Ambrose, Cynthia

OPINION ALCALA, J., delivered the opinion of the Court in which KELLER, P.J., MEYERS, JOHNSON, KEASLER, HERVEY, RICHARDSON, and NEWELL, J.J., joined. This case addresses accomplice-witness jury-charge error raised in a motion for new trial. Cynthia Ambrose, appellee, is a former kindergarten teacher at Salinas Elementary School who was convicted of the misdemeanor offense of official oppression after a jury trial at which the testimony of a purported accomplice, Ramirez, was presented. After the jury found her guilty, Ambrose filed a motion for new trial alleging that the jury instructions were erroneous in that they failed to in *590 struct the jury that Ramirez's accomplice-witness testimony had to be corroborated. See Tex.Code Ceim, PROC. art. 38.14. She further asserted that she was egregiously harmed by the error. The trial court agreed with Ambrose, and it made what it characterized as findings of fact and conclusions of law in support of its ruling. On appeal, the court of appeals assumed without deciding that the trial court properly determined that the jury instructions were erroneous, but, as to the matter of harm, it disregarded the trial court’s findings and conclusions and it instead determined that Ambrose was not egregiously harmed under the substantive application of the Almanza factors. State v. Ambrose, 457 S.W,3d 154, 162 (Tex.App.-San Antonio 2015). In her three grounds in her petition for discretionary review, Ambrose’s first two grounds contend that the appellate court erred by applying the Almanza egregious-harm standard and by failing to defer to the trial court’s findings of fact made after the motion-for-new-trial hearing. See Almanza v. State, 686 S.W.2d 157 (Tex.Crim.App.1984) (op.on.reh’g). Ambrose’s third ground for review asserts that the appellate court erred in concluding that the error was not egregiously harmful to her. 1 We conclude that the court of appeals properly held (1) that the Almanza harm standard applies to jury-charge error reviewed on appeal, even when the error was addressed in a motion for new trial, (2) that it was not required to defer to the trial court’s factual findings in this case, and (3) that the record failed to show that Ambrose was egregiously harmed by the error in the charge. We, therefore, affirm the judgment of the court of appeals. I. Background' Ramirez was a fellow kindergarten teacher at the same school as Ambrose. One of Ramirez’s students, A.N., was behaving disruptively in class and had hit another student. Pursuant to a school policy that directs a teacher to take a disruptive child into another teacher’s classroom, Ramirez took A.N. into Ambrose’s classroom. Ramirez asked Ambrose if A.N. could stay with her “[bjecause he was bullying the other, student.” Ambrose instructed A.N. to sit right beside her and asked the questions “Why are you bullying other students?” and “How would you like for other students to bully you?” Ami brose then said to the students in her classroom, “Come on, boys and gilds, let’s *591 line up and let’s bully [A.N.].” About seven students struck A.N. before the assault ended when one girl struck A.N. harder than the other students had, prompting Ambrose to say, “Okay, not that hard.” Ramirez then left Ambrose’s classroom and later sent one of her other students to retrieve A.N. About two weeks later, Ramirez reported the incident to Jeffrey Large, the school’s principal, and Gerrie Spellmann, the vice-principal. Large and Spellmann investigated the accusation, which included having a' discussion with Ambrose about the events. Ambrose was charged with the offense ,of official oppression under Texas Penal Code Section 39.03 for “intentionally subjecting] [A.N.] ... to mistreatment, to wit: directing and allowing others to strike the complainant, knowing it was unlawful.” 2 At Ambrose’s trial, other than Ramirez’s testimony, the jury heard testimony from Large and Spellmann, in addition to other witnesses. Large and Spellmann each recited statements made by Ambrose admitting that she had instructed students to hit A.N. and that she then observed two or three students hit A.N. before she stopped the assault on A.N. According to Spellmann, Ambrose “let us know that she had instructed her students to hit [A.N.] on the arm, but not too hard so that the student would know how it felt to be bullied.” Spellmann further testified, “[Am-brose] stated that — we asked how many students had [ ] struck [A.N.] and she stated only about two or three. But the next student hit too hard and that she had it stopped.” Like Spellmann, Large testified that, during the investigation of the incident, Ambrose stated that-“she instructed the students to hit the other student but not hard. She said that I believe two or three students hit the student and then a fourth student hit too hard and then she stopped it immediately after that.” In her defense at trial, Ambrose acknowledged the essence of the events as they had been described by Ramirez, Large, and Spellmann, but she minimized her culpability by claiming that she did not intend for any of the students to hit A.N. and that, she immediately stopped the assault when the first student suddenly struck A;N. Ambrose testified that Ramirez brought A.N. into her classroom and said that he was being a bully. Ambrose asked the other students what A.N.’s consequences should be, and they made different suggestions, one of which was that the child A.N. hit should be allowed to hit him back. Ambrose asked, “Does anybody want to show him what it ijeels like?” Before she knew what was happening, one student struck A.N., and. Ambrose immediately intervened, telling the other students to “get away from him.” When the prosecutor asked, “Did you ever intend for:one student to hit another student?” Ambrose responded, “No, that wasn’t my intention.” After the close of evidence, the trial court’s instructions to the jury did not include an accomplice-witness instruction, and Ambrose did not object to its absence. The jury found Ambrose guilty. The court assessed her punishment at one year of confinement, probated for two years, with *592 a 30-day jail sanction as a condition of her community supervision. Following her conviction, Ambrose filed a motion for a new trial that alleged, among other complaints, that the jury instructions were erroneous based on the trial court’s failure to instruct the jury regarding the law of accomplice-witness testimony. Ambrose claimed that Ramirez was an accomplice to the offense and that the jury should have been instructed that Ambrose could not be convicted based solely on Ramirez’s uncorroborated testimony. The trial court granted the motion for a new trial based on its determination that the absence of the accomplice-witness instruction was erroneous and that Ambrose was egregiously harmed by the error. The trial court made findings of fact and conclusions of law in support of its ruling. 3 The State appealed. On appeal, the State contended that the trial court erred by granting Ambrose’s motion for new trial. The State alternatively claimed that the jury instructions were correct because Ambrose was not entitled to an accomplice-witness instruction for Ramirez’s testimony, and that even if the instructions were erroneous on that basis, the record failed to show that Ambrose was egregiously harmed by the absence of the instruction. The court of appeals did not address the State’s first alternative argument, but it agreed with the second alternative under the assumption that error had occurred. The court of appeals held that the trial court erred by granting Ambrose’s motion for new trial because the record failed to show that she was egregiously harmed by the absence of an accomplice-witness instruction for Ramirez’s testimony. Am-brose, 457 S.W.3d at 162. The court of appeals applied the Almanza egregious-harm standard because Ambrose had not objected to the omission of the accomplice-witness instruction. Id. at 160. In analyzing the evidence for egregious harm, the court of appeals observed that “Ambrose agreed during her testimony that she was a public servant acting under the color of state law and that subjecting a child to corporal punishment through other students would constitute mistreatment. *593 Thus, the only contested element of the charged offense was whether Ambrose intended to subject A.N. to being struck by other students.” Id. The court of appeals therefore focused its analysis “on evidence ‘tending to connect’ Ambrose’s intent to the intent alleged in the charge — that Am-brose intended that other students strike AN.” Id. at 160-61. The court of appeals reviewed the record and observed that the “non-accomplice corroborative evidence ... is relatively strong.” Id. at 161. In particular, it noted that Ambrose “admitted to saying something that prompted students in her classroom to get up and strike AN.” Id. The court observed that, although Ambrose denied explicitly directing her students to hit A.N., “her own testimony tended to connect her intent to the charged intent of subjecting AN. to mistreatment by directing and allowing her students to strike him.” Id. The court also took note of the testimony from Large and Spellmann, both of whom indicated that Ambrose had admitted to instructing her students to hit AN. Id. Regarding the jury charge and the State’s arguments, the court reasoned that these factors “emphasized the. strength of the corroborating evidence and the weakness of Ambrose’s contradictory testimony.” Id. The court concluded that “[t]he inclusion of an accomplice-witness instruction would not have rendered the State’s case clearly and significantly less persuasive,” and, thus, it held that Ambrose was not “deprive[d][ ] of a fair and impartial trial” as a result of the omission of the accomplice-witness instruction from the charge. Id. at 162. II. Procedural Standards for Jury-Charge Error Raised in a Motion for New Trial Ambrose’s first two grounds for review challenge the procedural standards employed by the court of appeals. In her first ground, she contends that the court of appeals .erred by applying the Almanza egregious-harm standard on appellate review rather than applying the abuse-of-discretion standard usually employed by an appellate court reviewing a trial court’s ruling on a motion for a new trial. See State v. Thomas, 428 S.W.3d 99, 103-04 (Tex.Crim.App.2014). In her second ground, she contends that the court of appeals impermissibly substituted its own view of the evidence for that of the trial court by ignoring the trial court’s factual finding that Ambrose suffered egregious harm. After discussing the general requirements for accomplice-witness instructions, we address these two grounds in turn. A. General Requirements for Accomplice-Witness Instructions Because Ambrose complains about the absence of the accomplice-witness instruction, we begin by briefly reviewing that law. “A conviction cannot be had upon the testimony of an accomplice unless corroborated by other evidence tending to connect the defendant with the offense committed; and the corroboration is not sufficient if it merely shows the commission of the offense.” Tex.Code CRim. PRoc. art. 38.14. Therefore, if an accomplice to the offense testifies for the State, the accomplice’s testimony must be corroborated by non-accomplice evidence that tends to “connect the accused to the offense.” Smith v. State, 332 S.W.3d 426, 442 (Tex.Crim.App.2011). Such evidence may tie either direct or circumstantial. Cathey v. State, 992 S.W.2d 460, 462 (Tex.Crim.App.1999) (“It is not necessary that the corroborating evidence directly connect the defendant to the crime or that it be sufficient by itself to establish guilt; it need only tend to connect the defendant to the offense.”). Depending on the circumstances, an accomplice-witness instruction *594 may direct the-jury, that an accomplice is one as a matter of law or one as a matter of fact for the jury’s determination. Zamora v. State, 411 S.W.3d 504, 510 (Tex.Crim.App.2013). (“A- proper accomplice-witness instruction informs the jury either that a witness' is an accomplice as a matter of law or that-he is an. accomplice as-a matter of fact.. The evidence in each case will dictate the type of aceompliee-witness instruction that needs to be given.”), (citations omitted). Here, like the court of appeals, we will assume that the trial court erred by failing to include an accomplice-witness instruction for Ramirez’s testimony. When the issue is raised by the evidence, the trial court must instruct the jury on the accomplice-witness rule because it is the “law applicable to the case.” Id. at 513; Tex.Cqde ÜRiM. PROP. art. 36.14. B. Almanza Framework for Analyzing Harm Applies in a Motion for New Trial Because Ambrose’s complaint is one' of jury-charge error, the Almanza framework for assessing'harm applies. ' Under Almanza, 'the degree of harm required for reversal depends on whether the error was preserved in the trial court. Ngo v. State, 175 S.W.3d 738, 743 (Tex.Crim.App.2005); Almanza, 686 S.W.2d at 171. Where, as here, "the defendant did not raise a timely objection to the jury instructions, reversal is required only if the error was fundamental in the sense that it was so egregious and created such harm that the defendant was deprived of a fair and impartial trial. ■ ■ Villarreal v. State, 453 S.W.3d 429, 433 (Tex.Crim.App.2015). This “basic framework” for analyzing preserved and unpre-served claims of error in the charge is “not a court-made rule; it is based on this Court’s interpretation of Article 36.19.” Posey v. State, 966 S.W.2d 57, 60 (Tex.Crim.App.1998). 4 This Court has held that Almanza is the appropriate standard for conducting a harm analysis in cases reviewing jury-charge error, even when the error was first asserted in a motion for a new trial. Igo v. State, 210 S.W.3d 645, 647 (Tex.Crim.App.2006). In Igo, as in this case, the defendant had complained of an error in the charge for the first time in a motion for new trial. Id. at 646. After the trial court denied his motión, Igo’challenged this ruling on' appeal, and the court of appeals applied the Almanza egregious-harm standard and'concluded that the error was harmless. Id. On discretionary review in this Court, Igo asserted that Almanza was the improper standard for reviewing the trial court’s ruling on his motion for new trial and that the ruling should instead be reviewed for abuse of discretion. Id. In rejecting Igo’s complaint, this Court explained ’ that, “[e]ven though [Igo] characterized 'his claim as error in denying him a new trial, this case presents error in the charge,” and thus, “the proper standard is Article 36.19, as construed in Almanza.” Id. at 647. This Court further reasoned that “a statute *595 cannot be superceded by a rule,” and when a statute “directs what treatment an appellate court must give to a particular type of error, a rule of appellate procedure cannot be employed to circumvent the statutory requirement.” Id. Further, as a policy consideration, the Court- observed that, were it to accept Igo’s argument, “defendants would no longer be. required to preserve a jury-charge error at trial so long as the issue was raised in a motion for new trial because any error in the charge could be said to ‘misdirect’ the-jury.” Id. Such a rule would “contradict! ] the policy of encouraging the timely correction of errors, which is embodied both in Article 36.19 and in our own rules of appellate procedure.” Id. Moreover, the Court observed that Igo’s proposed rule would “essentially eviscerate the two-tiered harm analysis required by [Article 36.19] and. do away -with the requirement that egregious harm be shown when the defendant has failed to timely .urge an objection.” Id. For these reasons, the Court concluded that Igo could not “use a motion-for-new-trial appellate rule to circumvent the harm analysis imposed by Article 36.19.” Id. Thus, under Igo and Almanza, unpreserved jury-charge error does not require a new trial, even when the error is complained of in a motion for new trial, unless the error causes “egregious harm.” See id.; Al-manza, 686 S.W.2d at 171. Although she recognizes that this Court has already held in Igo that the Almanza egregious-harm standard, as opposed to an abuse-of-discretion standard, applies under these circumstances, she contends that Igo is distinguishable. In particular, she states that Igo involved a trial court’s denial of a motion for new trial, whereas this case involves a trial court’s granting of a new trial. This argument is without merit. Nothing in the language of Igo suggests that its holding is limited to review of rulings denying motions for new trial, and we can conceive of no reason why it should be so limited. Moreover, in a case subsequent to Igo that involved an appeal of a trial court’s ruling granting a motion for new trial,- this Court granted the petition, vacated the court of -appeals’s judgment, and remanded that case to the court of appeals for it to- reconsider its ruling in light of Igo. See State v. McKnight, 213 S.W.3d 915 (Tex.Crim.App.2007) (per cu-riam) (broadly describing the holding of Igo as being that “the Almanza harm analysis does apply to jury-charge errors presented • in a motion for new trial,” and remanding to the court of appeals “in light of our opinion in Igo”). We, therefore, disagree with Ambrose’s contention that Igo is inapplicable here. In addition to suggesting that Igo is distinguishable, Ambrose implicitly suggests that Igo was wrongly decided, at least insofar as it would apply to a situation in which the trial court made findings of fact and conclusions of law in support of its new-trial ruling. Specifically, she suggests that applying the Almanza egregious-harm 'standard on appeal under these circumstances is inconsistent with the rule that an appellate eourt must defer to a trial court’s factual findings made in support of its ruling, particularly when those findings involve evaluations of witness credibility and demeanor. See, e.g., Guzman v. State, 955 S.W.2d 85, 89 (Tex.Crim.App.1997). As we explain further below in our analysis of Ambrose’s second ground ‘ for review, we do not find any inherent inconsistency between an appellate court’s application of the Almanza egregious-harm standard under these circumstances and the rule that an appellate court owes deference to a trial court’s factual findings. The problem with Am-brose’s assertion in this regard is that a trial court’s egregious-harm determination is not a purely factual matter; rather, it is *596 a mixed question of law and fact to which deference is afforded only some of the time. See id. (explaining that rule of deference applies to trial court’s resolution of mixed question of law and fact only when the resolution of that ultimate question “turns on an evaluation of credibility and demeanor”). Finding that Ambrose has presented us with no compelling reason to revisit our holding in Igo on the basis of any conflict between the egregious-harm standard and the rule requiring deference to a trial court’s factual findings, we decline to do so. We hold that Igo is controlling here, ■ and on that basis, we conclude that the court of appeals properly applied the Almanza egregious-harm standard on appeal. Therefore, Ambrose’s first ground for review alleging that the court of appeals erred by applying the Almanza standard is overruled. C. Deference to Trial Court Not Required on Conclusion About Egregious Harm Ambrose’s second ground for review contends that, in deciding whether there is egregious harm in a case in which that question was considered by the trial court and the trial court made findings of fact and conclusions of law addressing it, appellate courts should defer to a trial court’s factual findings. In support, she cites cases indicating that a reviewing court is required to defer to the trial court’s factual findings that are supported by the record. See, e.g., Oregon v. Kennedy, 456 U.S. 667, 102 S.Ct. 2083, 72 L.Ed.2d 416 (1982); Ex parte Wheeler, 203 S.W.3d 317, 325-26 (Tex.Crim.App.2006). In general, we agree with Ambrose to the extent that a question of fact or a mixed question of law and fact that turns on an evaluation of credibility and demeanor should be reviewed with deference. Guzman, 955 S.W.2d at 87-89 (explaining that a reviewing court “should afford almost total deference to a trial court’s determination of the historical facts that the record supports especially when the trial court’s findings are based on an evaluation of credibility and demeanor”; deference is also warranted as to trial court’s resolution of mixed questions of law and fact that “turn[ ] on an evaluation of credibility and demeanor”). However, as we explain more fully below, the ultimate question of egregious harm in this case is a mixed question of law and fact that does not turn on an evaluation of credibility and demean- or; thus, contrary to Ambrose’s suggestion, the appellate court was not required to defer to the trial court’s egregious-harm determination. 5 An appellate court owes no deference to a trial court’s determinations when they are actually determinations as to questions of law or mixed questions of law and fact that do not turn on an evaluation of credibility and demeanor. See Martinez v. State, 348 S.W.3d 919, 923 (Tex.Crim.App.2011) (holding that a de novo standard of review applied to “mixed questions that do not depend on credibility determinations”); see also Loserth v. State, 963 S.W.2d 770, 772 (Tex.Crim.App. 1998) (holding that whether an in-court *597 identification has been tainted.is a mixed question of law and fact that does not'turn on an evaluation of credibility and therefore is subject to de. novo review). Applying -that rule here, we conclude- that the court of appeals properly declined to defer to the trial court’s determination-that. Am-brose was egregiously harmed by the absence of the. accomplice-witness instruction. Despite Ambrose’s characterization of the trial court’s determination of egregious harm as a factual finding, it was actually a mixed question of law and fact that did not turn on an evaluation of witness credibility and demeanor in this case. We also note that, regardless of how a trial court labels its findings of fact and conclusions of law, an appellate court must examine the substance of the findings and conclusions and treat them by their substance rather than by their label. See State v. Sheppard, 271 S.W.3d 281, 291-92 (Tex.Crim.App.2008). The trial court made its determination of egregious harm in its “Findings of Fact and Conclusions of Law.” The document consisted of two subheadings,- each with several subparagraphs. Under the first subheading, “Findings of Fact,” the trial court listed what Ramirez testified about, that Ramirez had initially been charged with official oppression, that the State had provided Ramirez with transactional immunity, that Ramirez was an accomplice as a matter of law, and that Ambrose had not requested an accomplice-witness instruction. The trial court did not make any factual determinations related to the credibility or demeanor of Ramirez or any of the other witnesses. Under the second subheading, “Conclusions of Law,” the trial court summarized the law regarding who qualifies as an accomplice, when the trial court is required to provide the accomplice-witness instruction, and it indicated that the defendant must show egregious harm when the omission of the instruction is unobjected to. In the final subpara-graph, the document stated, “This Court was present for all the testimony in this case, has reviewed the legal arguments presented by the State and Defense, and has reviewed the law on this issue. This Court concludes that the failure to include the jury instruction on the accomplice witness rule caused Ms. Ambrose egregious harm, thus requiring a new trial.” • Here, there were no factual findings or credibility determinations as to which a reviewing court should defer. Rather, here the purported factual findings were in actuality mixed questions of law and fact that did not turn on the credibility of the evidence. We, therefore, hold that the trial court’s determination that Ambrose suffered egregious harm is a legal conclusion to which no deference is required by the appellate court. Martinez, 348 S.W.3d at 923. Therefore, we overrule Ambrose’s second ground of error. III. Analysis of Whether Record Demonstrates Egregious Harm Ambrose’s third ground for review alleges that the court of appeals misapplied the Almanza standard in its analysis of whether the omission of the accomplice-witness instruction egregiously harmed Ambrose. The court of appeals held that the trial court’s conclusion that Ambrose was egregiously harmed was in error. We agree that Ambrose was not egregiously harmed, although we reach that conclusion through slightly different reasoning than that employed by the court of appeals, “Jury-charge error is egregiously harmful if it affects the very basis of the case, deprives the defendant of a valuable right, or vitally affects a defensive theory.” Marshall v. State, 479 S.W.3d 840, 843 (Tex.Crim.App.2016) (citation *598 omitted). Under the relevant standard, we have traditionally considered (1) the entirety of the jury charge, (2) the state of the evidence, (3) counsel’s arguments, and (4) any other relevant information revealed by the entire trial record. Id. “[Ojourts are required to review the relevant portions of the entire record to determine whether [a defendant] suffered actual harm, as opposed to theoretical harm, as a result of the error.” Id. (citing Arline V. State, 721 S.W.2d 348, 351-52 (Tex.Crim.App.1986)). “Under the egregious harm standard, the omission , of an accomplice witness instruction is generally harmless unless the corroborating (non-accomplice) evidence is ‘so unconvincing in fact as to render the State’s overall case for conviction clearly, and significantly less persuasive.’ ” Herron v. State, .86 S.W.3d 621, 632 (Tex.Crim.App.2002) (quoting Saunders v. State, 817 S.W.2d 688, 692. (Tex.Crim.App.1991)). In assessing ■ the strength of the non-accomplice evidence, we examine (1) its reliability or believability, and (2) the strength of its tendency to connect the defendant to the crime. Id. .In determining whether the omission of the accomplice-witness instruction egregiously harmed Ambrose, the court .of appeals conducted its harm analysis by examining the non-accomplice evidence for corroboration of Ramirez’s assertion ,that Ambrose intended for the other students to strike A.N. The court of appeals determined that “the only contested element of the charged offense was whether Ambrose intended to subject A.N.. to being struck by other students,” and it therefore focused its analysis on “evidence tending to connect Ambrose’s intent to the intent alleged in the.charge — that Ambrose intended that other students strike A.N.” Am-brose, 457- S.W.3d at 160-61. .However, the appropriate focus of a harm analysis in this case is to examine the non-accomplice evidence for' corroboration “tending to connect” Ambrose to the commission of the offense. Casanova v. State, 383 S.W.3d 530, 539 v (Tex.Crim.App.2012). The corroboration requirement in Article 38.14 does not apply separately to each element of the offense charged or to-each aspect of the accomplice’s testimony. If that were so, there would be no value in the testimony. at all. See Holladay v. State, 709 S.W.2d 194, 199 (Tex.Crim.App.1986) (“It is clearly unnecessary that the accomplice should be confirmed in every circumstance which he details in evidence; for there would-be no occasion to use him at all as a witness if his narrative could be completely proved by other evidence free from all suspicion.”). In- Vasquez v. State, this Court observed that Article 38.14, by its very terms, requires only that there “be some non-accomplice evidence tending to connect the. defendant to the crime, not to every element of the crime.” 56 S.W.3d 46, 48 (Tex.Crim.App.2001) (holding .that Article 38.14 does not require corroboration of. accomplice-witness testimony regarding a deadly-weapon finding). In Jou-bert v. State, this Court similarly explained that the “corroborating evidence under 38.14 need not be. sufficient, standing alone, to prove beyond a reasonable doubt that a defendant committed the offense”; “[t]here need be only some non-accomplice evidence tending to connect the defendant to the crime, not to- every element of the crime.” 235 S.W.3d 729, 731 (Tex.Crim.App.2007) (holding that Article 38.14 does not require corroboration as to defendant’s role in an offense as a principal or as a party). Thus, our analysis to determine whether Ambrose suffered egregious harm from’ the omission of the, accomplice-witness instruction is distinct from that of the eourt of appeals because we do not focus on. the intent element, but rather look only for evidence tending to connect Ambrose to the charged offense. *599 Because there was strong corroborative evidence tending to connect Ambrose to the offense, the totality of the record- fails to show that she was egregiously harmed by the omission of the accomplice-witness instruction. Casanova, 383 S.W.3d at 539 (“As the strength of the corroborating evidence increases, however, a reviewing court may no longer be able to declare that the lack of an accomplice-witness instruction resulted in egregious harm.”). In this case, the record has strong corroborative evidence of Ramirez’s testimony from the testimony by principal Large, assistant-principal Spellmann, and Ambrose herself. Spellmann testified, “[Ambrose] let us know she had instructed her students to hit the student on the arm, but not too hard so that the student would know how it felt to be bullied.” Large testified, “[Ambrose] told us that she instructed the students to hit the other student but not too hard.” Spellmann and Large each recounted that Ambrose stated that two or three students hit A.N. after Ambrose had instructed them to hit A.N. Furthermore, by her own testimony, Ambrose asked, “Does anybody want to show him what it feels like [to be bullied]?” ■ Also according to Ambrose’s own testimony, after she asked that question, a child struck A.N. In light of the testimony from Large, Spell-mann, and Ambrose herself that she either instructed or asked the students to hit A.N. and that A.N. was hit one or more times after that, there is an adequate amount of non-accomplice testimony that tends to connect Ambrose to the offense of official oppression. Unlike the court of appeals’s analysis, we do not additionally determine whether there is an adequate amount of non-accomplice testimony that tends to connect-Ambrose to the specific element of her intent to commit the offense because it is unnecessary to corroborate every element of an offense for the evidence to be adequate. Although we reach the same conclusion as the court of appeals by a slightly different analysis, we agree with that court that the totality of the record fails to show. egregious harm from the omission of -the accomplice-witness instruction. .We overrule Ambrose’s third ground for review. . IV. Conclusion We hold that the court of -appeals properly applied the Almanza egregious-harm standard in this case in which the jury-charge error was raised in a motion for new trial. We also hold that the court of appeals properly declined to defer to the trial court’s fact findings in this case because those findings were in actuality matters of law or mixed questions of law and fact that did not turn on the credibility of the evidence or demeanor of the witnesses. We further hold that the court of appeals correctly concluded that Ambrose did not suffer egregious harm from the omission of the accomplice-witness instruction. Therefore, we overrule all three of Am-brose’s grounds for review and affirm the judgment of the court of appeals. YEARY, J., did not participate. Footnotes [1] The three grounds in Ambrose’s petition state: (1) When a trial judge issues findings of fact and conclusions of law that find a defendant suffered egregious harm from unobjected to jury charge error, does applying the Almanza egregious harm standard on appellate review violate and conflict with Texas (Ex parte Wheeler, 203 S.W.3d 317 (Tex.Crim.App,2006)) and United States Supreme Court (Oregon v. Kennedy, 456 U.S. 667, 102 S.Ct. 2083, 72 L.Ed.2d 416 (1982)) precedent that a reviewing court must defer to a lower court’s factual findings? (2) Under the egregious harm standard, does an appellate court violate Texas (Ex parte Wheeler, 203 S.W.3d 317 (Tex.Crim.App.2006)) and United States Supreme Court (Oregon v. Kennedy, 456 U.S. 667, 102 S.Ct. 2083, 72 L.Ed.2d 416 (1982)) precedent when it ignores a trial court’s factual findings and substitutes its own view of the-evidence for that of the trial [court]? (3)If the egregious harm standard does apply on direct review in this case, did the appellate court correctly apply the ’ egregious harm standard when it only considered the testimony that supported the State’s case and not "the entire jury charge, the state of the evidence, including-the contested issues and weight of probative evidence, the argument of counsel and any other relevant information revealed by the record of the trial as a whole” as required by Almanza v. State, 686 S.W.2d 157 (Tex.Crim.App.1984) [2] Section 39.03 of the Texas Penal Code states in relevant part, (a) A public servant acting under color of his office of employment commits an offense if he: (1) intentionally subjects another to mistreatment .., that he knows is unlawful.... (b) For purposes of this section, a public servant acts under color of his office of employment if he acts or purports to act in an official capacity_ Tex. Penal Code § 39.03. [3] In its findings of fact and conclusions of law in support of its ruling, the trial court summarized the substance of Ramirez's testimony, but it did not make any finding regarding the credibility of that testimony. The trial court further found that Ramirez was provided transactional immunity in exchange for her testimony. In addition, based on evidence that was presented in the post-trial proceedings, the trial court found it to be true that Judson ISD police initially intended to file charges against Ramirez "for official oppression and [for] failing to report this event,” but later decided to pursue charges only against Ambrose. The findings further state, "When asked by this Court why Ms. Ramirez received immunity during Ms. Ambrose’s sentencing, the prosecutor for the State responded that [Ramirez’s] testimony was necessary for the State’s case because she was the only other adult to witness the events.” The findings indicate that Ramirez was an accomplice as a matter of law, or, "[a]t the very least,” she was an accomplice as a matter of fact. The findings note that, although the trial court had initially prepared jury instructions that included an accomplice-witness instruction, that instruction was not ultimately included in the instructions that were submitted to the jury. The findings state that "[n]o request for a jury instruction on accomplice as a matter of fact or as a matter of law was requested in this case,” and they further state that Am-brose’s trial counsel testified at the hearing on the motion for new trial that he "ultimately decided [not to] request an instruction for his own strategic purposes.” In its conclusions of law, the trial court summarized both the law on accomplice-witness testimony and the general law governing errors in the charge. Applying that law, the court concluded that the failure to include the accomplice-witness instruction had caused Ambrose egregious harm. [4] Article 36.19 of the Code of Criminal Procedure states, Whenever it appears by the record in any criminal action upon appeal that any requirement of Articles 36.14, 36.15, 36.16, 36.17 and 36.18 has been disregarded, the judgment,shall not be reversed unless the error, appearing from the record was calculated to injure the rights of defendant, or unless it appears from the record that the defendant has not had a fair and impartial trial. All objections to^ the charge and to the refusal of special charges shall be made at the time of the trial. Tex.Code Crim. Proc, art. 36.19. [5] Because we conclude here that the question of whether there is egregious harm is a mixed question of law and fact that does not turn on an evaluation of witness credibility and demeanor, we decide that the appellate court was not required to defer to the trial court’s egregious-harm determination in this case. We do not reach the more difficult question of whether a determination of egregious harm could ever be a mixed question of law and fact that does turn on an evaluation of credibility and demeanor, and, if so, whether under those circumstances the appellate court would be bound to defer to some or all of the trial court’s factual findings or its conclusions of law that turn on an evaluation of credibility and demeanor.