NIES, Circuit Judge.
Fischer & Porter Co., Inc. (F & P) appeals from the International Trade Commission’s final determination in In the Matter of Certain Unitary Electromagnetic Flowmeters, Investigation No. 337-TA-230, U.S.I.T.C.Pub. 1924 (Nov. 17, 1986) (Comm’n op.), holding that the importation and sale of certain unitary, flangeless, electromagnetic flowmeters made in West Germany by Krohne Messtechnik GmbH & Co. KG and sold in the United States by Krohne America, Inc. (collectively, Krohne) did not violate section 337 of the Tariff Act of 1930, as amended, 19 U.S.C. § 1337 (1982).1 Although the Commission found that the subject imports infringed F & P’s U.S. Patent No. 4,420,982 (’982) covering certain flowmeters, it found no effect or tendency to substantially injure a domestic industry as required for relief under section 337. Our jurisdiction over the appeal is found in 19 U.S.C. § 1337(c) (1982) and 28 U.S.C. § 1295(a)(6) (1982). We affirm the Commission’s finding of no injury and, thus, its ultimate determination of no violation of section 337. The Commission’s decision with respect to validity and infringement of the ’982 patent is vacated.
I
Following proceedings on a section 337 complaint filed by F & P, an Administrative Law Judge (AU) in an initial decision ruled that there was an unfair trade practice in the importation and sale of certain unitary, electromagnetic flowmeters by Krohne by reason of infringement of claims 1 through 5 of F & P’s 1982 patent.2 The AU also found that such unfair trade practice had the effect or tendency to destroy or substantially injure an industry, efficiently and economically operated, in the United States. Accordingly, the AU determined that section 337 had been violated by the Krohne imports and that an exclusion order should be entered.
Krohne petitioned the Commission to review the AU’s decision. The Commission chose to review only the issue of whether Krohne’s infringing flowmeters had the effect, or tendency, to destroy or substantially injure a domestic industry and the majority of the Commission (two members dissenting) determined that F & P had not established such injury, rejecting the AU’s contrary finding. 51 Fed.Reg. 40,270 (Oct. 30, 1986). Consequently, F & P obtained no relief under section 337 and appeals to this court for reinstatement of the AU’s decision.
II
F & P set forth the issue on appeal as whether the majority decision of the Commission was arbitrary, capricious, or otherwise not in accordance with law “when it reversed the AU.” It argues that the AU’s findings are unassailable under Fed.R.Civ.P. 52, that a jury could have found as did the AU, and that such a verdict would have survived a JNOV motion.
Throughout its briefs, F & P misdirects the focus of the appellate inquiry in disregard of the precedent of this court. In Corning Glass Works v. United States International Trade Commission, 799 [1577] F.2d 1559, 230 USPQ 822 (Fed.Cir.1986), this court undertook to set out in detail the appellate function in connection with appeals from the International Trade Commission and, particularly, in reviewing an injury determination. Id. at 1564-68, 230 USPQ at 825-28. Contrary to F & P’s understanding, we do not “review” the correctness of the AU’s initial decision or the correctness of the Commission’s “reversal” under Rule 52 or otherwise. The statute, 19 U.S.C. § 1337(c), directs that on appeal to this court, this court must review the “final determination of the Commission ... in accordance with chapter 7 of title 5 [i.e., the Administrative Procedure Act (APA) ].” The AU’s decision is, of course, part of the record and will be accorded “such probative force as it intrinsically commands.” Universal Camera Corp. v. NLRB, 340 U.S. 474, 495, 71 S.Ct. 456, 468, 95 L.Ed. 456 (1951); see also SSIH Equip. S.A. v. United States Int’l Trade Comm’n, 718 F.2d 365, 382, 218 USPQ 678, 692 (Fed.Cir.1983) (Nies, J., additional views). However, the issue before us is whether the “no injury” finding of the majority is arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law or unsupported by substantial evidence. 5 U.S.C. § 706 (1982). That the contrary determination of the ALT and the dissenting members of the Commission would be sustained if it were the decision under review does not mean that the majority’s determination must be overturned. Both decisions may be reasonable based on the entirety of the record, a concept some litigants find difficult to accept. In sum, this court may not substitute its judgment for the Commission’s final determination on the ground that the court believes the AU’s and the minority’s view is “more reasonable” than the majority’s view. Such an appellate evaluation is patently not a basis for reversal under 5 U.S.C. § 706.
Ill
COMMISSION’S INJURY ANALYSIS
The Commission’s analysis of the issue of injury begins with the correct statements of law that the section 337 requirement for injury to a domestic industry does not automatically follow from proof of infringement of an intellectual property right, citing Corning Glass, 799 F.2d at 1566, 230 USPQ at 826-27; Textron, Inc. v. United States International Trade Commission, 753 F.2d 1019, 1028, 224 USPQ 625, 631 (Fed.Cir.1985); [accord Akzo N.V. v. United States International Trade Commission, 808 F.2d 1471, 1486, 1 USPQ2d 1241, 1250 (Fed.Cir.1986), cert. denied, — U.S. -, 107 S.Ct. 2490, 96 L.Ed.2d 382 (1987) ]; the holder of an intellectual property right has a smaller quantum of proof of injury than would be required in a nonintellectual property-based case, Textron, 753 F.2d at 1029, 224 USPQ at 632; Bally/Midway Mfg. v. United States Int’l Trade Comm’n, 714 F.2d 1117, 1124, 219 USPQ 97, 102 (Fed.Cir.1983); the determination of injury to an industry is the type of question particularly within the expertise of the Commission, Corning Glass, 799 F.2d at 1568, 230 USPQ at 828; a determination of injury is dependent on the facts of the particular case, and, thus, not controlled by precedent, id.; and, finally, because the relevant market in the instant investigation contains noninfringing substitutes, it could not be assumed that sales by Krohne would have been made by F & P; rather F & P had to prove a causal connection between infringing imports and lost sales.
Reviewing the evidence, the Commission found that F & P did not carry its burden of proof that Krohne’s infringement caused significant injury to F & P or a tendency to injure. The Commission summarized the substance of its findings as follows:
Complainant [F & P] has not provided the Commission with specific domestic industry data, despite the fact that such data is within its control, and complainant does not explain why presenting the data in usable form to the Commission would impose any more of a hardship on it than it would on any complainant with a diverse product line. In addition, no shift of market share from complainant has been established in this investigation. Respondents [Krohne] were not found to [1578] be underselling. There were few lost sales established, and it is impossible to determine from the record whether those sales are significant. While excess domestic capacity has been established, that finding alone cannot support a determination of substantial injury.
Comm’n op. at 21.
IV
EFFECT TO SUBSTANTIALLY INJURE
To establish an effect to substantially injure the domestic industry, F & P sought to prove that F & P’s sales and profits of flowmeters within the scope of the domestic industry increased from 1982 to 1984 and declined in 1985, while the value of its inventory unfavorably rose in 1985; that Krohne’s market share increased during the period; and that F & P lost sales to Krohne in an amount which represented a significant percentage of F & P’s market. The Commission rejected each of these contentions for failure of proof. F & P argues that the failure of proof results from the Commission (1) improperly discounting, as lacking credibility, the testimony of F & P’s sales manager with respect to F & P’s sales volume of the patented flowmeters; and (2) improperly rejecting a 1983 business report (the Frost & Sullivan report) used as a basis for estimating a shift to Krohne of market share. In this connection, F & P asserts that the Commission failed to follow the precedent of this court in Textron, which held that a “lesser quantum” of proof of injury is required when unfair acts are based on infringement of intellectual property rights. We address each of these arguments in turn.
A
F & P manufactures two product lines of flangeless, magnetic flowmeters under the '982 patent, the MINI-MAG and K-MAG. These lines of flowmeters are made with flow tubes ranging in size from 1/10" — 4" in diameter, and F & P had originally asserted that all sizes were within the patent’s scope. Prior to the evidentiary hearing, however, F & P stipulated that only its MINI-MAG and K-MAG flow-meters with flow tubes of diameters greater than two inches were made in accordance with the ’982 patent, and, thus, defined the “domestic industry.” Despite the stipulation, F & P submitted proof of sales, profitability, and inventory for its total MINI-MAG and K-MAG product lines for the years 1982-1985. These data show that F & P’s dollar and unit sales of, and profits from, all sizes of MINI-MAG and K-MAG flowmeters increased from 1982 to 1984 and decreased in 1985, while the value of year-end inventory increased from 1984 to 1985.
None of F & P’s data was specific to the “domestic industry,” i.e., MINI-MAG and K-MAG meters with 2”-4” diameter flow tubes. F & P relies on the following testimony of its sales manager, Mr. Dimm, to convert the unit and dollar figures for all sizes to unit and dollar figures for the 2”-4” models:
Q Mr. Dimm, do you know what the approximate percent of your unit sales are made of MINI-MAGs and K-MAGs that are two inches or greater in diameter?
A Of just the MINI-MAG type products?
Q Yes.
A Larger than two inch?
Q Two inches and up.
A I don’t know an exact figure, but I would estimate that approximately 60 percent.
Q Would the same number hold for K-MAG?
A I couldn’t say for K-MAG at this point in time. We have had it in the market only since early 1985.
Applying Dimm’s sixty percent estimate across the board to all of its total dollar and unit figures, F & P “constructed data” for the 2”-4” meters for the 1982-1985 time period. Not surprisingly, sixty percent of each year’s total figures shows the same dollar and unit trend in sales, the same 1985 downward turn in sales and profits, and the same increased value of inventory in 1985 for 2”-4” units as for all units.
[1579] Having reviewed the entirety of its opinion, we can find nothing suggesting that the Commission made an adverse credibility determination with respect to the Dimm testimony. Specifically, the Commission stated:
We find that the record does not support the assumption that trends in the domestic industry and F & P’s entire product line are identical. Moreover, because F & P’s estimate of sales of the patented product concerned unit sales, we find that F & P’s estimate cannot be used to construct valid domestic industry data from total product line data provided by F & P in dollar amounts.
Comm’n op. at 9 (emphasis in original). F & P points to no evidence specifically directed to the trend in the sales of 2”-4” devices so that it would be appropriate to apply sixty percent across the board. Nor does the record contain any evidence that sixty percent of units translates to sixty percent of dollar sales. Indeed, the evidence on price differentials with respect to sizes would lead to the conclusion that it does not.
We conclude that F & P’s argument that the Commission rejected Dimm’s testimony as lacking credibility is without foundation. The Commission accepted Dimm’s testimony, but not the extrapolations therefrom that sixty percent was applicable to each year and that sixty percent of total sales dollars and profits represented sales and profits attributable to the patented products.
In addition, the Commission found that F & P had the data on sales figures for the patented products and, thus, F & P could not assert that it was impossible to present better information. The Commission stated:
[We] note[ ] that complainant [F & P] has expressed little interest in remedying the deficiencies of its domestic industry data. In its notice of review the Commission asked the parties to address the issue of whether a remand to the AU to take additional evidence would be helpful in concluding this investigation. 51 Fed. Reg. 33933, September 24, 1986. Complainant did not respond that such a remand would be helpful. Complainant reiterated in its brief on review that it does not keep its business records segregated by unit size and that it is not required to provide the Commission with such segregated data.
Comm’n op. at 11. Although F & P could have supplied the domestic industry data in its possession, it refused to do so even when given a second chance by the Commission. Thus, F & P’s assertion that the Commission’s finding of no injury turned on an improper credibility determination is rejected.
B
To show a shift in market share in the relevant market, which comprises the “domestic industry” and competitive devices, F & P relied on data from the Frost & Sullivan report on the flowmeter industry. That data indicated F & P’s share of the United States market in magnetic flow-meters, flanged and flangeless, for the year 1983. F & P then attributes this 1983 market share to the years 1984-85, relying on the assumption that its market share remained constant and disregarding any impact the ’982 patent (issued on December 20, 1983 and not reflected in the 1983 data) might have on market share. Moreover, F & P uses the report’s data to fix its market share in the relevant market. The report’s data were not broken down, however, with respect to flowmeter sizes or flangeless models, and did not include non-magnetic flowmeters which are competitive with F & P’s patented devices. To determine its market share in the relevant market, F & P thus further assumes that the report’s market share figure, covering magnetic, flanged, and flangeless flowmeters, fixes F & P’s market share in both the domestic industry, which includes only magnetic, flangeless, 2”-4” flowmeters, and the relevant market, which comprises the domestic industry and competitive devices.
With its share of the relevant market in hand, F & P turns to calculating the total relevant market and Krohne’s share of that market during the relevant years, as re[1580] quired to show a shift in market share. F & P provided data for its own unit sales of magnetic, flangeless meters — of all sizes— and applied the Dimm estimate to that data to obtain F & P’s unit sales of magnetic, flangeless, 2”-4” flowmeters (the domestic industry). Dimm’s estimate was an indefinite approximation and referred only to the MINI-MAG, not the K-MAG, product. Nevertheless, by dividing F & P’s unit sales of magnetic, flangeless, 2”-4” flow-meters by F & P’s previously determined market share, F & P calculated the total unit sales in the domestic industry. F & P then assumed that the total unit sales in the relevant market was the same figure.
F & P was now able to calculate Krohne’s market share. Krohne provided unit sales data for magnetic, flangeless, 2”-4” flowmeters. Krohne’s share of the relevant market is then simply the unit sales figure provided by Krohne divided by the total unit sales in the relevant market. Thus, by a comparison of F & P’s and Krohne’s market shares over the relevant time period, a shift appears in their respective market shares in the relevant market.
The Commission rejected this “proof” of a shift. However, this came not from a rejection of the Frost & Sullivan report as evidence, but from a refusal to make the assumptions which had to be made to show a shift in market share of the relevant market. In view of the assumption upon assumption apparent in F & P’s analysis, the Commission’s finding that the record does not support a finding of a shift in market share from F & P to Krohne must stand.
C
F & P argues that, even without the Frost & Sullivan report, it proved injury in that Krohne’s sales are “significant” because they amount to a large figure over a four-year period. However, out of that number, F & P established that it actually lost only three sales. The Commission found that it could not determine if that loss was significant without data on the relevant market. As this court held in Corning Glass, 799 F.2d at 1569, 230 USPQ at 828, the significance of sales cannot be determined in a vacuum. Thus, the extrapolations built from the Frost & Sullivan report were necessary but were no more acceptable in this connection than to show a shift in market share.
D
To overcome the above defects in its evidence, F & P attempts to invoke the ruling in the Textron decision that a lesser “quantum of proof” of injury is required in a patent infringement investigation than in other situations. 753 F.2d at 1029, 224 USPQ at 632. F & P urges that, contrary to Textron, the Commission imposed an unreasonably high burden of proof when it rejected the constructed data based on the 1983 Frost & Sullivan report and the Dimm testimony.
F & P’s reliance on Textron is misplaced. The Textron decision spoke of “quantum of proof” in the sense that fewer incidents of lost sales might be sufficient to establish injury. The ruling was not directed to lessening the quality and quantum of evidence necessary to establish the facts on which the injury determination is based. The quality and quantum of evidence necessary to establish the facts is fixed by statute.
Under 5 U.S.C. § 556(d), F & P had the burden of proof of injury as “the proponent of ... [an] order.” That section further precludes the Commission from issuing an order unless “supported by and in accordance with the reliable, probative, and substantial evidence.” 5 U.S.C. § 556(d) (1982). The phrase “in accordance with ... substantial evidence” in section 556(d) is distinctly different from the “unsupported by substantial evidence” standard which this appellate court must apply in this appeal. 5 U.S.C. § 706(2)(E) (1982). In the landmark ease of Steadman v. Securities & Exchange Commission, 450 U.S. 91, 99-100, 101 S.Ct. 999, 1006-07, 67 L.Ed.2d 69, reh’g denied, 451 U.S. 933, 101 S.Ct. 2008, [1581] 68 L.Ed.2d 318 (1981), the Supreme Court discussed the difference as follows:
[T]he adjudicating agency must weigh the evidence and decide, based on the weight of the evidence, whether a disciplinary order should be issued. The language of § 7(c), therefore, requires that the agency decision must be “in accordance with” the weight of the evidence, not simply supported by enough evidence “ ‘to justify, if the trial were to a jury, a refusal to direct a verdict when the conclusion sought to be drawn from it is one of fact for the jury’ ” Consolo v. FMC, 383 U.S. 607, 620 [86 S.Ct. 1018, 1026, 16 L.Ed.2d 131] (1966), quoting NLRB v. Columbian Enameling & Stamping Co., 306 U.S. 292, 300 [59 S.Ct. 501, 505, 83 L.Ed. 660] (1939). Obviously, weighing evidence has relevance only if the evidence on each side is to be measured against a standard of proof which allocates the risk of error. See Addington v. Texas, 441 U.S. 418, 423 [99 S.Ct. 1804, 1807, 60 L.Ed.2d 323] (1979). Section 10(e), by contrast, does not permit the reviewing court to weigh the evidence, but only to determine that there is in the record “ ‘such relevant evidence as a reasonable mind might accept as adequate to support a conclusion,’ ” Consolo v. FMC, supra, at 620, [86 S.Ct. at 1026] quoting Consolidated Edison Co. v. NLRB, 305 U.S. 197, 229 [59 S.Ct. 206, 217, 83 L.Ed. 126] (1938). It is not surprising, therefore, in view of the entirely different purposes of § 7(c) and § 10(e), that Congress intended the words “substantial evidence” to have different meanings in context.
The Court further held that in enacting section 556(d) Congress was primarily attempting to eliminate agency decisionmaking premised on evidence of poor quality (irrelevant, immaterial, unreliable, and non-probative) and of insufficient quantity (not substantial, i.e., not a preponderance). Id. at 102, 101 S.Ct. at 1007.
Textron could not and did not change the burden of proof of facts under section 556(d) established in Steadman. Thus, F & P was required to support its assertion of injury with evidence of good quality and sufficient quantity to amount to a preponderance. The Commission simply found its evidence not sufficiently probative. Contrary to F & P’s argument, nothing said by the Commission indicates that the Commission was looking for mathematical precision. Under section 556(d), however, the Commission could not rely on unsupported assumptions, speculations, and extrapolations. Nor could it place the burden on Krohne to disprove injury to F & P by compiling the actual data from F & P’s invoices. The burden was on F & P to prove injury. F & P had the data and it is no excuse for its failure to put it in usable form that it would have been an expensive undertaking.
Accordingly, we reject F & P’s arguments that the Commission departed from the holding of this court in Textron with respect to the quantum of proof.
V
TENDENCY TO SUBSTANTIALLY INJURE
We have considered F & P’s analysis on the tendency to substantially injure and are not persuaded of error in the Commission’s contrary finding. F & P argues that “if U.S. demand increased, Krohne would likely increase its exports to the U.S.” The Commission, in finding no proof of a tendency to substantially injure, noted, inter alia, that there is no evidence of record that foreign markets are saturated or that the U.S. market is being targeted. Comm’n op. at 22-23. F & P argues that circumstances other than foreign market saturation or U.S. market targeting can reasonably lead to a finding of a tendency to substantially injure. We agree, but “mere suspicions” that something will occur are not enough under 5 U.S.C. § 556(d). NLRB v. Columbian Enameling & Stamping Co., 306 U.S. 292, 300, 59 S.Ct. 501, 505, 83 L.Ed. 660 (1939).
[1582] VI
Having considered all of F & P’s arguments, we see no basis to overturn the Commissioner’s finding of no effect or tendency to substantially injure. Accordingly, the final determination of the Commission that F & P failed to establish a violation of 19 U.S.C. § 1337 is affirmed. The patent issues being mooted by affirmance on that ground, we vacate the portion of the Commission’s decision concerning the patent issues.
AFFIRMED IN PART; VACATED IN PART.