OPINION OF THE COURT
FISHER, Circuit Judge.These consolidated appeals arise from administrative proceedings and subsequent litigation concerning the provision of special education services by the Mars Area School District to J.R., a learning disabled child covered by the Individuals with Disabilities Education Act (IDEA), 20 U.S.C. *115§ 1400 et seq. In the case at Docket No. 07-2440, J.R.’s parents appeal from the District Court’s May 4, 2007 judgment in favor of the School District, challenging the District Court’s conclusion that the School District did not violate the IDEA’S “stay-put” provision, 20 U.S.C. § 1415(j). In the related case at Docket No. 07-2753, the parents appeal from an order entered by the District Court on June 5, 2007, in which the District Court denied their motion for prevailing party attorney’s fees in that case. For the reasons that follow, we will affirm the District Court in both instances.
I. BACKGROUND
We write exclusively for the parties, who are familiar with the factual context and legal history of this ease. Therefore, we will set forth only those facts necessary to our analysis.
A. No. 07-2440
In late 2003, when the dispute underlying the appeal at No. 07-2440 arose, J.R. was a fifth-grade student at Adams Elementary School in the Mars Area School District. J.R. had previously been identified by the School District as a student with a “specific learning disability” and, as such, the School District was required under the IDEA to develop an individualized education program (IEP) for him and to review it once a year. See 20 U.S.C. § 1414(d)(2), (4); Schaffer v. Weast, 546 U.S. 49, 53, 126 S.Ct. 528, 163 L.Ed.2d 387 (2005) (discussing 20 U.S.C. § 1414(d)). In May 2001 the School District and J.R.’s parents agreed upon an IEP to govern J.R.’s education during the 2001-2002 school year, his third-grade year. Under that IEP, J.R. spent most of his school day in a regular classroom but also received learning support for approximately one hour a day in a resource room. The 2001-2002 IEP was not intended to serve J.R. beyond his third-grade year, but it was the last IEP to which J.R.’s parents and the School District were able to agree, despite having met numerous times to discuss various program proposals for the subsequent 2002-2003 and 2003-2004 school years.
On August 6, 2003, the School District presented yet another proposed IEP to J.R.’s parents, which it detailed in a Notice of Recommended Educational Placement (NOREP). That proposed IEP was intended for J.R.’s 2003-2004 school year, and among its many provisions was a recommendation that, going forward, he receive “itinerant” learning support, i.e., learning support provided primarily in the regular classroom, in place of the daily hour of resource room support specified in his 2001-2002 IEP.1 A few days later, on August 11, 2003, the School District sent a form letter to J.R.’s parents informing them that it would offer inclusion classes to all of its elementary students, including J.R., starting in the 2003-2004 school year. After receiving both of these documents, J.R.’s parents, apparently displeased with the prospect that J.R.’s learning support services would be provided in an inclusion classroom instead of in a resource room, decided not to approve the NOREP, returning it unsigned to the School District. Instead, they requested a due process hearing to assess two related issues: (1) whether the School District had denied J.R. a free appropriate public education *116(FAPE) by failing to implement a mutually agreed upon IEP for the 2003-2004 school year and (2) whether the School District’s decision to provide learning support services to J.R. in an inclusion classroom instead of in a resource room amounted to a change in his educational placement.
At the due process hearing, the hearing officer heard the testimony of three witnesses for the School District—J.R.’s fifth-grade teacher, his special education teacher, and the district-wide director of special education—and one witness for J.R.—his mother. The School District’s witnesses testified that J.R. continued to receive in the inclusion classroom all the services specified in his 2001-2002 IEP, that he was making academic progress in the inclusion classroom, and that he was passing all of his subjects. J.R.’s mother, for her part, testified that she did not know, nor had she been informed of, the amount of time J.R. was spending with a special education teacher each day in the inclusion classroom.
After considering this testimony and the other evidence presented by the parties, the hearing officer held that the School District had neither deprived J.R. of a FAPE nor unilaterally changed his educational placement, concluding that “[t]here was no evidence presented or testimony that stated [J.R.] was not receiving the needed special education services.” J.R.’s parents administratively appealed2 and the state appeals panel reversed the hearing officer’s decision in part, holding that “the [School] District’s automatic decision to educate [J.R.] in an inclusion classroom ... constituted a unilateral placement” and ordering the School District to provide J.R. with one hour of compensatory education in learning services for each school day he attended during the 2003-2004 school year.
The School District then brought this action challenging the appeals panel’s decision. After reviewing the administrative record, the District Court entered judgment in favor of the School District on May 4, 2007, concluding that J.R.’s parents had not demonstrated that the School District’s unilateral relocation of J.R. to an inclusion classroom constituted a change in his educational placement.3 On May 10, 2007, J.R.’s parents filed a notice of appeal from that judgment, which the Clerk of the Court docketed at No. 07-2440.
B. No. 07-2753
The appeal at No. 07-2753 arises from an earlier due process proceeding initiated by J.R.’s parents to determine whether the School District had denied J.R. a FAPE during his third- and fourth-grade years. The hearing officer in that proceeding held that the School District had denied J.R. a FAPE during those school years and or*117dered it to provide J.R. with 283.5 hours of compensatory education, which an appeals panel subsequently increased to 306.5 hours. The parties eventually settled the matter during a pretrial conference before the District Court and, although the School District agreed in the settlement to provide J.R. with nothing more than the 306.5 hours of compensatory education ordered by the appeals panel, J.R.’s parents filed a motion seeking prevailing party attorney’s fees.
In an order dated May 4, 2007, the District Court determined “categorically” that J.R.’s parents could not be considered prevailing parties for purposes of the federal action because they “occupied] precisely the same legal position [as a result of the settlement] which they were in immediately following the decision of the Appeals Panel in 2003,” having “achieved absolutely nothing more than had been awarded by the Appeals Panel.” They could, however, be considered prevailing parties at the state administrative level, the District Court concluded, because of their success with respect to certain issues raised in that forum, and thus were eligible to recover reasonable attorney’s fees “specifically related to that success throughout the administrative proceedings.”
But the District Court, that conclusion notwithstanding, was unable to determine the amount of attorney’s fees to which J.R.’s parents were entitled because the “extremely generic and vague” supporting documentation submitted by their attorney, Eugene A. Lincoln, Esq., fell “far short of providing sufficient detail to permit the Court to determine whether each claimed legal activity was related to the specific issues on which [the parents] prevailed.” Rather than simply deny altogether the parents’ motion for fees, however, the District Court “held in abeyance” the question of their right to attorney’s fees “incurred in the administrative process” and instead directed Mr. Lincoln “to submit a revised fee petition [containing] sufficient detail to enable the Court to evaluate the relationship between counsel’s legal actions and the issue(s) on which [J.R.’s parents] prevailed at the administrative level.” Mr. Lincoln did not submit a revised petition, however, and the District Court ultimately denied the parents’ petition in full in an order dated June 5, 2007, concluding that they had not met their burden to submit sufficient evidence “to demonstrate which legal services are compensable for the ... issues on which they prevailed” at the administrative level. The parents filed a notice of appeal from that order on June 11, 2007, which the Clerk docketed at No. 07-2753 and consolidated with No. 07-2440.
II. JURISDICTION AND STANDARD OF REVIEW
The District Court had jurisdiction under 20 U.S.C. § 1415(i)(3) and 28 U.S.C. § 1331. We have appellate jurisdiction under 28 U.S.C. § 1291.
A district court must review an IDEA decision by a state administrative agency under a modified de novo standard. L.E. v. Ramsey Bd. of Educ., 435 F.3d 384, 389 (3d Cir.2006). Under this unusual standard, a district court must afford “due weight” to the factual findings from the administrative proceedings, considering those findings to be “prima facie correct.” See id.; Shore Reg’l High Sch. Bd. of Educ. v. P.S., 381 F.3d 194, 199 (3d Cir.2004); see also Bd. of Educ. of Hendrick Hudson Cent. Sch. Dist. v. Rowley, 458 U.S. 176, 206, 102 S.Ct. 3034, 73 L.Ed.2d 690 (1982); Carlisle Area Sch. v. Scott P., 62 F.3d 520, 524 (3d Cir.1995) (“The district court may reach an independent decision, except that it must accord the deci*118sion of the state agency ‘due weight’ in its consideration.”). At the same time, the district court must “hear additional evidence at the request of a party,” 20 U.S.C. § 1415(i)(2)(C)(ii), and must also “make its own findings by a preponderance of the evidence.” Shore Reg’l, 381 F.3d at 199; see 20 U.S.C. § 1415(i)(2)(C)(iii). But “where the District Court does not hear additional evidence it must find support for any factual conclusions contrary to [those of the administrative agency] in the record before it.” S.H. v. State-Operated Sch. Dist. of Newark, 336 F.3d 260, 270 (3d Cir.2003). In contrast, the “due weight” to be afforded to the administrative proceedings is “not implicated with respect to issues of law, such as the proper interpretation of the [IDEA] and its requirements”; that is, the district court owes no deference to conclusions of law drawn by a state or local educational agency. Lillbask ex rel. Mauclaire v. Conn. Dep’t of Educ., 397 F.3d 77, 82 (2d Cir.2005); see Alex R. v. Forrestville Valley Cmty. Unit Sch. Dist. # 221, 375 F.3d 603, 611 (7th Cir.2004); cf. Carlisle Area Sch., 62 F.3d at 528 n. 3 (“Obviously, conclusions of law receive plenary review.”).
We, in turn, “‘exercise plenary review with respect to the question whether the District Court applied the correct legal standards under the IDEA, but we review the District Court’s factual findings for clear error.’ ” Ramsey Bd. of Educ., 435 F.3d at 389 (quoting Shore Reg’l, 381 F.3d at 199); see Lauren W. v. DeFlaminis, 480 F.3d 259, 266 (3d Cir.2007); Carlisle Area Sch., 62 F.3d at 526. “ ‘A finding of fact is clearly erroneous when, after reviewing the evidence, the court of appeals is left with a definite and firm conviction that a mistake has been committed.’ ” Shore Reg’l, 381 F.3d at 199 (quoting Oberti v. Bd. of Educ. of Clementon Sch. Dist., 995 F.2d 1204, 1220 (3d Cir.1993)).
We review a district court’s denial of attorney’s fees under the IDEA’S fee-shifting provision, 20 U.S.C. § 1415(i)(3)(B), under an abuse of discretion standard. Pardini v. Allegheny Intermediate Unit, 524 F.3d 419, 422 (3d Cir.2008) (Pardini II); P.N. v. Clementon Bd. of Educ., 442 F.3d 848, 852 (3d Cir.2006). A district court abuses its discretion when it rests its decision “ ‘upon a clearly erroneous finding of fact, an errant conclusion of law or an improper application of law to fact.’ ” Pardini II, 524 F.3d at 422 (quoting Clementon Bd. of Educ., 442 F.3d at 852); see Rode v. Dellarciprete, 892 F.2d 1177, 1182-83 (3d Cir.1990) (“An abuse of discretion can occur when no reasonable person would adopt the district court’s view. Whether the district court applied the proper standards or procedures is a question of law subject to plenary review. The district court’s factual findings are reviewed under a clearly erroneous standard.”) (citations omitted).
III. DISCUSSION
A. Change in Educational Placement
The principal question in the appeal at No. 07-2440 is whether the District Court erred in concluding that the School District’s unilateral relocation of J.R. out of the resource room and into an inclusion classroom during the pendency of proceedings amounts to a change in his educational placement within the meaning of the IDEA’S “stay-put” provision, 20 U.S.C. § 1415(j). Section 1415(j) provides, in relevant part, that “during the pendency of any proceedings conducted pursuant to this section, unless the State or local educational agency and the parents otherwise agree, the child shall remain in the then-current educational placement of the child ... until all such proceedings have been completed.” The Supreme Court has explained that the main purpose of the stay-*119put provision is “to strip schools of the unilateral authority they had traditionally employed to exclude disabled students ... from school.” Honig v. Doe, 484 U.S. 305, 323, 108 S.Ct. 592, 98 L.Ed.2d 686 (1988).
The IDEA does not define the term “educational placement” and thus “identifying a change in this placement is something of an inexact science.” Bd. of Educ. of Cmty. High Sch. Dist. No. 218 v. Ill. State Bd. of Educ., 103 F.3d 545, 548 (7th Cir.1996). Nonetheless, our precedent provides some guidance in this regard. In DeLeon v. Susquehanna Community School District, 747 F.2d 149 (3d Cir.1984), a case relied on heavily by the District Court, we discussed our understanding of what constitutes a “change in educational placement” in the context of a predecessor statute to the IDEA. As we explained in that case, “[t]he question of what constitutes a change in educational placement is, necessarily, fact specific” and thus, “in determining whether a given modification in a child’s school day should be considered a ‘change in educational placement,’ ” the “touchstone” is whether the modification “is likely to affect in some significant way the child’s learning experience.” Id. at 153.
Here, the District Court properly focused its inquiry on whether the move to an inclusion classroom was likely to significantly affect J.R.’s ability to learn. Reviewing the administrative record, the District Court found that it contained “no evidence that the move to an inclusion classroom was likely to affect [J.R.’s] learning experience in any ‘significant way.’ ... The School District provided the same educational services to [J.R.] on a daily basis, with the same special education teacher, but in a different room.” The District Court also accepted the sparse findings made during the administrative proceedings. Having reviewed the record, we conclude that the District Court’s findings were not clearly erroneous and that it did not err in concluding that J.R.’s relocation did not amount to a change in his educational placement within the meaning of the stay-put provision.
J.R.’s parents assert that the District Court, in reaching its decision, misapplied the modified de novo standard of review by according due weight to the hearing officer’s decision instead of the appeals panel’s decision. We disagree. It is not at all clear that the District Court deferred to the hearing officer’s findings to the exclusion of those of the appeals panel. The District Court explicitly stated that it had considered the decisions of both the hearing officer and the appeals panel and that it was accepting the “factual findings made during the state administrative proceedings,” not simply those of the hearing officer. Indeed, the District Court clearly viewed the appeals panel’s findings, insofar as it made any, as compatible with those of the hearing officer, and we concur: Nothing in the record indicates that the appeals panel had a view of the underlying facts, or of the credibility of any witness, that differed in a significant way from that of the hearing officer. To the contrary, as the District Court noted, the appeals panel apparently accepted the hearing officer’s findings but reached a different legal conclusion based on those findings. Nor is there any indication that the District Court took a significantly different view of the underlying facts. Inasmuch as the issue before the District Court was one of law, it did not misapply the modified de novo standard of review. Cf. Pardini v. Allegheny Intermediate Unit, 420 F.3d 181, 192 n. 13 (3d Cir.2005) (Pardini I) (noting that “the interpretation of § 1415(j) ... is a purely legal one”); Drinker v. Colonial Sch. Dist., 78 F.3d 859, 865 (3d Cir.1996) (holding that the question of what consti*120tutes a student’s “current educational placement” is one of law).
B. Attorney’s Fees
We turn next to the appeal at No. 07-2753, where J.R.’s parents challenge the District Coui't’s decision denying their request for attorney’s fees in that case. Specifically, the parents contend that the settlement they reached with the School District was sufficient to confer “prevailing party” status on them and that the District Court abused its discretion in finding to the contrary. We disagree.
The IDEA’S fee-shifting provision, 20 U.S.C. § 1415(i)(3)(B), provides in relevant part that “the court, in its discretion, may award reasonable attorneys’ fees as part of the costs ... to a prevailing party who is the parent of a child with a disability.” For purposes of the fee-shifting provision, “a plaintiff ‘prevails’ when actual relief on the merits of his claim materially alters the legal relationship between the parties by modifying the defendant’s behavior in a way that directly benefits the plaintiff.” Clementon Bd. of Educ., 442 F.3d at 855 (quoting Farrar v. Hobby, 506 U.S. 103, 111-12, 113 S.Ct. 566, 121 L.Ed.2d 494 (1992)) (internal quotation marks omitted); see Hensley v. Eckerhart, 461 U.S. 424, 433, 103 S.Ct. 1933, 76 L.Ed.2d 40 (1983) (“[P]laintiffs may be considered ‘prevailing parties’ for attorney’s fees purposes if they succeed on any significant issue in litigation which achieves some of the benefit the parties sought in bringing suit.”).
J.R.’s parents cannot be considered prevailing parties at the district court level. After four years of federal litigation, they agreed to a settlement that did not provide them with any relief beyond that which they had already achieved in the administrative proceedings, abandoning “all other claims” raised in their complaint in doing so. As the District Court noted, J.R.’s parents “occup[ied] precisely the same legal position [after the settlement] they were in when the case started.” On these facts, we cannot say that the settlement materially altered the legal relationship between the parties, either in a way that directly benefits J.R.’s parents or otherwise, and we see no reason to disagree with the District Court’s decision.
Moreover, the parents seemingly restrict their argument on appeal to the prevailing party issue and do not acknowledge, much less attempt to justify or explain, Mr. Lincoln’s failure to comply with the District Court’s order that he submit a revised fee petition. To the extent they nonetheless intend to challenge the District Court’s ruling that they failed to meet their burden to submit sufficient evidence “to demonstrate which legal services are compensable for the ... issues on which they prevailed” at the administrative level, we find no abuse of discretion. See Hensley, 461 U.S. at 433, 103 S.Ct. 1933 (“Where the documentation of hours is inadequate, the district court may reduce the award accordingly.”); Loughner v. Univ. of Pittsburgh, 260 F.3d 173, 178 (3d Cir.2001) (holding that a district court “should reduce the hours claimed by the number of hours ... for which the fee petition inadequately documents the hours claimed”); Rode, 892 F.2d at 1190 (“A fee petition is required to be specific enough to allow the district court to determine if the hours claimed are unreasonable for the work performed.”) (internal quotation marks omitted).4
*121IV. CONCLUSION
For the foregoing reasons, we will affirm the District Court’s judgment of May 4, 2007, in the case at Docket No. 07-2440, and its order of June 5, 2007, in the case at Docket No. 07-2753.
. In the version of the Pennsylvania Code in effect during the time period relevant to this appeal, "itinerant” learning support is defined as "[r]egular classroom instruction for most of the school day, with special education services and programs provided by special education personnel inside or outside of the regular class for part of the school day.” 22 Pa.Code § 14.141 (2003).
. During the time period relevant to this case, Pennsylvania had a two-tiered administrative system of review in IDEA cases, in which the initial hearing took place before a hearing officer at the local educational agency level followed by an independent review of the hearing officer's decision by an appeals panel at the state educational agency level. See 22 Pa.Code § 14.162(d), (o) (2003); cf. Carlisle Area Sch. v. Scott P., 62 F.3d 520, 526-27 (3d Cir.1995) (discussing Pennsylvania's two-tiered system). Pennsylvania subsequently amended its administrative procedures to allow a hearing officer’s decision to be appealed directly to a court of competent jurisdiction. See 22 Pa.Code § 14.162(o) (2008). That amendment took effect on July 1, 2008, and does not apply to this case.
. Although the appeals panel affirmed the hearing officer's decision that the School District had not denied J.R. a FAPE by failing to implement a mutually agreed upon IEP for the 2003-2004 school year, J.R.'s parents abandoned their challenge to this aspect of the administrative decision in the District Court, and we do not address it here.
. The parents raise two additional issues in this appeal, neither of which has any merit. First, their contention that the District Court was required to obtain their consent before referring nondispositive pretrial matters to a Magistrate Judge is contradicted by, inter alia, our precedent interpreting the Federal Magistrates Act, 28 U.S.C. § 636. See, e.g., *121Beazer E., Inc. v. Mead Corp., 412 F.3d 429, 438 (3d Cir.2005) ("The Magistrates Act authorizes district courts to appoint magistrate judges to consider pretrial matters without regard to the parties' consent.”); In re U.S. Healthcare, 159 F.3d 142, 145 (3d Cir.1998) (“In general, a magistrate judge, without the consent of the parties, has the power to enter orders which do not dispose of the case.”). Second, we reject the notion that the District Court somehow "abused” Mr. Lincoln by ostensibly making certain negative remarks about him or otherwise. We note that Mr. Lincoln apparently makes these spurious (and somewhat incoherent) allegations on his own behalf; he does not argue any prejudice to his client, only an affront to his personal sensibilities. In any event, on the record before us Mr. Lincoln's accusations provide no basis for appellate intervention. Cf. Logue v. Dore, 103 F.3d 1040, 1047 (1st Cir.1997) ("[A]ppellate courts review orders and judgments, not judge[s'] statements.”).