by the evidence. See Mitsubishi Elec. Corp. v. United States, 12 CIT 1025, 1050-51, 700 F.Supp. 538, 558 (1988), aff'd, 898 F.2d 1577 (Fed.Cir.1990).

The Commerce regulation governing the selection of surrogate countries gives only nebulous guidance as to the factors to be considered by the ITA. The regulation refers to “generally recognized criteria, including per capita gross national product and infrastructure development (particularly in the industry producing such or similar merchandise).” 19 C.F.R. § 353.8(b)(1). In choosing to consider Yugoslavia, Portugal, Mexico, South Korea and Brazil as potential surrogates for Romania prior to the preliminary determinations, the ITA examined gross national product (“GNP”), gross domestic product (“GDP”) and the distribution of labor between agricultural and non-agricultural sectors of the countries for 1985. 2 RR (Pub.) Doc. 94 at 3. The figures for Yugoslavia were closest to those for Romania, and the ITA relied primarily on Yugoslavian factors of production in the preliminary determinations. Id. See also 53 Fed.Reg. at 45,327.

Although the record indeed indicates that among the five suitable surrogates, Yugoslavia was comparable to Romania, it is manifest from the evidence that Portugal too was economically comparable to Romania. See RR (Pub.) Doc. 94. The statistics supplied by Commerce show that in terms of gross domestic product, gross national product and labor distribution, Portugal and Romania were at a similar level of economic development. Id. Additionally, the law does not require the ITA to choose the most comparable economy, but rather a comparable economy. Clearly, the ITA has done so in this case. As both Portugal and Yugoslavia reasonably could have been selected as surrogates, the Court will not disturb the ITA’s finding merely because Yugoslavia’s figures were slightly closer to Romania’s than Portugal’s were. 3

*1176 Furthermore, it does not appear from the record that the decision to change surrogates was made arbitrarily. Commerce stated in the final determinations that it was “unable to obtain adequate pricing data” from Yugoslavia, and “[d]ue to the lack of information from Yugoslavia, we have chosen Portugal as the surrogate country for purposes of valuing the factors of production in these final determinations.” 54 Fed.Reg. at 19,110. In fact, Tehnoimportexport’s counsel acquiesced in a letter to Commerce on February 7, 1989, wherein counsel wrote that “to date it has been impossible to obtain steel prices in Yugoslavia.” RR (Pub.) Doc. 249 at 1.

Lastly, plaintiffs assert that Portugal was a poor choice because there was more information publicly available from Yugoslavia. Plaintiffs ’ Memorandum at 27-28. Though such information is preferred, it is not imperative that the ITA choose the country with the most publicly available data. Regardless, the difference between the amount of public data obtainable from Yugoslavia as opposed to that from Portugal is not so substantial as to warrant a reversal of the agency’s selection. 4

Accordingly, the Court finds that Commerce’s decision to use Portugal as a surrogate country for valuation of the factors of production in Romania was supported by substantial evidence and was otherwise in accordance with law, and is affirmed.

II. Use of Data from Firms Under Investigation

Tehnoimportexport also contests the determinations on the basis that the ITA improperly used data on overhead and packing costs from companies which either were under investigation themselves or were related to companies under investigation. Plaintiffs cite a longstanding Commerce policy not to use pricing data from companies under investigation when determining FMV for a company from a state-controlled economy country. Tapered Roller Bearings and Parts Thereof, Finished or Unfinished, From the Hungarian People’s Republic, 52 Fed.Reg. 17,428, 17,430 (1987). The reason for this policy is that such data is subject to manipulation.

A. Overhead Data

During the period between the preliminary and final determinations, the ITA did not receive any responses to its questionnaires from the relevant companies in the potential surrogate countries regarding their overhead costs. The Tariff Act of 1930, as amended, provides that “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, [the ITA shall] use the best information otherwise available.” 19 U.S.C. § 1677e(c).

Since those questioned did not provide the information Commerce sought, the ITA based overhead costs on best information available (“BIA”), pursuant to 19 U.S.C. § 1677e. Defendant’s Memorandum at 36-37. Information regarding overhead costs was taken from a Portuguese company related to a German AFB manufacturer under investigation, an apparent violation of Commerce’s policy against using costs from such companies.

However, the related firm’s costs were not used as direct price surrogates, but rather as BIA pursuant to 19 U.S.C. § 1677e. The policy proscribing use of a related firm’s costs clearly does not apply to situations where that data is used as BIA. Commerce has much greater latitude under the BIA provisions than under normal conditions, that is, when the questioned firms respond to Commerce’s request for information. See N.A.R., S.p.A. v. United *1177 States, 14 CIT —, 741 F.Supp. 936 (1990). Indeed, such information is not even required to be verified by the ITA. Timken, 12 CIT at 960, 699 F.Supp. at 305; see also Ansaldo Components S.p.A. v. United States, 10 CIT 28, 34, 628 F.Supp. 198, 203 (1986). Therefore, provided the data is reasonable, it may be used to calculate FMV.

The overhead costs from Portugal were based on the undertakings of a Portuguese AFB producer, and were “estimated from the available data about the only Portuguese (German owned) company in the sector.” RR (Pub.) Doc. 185; RR (Pub.) Doc. 298 at 3. Additionally, plaintiffs have offered no evidence that this data was inaccurate or had been manipulated. Therefore, the Court finds that the ITA’s use of the overhead cost data of a firm related to a firm under investigation was in accordance with law, because none of the companies surveyed responded to Commerce’s questionnaires, and the costs were calculated pursuant to the best information available provisions of the Tariff Act, to which the policy against use of such data is not applicable.

B. Packing Data

Tehnoimportexport also challenges the use of packing cost data from a Swedish firm under investigation. In addition to the arguments, stated above, regarding the fact that the firm was under investigation, the packing data is opposed on the grounds that the ITA used the Swedish firm’s confidential packing figures in contravention of agency policy. Plaintiffs’ Memorandum at 39-40.

Regarding the use of a related firm’s cost data, the Court’s analysis of overhead costs above is also applicable to packing costs, inasmuch as the ITA used that data as BIA pursuant to 19 U.S.C. § 1677e when the questioned firms did not respond to requests for information. However, the use of confidential data without the communicated consent of the company from which the data is compiled is contrary to law and established ITA policy, a fact which the Government also acknowledges. Defendant’s Memorandum at 2-3; see 19 U.S.C. § 1677f(b)(1) (1988); China Nat’l Metals & Minerals Import & Export Corp. v. United States, 11 CIT 859, 861, 674 F.Supp. 1482, 1484 (1987). Hence, the case is remanded to the agency in order for the ITA to recalculate packing costs in accordance with law and its own regulations and policy. Any consequent changes in the dumping margins are to be reported to this Court within sixty (60) days.

III. Correction of Clerical Errors

Plaintiffs’ final complaint is that the ITA improperly failed to correct clerical errors regarding the weights of rivets on certain types of bearings. 5 The contested types are series 6200, 6201, 6202, 6203, 6204 and 6304. Plaintiffs’ Memorandum at 74. In a letter to Commerce on March 13, 1989, Tehnoimportexport explained that it had inadvertently overstated the weights of certain rivets in its questionnaire response and requested that the ITA correct them in the final determinations. RR (Pub.) Doc. 291; RR (Conf.) Doc. 41. Plaintiffs claim that correction of these mistakes could “result in an overall reduction of dumping margins by 5-10%.” Plaintiffs’ Memorandum at 76. The Government’s response is that, since the corrected submissions were received after verification and only eleven days prior to the issuance of the final determinations, “the ITA had no way of determining the accuracy of any calculation errors ... [and] no further adjustments were made for the final determination.” Defendant’s Memorandum at 39.

The ITA’s regulations state that parties will be allowed to rectify a faulty submission provided “the corrected submission is received in time to permit proper analysis and verification of the information concerned; otherwise no corrected submission will be taken into account.” 19 C.F.R. § 353.51(b) (1988). The correction in this *1178 case was submitted more than three months after the ITA’s on-site verification in Romania and eleven days prior to the issuance of the final determinations. Clearly, the correction was tendered at the last minute and the ITA acted according to its regulations in rejecting it. However, if the error was so egregious and so obvious that the failure to correct it was an abuse of discretion and undermined the interests of justice, the Court may remand the case to the ITA for adjustment of the calculations.

The error here concerns the weight of rivets, which are components of antifriction bearings. On December 5-9, 1988, an ITA investigative team conducted an on-site verification of respondent’s questionnaire responses in Romania. During verification, the ITA found that the weight of the rivets on bearing model 6203 2RS had been overstated by ten times in Tehnoimportexport’s response. RR (Conf.) Doc. 44 at 2. Though this mistake was corrected, the ITA staff accountant stated that he thought it was an “isolated incident” and there was no reason to believe the weights of the other rivets were mistaken as well. Id.

While in Romania, the ITA verified data regarding nine different types of plaintiffs’ bearings, which bearings were manufactured in two of plaintiffs’ four plants. RR (Conf.) Doc. 33 at 7. Plaintiffs are contesting the reported rivet weights of five types of bearings, only one of which was among the nine types verified by the ITA in its journey to Romania. Hence, the Government’s claim that it could not amend its calculations because there was no way of verifying the information is disingenuous; the information on which the ITA itself relied for most of the five contested types of bearings was not verified. If, by mere observation of the ITA’s lists of the weights of the factors of production, the error regarding rivet weights was so obvious that to have left it unamended was plain error and an abuse of discretion, it must be corrected.

A review of five of the nine types of plaintiffs’ bearings which the ITA verified reveals that rivets were assigned a weight of 0.00kg in four cases, meaning they weighed less than 0.01kg. RR (Conf.) Doc. 44 at Ex. II. In the fifth case, rivets were reported to weigh 0.08kg. Where they weighed less than 0.01kg, they apparently were not factored into the total weight of the bearings. 6 In the fifth case, the rivets weighed 14% of the total gross weight of the bearing. Id. Furthermore, the Court observed that, among the many other non-contested types, a random sampling of the proportionate weight of the rivets disclosed weights ranging from 10.8% to 19.5% of the total gross weight. 7 Id. Therefore, there appears to be a significant range of proportionate weight assigned to rivets which plaintiffs do not challenge.

Turning now to the contested bearings, the record shows that five of the twenty eight types within the contested series were assigned a weight of 0.00kg. Id. The remaining rivets were reported to comprise between 9% and 22% of the total gross weight of the bearings, except for one type which was reported to constitute 63.2% of the total gross weight of the bearing. Id. In the latter case, model 6203 2RSNR, the rivet was shown to make up more than three times as great a percentage of the total gross weight as the outer ring, which in virtually every other type of bearing forms a much more substantial proportion of the bearing.

Since the calculations of the rivet weights in most of the contested models generally coincide with the relative weights of rivets in the uncontested models, the Court finds that the mistake was not so obvious or egregious that the ITA should have known it was error and corrected it. Given the dilatory nature of plaintiffs’ corrections, and the need to complete the final determinations, the Court finds that the *1179 ITA acted within its discretion in refusing to correct most of plaintiffs’ mistakes. However, the mistake as to model 6203 2RSNR was so obvious that the failure to correct it did constitute an abuse of discretion. Therefore, the Court orders the ITA to correct that error and make any further adjustments as that change may require. Since the case is being remanded for recalculation of packing costs, the Court finds that this additional calculation will not overburden the agency and will advance the interests of justice and yield a more accurate result.

Conclusion

The Court holds that the ITA’s choice of Portugal as the surrogate for Romania and the calculation of overhead costs were supported by substantial evidence and were otherwise in accordance with law. The Court finds that the calculation of packing costs was not in accordance with law because the costs were taken from the confidential data of a firm without its permission.

Lastly, the Court finds that the calculation of the weights of rivets was in accordance with law, except that the calculation of the weight of the rivets on bearing model 6203 2RSNR was not in accordance with law for the reasons discussed above.

Accordingly, this ease is remanded to the ITA for recalculation of packing costs, and recalculation of the rivet weight in the above-mentioned bearing, and the ITA shall make any other adjustments in its determinations as these changes may require. Commerce shall report the results of the remand determination to this Court within sixty (60) days of the date this opinion is entered.

. A revised version of the regulation regarding calculation of foreign market value from state-controlled-economy countries is now found at 19 C.F.R. § 353.52 (1990).

. As a threshold issue, plaintiffs contest the ITA’s use of 1985 data since the period of investigation was October 1987 to March 1988. However, the Government stated that 1985 was the closest year to the investigation for which data was available for all the countries involved. Comparing the data of various countries from different years clearly would have yielded skewed results; thus the Court finds no error in this decision.

. Plaintiffs go to considerable lengths to establish that the labor rates in Yugoslavia were significantly lower than the Portuguese rates. Plaintiffs’ Memorandum at 18-22. Labor rates are not a factor to be considered when assessing *1176 comparability for these purposes, nor should they be. The cost of labor is but one tile of the mosaic which is reflected by the GNP. Selecting or disqualifying a potential surrogate on the basis of its labor rates would be fallacious, and would run contrary to the intent of the statute and regulations.

. The parties agree that Yugoslavian information was available for 69.7% of the factors of production, while Portuguese information was available for 59.6% of the factors of production. Plaintiffs’ Memorandum at 27-28; Defendant’s Memorandum at 25.

. Plaintiffs incorrectly describe the errors as "ministerial." The mistakes in this case were made by plaintiffs, not the agency, hence they were not "ministerial” errors, but simply clerical errors made by a respondent in a submission to the ITA. The fact that the agency then adopted those figures does not convert respondent’s mistake to ministerial error.

. One of these four was the 6203 2RS type, whose weight the ITA corrected. RR (Conf.) Doc. 44.

. The Court randomly sampled bearing models 99502, FH 204-12G, FHS 204-12, and NJ 1040M.