OPINION

CALLAHAN, Circuit Judge:

In 2004, before E.M. entered the fourth grade, he was first tested for a learning disability. Through this lengthy litigation it has been established that E.M. has an auditory processing disorder or a central auditory processing disorder. However, in the fall of 2004 and the spring of 2005, E.M.'s school district, the Pajaro Valley Unified School District ("PVUSD") tested E.M. and determined that, despite his learning disability, E.M. was not eligible for special education services. Subsequently, as a result of further testing procured by E.M.'s parents, PVUSD determined in February 2008 that E.M. did qualify for special education. Shortly thereafter, E.M. moved to another school district which also recognized that he qualified for special education.

Meanwhile, E.M. filed an administrative complaint with the Special Education Division of the California Office of Administrative Hearings. When the Administrative Law Judge ("ALJ") issued a decision in favor of PVUSD, E.M., through his parents (the "Plaintiffs"), filed a complaint in the United States District Court for the Northern District of California alleging that E.M. had been denied a "Free and Appropriate Public Education" as set forth in the Individuals with Disabilities Education Improvement Act of 2004 ("IDEA"). 20 U.S.C. §§ 1400, et seq. The district court granted summary judgment in favor of PVUSD, Plaintiffs appealed, and we issued an opinion affirming in part, reversing in part and remanding. E.M. v. Pajaro Valley Unified Sch. Dist., 652 F.3d 999 (9th Cir.2011). On remand the district court again denied Plaintiffs any relief and further ruled that E.M.'s central auditory processing disorder could not be considered an "other health impairment" under the applicable federal and state regulations. See 34 C.F.R. § 300.7(c)(9) (2005); Cal.Code Regs. Tit. 5, § 3030(f) (2005).1

On this appeal we address three primary issues. First, we conclude that Plaintiffs have failed to show that PVUSD acted unreasonably in determining in 2005 that E.M. did not qualify for special education services under the "specific learning disability" category. See 20 U.S.C. § 1403(1)(A). Second, we conclude that the Department of Education's position that a central auditory processing disorder is eligible for consideration for benefits under the "other health impairment" category merits deference. Finally, we determine that Plaintiffs have failed to show that PVUSD acted unreasonably in not considering E.M. for benefits under the "other health impairment" category in 2005. Accordingly, we affirm the denial of relief to Plaintiffs.

I

A. PVUSD's Initial Assessment of E.M.

E.M. enrolled in PVUSD as a kindergarten student in 1999. Plaintiffs assert that E.M. struggled at school and that PVUSD should have referred him for a special education assessment as early as December 2002, pursuant to its "child find" obligation. This provision of the IDEA requires school districts to identify children with disabilities and to ensure that each child is evaluated and provided appropriate special education services.2

In the summer of 2004, before E.M. entered the fifth grade, Plaintiffs had E.M. tested by psychologist Dr. Roz Wright, who administered the Weschsler Intelligence Scale for Children (3d ed.) and the Woodcock Johnson Tests of Achievement-III ("WISC"). Dr. Wright estimated E.M.'s intelligence quotient ("IQ") to be 104, based on the test. Plaintiffs then requested that PVUSD evaluate E.M. and submitted Dr. Wright's assessment.

In October 2004, PVUSD convened a meeting of E.M.'s Individualized Education Program ("IEP") team. In addition to Dr. Wright's assessment, the IEP team considered the results of additional tests administered by Leslie Viall, PVUSD's psychologist.

Ms. Viall, who had more than fifteen years of experience administering educational assessments of children, testified that she thought the WISC score of 104 was a valid measure of E.M.'s intellectual ability. She stated that in October 2004, she had given E.M. the Kaufman Assessment Battery for Children test ("K-ABC" test) and that E.M. had obtained a higher score of 111. Ms. Viall explained that she administered the K-ABC test because the parents' assessor, Dr. Wright, had recently administered the WISC test and that re-administering the same test less than four months later would have produced an invalid score. When the K-ABC test produced a significantly higher score, Ms. Viall administered a third intelligence test, the Test of Nonverbal Intelligence ("TONI"), on which E.M. scored a 98. Because E.M.'s TONI score was consistent with his performance on the WISC, rather than the higher score on the K-ABC, Ms. Viall determined that 104 was the most reliable measure of E.M.'s intellectual ability.

In 2005, to qualify for special education under the "specific learning disability" (sometimes referred to as "SLD") category in California, a child had to meet three requirements: (1) "there must be a severe discrepancy between intellectual ability and achievement in oral expression, listening comprehension, written expression, basic reading skills, reading comprehension, mathematics calculation, or mathematical reasoning"; (2) "the severe discrepancy must be due to a disorder in one or more of the basic psychological processes and must not be primarily the result of an environmental, cultural, or economic disadvantage"; and (3) "the discrepancy cannot be ameliorated through other regular or categorical services offered within the regular education program." Cal. Educ.Code § 56337 (2005).

PVUSD determined that E.M. had not demonstrated the requisite "severe discrepancy between intellectual ability and achievement." The applicable California regulations defined a severe discrepancy as a difference of at least 22.5 points, adjusted by 4 points, between a child's ability and performance. Faced with three scores, 111 on the K-ABC, 104 on the WISC, and 98 on the TONI, PVUSD opted to use the middle score, 104 on the WISC. E.M.'s lowest standard score in any academic area was 87 on listening comprehension. The discrepancy between 87 and 104 was only 17 points, not sufficient to constitute a severe discrepancy.

B. Plaintiffs' Initial Proceedings Before the Administrative Law Judge and the District Court

When PVUSD denied E.M. special education benefits, Plaintiffs filed an administrative complaint with the Special Education Division of the California Office of Administrative Hearings. A hearing was held, and on May 2006, the ALJ issued a final decision denying Plaintiffs any relief.

Plaintiffs then commenced this action in the United States District Court for the Northern District of California. In October 2007, the district court denied cross-motions for summary judgment and remanded the case to the ALJ. The ALJ was asked to "set forth more completely his reasoning as to why the WISC test was favored over the K-ABC, as well as his approach to evaluating all of the quantitative test data in light of the mixed results of that data."

Meanwhile, Plaintiffs had E.M. tested by Dr. Cheryl Jacques, who estimated his IQ to be 110. PVUSD then retested E.M. for eligibility for special education and found E.M.'s IQ to be 114. This led PVUSD to determine in February 2008 that E.M. was eligible for special education benefits. Shortly thereafter, E.M. moved to the Fullerton Joint Union High School District, which also determined that he was eligible for special education services.

On remand, the ALJ again determined that Plaintiffs were not entitled to any relief. Plaintiffs appealed to the district court.

On August 27, 2009, the district court granted PVUSD's motion for summary judgment. In doing so, the court first agreed with the ALJ that Ms. Viall was credible and her reasoning persuasive.3 The court noted the irony that PVUSD relied on the diagnostic score provided by Plaintiffs, while Plaintiffs claimed that PVUSD should have used its own KABC scores. The district court further agreed with the ALJ that PVUSD had administered multiple tests to E.M. and had used the totality of the results to arrive at its ultimate determination of ineligibility.

The district court further noted that "viewed as a whole, the observational and anecdotal evidence describes a student who was distracted easily but who also responded to various forms of classroom intervention." It opined that had E.M. "been able to complete assignments and homework on a more consistent basis, it seems likely that he would have been a consistently average to above-average performer."

Finally, addressing Plaintiffs' allegation that PVUSD failed to perform assessments with respect to E.M.'s auditory processing, hearing and behavior, the district court commented that at least one auditory processing test was administered by Ms. Viall, and that PVUSD's resource specialist "conducted the Brigance test in both Spanish and English as part of the initial assessment, and this test arguably addresses auditory processing through a subtest involving sentence repetition."4

C. Plaintiffs' Initial Appeal to the Ninth Circuit

Plaintiffs appealed, and we issued an opinion affirming in part and reversing in part. E.M. v. Pajaro Valley Unified Sch. Dist., 652 F.3d 999 (9th Cir.2011). We recognized that "school districts have discretion in selecting the diagnostic tests they use to determine special education eligibility." Id. at 1003. Noting the different tests used to evaluate E.M., we held that a school district, "considering all relevant material available on a pupil, must make a reasonable choice between valid but conflicting test results in determining whether a `severe discrepancy' exists." Id. at 1004.

We did not determine whether PVUSD's choice was reasonable because we determined that the district court had erred in excluding Dr. Jacques's 2007 report.5 Id. at 1006. Accordingly, the district court was instructed on remand to consider whether Dr. Jacques's report, as well as PVUSD's 2008 assessment of E.M., were "relevant to the determination whether PVUSD met its obligations to E.M." Id.

We then held, over a dissent, that Plaintiffs had not waived their assertion that the district court should have considered whether E.M.'s auditory processing disorder qualified him for special education as a child with an "other health impairment." Id. at 1006. We remanded the case to the district court "for a determination whether, during all relevant times, PVUSD met its affirmative obligation to locate, evaluate, and identify E.M. as a child with an other health impairment or a specific learning disability related to his auditory processing disorder." Id. at 1007.

D. The District Court's Opinion on Remand

On remand, the district court read our opinion as holding that "E.M. had a `disorder in a basic psychological process,' specifically, `an auditory processing disorder.'" However, the court found that we had not reached "the issue of whether PVUSD's choice among test scores was reasonable; rather [we] remanded the matter for further consideration of that issue." The district court proceeded to determine whether Plaintiffs had shown that there was a "severe discrepancy" between E.M.'s intellectual ability and his achievement.

The district court noted that all agree that E.M.'s lowest academic standard score was 87. The court then reviewed the three test scores, and concluded that the ALJ's use of the WISC's score of 104, and the consequential finding that there was no severe discrepancy (only 17 points difference), were "thorough and careful" and entitled to deference. The court further conducted its own de novo review of the evidence in the administrative record, and concluded that Plaintiffs had not met their burden of showing that it was unreasonable for PVUSD to use the WISC test score.

The district court agreed with the ALJ that the school psychologist's testimony was more persuasive than Dr. Wright's perspective because of her experience administrating educational assessments to children and her actual knowledge of E.M.6 The court further found that neither Dr. Jacques's report nor the PVUSD's 2008 assessment of E.M. altered its determination that PVUSD's 2005 assessment of E.M. was not unreasonable.

Turning to the issue of whether E.M. could qualify for special education on the basis of having an "other health impairment," the district court noted that 20 U.S.C. § 1401(3)(A)(i) listed nine defined categories such as "intellectual disabilities," "autism," and "specific learning disabilities," and a tenth category described broadly as "other health impairment." At the time of the PVUSD assessment, "other health impairment" (sometimes referred to as "OHI") was defined as follows:

34 C.F.R. § 300.7(c)(9)).

Despite Plaintiffs' contrary assertion, the district court did not find any decisions by courts or hearing officers specifically holding that auditory processing disorders qualify as OHIs. Accordingly, the court approached the question as a matter of first impression, using canons of construction. The court determined that "specific learning disability" and "other health impairment" concerned two different categories of impairment.7 The district court, noting that the statute included a non-exhaustive list, employed the dictionary meaning of "other" as "another," and concluded that because a qualifying auditory processing disorder is a "specific learning disability," "it necessarily follows that an auditory processing disorder cannot at the same time be an `other health impairment.'" The court expressed concern that a contrary finding would render superfluous the requirement of showing severe discrepancy to qualify for benefits under the "specific learning disability" category.8

Having concluded that PVUSD had reasonably determined that Plaintiffs had failed to show a "severe discrepancy" between E.M.'s intellectual ability and academic achievement in 2005, and that E.M.'s auditory processing disorder could not be an "other health impairment," the district court granted judgment in favor of PVUSD. Plaintiffs filed a timely notice of appeal.

II

A district court's compliance with our mandate is reviewed de novo. United States v. Paul, 561 F.3d 970, 973 (9th Cir.2009); United States v. Kellington, 217 F.3d 1084, 1092 (9th Cir.2000). We also review de novo "the district court's decision that the school district complied with the IDEA." K.D. v. Dep't of Education, 665 F.3d 1110, 1117 (9th Cir.2011); N.B. v. Hellgate Elementary Sch. Dist., 541 F.3d 1202, 1207 (9th Cir.2008). However, we give "due weight to judgments of education policy when reviewing state hearings and must take care to not substitute [our] own notions of sound educational policy for those of the school authorities [we] review." K.D., 665 F.3d at 1117 (internal quotation marks omitted). Although "[t]he extent of deference given to the state hearing officer's determination is within our discretion," "[w]e give deference to the state hearing officer's findings particularly when, as here, they are thorough and careful." Id.; see also Union Sch. Dist. v. Smith, 15 F.3d 1519, 1524 (9th Cir.1994).

In K.D., we further reiterated that: (1) we review "the district court's factual determinations for clear error, even when based on the administrative record"; (2) a "finding of fact is clearly erroneous when the evidence in the record supports the finding but the reviewing court is left with a definite and firm conviction that a mistake has been committed"; and (3) the party "challenging the district court's ruling, bears the burden of proof on appeal." 665 F.3d at 1117 (internal quotation marks omitted).

III

A. The District Court Complied with Our Mandate

Initially, we affirm that the district court order is consistent with our mandate. Plaintiffs argue that we had found that E.M. had a specific learning disability, that we held that the district court should apply more of a de novo standard of review, and that the ALJ should not have relied on the testimony of PVUSD's psychologist. We held that E.M. had alleged an auditory processing disorder, but we did not reach the question of whether Plaintiffs had shown that E.M. had qualified for special education benefits under the "specific learning disability" category. Our opinion did not alter the standard of review or make any factual determinations as to any witness's credibility. Rather, we remanded for a determination whether "PVUSD met its affirmative obligation to locate, evaluate, and identify E.M. as a child with an other health impairment or a specific learning disability related to his auditory processing disorder." E.M., 652 F.3d at 1007. The district court did this in compliance with our mandate.

B. Plaintiffs Have Not Shown that PVUSD Unreasonably Found that E.M. Lacked the Severe Discrepancy Between His Achievement and Academic Test Scores Then Required to Qualify for Benefits Under the "Specific Learning Disability" Category

In Schaffer v. Weast, 546 U.S. 49, 56-58, 126 S.Ct. 528, 163 L.Ed.2d 387 (2005), the Supreme Court clarified that under the IDEA, the burden of persuasion rests with the party seeking relief. Here, all appear to agree that E.M.'s achievement score in 2004 was 87, and that then applicable state regulations required a difference of 22.5 points between E.M.'s achievement and ability scores. Thus, to prevail on their claims that E.M. was entitled to special education benefits under the "specific learning disability" category, Plaintiffs have to show that it was unreasonable for PVUSD to use any test results other than E.M.'s score on the K-ABC test. This they have failed to do.

In challenging PVUSD's use of the WISC test, Plaintiffs argue that: (1) the school psychologist, Ms. Viall testified that the K-ABC test was a good cognitive test; (2) although Ms. Viall testified that other colleagues thought the scores on the K-ABC test can be inflated, she "was never able to identify which colleagues and what their credentials were"; and (3) Ms. Viall's belief that E.M.'s score on the K-ABC test was high was a product of her unreasonably low expectations. Plaintiffs assert that at least one authoritative article in a peer reviewed journal identified the K-ABC test as the best predictor of achievement of all cognitive tests. They also submitted a declaration from Dr. Kaufman, who authored both portions of the WISC test and the K-ABC test, favoring the use of the K-ABC test and noting that it was not appropriate to substitute a brief test such as the TONI for comprehensive tests such as the K-ABC. In addition, Dr. Wright, who administered the WISC test, testified that E.M. had been unusually distracted when he took the test.

Plaintiffs also contend that Dr. Jacques's report supports their positions that: (1) E.M. had a long history of auditory processing disorder symptoms; (2) PVUSD was on notice that E.M. had a learning disability; and (3) E.M. had a long history of school failures. Plaintiffs point to Dr. Jacques's statement that she found it "puzzling" that the district did not find E.M. eligible for special education services in 2005.

Plaintiffs have shown that PVUSD could have used E.M.'s K-ABC score, but they have not shown that PVUSD acted unreasonably in using his WISC score. The record shows that Dr. Wright gave E.M. the WISC test in the summer of 2004 and that E.M. scored a 104 on that test. Plaintiffs then asked PVUSD to test E.M. for a learning disability. PVUSD did so. The school psychologist administered the K-ABC test because re-administering the WISC test would not have produced a reliable score. E.M. scored 111 on the K-ABC test. Ms. Viall, noting the disparity between the test scores and having concerns both about KABC test scores in general and E.M.'s score in particular, administered a third test. On the TONI test, E.M. scored 98. PVUSD considered all three test results and then decided to use the middle score, the one submitted by Plaintiffs. This course of action has the indicia of reasonableness.

Plaintiffs' evidence and arguments do not undermine the reasonableness of PVUSD's decision. Plaintiffs presented evidence such as Dr. Kaufman's declaration praising the KABC test, but not evidence that the other two tests were not well-respected tests for cognitive ability or that it was unreasonable to average test scores from different tests. Moreover, none of the later developed information — Dr. Jacques's report, the 2008 assessment, or the later assessments by E.M.'s new school district — bear on PVUSD's 2005 determination because they do not undermine E.M.'s test scores on the WISC and TONI. The later developed evidence does indicate that E.M. had a learning disability in 2004, but PVUSD did not deny that he had a disability. Rather, it denied relief because there was not a 22.5 point discrepancy between E.M.'s tested ability and performance. Subsequently, when E.M. was retested and reevaluated, PVUSD in 2008 determined that he was eligible for special educational benefits.9

In Schaffer, 546 U.S. at 62, 126 S.Ct. 528, the Supreme Court held that the party challenging the district court's ruling bears the burden of proof on appeal, and in K.D., 665 F.3d at 1117, we held that we review the district court's factual determinations for clear error. In E.M., we reiterated that "school districts have discretion in selecting the diagnostic tests they use to determine special education eligibility." 652 F.3d at 1003. Applying these standards, we conclude that the record, developed over at least seven years, does not show that PVUSD unreasonably denied E.M. special education benefits in 2005 under the "specific learning disability" category. Accordingly, the district court's determination of this issue must be affirmed.

C. We Defer to the Department of Education's Position that a Child With a Disability May Be Eligible for Special Educational Benefits Under More Than One Category

Although we held in E.M., 652 F.3d at 1007, that Plaintiffs had not waived their contention that E.M.'s auditory processing disorder could qualify him for special education as a child with an "other health impairment," the merits of this contention had not been previously addressed. In addressing the contention in the first instance, the district court did not have the benefit of the perspective of the Department of Education ("DOE"). On appeal, the DOE has participated as an amicus curiae. Thus, in reviewing the district court's reading of 20 U.S.C. § 1401(3)(A)(i), we have the benefit of the views of the agency charged by Congress with administering the IDEA. See 20 U.S.C. §§ 1406, 1416.10

In 1991, the DOE issued a Joint Policy Memorandum that explained that a child with attention deficit disorder or attention deficit hyperactivity disorder might qualify for special education benefits under one of three categories of the IDEA's definition of "child with a disability" — "other health impairment," "specific learning disability," or "serious emotional disturbance." 18 IDELR 116 (Sept. 16, 1991). In 1994, the DOE's Office of Special Education Programs issued a letter explaining that a child with chronic fatigue syndrome could qualify for special education under the "other health impairment" category or under another category if the child met the criteria for that category. Letter to Fazio, 21 IDELR 572 (Apr. 26, 1994). The DOE asserts that while these documents do not address auditory processing disorders, they reflect the Secretary's position that a particular condition may qualify for benefits under more than one of the IDEA categories.

The DOE asserts that its interpretation of a "child with a disability" is consistent with the history and purpose of the IDEA. Congress first enacted the IDEA in 1970 "to reverse this history of neglect" of disabled children in the United States. Schaffer, 546 U.S. at 52, 126 S.Ct. 528. Congress subsequently expanded the definition of children with disabilities to include specific learning disabilities,11 autism and traumatic brain injury,12 and children between the ages of three and nine who experienced developmental delays.13 These amendments furthered the IDEA's overarching substantive goal "to ensure that all children with disabilities have available to them a free appropriate public education that emphasizes special education and related services designed to meet their unique needs." 20 U.S.C. § 1400(d)(1)(A); Forest Grove Sch. Dist. v. T.A., 557 U.S. 230, 244-45, 129 S.Ct. 2484, 174 L.Ed.2d 168 (2009) (noting the IDEA's express purpose as set forth in the statute and holding that "[a] reading of the Act that left parents without an adequate remedy when a school district unreasonably failed to identify a child with disabilities would not comport with Congress' acknowledgment of the paramount importance of properly identifying each child eligible for services.").

The DOE further claims that its perspective is consistent with a State and local school district's duty under the "child find" provisions of the IDEA. See 20 U.S.C. § 1412(a)(3). The DOE argues that considering a child's condition under only one possible category of disability, when more than one might apply, elevates a myopic concern with the child's specific classification over determining the child's actual educational needs. See Heather S. v. Wisconsin, 125 F.3d 1045, 1055 (7th Cir.1997) (noting "whether Heather was described as cognitively disabled, other health impaired, or learning disabled is all beside the point. The IDEA concerns itself not with labels, but with whether a student is receiving a free and appropriate education."); see also 20 U.S.C. § 1412(a)(3)(B) ("Nothing in this chapter requires that children be classified by their disability so long as each child who has a disability listed in section 1401 of this title and who, by reason of that disability, needs special education and related services is regarded as a child with a disability under this subchapter.").

Where a statute speaks clearly to the precise question at issue, we "must give effect to the unambiguously expressed intent of Congress." Chevron, U.S.A., Inc. v. Natural Res. Def. Council, Inc., 467 U.S. 837, 842-43, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). However, "if the statute is silent or ambiguous with respect to the specific issue, the question for the court is whether the agency's answer is based on a permissible construction of the statute." Id. at 843, 104 S.Ct. 2778.

The Supreme Court has noted that deference may be extended to an agency's perspective not only when it exercises its rulemaking authority, but also when an agency authorized to administer a statute interprets its own regulation or the statute by other means. In Chase Bank USA, N.A. v. McCoy, ___ U.S. ___, 131 S.Ct. 871, 178 L.Ed.2d 716 (2011), the federal agency presented its position in an amicus brief and the Supreme Court held: "we defer to an agency's interpretation of its own regulation, advanced in a legal brief, unless that interpretation is `plainly erroneous or inconsistent with the regulation.'" Id. at 880 (quoting Auer v. Robbins, 519 U.S. 452, 461, 117 S.Ct. 905, 137 L.Ed.2d 79 (1997)). In Capistrano Unified School District v. Wartenberg, 59 F.3d 884 (9th Cir.1995), the DOE clarified its position in a "letter to all chief state school officers," and we held that the agency was "entitled to deference in its interpretation of the statute, because the interpretation is based on a permissible construction of the existing statutory language." Id. at 894.

The Supreme Court has recognized that even where the express delegation of specific interpretive authority is implicit and the agency has not engaged in the process of rulemaking or adjudication, an agency's decision may still be entitled to Chevron deference. United States v. Mead Corp., 533 U.S. 218, 229-30, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001). Moreover, even when an agency's decision does not qualify for Chevron deference, "an agency's interpretation may merit some deference whatever its form, given the `specialized experience and broader investigations and information' available to the agency, ... and given the value of uniformity in its administrative and judicial understandings of what a national law requires." Id. at 234, 121 S.Ct. 2164 (quoting Skidmore v. Swift, 323 U.S. 134, 139-40, 65 S.Ct. 161, 89 L.Ed. 124 (1944)). We need not determine whether DOE's policy letters and amicus brief command Chevron deference as we find its interpretation of the statute persuasive under Skidmore.

Here, as the district court's resort to a canon of construction implicitly admits, Congress' intent is not clear. Title 20 U.S.C. § 1401(3)(A)(i) offers a number of ways in which an individual can qualify as a "child with a disability." Some of the proffered categories are quite specific, for example: "orthopedic impairments," "autism," and "traumatic brain injury." Other categories appear to be relatively broad, such as "intellectual disabilities," "hearing impairments," and "serious emotional disturbance." It is not clear from the statute whether the category "other health impairments" was intended as an alternate category or an additional category. In other words, Congress did not indicate whether "other health impairments" was limited to disabilities that did not fit into any of the other listed categories or included disabilities that might also fit within another category.

Because Congress was not clear, we must consider the DOE's interpretation. We find neither of the grounds advanced in support of a restricted interpretation of the statute to be persuasive. Certainly, the application of a canon of construction should yield to Congress' purpose in passing the IDEA of ensuring that all children with disabilities have available to them a free appropriate public education. 20 U.S.C. § 1400(d)(1)(A). Children with disabilities will be disadvantaged if they have to select one category to the exclusion of any other category. In many instances, neither the child nor the parents will initially know which category encompasses the child's disability. Indeed, compelling a selection of one category seems contrary to the school district's child find duty.

Upon further inspection, the second proffered ground, a fear that allowing a disability to qualify under more than one category will "negate and render superfluous" the distinct requirements for various categories, proves to be unfounded. As the DOE asserts and the district court found, the regulations that defined "specific learning disability" and "other health impairments" in California in 2005 pertained to two different categories of impairment with distinct criteria. Viewing the requirements side by side reveals their distinctiveness.

A severe discrepancy, which all parties agreed in 2005 required a difference of 22.5 points between tested ability and performance, is not the same thing as a condition that limits "strength, vitality or alertness." Also, it appears that a "disorder of the basic psychological processes" is distinct from "chronic or acute health problems." Of course, a "disorder" could also be a "health problem," but presumably a child could be otherwise very healthy and still have a "disorder of the basic psychological processes." The third criterion was also different. An "other health impairment" only required a showing that the condition adversely affects the child's educational performance, whereas a "specific learning disability" required a showing that other educational tools were inadequate. Perhaps, as the district court found, the third criterion for an "other health impairment" might be easier to meet than the third criterion for a "specific learning disability," but the different provisions of the categories' other criteria indicate that an "other health impairment" is not necessarily easier to show than a "specific learning disability." Regardless of the comparative difficulty of qualifying for benefits under the different categories, the two categories definitely have different requirements and appear to address different facets of disabilities. Thus, the fact that a particular child might qualify under both categories is in no way contrary to or inconsistent with Congress' purposes in enacting the IDEA. A contrary position would create the possibility that a child with a disability could be denied special education benefits not because he did not qualify for benefits, but because the child, his parents, or the school district's initial selection of one category barred consideration of a more appropriate category.

The district court, faced with a question of first impression, reasonably turned to a canon of construction to interpret an ambiguous statute. On appeal we have the benefit of a presentation by the DOE, which is charged by Congress with enforcing the IDEA. Because Congress did not clearly address the issue, and because we determine that the DOE's interpretation of the statutes and regulations is reasonable and furthers the overall intent of the IDEA, we defer to the agency's interpretation. Accordingly, we hold that a "child with a disability" may seek to qualify for special education benefits under more than one of the categories listed in 20 U.S.C. § 1401(3)(A)(i).

D. Plaintiffs Have Not, and Cannot, Show that PVUSD Unreasonably Denied E.M. Special Education Benefits in 2005 Under the "Other Health Impairment" Category

Our decision that E.M. may qualify for special education services under the other health impairment category does not answer the question whether he did qualify for services in 2005, or more to the point, whether Plaintiffs can show that PVUSD unreasonably failed to extend special education benefits to E.M. in 2005 based on his "other health impairment." In a usual case, we would remand for the district court or the ALJ to determine such a factual question in the first instance. However, over the last eight years this matter has been before the ALJ twice, before the district court thrice, and is now before us a second time. E.M. has graduated from high school. Accordingly, judicial efficiency and fairness to all concerned recommend that we review the existing record to consider whether a remand would be futile and would needlessly prolong this litigation.

The record is not clear as to when the possibility of E.M. qualifying for educational benefits under the OHI category first arose. There is no indication that this possibility was specifically mentioned by anyone in 2004 or 2005. As we noted in our prior opinion, Plaintiffs' prayer in their January 2006 filing with California's Office of Administrative Hearings included the words "other health impairment." E.M., 652 F.3d at 1006. However, the filing as a whole does not present any evidence or arguments that E.M. met the criteria for qualifying under the other health impairment category.15

A review of the ALJ's decisions show that all parties were focused on E.M.'s auditory processing disorder. The issues presented were broad, including whether PVUSD fulfilled its child find and search and serve obligations, whether PVUSD denied E.M. a free and appropriate public education and whether PVUSD failed to assess E.M. in all areas of suspected disability.

The ALJ's report concentrates on Plaintiffs' claim that E.M. was eligible for services under the "specific learning disability" category, but it also considered Plaintiffs' allegations that PVUSD failed to assess E.M. "[i]n the areas of auditory processing, hearing and behavior." The ALJ found that Ms. Viall administered the Spanish and English versions of a test that included a subtest for auditory processing and that Plaintiffs had failed to establish that PVUSD failed to assess E.M. "in the suspected area of disability of auditory processing."16 As to testing for hearing, the ALJ noted that E.M.'s "initial evaluation report dated October 13, 2004, states that [E.M.] was screened for hearing problems" and that E.M. "passed the hearing screening." The ALJ found that Plaintiffs had failed to show that PVUSD had failed to assess E.M. in the area of hearing and commented: "[w]hile [E.M.] listed this as an issue, he presented no evidence in support of his claim that [PVUSD] failed to screen his hearing."

As noted, the criteria for qualifying for special education benefits under the "other health impairment" category were (1) limited strength, vitality or alertness (2) due to chronic or acute health problems, that (3) adversely affects the child's educational performance. Here, there is no suggestion that E.M. had limited strength or vitality, but his auditory processing disorder might well have limited his "alertness." However, the record, rather than supporting this possible connection, indicates that when E.M. was tested for hearing, the results were normal, and that Plaintiffs failed to proffer any contrary evidence.

It is now too late to develop new evidence as to E.M.'s "alertness" in 2005. The existing evidence suggests that E.M. did not have limited "alertness." Ms. Viall and Nancy Navarro, the resource specialist who assessed E.M., reported that he was alert and responsive during assessment. E.M.'s fourth and fifth grade teacher testified that she believed E.M. was no more distractable than her other students, and his sixth grade teacher reported that after she worked with E.M. on his attention, his attention to tasks improved significantly. Moreover, there was evidence that none of his teachers, nor the speech and language therapists, thought that E.M. had trouble following oral directions. This evidence might not prove that E.M. was alert, but is more than sufficient, absent any contrary evidence from 2004 and 2005, to compel a finding that in 2005 PVUSD did not unreasonably fail to diagnose E.M. as having limited alertness.

Limited alertness is the criteria for eligibility for benefits under the "other health impairment" category that E.M. was most likely to meet. Because Plaintiffs have failed to show that PVUSD unreasonably failed to diagnose limited alertness, we need not consider whether there was evidence that E.M. met the other criteria for eligibility under the OHI category. Nonetheless, we note that our review of the record reveals nothing to suggest that E.M. suffered from chronic or acute health problems. Furthermore, even assuming that E.M. had limited alertness, there is scant evidence that this, rather than other causes, such as his failure to complete his homework, adversely affected his educational performance.

IV

We can hope that today, with the evolution of the law and improved testing, a child with a disability, such as E.M., will not have to wait three years to be determined eligible for special educational services. However, our task is to determine whether PVUSD's past determinations were unreasonable. We conclude that they were not.

PVUSD was not insensitive to Plaintiffs' request that E.M. be assessed. It formed an IEP team and had E.M. tested and evaluated. Morever, PVUSD did not deny E.M. benefits on the basis of some subjective evaluation or opinion, but because E.M.'s test scores did not show the severe discrepancy between his ability and achievement then required. Plaintiffs have not shown that PVUSD's decision was unreasonable.

We do agree with Plaintiffs and the Department of Education that a child with an auditory processing disorder, such as E.M., may seek special education services pursuant to more than one of the categories listed in 20 U.S.C. § 1401(3)(A). The DOE is charged by Congress with administering the IDEA and its interpretation of the statute is permissible and furthers Congress' intent in enacting the IDEA. Accordingly, we defer to its position. See Mead Corp., 533 U.S. at 234-35, 121 S.Ct. 2164.

Finally, while we recognize that a child with an auditory processing disorder may qualify for special educational services under the "other health impairment" category, we conclude that Plaintiffs cannot show that PVUSD was unreasonable in 2005 in failing to diagnose E.M. under the OHI category. Our review of the record reveals a dearth of any evidence that in 2005 E.M.'s auditory processing disorder manifested itself by limiting E.M.'s alertness or that the disorder was due to chronic or acute health problems. Plaintiffs over the last eight years have broadly challenged PVUSD's alleged failure to fulfill its child find obligations and failure to assess E.M. in all areas of suspected disability. We doubt that Plaintiffs have any additional evidence concerning E.M.'s "other health impairment" in 2005 and question whether such evidence, if it exists, could now be admitted.

Accordingly, we AFFIRM the district court's judgment in favor of PVUSD.

Each side shall bear its own costs.