BIBAS, Circuit Judge.

Catherine M. Reisman, Esq. [ARGUED], Judith A. Gran, Sarah E. Zuba, Reisman Carolla & Gran, 19 Chestnut Street, Haddonfield, NJ 08033, Counsel for Appellants

Kevin A. Golembiewski, Esq., David Berney, Berney & Sang, 1628 John F. Kennedy Boulevard, 8 Penn Center, Suite 1000, Philadelphia, PA 19103, Ellen M. Saideman, Esq., Selene Almazan-Altobelli, 7 Henry Drive, Barrington, RI 02806, Counsel for Amici Curiae Council of Parent Attorneys and Advocates, Education Law Center, New Jersey Special Education Practitioners, and National Center for Learning Disabilities in Support of Appellants

Robert M. Abrahams, Esq., Schulte Roth & Zabel, 919 Third Avenue, New York, NY 10022, Eric A. Bensky, Esq., Schulte Roth & Zabel, 1152 15th Street, NW, Suite 850, Washington, DC 20005, Counsel for Amici Curiae Former Officials of the Department of Education, National Center for Youth Law, Judge David L. Bazelon Center for Mental Health Law, and Disability Rights Network of Pennsylvania in Support of Appellants

Karl A. Romberger, Jr., Esq. [ARGUED], Sweet Stevens Katz & Williams, 331 East Butler Avenue, P.O. Box 5069, New Britain, PA 18901, Counsel for Appellee

Mark G. Morford, Esq., Kevin M. McKenna, Esq., Nicole D. Snyder, Esq., Sarah B. Dragotta, Esq., Latsha Davis & McKenna, 350 Eagleview Boulevard, Suite 100, Exton, PA 19341, Counsel for Amici Curiae 21st Century Cyber Charter School, Agora Cyber Charter School, Arts Academy Charter School, Chester Community Charter School, Collegium Charter School, New Foundations Charter School, PA Distance Learning Charter School, Pennsylvania Leadership Charter School, and Pennsylvania Virtual Charter School in Support of Appellee

Craig D. Ginsburg, Esq., Anne E. Hendricks, Esq., Levin Legal Group, 1800 Byberry Road, 1301 Masons Mill Business Park, Huntingdon Valley, PA 19006, Counsel for Amici Curiae Pennsylvania School Boards Association in Support of Appellee

Before: GREENAWAY, JR., RESTREPO, and BIBAS, Circuit Judges

OPINION OF THE COURT

BIBAS, Circuit Judge.

When schools use their expertise to address each child's distinct educational needs, we must give their judgments appropriate deference. Endrew F. ex rel. Joseph F. v. Douglas Cty. Sch. Dist. RE-1 , ––– U.S. ––––, 137 S.Ct. 988, 1001-02, 197 L.Ed.2d 335 (2017). The Individuals with Disabilities Education Act (IDEA) "requires an educational program reasonably calculated to enable a child to make progress appropriate in light of the child's circumstances." Id. at 1001. But we may not "substitute [our] own notions of sound educational policy for those of the school authorities which [we] review." Id. (internal quotation marks omitted). Our precedents already accord with the Supreme Court’s guidance in Endrew F. , so we continue to apply them. Under both Endrew F. and our precedents, Downingtown Area School District followed the law in educating K.D. So we will affirm.

I. BACKGROUND

A. Facts

1. Kindergarten and testing. K.D. attended public school in the Downingtown Area School District from preschool through the first semester of third grade. Halfway through kindergarten, Downingtown assigned an Instructional Support Team to monitor K.D.'s educational progress and give her extra support.

After kindergarten, over the summer of 2012, K.D., her parents, and her teachers completed a battery of tests. The psychologist found that K.D. had a low-average IQ (87) and Attention Deficit Hyperactivity Disorder (ADHD). K.D. scored below average in early reading skills, basic reading, total reading, writing, and math, and average in oral language. She could not read any common grade-level sight words nor the oral-reading passages. In writing letters of the alphabet, she scored in the first percentile. She scored much lower than average on executive function, and struggled with impulsivity and organization.

2. The first IEP . After completing these tests, Downingtown offered K.D. an individualized education program (IEP) in 2012, at the start of first grade. It set measurable goals for letter naming, letter sounds, writing, rhyming, reading comprehension, math, and on-task behavior. The program's specially designed instructions provided for audiobooks, extra time for tests and quizzes, a quiet place to take tests, and using visual images and thinking aloud to promote recall of text. It also provided for three hours of learning-support instruction every school day.

3. First grade . K.D. started first grade. She spent part of her time with the regular teacher and part with her special education teacher, Ms. Smith. Ms. Smith was unhappy with K.D.'s progress in naming and sounding out letters, so she changed K.D.'s homework and sent home a packet of stories to help her improve. Because K.D.'s visual and motor skills were lagging, Downingtown asked for an occupational-therapy screening. And to keep K.D. from regressing over the summer, the school changed the first IEP, arranging for three hours of academic support, three days per week, during July.

4. The second IEP . The summer before second grade, in 2013, Downingtown developed K.D.'s second IEP. It increased her baselines for letter naming, letter sounds, reading comprehension, writing, and math calculation. Her goals for writing letters, rhyming, math facts, and on-task behavior remained unchanged. Downingtown added "an evidence based multi sensory reading and writing program" for two and a half hours. JA 98. It retained her supplemental learning support and extended-school-year services.

5. Summer before second grade . K.D.'s parents were dissatisfied with K.D.'s summer schooling. They asked about testing K.D. for dyslexia and dysgraphia, and about the Wilson reading program for struggling readers. Ms. Smith replied that school psychologists do not diagnose those conditions, but offered to put them in touch with the school psychologist anyway. She also said that Downingtown did not (yet) offer Wilson before middle school, but that K.D. would receive a similar program geared toward elementary-school students.

6. Second grade and updating the second IEP . Just as K.D. started second grade, Downingtown switched to the Wilson program for kindergarteners through third graders. K.D. mastered 4 of 11 units in Wilson's Level 1 by the end of second grade. Downingtown also updated K.D.'s second IEP to reflect the results of her occupational-therapy evaluation.

7. The third IEP . At the end of second grade, in 2014, Downingtown developed K.D.'s third IEP, for third grade. Reflecting K.D.'s progress, it increased her goals or baselines for letter naming, reading, writing, comprehension, and on-task behavior. And it kept her occupational-therapy goals and specially designed instruction.

K.D.'s parents were dissatisfied with the new IEP, so they met with school officials to discuss it. They did not reject it after the meeting, so the IEP took effect. They also hired Ms. Smith to tutor K.D. over the summer, while K.D. continued in the school's extended-school-year program.

8. Dr. Kelly's independent evaluation . In July 2014, K.D.'s parents hired Dr. Karen Kelly to do a neuropsychological evaluation. Dr. Kelly diagnosed K.D. with dyslexia, ADHD, "mathematics disorder, ... organizational deficits, memory impairment, [and] executive function[ ] impairments." JA 192. She also found that K.D. was reading below first-grade level.

Beyond diagnosing K.D., Dr. Kelly criticized Downingtown's programming. She stated that K.D.'s poor achievement showed that K.D. could not benefit from the school's program, evidencing the school's "global disregard for this level of impairment." JA 191. K.D.'s parents did not immediately notify Downingtown of the evaluation.

9. Third grade and updating the third IEP . To prepare for third grade, Downingtown tested K.D. again. She had advanced in all aspects of reading and writing. It also tested her vision and found that she qualified for vision services. Two months after the fact, K.D.'s parents told Downingtown that Dr. Kelly had evaluated K.D. and that they had hired an educational advocate.

Downingtown met with K.D.'s parents to discuss the up-coming school year. It then performed more evaluations, added vision services, and offered a one-on-one aide. K.D.'s parents rejected the aide, for fear that it would make K.D. stand out. The latest IQ test showed that K.D.'s IQ had risen into the average range.

Downingtown presented K.D.'s parents with the IEP as modified. They checked both the boxes for approving and for disapproving the IEP. They did not explain which parts they disliked, but expressed both hope for progress and concern about how appropriate her programming was.

10. The fourth IEP and withdrawal . In the middle of third grade, Downingtown’s team met again. Based on their own and Dr. Kelly's evaluations, as well as K.D.'s progress, Downingtown increased her goals for writing and on-task behavior. It added new goals for math, reading fluency, and reading comprehension. It added an hour of direct math instruction, forty-five minutes of direct writing instruction, and fifty-five minutes of "multisensory reading instruction" per day, all in "evidence based" programs. JA 104. Downingtown also took Dr. Kelly's advice to replace Wilson with "SRA/Corrective Reading and FastForward," two other "research-based programs that provide phonics and reading comprehension instruction." Id.

In December 2014, midway through third grade, Downingtown offered K.D.'s parents the new program. But they rejected it, withdrew K.D., and placed her in private school.

B. Procedural History

1. The administrative hearing . K.D.'s parents filed a complaint with Pennsylvania's Office of Dispute Resolution, seeking reimbursement for private-school tuition. They argued that Downingtown had denied K.D. a free appropriate public education under the IDEA. They also alleged that, by not adequately addressing K.D.'s needs, Downingtown had discriminated against K.D. based on her disability, in violation of the Rehabilitation Act and the Americans with Disabilities Act (ADA).

The administrative officer found that the IEPs were adequate and that Downingtown had provided K.D. with a free appropriate public education. Because the officer decided the case before Endrew F. , he applied the Third Circuit's meaningful-benefit test. See Ridley Sch. Dist. v. M.R. , 680 F.3d 260, 268 (3d Cir. 2012) (explaining meaningful-benefit test). He found that Downingtown remained aware of K.D.'s slow progress and kept trying to improve her programming in response to K.D.'s performance and Dr. Kelly's report. And while it repeated some goals, Downingtown "did not simply hand out the same IEP year after year," but repeated foundational skills where needed to address "the challenge of teaching even fundamental skills to [K.D.]." JA 115. Downingtown had explained clearly why it chose the programs it did and how they addressed K.D.'s needs.

The officer disagreed with Dr. Kelly's criticisms of Downingtown. He found that Downingtown had acted reasonably in giving the Wilson program time to work and in pursuing occupational therapy and vision services. He held that all the IEPs were "reasonably calculated to provide a meaningful educational benefit to [K.D.] when they were issued." JA 116. So he rejected the claims based on the IDEA. The ADA and Rehabilitation Act claims rested on the same theory as the IDEA claim, so the officer rejected those claims as well.

2. The District Court . K.D.'s parents then filed a complaint in District Court, bringing the same three claims. K.D. v. Downingtown Area Sch. Dist. , No. 16-0165, 2016 WL 4502349, at *3 (E.D. Pa. Aug. 29, 2016). They moved to supplement the administrative record with new evidence, including AIMSweb reports comparing K.D. with her peers, Downingtown's interrogatory answers, and a Wilson program teacher's manual. Id. at *2-3.

The District Court denied the motion. It reasoned that the ADA and Rehabilitation Act claims rested on the same grounds as their IDEA claim, that the new evidence was only minimally relevant, and that K.D.'s parents should have introduced the evidence before the hearing officer. Id .

On cross-motions for judgment on the administrative record, the District Court granted judgment for Downingtown. K.D. v. Downingtown Area Sch. Dist. , No. 16-0165, 2017 WL 3838653, at *13 (E.D. Pa. Sept. 1, 2017). Because Endrew F. came down before it decided the case, the District Court first held that Endrew F. "simply confirm[ed] the standard that has been used in the Third Circuit for years." Id. at *7 n.7. It went on to apply Endrew F. alongside our precedents, holding that "the IEPs contained meaningful changes" and that "in light of her circumstances, K.D. made appropriate and meaningful progress." Id. at *8-9 (capitalization removed).

The District Court had jurisdiction under 28 U.S.C. §§ 1331 and 1343. We have jurisdiction under 28 U.S.C. § 1291. II. DISCUSSION

Downingtown complied with the IDEA. It gave K.D. a free appropriate public education by developing tailored IEPs. And though K.D.'s parents claim disability discrimination under the ADA and Rehabilitation Act, their theory is indistinguishable from their IDEA claim. So all three claims fail together.

A. Downingtown complied with the IDEA

First, K.D.'s IDEA claim fails. K.D.'s parents argue that the Supreme Court, in Endrew F. , implicitly overruled the Third Circuit's meaningful-benefit test. And they argue that under Endrew F. , Downingtown did not do enough with its IEPs to provide a free and appropriate public education. But Endrew F. did not overrule our precedent. And their claim fails under Supreme Court and Third Circuit decisions.

Whether Endrew F. implicitly overruled Third Circuit precedent is a question of law, which we review de novo. Ridley , 680 F.3d at 268. Whether an IEP is appropriate is a question of fact, which we review for clear error. Id.

1. Endrew F. did not overrule Third Circuit precedent . In Endrew F. , the Tenth Circuit had read the IDEA to require only that students make "merely ... more than de minimis " progress. Endrew F. ex rel. Joseph F. v. Douglas Cty. Sch. Dist. , 798 F.3d 1329, 1338 (10th Cir. 2015) (internal quotation marks omitted). The Supreme Court rejected the Tenth Circuit's standard, not ours. See 137 S.Ct. at 1000-01. On the contrary, Endrew F. 's language parallels that of our precedents.

The Court held that the IDEA "requires an educational program reasonably calculated to enable a child to make progress appropriate in light of the child's circumstances." Id. at 1001. That language mirrors our longstanding formulation: the educational program "must be reasonably calculated to enable the child to receive meaningful educational benefits in light of the student's intellectual potential and individual abilities." Ridley , 680 F.3d at 269 (internal quotation marks and citation omitted). Our test requires an educational program "likely to produce progress, not regression or trivial educational advancement." Id. (internal quotation marks omitted).

Like our precedents, Endrew F. treated a child's intellectual abilities and potential as among the most important circumstances to consider. 137 S.Ct. at 999. And we have contrasted our standard with that applied by the Tenth Circuit: "the provision of merely more than a trivial educational benefit does not meet the meaningful benefit requirement ...." L.E. v. Ramsey Bd. of Educ. , 435 F.3d 384, 390 (3d Cir. 2006) (internal quotation marks omitted); see also T.R. v. Kingwood Twp. Bd. of Educ. , 205 F.3d 572, 577 (3d Cir. 2000) ("[A] satisfactory IEP must provide significant learning and confer meaningful benefit." (internal quotation marks omitted) ). So we see no conflict between Endrew F. and our precedent.

2. K.D.'s IEPs were reasonably calculated to enable her to make appropriate progress . The IDEA required Downingtown to work with K.D.'s parents to develop IEPs that "aim[ed] to enable [K.D.] to make progress." Endrew F. , 137 S.Ct. at 999. Those aims must be "reasonably calculated" and formulated "in light of the child's circumstances." Id. Downingtown did so.

Downingtown had significant foundational work to do with K.D. She had ADHD, vision problems, and poor motor skills. She was quite challenged in perceptual reasoning and processing speed. Her reading, writing, and math skills were well below average. And she suffered from dyslexia and mathematics disorder. Given her impairments and circumstances, the District Court did not clearly err in finding that "this kind of fragmented progress could reasonably be expected." 2017 WL 3838653, at *12.

i. IEPs must be reasonable, not ideal. Though her parents argue otherwise, K.D.'s slow progress does not prove that her IEPs were deficient. "Any review of an IEP must appreciate that the question is whether the IEP is reasonable , not whether the court regards it as ideal." Endrew F. , 137 S.Ct. at 999 (emphasis in original). "The IEP must aim to enable the child to make progress." Id. (emphasis added). We may not rely on hindsight to second-guess an educational program that was reasonable at the time.

While courts can expect fully integrated students to advance with their grades, they cannot necessarily expect the same of less-integrated students. As Endrew F. explained, "for a child fully integrated in the regular classroom, an IEP typically should ... be reasonably calculated to enable the child to achieve passing marks and advance from grade to grade." 137 S.Ct. at 999 (internal quotation marks omitted). But the District Court found that K.D. was not fully integrated into the regular classroom. 2017 WL 3838653, at *2-3, *12. Instead, she received supplemental learning support for much of the day. So there is no reason to presume that she should advance at the same pace as her grade-level peers.

Still, K.D.'s parents seek to extend the presumption beyond fully integrated students. They point to a regulatory guidance letter from the Department of Education's Office of Special Education and Rehabilitative Services. It states: "Research has demonstrated that children with disabilities who struggle in reading and mathematics can successfully learn grade-level content and make significant academic progress when appropriate instruction, services, and supports are provided." U.S. Dep't of Educ., Dear Colleague Ltr., at 1 (Nov. 16, 2015). It also instructs that "the annual goals ... should be sufficiently ambitious to help close the gap" between the child's current and grade-level achievements. Id. at 5.

K.D.'s parents overread the letter. The letter sets forth research and aspirational goals, which may be helpful for some children. But while it aspires to "close the gap," it does not specifically require grade-level goals for children who are not and cannot be fully integrated into regular classrooms. It never mentions a presumption. Nor does it suggest that all (or even most) disabled children can advance at a grade-level pace.

Even if the letter could be read as relevant, it would neither bind nor persuade us. Guidance letters do not enjoy Chevron deference. Christensen v. Harris County , 529 U.S. 576, 586-87, 120 S.Ct. 1655, 146 L.Ed.2d 621 (2000) (discussing Chevron U.S.A. Inc. v. NRDC , 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984) ). And this guidance letter does not address the IDEA's language, let alone parse it. The IDEA contemplates educational programs tailored to "how the child's disability affects the child's involvement and progress in the general education curriculum." 20 U.S.C. § 1414(d)(1)(A)(i)(I)(aa). Rather than presuming grade-level advancement, the Act requires revisions to education programs "as appropriate to address any lack of expected progress toward the annual goals and in the general education curriculum, where appropriate ." Id. § 1414(d)(4)(A)(ii), (ii)(I) (emphases added). Because the letter neither "thorough[ly] ... consider[s]" nor "valid[ly] ... reason[s]" about the meaning of the statute, we find it unpersuasive on this issue. Skidmore v. Swift & Co. , 323 U.S. 134, 140, 65 S.Ct. 161, 89 L.Ed. 124 (1944).

ii. Downingtown reviewed and revised the IEPs to keep them appropriately rigorous . K.D.'s slow progress does not prove that her IEPs were not challenging enough or updated enough. The hearing officer found that Downingtown did not simply repeat educational programs. The District Court agreed. The Court also rejected Dr. Kelly's assertion that K.D. was not making meaningful progress. 2017 WL 3838653, at *9-12. We defer to both sets of findings on appeal. See S.H. v. State-Operated Sch. Dist. of the City of Newark , 336 F.3d 260, 270 (3d Cir. 2003) (treating the hearing officer's factual findings as "prima facie correct"); Ridley , 680 F.3d at 268 (reviewing the District Court's findings for clear error).

Both the hearing officer and the District Court found that Downingtown was willing and able to review and revise K.D.'s IEPs throughout her education. After K.D.'s parents notified Downingtown of Dr. Kelly's evaluation and recommendations, Downingtown responded within a week. It scheduled a meeting, sought more assessments, and offered a one-on-one aide. And it developed a fourth IEP, which incorporated many of Dr. Kelly's recommendations, including adopting a new reading program.

Finally, K.D.'s parents advance arguments not made below. They claim that Downingtown did not offer K.D. an IEP for 2015. And, at oral argument, they asserted that K.D.'s IEPs were not intense enough and did not strike the right balance between regular and special education. But "[a]bsent exceptional circumstances, this Court will not consider issues raised for the first time on appeal." Del. Nation v. Pennsylvania , 446 F.3d 410, 416 (3d Cir. 2006). We see no exceptional reason to excuse their failure. And in any event, Downingtown offered K.D. a fourth IEP in December 2014, which would have run for almost all of 2015.

In sum, the District Court did not err in finding that Downingtown set appropriately challenging goals for K.D.

B. No basis to supplement the record

Nor did the District Court abuse its discretion in rejecting irrelevant and cumulative evidence. D.K. v. Abington Sch. Dist. , 696 F.3d 233, 253-54 (3d Cir. 2012). Though the AIMSweb evidence charted K.D.'s progress on school-district benchmarks, the administrative record already contained ample evidence of how K.D. compared to her peers. Downingtown's interrogatory answers add no facts to what is elsewhere in the record. And K.D.'s parents should have introduced the Wilson teacher's manual earlier, before the hearing officer. They gave no good reason for not doing so.

C. No disability discrimination under the ADA or Rehabilitation Act

K.D.'s parents also assert disability discrimination under the Rehabilitation Act and ADA. They allege that Downingtown did not use "appropriate research-based interventions" to "teach a student like K.D. to read." Appellants' Br. 45. Though they deny it, their allegations simply repackage those underlying the IDEA claim. So the District Court properly rejected these claims when it rejected the IDEA claim.

* * * * *

K.D.'s parents understandably want only the best opportunities for their daughter. But Downingtown followed the law by individualizing her education programs to help her make progress appropriate to her circumstances. So we will affirm the District Court's judgment in favor of Downingtown.