An all-white jury in Black Hawk County convicted the defendant, a black man, of one count of harassment in the first degree, an aggravated misdemeanor under Iowa Code section 708.7(1)(6) (2015). On appeal, the defendant raises several claims of error, including that the district court erred in using only the absolute disparity method for assessing the representativeness of a jury pool when considering whether the racial composition of the jury pool violated the defendant’s Sixth Amendment right to an impartial jury. We abandon the exclusive use of absolute disparity as a test for jury representativeness under the Sixth Amendment and permit absolute disparity, comparative disparity, and standard deviation analyses to be used. Because the defendant’s other claims of error fail, we conditionally affirm his conviction and remand for further proceedings on the issue of whether the racial composition of the jury pool violated his Sixth Amendment right to an impartial jury.
I. Factual and Procedural Background.
Kelvin Plain Sr. and Randy Gray were neighbors with an acrimonious relationship who lived in a small apartment building in Waterloo, Iowa. Around 11:00 p.m. on July 10, 2015, Gray and his spouse—both Caucasians—were asleep in their apartment when they were awakened by the sound of Plain—an African-American—sweeping a stairwell in the common area of the apartment building. Gray left his apartment to tell Plain to stop sweeping. When Plain refused, an argument ensued.
At trial, Gray testified that during the argument Plain went back to his apartment while he remained in the common area. Plain then returned with “a black handle of something in his right pocket” and the two men continued to argue. According to Gray, Plain then told him he had a gun and knew his mother-in-law’s car was parked outside. At that point, Plain exited the building and Gray followed, brandishing a golf club.
*810Gray testified that as he stepped - outside, Plain threw a pair of bolt cutters at his head. Gray averred the bolt cutters hit the side of the apartment building so loud they made a sound he thought was. a gunshot.
At that point, Gray picked up the bolt cutters and stood on the porch of the apartment building with the golf club and prevented Plain from going back inside. Gray testified Plain then threatened him repeatedly,
Gray’s spouse, who had been on the phone making a 911 call during the argument, followed the two men outside. She testified that Plain threw the bolt cutters and that he threatened to shoot, cut, and stab her husband.
When law enforcement officers arrived at the scene, they found Gray on the front porch and Plain in the yard. Plain told them he had been sweeping the building when Gray came out of his apartment waiving a golf club at him. Gray and his spouse told the officers Plain had thrown the bolt cutters. After interviewing everyone involved, officers seized the bolt cutters and arrested Plain.
On August 19, 2015, Plain was charged by trial information with harassment in the first degree in violation of Iowa Code section 708.7(2), an aggravated misdemeanor.
On the first day of trial, Plain objected to the racial composition of the jury pool, alleging a violation of his Sixth Amendment right to an impartial jury. Although African-Americans represent 8.9% of the population of Black Hawk County, the pool of potential jurors included only one African-American man among fifty-six potential petit jurors—or 1.8% .of the group. Plain did not present any evidence of systematic exclusion and conceded this meant he could not prove a prima facie case; however, he asserted this was because the jury manager did not provide him with the six months’ worth of data on jury pools that he requested.
During trial, Plain raised a hearsay objection to testimony from the officer about what the officer learned from the alleged victim and his spouse, but the court concluded the testimony was not hearsay and admitted it. Plain requested a mistrial after determining the 911 recording in evidence contained references to his criminal history, but the judge denied the motion and gave a cautionary instruction instead. Plain objected to the prosecutor’s repeated reference to Gray as the “victim” during closing argument, but .the court overruled the objection. The court also denied- Plain’s request for a jury instruction addressing implicit racial bias.
The jury convicted Plain of one count of harassment in the first degree, an aggravated misdemeanor' in violation of Iowa Code sections 708.7(1)(6) and (2). The court imposed a two-year prison sentence, but suspended it and ordered a term of probation running consecutively after a sentence on a parole violation. After filing a motion for a new trial, which was denied, Plain appealed. We retained the appeal.
II. Standard of Review.
We review constitutional issues de novo. State v. Chidester, 570 N.W.2d 78, 80 (Iowa 1997). We review the admission of evidence' challenged as hearsay for the correction of errors at law. State v. Dudley, 856 N.W.2d 668, 675 (Iowa 2014). Improperly admitted hearsay constitutes grounds for reversal unless the proffering party establishes the error was not prejudicial. Id.
Plain’s remaining claims are reviewed for an abuse of discretion. “Trial courts have broad discretion in ruling on claims of prosecutorial misconduct and we review such rulings for an abuse of discre*811tion,” State v. Jacobs, 607 N.W.2d 679, 689 (Iowa 2000). We review denials of a mistrial and the giving of a cautionary instruction for an abuse of discretion. State v. Wade, 467 N.W.2d 283, 285 (Iowa 1991). Finally, we generally review a district court’s refusal to give a requested jury instruction for errors at law; however, if the jury instruction is not required but discretionary, we review for an abuse of discretion. Alcala v. Marriott Int'l, Inc., 880 N.W.2d 699, 707-08 (Iowa 2016); Herbst v. State, 616 N.W.2d 582, 585 (Iowa 2000).
When assessing a district court’s decision for abuse of discretion, we Only reversé if the' district court’s decision rested on grounds or reasoning that were clearly untenable or clearly unreasonable. Dudley, 856 N.W.2d at 675. Grounds or reasons are untenable if they are “based on an erroneous application of the law or not supported by substantial evidence.” Id.
III. Analysis.
Plain raises five .issues on appeal. First, he argues the judge admitted prejudicial hearsay. Second, he asserts a reference in the 911 recording to his past criminal conduct was grounds for a mistrial. Third, he contends the judge improperly denied his request for an implicit-bias jury instruction. Fourth, he insists the prosecutor’s repeated references to Gray as a victim constituted prosecutorial error and violated his due process rights. Finally, he- asserts that the racial composition of the jury pool violated his Sixth Amendment right to an impartial jury.
A. Prejudicial Hearsay. The first issue on appeal is whether the district' court’s admission of testimony, from an officer about what he learned from two witnesses—Gray and his spouse—constituted reversible error. We begin with the testimony at issue:
Q. Based on your or Officer Shaaf s conversation with Mr. [Gray] and [his spouse], did you learn what caused the mark on the [w]all? A. Yes, we did.
Q. And what caused the mark? A. It was the bolt cutters hitting the wall after being thrown. , . .
The district court admitted this testimony after instructing the jury it could only be used as evidence-,of the officer’s subsequent course of conduct, not as evidence that the bolt cutters caused the mark on the wall after being thrown.
1. Hearsay. We must first deter-miné whether the investigating officer’s testimony about what he learned from the alleged victim and his spouse about the cause of the mark on the wall constitutes hearsay. We conclude that it does.
Under Iowa law, hearsay is defined as “a statement, other than one made by the declarant while testifying at the trial or hearing, offered in evidence to prove the truth of the matter asserted.” Iowa R. Evid. 5.801(c) (2016).1 We generally exclude hearsay from admission to the record, subject to the exceptions and exclusions set forth in rules 5.802 and 5.803, because we deem it inherently unreliable— the declarant typically does not speak under oath and the party against whom the evidence is offered does not have the opportunity to cross-examine the declarant in order to test the declarant’s perception, memory, or narration. See State v. Smith, 876 N.W.2d 180, 185 (Iowa 2016).
*812Evidence is not hearsay if it is not offered to show the truth of the matter asserted. State v. Mitchell, 450 N.W.2d 828, 832 (Iowa 1990). An out-of-court statement offered only to explain responsive conduct that is relevant to an aspect of the state’s case is not offered to prove the truth of the matter asserted and is therefore not hearsay. Id. But “if the evidence is admitted, the court must limit its scope to that needed to achieve its purpose.” McElroy v. State, 637 N.W.2d 488, 502 (Iowa 2001).
In deciding whether an out-of-court statement is offered to explain responsive conduct, the court considers “whether the statement is truly relevant to the purpose for which it is being offered, or whether the statement is merely an attempt to put before the fact finder inadmissible evidence.” Mitchell, 450 N.W.2d at 832. In several cases, we have concluded out-of-court statements made to law enforcement officers were admissible because they were truly relevant for a purpose other than to prove the truth of the matter asserted. In State v. Mann, for example, we concluded an officer’s testimony about statements from other officers that led her to believe someone was a victim of sexual assault was not hearsay when used to explain why the officer transported the alleged victim to a hospital for an examination. 512 N.W.2d 528, 536 (Iowa 1994). More recently, in State v. DeWitt, we concluded an officer’s testimony about information he received from a confidential source was not hearsay when offered to explain why the police went to a store and approached the defendant “rather than another customer displaying similar behavior.” 811 N.W.2d 460, 477 (Iowa 2012).
In other cases, we have rejected claims that out-of-court declarations were admissible because they explained responsive conduct of law enforcement officers. For example, in State v. Tompkins, an officer testified that a witness stated a domestic abuse defendant pushed his girlfriend. 859 N.W.2d 631, 636 (Iowa 2015). We concluded the officer’s account of the declarant’s out-of-court-statement constituted inadmissible hearsay because it “went beyond the mere fact that a conversation occurred and instead actually stated what the witness said.” Id. at 643. We further concluded the out-of-court declaration recounted by the officer “did not merely explain the investigation” but instead “directly challenged [defense] counsel’s assertion” to the contrary. Id. In State v. Elliott, we opined that when the “investigating officer specifically repeats a victim’s complaint of a particular crime, it is likely that the testimony will be construed by the jury as evidence of the facts asserted.” 806 N.W.2d 660, 667 (Iowa 2011) (quoting State v. Mount, 422 N.W.2d 497, 502 (Iowa 1988), overruled on other grounds by State v. Royer, 436 N.W.2d 637, 639-40 (Iowa 1989)).
Here, Plain’s overarching argument is that the State solicited the testimony at issue not to explain the officer’s conduct but instead to use the weight of the officer’s authority- to bootstrap statements from the alleged victim and his wife about the cause of the mark on the wall. The State responds that the officer’s testimony is relevant to explain the officer’s subsequent decision to take the bolt cutters into evidence and Plain into custody.
We believe this case most closely resembles the situation in Tompkins. Here, as in Tompkins, an officer did not merely testify that a conversation occurred and about the conclusions he reached as a result. See Tompkins, 859 N.W.2d at 643. The officer testified that he learned from the alleged victim and his wife that the bolt cutters had been thrown, causing a mark in the wall. The jury is likely to consider such a statement to be evidence of the fact assert*813ed. See Elliott, 806 N.W.2d at 667. Moreover, the State did not ask why the officer took the bolt cutters into evidence but instead asked what caused the mark. We think this fact is telling and conclude “the statement is merely an attempt to put before the fact finder inadmissible evidence.” See Mitchell, 450 N.W.2d at 832. Because the primary purpose of the testimony was to prove the truth of the matter asserted, as in Tompkins, we find the officer’s testimony about what the alleged victim and his wife told him about the cause of the hole in the wall to be inadmissible hearsay.
2. Prejudice. We must also consider whether Plain suffered prejudice from the admission of the hearsay evidence. In assessing prejudice, we place the burden on the state to affirmatively establish that the admission of hearsay evidence over proper objection was not prejudicial. Elliott, 806 N.W.2d at 669. The burden to affirmatively establish lack of prejudice is met “if the record shows the hearsay evidence did not affect the jury’s finding of guilt.” Id. Tainted evidence that is merely cumulative does not affect the jury’s finding of guilt. Id.
The evidence tending to prove Plain threw the bolt cutters was strong. The State offered physical evidence that the bolt cutters were thrown against the wall. The evidence included a photograph of the hole in the wall and the bolt cutters. The bolt cutters were damaged and had paint markings consistent with the color of the house. Notably, the physical evidence and the photograph came into the record without objection, see State v. McGuire, 572 N.W.2d 545, 547-48 (Iowa 1997) (concluding there was no prejudice where “substantially the same evidence [came] into the record without objection”), and it was consistent with the testimony given by Gray and his wife. Accordingly, we conclude the challenged hearsay evidence was merely cumulative and therefore not prejudicial. See State v. Hildreth, 582 N.W.2d 167, 170 (Iowa 1998).
In addition, the trial court gave a jury instruction limiting the purposes for which the hearsay testimony could be used. We have held that an instruction limiting the “purposes for which this evidence [can] be used” may serve as “an antidote for the danger of prejudice.” See State v. Putman, 848 N.W.2d 1, 15-16 (Iowa 2014). In this case, the district court instructed the jury not to consider the out-of-court declaration as evidence of the truth of the matter asserted and did so prior to the testimony being given.
Because the officer’s testimony was cumulative of other evidence and the court gave an instruction limiting the purposes for which the evidence could be used, we find no reversible error on this issue.
B. Reference to Past Conduct. The second issue on appeal is whether the district court abused its discretion when it denied Plain’s motion for a mistrial after a 911 call was played to the jury that allegedly contained prejudicial references to Plain’s probation status and status as a felon.2 We begin with the evidence at issue.
1. The evidence at issue. During the first day of trial, the State played for the jury a redacted recording of a 911 call Gray’s spouse made during the incident. *814After the trial adjourned for the night, defense counsel listened to the recording and heard statements on the tape that Plain was wearing a GPS monitoring device and was not afraid to go “back, to prison.” The following morning, Plain requested a mistrial.
The district court denied the request for a mistrial', noting it had “not hear[d] any reference to going back to prison or the GPS device” when the recording was played for the jury.3 The court offéred instead to give a cautionary limiting instruction to the jury to alleviate any potential harm, if Plain wished. But the cohrt cautioned that “from [its] listening to the 911 call, ... the cautionary instruction would almost be bringing attention to something that they didn’t hear.”
Defense counsel subsequently requested—and the court gave—a cautionary instruction. The instruction stated,
As to ... the 911 recording, there may have been references to a GPS monitoring device and/or time spent in prison. The jury is to disregard any statement made regarding a GPS monitoring device and/or time spent in prison during the course of your deliberation and in reaching your verdict.
2. Discussion. On appeal, Plain asserts the court abused its discretion in giving a cautionary limiting instruction instead of granting his motion for, a mistrial.
We must first determine whether the evidence was inadmissible. Under Iowa Rule of Evidence 5.404(6) (2015),
Evidence of other crimes, wrongs, or acts is not admissible to prove the character of a person in order to show that the person acted in conformity therewith. It may, however, be admissible for other purposes, such as proof of motive, opportunity, intent, preparation, plan, knowledge, identity, or absence of mis-, take or accident.
Rule 5.404(6) excludes evidence of other crimes not on grounds of relevance1 but “based On the premise that a jury will tend to give other crimes, wrongs, or acts evidence excessive weight and the belief that a jury should not convict a person based on his or her previous misdeeds.” State v. Nelson, 791 N.W.2d 414, 425 (Iowa 2010).
Rule 5.404(6) expressly permits, evidence of other crimes, wrongs, or acts for “proof of motive, opportunity, intent, preparation, plan, knowledge, identity, or absence of mistake ■ or accident.” Iowa R. Evid. 5.404(6). This list is not exclusive; rather, we admit evidence of other crimes, wrongs,- or acts if there is a noncharacter theory of relevance and the evidence is material -to a- legitimate issue other than the defendant’s general criminal disposition. Nelson, 791 N.W.2d at 425. Even if there is a noncharacter theory of relevance, “the probative value of the evidence [must]. not [be] substantially outweighed by the danger of unfair prejudice to the defendant.” State v. Cox, 781 N.W.2d 757, 761 (Iowa 2010).
. In this: Case, the State did not offer a noncharacter theory of relevance that is probative of a- factor other than a general propensity to commit wrongful acts.- Nor *815does the evidence fit into a category expressly permitted by rule 5.404(6). Thus, the evidence of Plain’s prior criminal history is inadmissible under Iowa law.
Because the evidence was inadmissible, we must determine whether the court abused its discretion by granting a cautionary limiting instruction rather than a mistrial. Generally, a district court’s decision not to grant a mistrial but to offer a cautionary instruction instead is entitled to broad deference. See Wade, 467 N.W.2d at 285. Cautionary instructions are sufficient to mitigate the prejudicial impact of inadmissible evidence “in all but the most extreme cases.” State v. Breitbach, 488 N.W.2d 444, 448 (Iowa 1992).
In State v. Belieu, we identified several important considerations for determining whether a cautionary instruction can adequately mitigate the prejudicial impact of inadmissible evidence. 288 N.W.2d 895, 901 (Iowa 1980). The first Belieu consideration is whether the “defendant [can] combat the evidence without compounding the prejudice.” Id. The second consideration is how extensive the evidence is and the promptness with which it was addressed. Id. at 901-02. Finally, we assess prejudice—the stronger the State’s evidence of Plain’s guilt is, the less prejudicial the effect of the challenged testimony. Id. at 900-01.
Applying these three considerations to the case’at hand, we conclude the cautionary instruction adequately mitigated any prejudicial impact of the evidence. Although Plain’s counsel had opportunity to listen to the redacted copy of the recording before the trial, and apparently did listen to at least portions of it prior to trial, once the recording was played for the jury, Plain lacked the ability to combat the evidence without compounding the prejudice. See id. at 901.
However, the evidence was not .extensive and the district court promptly addressed the matter. See id. at 901-02. Unlike in Belieu where we found numerous references to a defendant’s .criminal history to be “so pervasive .and central to the defenses that its prejudicial effect against this defendant could not reasonably be cured by a limiting instruction,” see id., the references here were brief and inadvertently included in the redacted version of the recording. In such circumstances of brief and inadvertent presentations of evidence, we have concluded a cautionary instruction sufficiently managed the risk of prejudice. See, e.g., id. at 901-02. Furthermore, the statements were not the substantive focus of the recording or very obvious; they were presented with the background noise accompanying a phone conversation between a witness at the scene and a 911 operator. Indeed, neither the district court nor the prosecutor recalled having heard the statements when the recording was played for the jury. Upon learning the recording contained the objectionable statements, the district court acted swiftly. Although the court concluded a mistrial was unwarranted, the court proposed and gave a cautionary limiting instruction.
Finally, the State’s evidence was strong. As we noted above, the State presented testimony and corroborative physical evidence and photographs all tending to establish that the bolt cutters were thrown against the wall. Although the statements pertaining to Plain’s prior criminal history included in the 911 recording were inadmissible under rule 5.404(6), they were brief, inadvertent, and did not play a major part in the State’s case. Because the State’s evidence on the contested point was strong, the prejudicial effect of the challenged testimony is minimal. See id. We find .no abuse of discretion in the district court’s ruling denying the motion for *816mistrial or in giving the cautionary limiting instruction.
C. Jury Instruction on Implicit Bias. The third issue on appeal is whether the district court’s refusal to include an anti-discrimination jury instruction requested by Plain constituted reversible error. We generally review refusals to give jury instructions for errors at law; however, if the requested jury instruction is not required or prohibited by law, we review for abuse of discretion. Alcala, 880 N.W.2d at 707-08.
“Iowa law requires a court to give a requested jury instruction if it correctly states the applicable law and is not embodied in other instructions.” Id. at 707 (quoting Sonnek v. Warren, 522 N.W.2d 45, 47 (Iowa 1994)). An issue is material if it is outcome determinative. See C & J Vantage Leasing Co. v. Outlook Farm Golf Club, LLC, 784 N.W.2d 753, 756 (Iowa 2010) (“An issue is ‘material’ [for purposes of summary judgment] if it might affect the outcome of the suit.”). “Parties to lawsuits are entitled to have their legal theories submitted to a jury if they are supported by the pleadings and substantial evidence in the record.” Herbst, 616 N.W.2d at 585 (quoting Sonnek, 522 N.W.2d at 47).
Iowa law permits—but does not require—cautionary instructions that mitigate the danger of unfair prejudice. See Wade, 467 N.W.2d at 285 (recognizing decision to give cautionary instruction is discretionary); see also Breitbach, 488 N.W.2d at 448 (noting limiting instructions may be used to combat potential prejudice from improper evidence). Cautionary instructions include instructions that recommend a course of action to a jury or that limit the jury’s considerations of certain facts. See Breitbach, 488 N.W.2d at 448. We review the issuance or denial of a requested cautionary instruction for abuse of discretion and only reverse if the district court’s decision rested on grounds or reasoning that were clearly untenable or clearly unreasonable. See State v. Newell, 710 N.W.2d 6, 18, 32-33 (Iowa 2006) (concluding issuance of cautionary instruction did not amount to unreasonable exercise of discretion). Grounds or reasons are untenable if they are “based on an erroneous application of the law or not supported by substantial evidence.” Dudley, 856 N.W.2d at 675.
In this case, the requested antidis-crimination jury instruction read,
Reach your verdict without discrimination. In reaching your verdict, you must not consider the defendant’s race, color, religious beliefs, national origin, or sex. You are not to return a verdict for or against the defendant unless you would return the same verdict without regard to his race, color, religious belief, national origin, or sex.
Although the requested instruction correctly states the law, it does not concern a material issue because it is not outcome determinative; therefore, it is not required. On the contrary, the requested instruction is a cautionary instruction because it limits the jury’s consideration of certain facts and recommends a course of action. See Breitbach, 488 N.W.2d at 448. The instruction recommends a course of action by asking the jurors to consciously reflect on their decisionmaking process and limits consideration of facts related to race, color, religious belief, national origin, or sex. Because the requested instruction is a cautionary instruction, the district court’s refusal to give it is reviewed for abuse of discretion. Newell, 710 N.W.2d at 32-33.
The denial of a cautionary instruction constitutes an abuse of discretion if the district court’s decision rested on clearly untenable or unreasonable grounds, *817such as an erroneous application of law. Dudley, 856 N.W.2d at 675; see also Newell, 710 N.W.2d at 18. Here, the district court declined to give the requested implicit-bias instruction because it knew of no authority approving or requiring the instruction and because the instruction was not included in the Iowa State Bar Association’s model instructions.
As explained above, Iowa law permits cautionary instructions designed to mitigate the danger of unfair prejudice. See Wade, 467 N.W.2d at 285; see also Breitbach, 488 N.W.2d at 448. Thus, the cautionary instruction, which is a correct statement of antidiscrimination principles, would have been permitted under Iowa law. The district court, however, refused to give the instruction because it erroneously believed it lacked authority from our court to give the instruction. Because the court’s decision rested on an error of law, it constituted an abuse of discretion.
Our inquiry does not end there. “Error in giving or refusing to give a jury instruction does not warrant reversal unless it results in prejudice to the complaining party.” State v. Hoyman, 863 N.W.2d 1, 7 (Iowa 2015) (quoting State v. Cordero, 861 N.W.2d 253, 257-58 (Iowa 2015)). “When the error is not of constitutional magnitude, the test of prejudice is whether it sufficiently appears that the rights of the complaining party have been injuriously affected or that the party has suffered a miscarriage of justice.” State v. Marin, 788 N.W.2d 833, 836 (Iowa 2010) (quoting State v. Gansz, 376 N.W.2d 887, 891 (Iowa 1985)), overruled on other grounds by Alcala, 880 N.W.2d 699. ‘We [do] not reverse for marginal or technical omissions.... ” Sonnek, 522 N.W.2d at 47. Here, Plain asserts an abuse of discretion, not a violation of the constitution. Under the facts of this case, there was strong evidence of guilt. We therefore conclude the error was not prejudicial.
While there is general agreement that courts should address the problem of implicit bias in the courtroom, courts have broad discretion about how to do so. One of the ways courts have addressed implicit bias is by giving jury instructions similar to the one proposed by Plain in this case. We strongly encourage district courts to be proactive about addressing implicit bias; however, we do not mandate a singular method of doing so. As we conclude Plain was not prejudiced by the denial of the requested instruction in this case, we affirm on this issue.
D. Prosecutor’s Use of the Term “Victim.” The fourth issue on appeal is whether Plain was denied his constitutional right to a fair trial due to prosecutorial misconduct or error. During the State’s closing argument, the district court permitted the prosecutor to refer to the complaining witness as the victim, over Plain’s objection, no fewer than thirty-seven times. In her rebuttal argument, the prosecutor used the word eleven more times. Plain argues the prosecutor’s repeated use of the term with the court’s tacit approval violated his constitutional right to a fair trial under article I, section 9 of the Iowa Constitution and the Fifth and Fourteenth Amendments of the United States Constitution.
In his motion for new trial and in arrest of judgment, Plain asserted the State’s repeated references to the complaining witness as the victim justified a new trial. At the hearing on the motion, the prosecutor defended her actions:
[PROSECUTOR]: It was not improper for me to repeatedly refer to Mr. Gray as the victim during closing arguments. The testimony and evidence in this case showed that Mr. Gray was in fact a victim. And because the testimony *818showed that, I am allowed to argue that in closing arguments.
The court denied the motion.
We consider a claim that the defendant was deprived of a fair trial under our doctrines of prosecutorial error and prosecutorial misconduct. See State v. Schlitter, 881 N.W.2d 380, 393-94 (Iowa 2016) (distinguishing between claims of prosecutorial error and prosecutorial misconduct); State v. Graves, 668 N.W.2d 860, 869 (Iowa 2003) (synthesizing our prosecu-torial misconduct doctrine). In order to establish a violation of the right to a fair trial, a defendant must show both (1) error or misconduct and (2) prejudice. State v. Krogmann, 804 N.W.2d 518, 526 (Iowa 2011).
1. Prosecutorial behavior. The defendant must first establish the prosecutor violated a prosecutorial duty. Prosecutors have a special role in our criminal justice system. A prosecutor is not an ordinary advocate, at least in the sense that “a prosecutor owes a duty to the defendant as well as to the public.” Graves, 668 N.W.2d at. 870. Although a prosecutor should zealously and lawfully advocate on behalf of the state, “the prosecutor’s primary interest should be to see that justice is done, not to obtain a conviction.” Id. Thus, we impose special duties on prosecutors to ensure they act in accordance with the special role with which they are entrusted. See, e.g., id. (“[I]t is improper to ask the defendant whether another witness has lied.”); DeVoss v. State, 648 N.W.2d 56, 64 (Iowa 2002) (noting the prosecutor must “assure the defendant a fair trial”).
‘ We set forth standards governing a prosecutor’s duty to the defendant during closing arguments in Graves. See 668 N.W.2d at 874. There we stated that although we accord prosecutors “some latitude" to analyze evidence and argue “reasonable inferences and conclusions to be drawn from the evidence,” we do not permit a prosecutor to “express his or her personal beliefs.” Id. We emphasized that prosecutors may not “vouch personally as to a defendant’s guilt or a witness’s credibility.” Id. (quoting State v. Williams, 334 N.W.2d 742, 744 (Iowa 1983)). We stressed that
[t]his is true whether the personal belief is purportedly based on knowledge of facts not possessed by the jury, counsel’s experience in similar cases, or any ground other than the weight of the evidence in the trial. A defendant is entitled to have the case decided solely on the evidence.
Id. (emphasis added)' (quoting Williams, 334 N.W.2d at 744). “In addition, the prosecutor is ‘not allowed to make inflammatory or prejudicial statements regarding a defendant in a criminal action.’ ” Id. (quoting State v. Leiss, 258 Iowa 787, 792, 140 N.W.2d 172, 175 (1966)).
A defendant’s constitutional right to a fair trial is violated if a prosecutor fails to comply with the requirements of due process at the trial, whether by virtue of prosecutorial error or misconduct.4 Id. at 870. In addressing fair-trial *819challenges based on prosecutorial behavior, we first determine whether the prosecutor violated a duty to the defendant. See id. If so, we consider whether that violation was intentional or reckless. Schlitter, 881 N.W.2d at 394. An intentional or-reckless violation amounts to prosecutorial misconduct while an unintentional violation amounts only to prosecutorial error. Id. We then determine whether the error caused prejudice.
The behavior at issue here is the prosecutor’s use of the term “victim” twenty-five times during her closing argument over Plain’s objection to refer to Gray, the State’s complaining witness. The prosecutor used the term ten more times during her rebuttal argument. Plain contends this repeated use of the “victim” descriptor denied him a fair trial because it impermis-sibly expressed the prosecutor’s opinion that a crime was, in fact, committed and that Gray was Plain’s victim. The State rejoins that the prosecutor violated no rule when she engaged in spirited advocacy drawing conclusions and arguing inferences reasonably flowing from the record pertaining to Gray’s role in the occurrence that was the subject of the trial.
Other courts have confronted the question of whether a state’s use of the “victim” descriptor interferes with a defendant’s right to a fair trial. The decisions of those courts reveal that the propriety of -the use of the descriptor depends in part on the stage of the trial and the context in which it was used. The Court of Appeals of Texas has concluded a trial court’s use of the term “victim” rather than “alleged victim” in jury instructions constituted reversible error in a case in which the sole issue was whether the complainant consented to sexual intercourse. Talkington v. State, 682 S.W.2d 674, 675 (Tex. Ct. App. 1985); see also Veteto v. State, 8 S.W.3d 805, 816 (Tex. Ct. App. 2000) (concluding use of the term “victim” rather than “alleged victim” in describing the complaining witness in a jury instruction constituted an “improper- comment on the weight of the evidence by the court”), abrogated by State v. Crook, 248 S.W.3d 172, 176-77 (Tex. Crim. App. 2008).
A prosecutor’s use of the word “victim” has not, however, been deemed by other courts to be categorically improper when used in other contexts in criminal cases. The Delaware Supreme Court has noted, for example, that the term “victim” is routinely used in indictments in that state to describe the state’s complaining witness. Jackson v. State, 600 A.2d 21, 24 (Del. 1991). The same court rejected a claim that a trial court committed plain error in allowing a police officer and an FBI forensics expert called by the state to usé the word “victim” in their testimony because law enforcement officers use the descriptor as a “term of art synonymous with ‘complaining witness.’ ” Id. at 24-25.
But the use of the “victim” descriptor by prosecutors during criminal trials has been found improper by other courts in certain contexts. Although infrequent or isolated utterance of the word “victim” by a prosecutor in conveying the state’s position that specific evidence supports'a conviction has been found permissible, pervasive use of the word by a prosecutor during the presentation of evidence and during closing argument has been disapproved. Compare State v. Warholic, 278 Conn. 354, 897 A.2d 569, 584 (2006) (concluding prosecutor’s two closing argument references to a child *820as the victim were likely understood by the jury as reflecting the state’s contention “that, based on the state’s evidence, the complainant was the victim of the alleged crimes” and, therefore, not improper), with State v. Albino, 130 Conn.App. 745, 24 A.3d 602, 617 (2011) (concluding prosecutor’s twenty-seven references to “victim” and repeated references to “murder” and “murder weapon” during evidentiary' phase of trial and in closing argument were improper in a ease in which the defendant asserted self-defense).
Emphasizing the importance of the context in which the word is used, the Delaware Supreme Court has concluded the state’s use of the term during trial causes no cognizable harm “when there is no doubt that a crime was committed and simply the identity of the perpetrator is in issue.” Jackson, 600 A.2d at 24. However, the court cautioned that a prosecutor’s repeated use of the term is improper “in a case where consent [is] the sole defense, and the principal issue one of credibility.” Id.'at 25, clarified on rehearing en banc. The court reasoned that in cases in which the parties dispute whether a crime has been committed, “it is incompatible with the presumption of innocence for the prosecutor to refer to the complaining witness as the ‘victim,’ just as it is to refer to the defendant as a ‘criminal.’ ” Id.
With these principles in mind, we turn to the record. Plain’s defense theory was to characterize the transaction as a mere argument between neighbors. Through cross-examination and argument, defense counsel sought to persuade the jury that the neighbors’ argument was initiated and continued by Gray. Counsel further posited that the State’s evidence tending to prove Plain threatened Gray or threw the bolt cutter at him was incredible and that no crime was committed. The prosecutor consistently and repeatedly referenced Gray as the victim during her closing and rebuttal arguments, making no discernible effort to link that descriptor to any specific evidence in the record.
Although we conclude the prosecutor erred during closing argument in persistently using the term “victim” to refer to the complaining witness under- the circumstances of this case, we do not find that the prosecutor intentionally violated her duty. At the hearing on the motion for new trial and in arrest of judgment, the prosecutor indicated she believed her use of the term was justified because the evidence and testimony established Gray was a victim under the circumstances of this case. Because the limitations on the proper use of the term “victim” by prosecutors in this jurisdiction were not clear at the time of Plain’s trial, we cannot conclude the prosecutor committed misconduct.
2. Prejudice. We next consider whether the prosecutor’s error “resulted in prejudice to such an extent that the defendant was denied a fair trial.” Graves, 668 N.W.2d at 869. In determining prejudice, we consider
(1) the severity and pervasiveness of the misconduct; (2) the significance of the misconduct to the central issues in the case; (3) the strength of the State’s evidence; (4) the use of cautionary instructions or other curative measures; and (5) the extent to which the defense invited the misconduct.
Id. (citations omitted).
Applying these factors to the case at hand, we find insufficient prejudice to merit reversal on this issue. A couple of the factors weigh in favor of a finding of prejudice. In particular, we note the defendant did nothing to invite the prosecutor’s behavior and affirmatively objected to it without avail. Also, whether Gray was a *821victim was highly central to Plain’s defense that no crime was committed.
Other factors weigh more strongly against a determination that Plain suffered prejudice. First, we note the- prosecutor’s use of the term “victim” was limited to closing arguments and the district court instructed jurors that “[the] summations and closing arguments of counsel are not evidence,” thus mitigating the term’s prejudicial effect. Second, we find the physical and photographic evidence in this case, the direct evidence of verbal threats made by Plain and recorded on the 911 call, and the consistent witness testimony presented by the State are strong evidence in support of Plain’s conviction. Accordingly, we conclude the prosecutorial error was not prejudicial and thus do not reverse on this basis.
E. Sixth Amendment Right to Impartial Jury. The final issue on appeal is whether the racial composition of the jury pool violated Plain’s Sixth Amendment right to an impartial jury. As we have noted, the jury pool in Plain’s case only contained one African-American man among fifty-six potential petit jurors.5 Although African-Americans comprise 8.9% of Black Hawk County’s population overall, they represented only 1.8% of the jury pool summoned for Plain’s trial. Plain, who is African-American, objected to the composition of the jury pool and requested a new panel. The district court overruled his objection and denied his request.
1. The right to an impartial jury. The Sixth Amendment to the United States' Constitution provides that “[i]n all criminal prosecutions, the accused shall enjoy the right to a speedy and public trial, by an impartial jury of the State and district wherein the crime shall have been committed.”6 U.S. Const, amend. VI. The right to án impartial jury entitles the criminally accused to a jury drawn from a fair cross-section of thé community. Taylor v. Louisiana, 419 U.S. 522, 530, 95 S.Ct. 692, 697-98, 42 L.Ed.2d 690 (1975).
A jury that represents a fair cross-section of the community enables “the com-monsénse judgment of the community [to serve] as a hedge against the overzealous or mistaken prosecutor and in preference to the professional or perhaps overcondi-tioned or biased response of a judge.” Id. at 530, 95 S.Ct. at 698. It also helps legitimize the legal system and is “critical to public confidence in the fairness of the criminal justice system.” Id. Finally, it encourages civic participation through the shared administration of justice. Id.
2. The Duren three-part test. In Duren v. Missouri, the Supreme Court defined a three-part test for establishing a violation of the fair cross-section requirement. 439 U.S. 357, 364, 99 S.Ct. 664, 668, 58 L.Ed.2d 579 (1979); see also State v. Watkins, 463 N.W.2d 411, 414 (Iowa 1990). Under this three-part test, a defendant can establish a prima facie violation of the fair cross-section requirement by showing
*822(1) that the group alleged to be excluded is a “distinctive” group in the community; (2) that the representation of this group in venires from which juries are selected is not fair and reasonable in relation to the number of such persons in the community; and (3) that this un-derrepresentation is due to systematic exclusion of the group in the jury-selection process.
Duren, 439 U.S. at 364, 99 S.Ct. at 668. If the defendant establishes a prima facie case, the burden shifts to the state to justify the disproportionate representation by proving “a significant state interest” is “manifestly and primarily advanced” by the causes of the disproportionate exclusion. Id. at 367-68, 99 S.Ct. at 670.
Under the first Duren prong, a defendant must-establish membership in a distinctive group under community standards, id. at 364, 99 S.Ct. at 668, meaning a community group with “a definite, objectively ascertainable membership” that “constitutes a substantial segment of the population” and has “common and unique opinions, attitudes, and experiences” that cannot be adequately represented by members of the general population. Thomas M. Fleming, Age Group Underrepresentation in Grand Jury or Petit Jury Venire, 62 A.L.R. 4th 859, 867 (1988). In other words, a defendant must show she has “characteristics that are relevant to constituting a jury venire that is representative of the community.” David M. Coriell, Note, An (Un)fair Cross Section: How the Application of Duren Undermines the Jury, 100 Cornell L. Rev. 463, 480 (2015) [Hereinafter Coriell]. Under Iowa and federal law, these characteristics might include race, color, religion, sex, national origin, or economic status. Compare Iowa Code §§ 607A.2, - .4, .10, .19-.22, with 28 U.S.C. §§ 1862,1864 (2012).
Under the second Duren prong, a defendant must establish the proportion of group members in the jury pool is under-representative of the proportion of group members in the community. See, e.g., Duren, 439 U.S. at 364-66, 99 S.Ct. at 669. The second prong “distinguishes between acceptable and unacceptable levels of deviation for the representation of a distinctive group on a jury venire from their representation in the community.” Coriell, 100 Cornell L. Rev. at 480. To measure representation, jurisdictions generally apply one Or more of the following statistical tests: (1) absolute disparity, (2) comparative disparity, and/or (3) standard deviation. United States v. Hernandez-Estrada, 749 F.3d 1154, 1160 (9th Cir. 2014).7
The first test, absolute disparity, is useful as a quick dipstick for providing a rough gauge of the representativeness of a jury pool. United States v. Rogers, 73 F.3d 774, 777 (8th Cir. 1996) (panel decision). Absolute disparity is calculated “by taking the percentage of the distinct group in the population and subtracting froth it the percentage of that group represented in the jury panel.” State v. Jones, 490 N.W.2d 787, 793 (Iowa 1992). The lower the resulting percentage, the more representative the jury pool.8
*823The absolute disparity formula does not account for the relative size of the minority group in the general population. It instead excludes any minority population that makes up a percentage of the population that is lower than the permissible amount for absolute disparity. For example, if absolute disparity is established at ten percent, see id., then it excludes any minority population that accounts for less than ten percent of the population from the Sixth Amendment’s protections.
The second test, comparative disparity, is useful in examining the relative size of the minority group in the general population. Rogers, 73 F.3d at 777 (“[I]t calculates the representation of African Americans in jury pools relative to the African-American community rather than relative to the entire population.”). Comparative disparity is calculated by dividing the absolute disparity by the percentage of the population represented by the group in question. See United States v. Sanchez, 156 F.3d 875, 879 n.4 (8th Cir. 1998). The higher the comparative disparity percentage, the less representative the jury pool.
The primary shortcoming of the comparative disparity test is that it can overstate underrepresentation for groups with a small population percentage. Hernandez-Estrada, 749 F.3d at 1163 (“Let’s assume, for example, that a group constitutes .1% of the population, with no representation in the jury pool. The comparative disparity of that group would be 100%. Few would argue that the absence of a group representing just 0.1% of the population violates the fair cross-section requirement, yet comparative disparity analysis would suggest otherwise.”). In addition, with very large groups, comparative disparity tends to validate deviations that are not produced by chance even though that can alter the repi’esentativeness of the average jury significantly. Id.
The final test, standard deviation, is useful to measure “predicted fluctuations from the expected value.” Castaneda v. Partida, 430 U.S. 482, 496 n.17, 97 S.Ct. 1272, 1281 n.17, 51 L.Ed.2d 498 (1977). Standard deviation is calculated by analyzing a sample taken from the voter wheel and analyzing it for randomness and fluctuations. Id.; see also United States v. Weaver, 267 F.3d 231, 238-39 (3d Cir. 2001).
Standard deviation, too, is an imperfect test for assessing disproportional minority representation. Measures of the standard deviation presume randomness; however, the chances of drawing a particular jury composition are not random, in part because “the characteristics of the general population differ from a pool of qualified jurors.” Hernandez-Estrada, 749 F.3d at 1163.
Under the third Duren prong, a defendant must establish that systematic exclusion of the group caused the under-representation of the group.9 See Duren, *824439 U.S. at 364, 99 S.Ct. at 668. The third prong “distinguishes between situations where a particular jury venire is nonrepre-sentative and those situations where the jury venires in a district are continuously nonrepresentative of the community.” Coriell, 100 Cornell L. Rev. at 481. To establish systematic exclusion, a defendant must establish the exclusion is “inherent in the particular jury-selection process utilized” but need not show intent. Duren, 439 U.S. at 366, 99 S.Ct. at 669. In other words, the defendant must show evidence of a statistical disparity over time that is attributable to the system for compiling jury pools. Coriell, 100 Cornell L. Rev. at 481. “If there is a pattern of underrepresentation of certain groups on jury venires, it stands to reason that some aspect of the jury-selection procedure is causing that under-representation.” Id.; accord Duren, 439 U.S. at 366, 99 S.Ct. at 669-70.
3. State v. Jones. We initially endorsed the absolute disparity test and rejected the comparative disparity test in Jones. See 490 N.W.2d at 793. In Jones, a five-justice panel of our court considered whether the Scott County procedure for compiling a jury list that relied solely on voter registration lists and drivers’ license records violated the fair cross-section requirement of the Sixth Amendment. Id. at 791-92. The jury panel at issue in that case consisted of less than 3.5% African-Americans while Scott County was 4.1% African-American. Id.
Concluding that the United States Supreme Court had adopted the absolute disparity test, we expressly rejected the comparative disparity test. Id. at 793. Citing precedents from the United States Courts of Appeals for the Eighth and Ninth Circuits, we concluded an absolute disparity of 1.5% was not unfair or unreasonable under Duren. Id.
We affirmed two other cases in which the absolute disparity test was utilized in the early 1990s. In State v. Huffaker, we concluded an absolute disparity of 2.85% between the number of African-Americans in Polk County (4.52%) and the percentage of African-Americans in the jury venire (1.67%) was not sufficient to constitute a violation of Duren’s second prong. See 493 N.W.2d 832, 833-34 (Iowa 1992). In Thongvanh v. State, we decided a 0.18% absolute disparity between Asians selected for jury duty (0%) and Asians in the general population of a county (.18%) did not *825violate the fair cross-section requirement. 494 N.W.2d 679, 683 (Iowa 1993).
i. Courts criticize the absolute disparity test. At the time they were decided, the scope of the Supreme Court’s decisions in Swain v. Alabama, 380 U.S. 202, 85 S.Ct. 824, 13 L.Ed.2d 759 (1965), and Castaneda, 430 U.S. 482, 97 S.Ct. 1272, was unclear. We mistakenly viewed the Supreme Court’s acknowledgement of the absolute disparity test in Swain and Castaneda as tacit approval of the absolute disparity model of statistical computation. See Jones, 490 N.W.2d at 792-93 (discussing Swain, 380 U.S. at 208-09, 85 S.Ct. at 829, and Castaneda, 430 U.S. at 495, 97 S.Ct. at 1280). After Jones was decided, the Supreme Court acknowledged the comparative disparity formula and followed a case-by-case approach permitting consideration of the results of multiple analytical models. Berghuis v. Smith, 559 U.S. 314, 329-30, 130 S.Ct. 1382, 1393, 176 L.Ed.2d 249 (2010) (discussing People v. Smith, 463 Mich. 199, 615 N.W.2d 1, 2 (2000), and Smith v. Berghuis, 543 F.3d 326, 338 (6th Cir. 2008), rev’d and remanded on other grounds, 559 U.S. 314, 130 S.Ct. 1382, 176 L.Ed.2d 249 (2010)). Thus, we can no longer say that the Supreme Court’s acknowl-edgement of the absolute disparity model is a preference for that approach.
Further, our decision in Jones relied on authorities from the United States Courts of Appeals for the Eighth and Ninth Circuits. Shortly after we decided Jones, the Eighth Circuit affirmed a defendant’s conviction under Eighth Circuit precedent but sharply criticized the circuit’s reliance on absolute disparity analysis and opined that the comparative disparity model should also be used. Rogers, 73 F.3d at 775-76. Subsequently, the Ninth Circuit expressly overruled its prior precedent that had deemed the absolute disparity formula to be the exclusive analytical tool. See Hernandez-Estrada, 749 F.3d at 1163.
ii. Practical issues with Jones. Exclusive use of the absolute disparity test creates problems of constitutional significance in Iowa. The test offers less protection for a minority group as the group’s percentage of the community’s total population decreases. Id. at 1162. In adopting the absolute disparity test in Jones, we noted that the Supreme Court had determined that “the underrepresentation of as much as ten percent” did not establish a prima facie case for the second Duren prong. 490 N.W.2d at 793.
The shortcoming of the absolute disparity formula is demonstrated by the fact that African-Americans do not represent more than ten percent of the population in any county in Iowa. See Iowa Black Population Percentage, 20IS by County, www. indexmundi.com/facts/united-states/quick-faets/iowa/black-population-percentage#chart (last visited June 27, 2017). In the seven Iowa counties with the highest African-American population— Black Hawk (8.9%), Scott (7.4%), Polk (6.5%), Des Moines (5.7%), Johnson (5.5%), Linn (4.3%), and Webster (4.1%)—an African-American could not establish a racially unrepresentative jury using the absolute disparity model under the Sixth Amendment even if the exclusion of African-Americans was total and systematic. Outside of the Hispanic and Latino population, no Iowa county has a minority population that rises above the ten-percent floor referenced in Jones. Such a result is at odds with the Sixth Amendment and the purposes of the fair cross-section requirement.
A test without teeth leaves the right to an impartial jury for some minority populations without protection. Empirical evidence overwhelmingly shows that having just one person of color on an otherwise all-white jury can reduce disparate rates of *826convictions between black and white defendants. For example, when researchers at Duke University compared data on conviction rates by race in over 700 criminal trials over a ten-year period, they, found that where there was one or more black jurors, black and white defendants had roughly equal rates of conviction; however, all-white juries convicted Afriean-AmerR can defendants 81% of the time and white defendants only 66% of the time. Shamena Anwar, et al., The Impact of Jury Race in Criminal Trials, 127 Q.J. Econ. 1017, 1027-28,1032 (2012).
The data evidencing the impact of racially unrepresentative juries on izase outcomes are especially troubling given that Iowa ranks worst in the nation for the percentage of our prison population that is African-American (more than 26%), while African-Americans represent just 3.3% of the state’s population. Kyle Munson, Black-White Disparities Persist in lerna, Des Moines Register, http://www. desmoinesregister.com/story/newsAocal/ kyle-munson/2015/07/12/black-iowa-statistics-economies-incarceration/30059517 (last updated July 12, 2016). Troubling, too, -is the fact that African-Americans in Iowa are ten times more likely to be arrested than persons of other races; and Iowa ranks third worst in the nation for our incarceration rate for black men (9.4%).10 Id. A sound formula for assessing underrepresentation of minorities in our jury pools must provide meaningful protections of the right to an impartial jury.
4. The second Duren prong. Our decision to adopt absolute disparity as' the exclusive test and' to' reject comparative disparity in Jones rested upon an error of law and on cases from other jurisdictions that have since been overruled or criticized. After surveying the various tests, and bearing in mind the practical problems associated with the use of the absolute disparity test in Iowa, we conclude it is no longer appropriate to rely exclusively upon the absolute disparity test as an indicator of representativeness. We therefore overrule Jones, 490 N.W.2d at 792-93.
A sizeable number of jurisdictions ■ consider more than one analytical model when making determinations about whether a jury pool constitutes a fair and reasonable cross-section of the community.11 Like the Ninth Circuit, we decide “not *827[to] prescribe an alternative exclusive analysis to be applied in every case” and to follow the Supreme Court in permitting multiple analyses to be used that are appropriate to the circumstances of each case. See Hernandez-Estrada, 749 F.3d at 1164. Parties challenging jury' pools on the ground that they are unrepresentative may base their challenges on multiple analytical models. The district court may use multiple analytical models in its analysis, taking into account the various strengths and weaknesses of each test when determining whether jury pools comport with the Sixth Amendment mandate of representativeness.
This flexible approach is consistent with the nature of the Sixth Amendment right to have a jury pool made up of a fair cross-section of the community. What constitutes a fair cross-section of the community is subject to change over time. See Taylor, 419 U.S. at 537, 95 S.Ct. at. 701 (noting the definition of “community” is subject to change because “[cjommunities differ at different times and places”). “What is a fair cross section at one time or place is not necessarily a fair cross section at another time or a different place.” Id. (noting that although the total exclusion of women was permitted at one time, women could no longer be excluded at all on account of their sex). Because what constitutes a fair cross-section of the community is a fluid concept, a flexible approach for determining when, a racial disparity rises to the level of a constitutional violation is warranted.
5. The third Daren prong. The State argues that even if we are convinced that the problems with the absolute disparity calculation and the benefits of the comparative disparity calculation justify overruling Jones, we should not change course in this case because Plain did not present evidence under the third factor, which is required to establish a prima facie case. Plain contends that any failure to present a prima facie case on the third prong was caused by the jury manager’s denial of his request for access to . the information necessary to compile historical data on the composition of juries in Black Hawk County.
■■ Plain’s counsel recounted, during a colloquy with the court, her effort to obtain the relevant historical data.
[F]or purposes of preserving the record at least ... I also have what I was hoping would- be better[, which] I’ve marked as Court’s Exhibit 2. That is the report that I had our jury manager Billie Treloar print off for me.
.1 thought that she might be able to get me some, statistics for the last six months. She says she can .only print off today’s, panel in this courtroom. So I guess I just want to enter these for purposes of the record. I understand that I haven’t met my burden with the prima facie case. I’d concede that. But since I’ve tried to make the motion, I’d at least 'like to supplement the record.
In Iowa, the constitutional right to an impartial jury is enshrined in and implemented by statute. See Iowa Code §§ 607A.1-.47 (2016), Iowa’s jury laws are loosely based on the federal jury laws. Compare Iowa Code §§ 607A.1-.47, with 28 U.S.C. §§ 1861-78. The statutes in the state and federal systems, describe in detail the procedures and standards for compiling jury pools. In each instance, their essential purpose is to implement the constitutional right to a jury.
Importantly, the state and federal statutes require the selection authorities to maintain all records, and lists concerning *828the selection of jurors for a period of at least four years. Compare Iowa Code § 607A.26, with 28 U.S.C. § 1868. Unlike the federal law, however, which requires the selection authority to make the records and papers it compiled and maintained “available for public inspection for the purpose of determining the validity of the selection of any jury,” 28 U.S.C. § 1868, the Iowa Code does not expressly require the state to provide access to the data and information that must be maintained, see Iowa Code § 607A.26.
In at least four jurisdictions that lack a state statute establishing the right and procedures for obtaining the information necessary to enforce an individual’s Sixth Amendment rights, courts have determined that the constitution mandates access even if statutes do not. See Nina W. Chernoff, No Records, No Right: Discovery & the Fair Cross-Section Guarantee, 101 Iowa L. Rev. 1719,1763 (2016) [hereinafter Chernoff II] (reviewing cases).
The Supreme Court of Missouri, for example, reasoned that the “cross-section requirement would be without meaning if a defendant were denied all means of discovery in an effort to assert that right.” State ex rel. Garrett v. Saitz, 594 S.W.2d 606, 608 (Mo. 1980) (en banc). In holding that a defendant “is entitled to information relating to the racial composition of the grand jury so that he may assess whether he has a viable constitutional challenge,” the Nevada Supreme Court reasoned that “[without this information, [the defendant's ability to show a potential violation of his constitutional right to a grand jury drawn from a fair cross-section of the community is limited.” Afzali v. State, 326 P.3d 1, 1, 3 (Nev. 2014). The Superior Court of New Jersey, Appellate Division, determined “[i]t would be virtually impossible for defendants who are endeavoring to ascertain if a successful attack on the grand jury selection process can be advanced if the facts necessary to prove a defect in the selection process are withheld.” State v. Ciba-Geigy Corp., 240 N.J.Super. 511, 573 A.2d 944, 950 (1990); accord Mobley v. United States, 379 F.2d 768, 773 (5th Cir. 1967).
Like the courts in Missouri, Nevada, and New Jersey, we conclude the constitutional fair cross-section purpose alone is sufficient to require access to the information necessary to prove a prima facie case. Those courts
were each faced with a criminal defendant’s request for data on the racial and ethnic composition of the jury pool. Each court acknowledged that their state had no statutory equivalent to the [Federal Jury Selection and Service Act (JSSA) ], and each granted the defendant’s request for jury data on purely constitutional grounds, recognizing that access was necessary to enforcement of the fair cross-section right.
Chernoff II, 101 Iowa L. Rev. at 1763 (citations omitted).
Defendants are entitled to access the information needed to enforce their constitutional right to a jury trial by a representative cross-section of the community. In this case, Plain attempted to obtain the information he is entitled to receive. Because our statutes do not specify a procedure for accessing the information, he took what we view to be a reasonable approach—he asked the jury manager to provide it. The jury manager did not produce the information, citing a lack of access to the information the state is constitutionally required to maintain. To the extent Plain did not meet his prima facie case with respect to the third prong of the test, we conclude he lacked the opportunity to do so because he was not provided access to the records to which he was entitled.
*829Because we find no reversible error on the other issues raised on appeal, we conditionally affirm Plain’s conviction and remand to the district court for development of the record on the Sixth Amendment challenge. Following proper development of the record pertaining to that challenge, the district court shall determine whether Plain’s right to a representative jury under the Sixth Amendment was violated. If the court concludes that right,was violated, the court shall grant a new trial.
IV. Disposition.
We conclude that the district court’s admission of hearsay evidence through the testimony of an investigating officer about what caused the hole in the side of the house did not constitute reversible error because it was merely cumulative and thus not prejudicial. We further conclude the district court did not abuse its discretion in using a cautionary instruction to address a reference to Plain’s past criminal conduct on the 911 tape. Although the district court’s denial of Plain’s request for an implicit-bias jury instruction rested upon an error of law and thus constituted an abuse of discretion, we conclude the denial did not prejudice Plain. Although the prosecutor committed error in repeatedly using the word “victim” to refer to the complainant in this case, we find no prejudicial error resulted. Finally, we conclude the district court made an error of law in concluding the absolute disparity test must be used in deciding whether the jury pool was drawn from a fair cross-section of the community. We conditionally affirm Plain’s conviction and remand for development of the record on his Sixth Amendment challenge. On remand, the State shall provide the defendant reasonable access to the records necessary to evaluate whether African-Americans were systematically underrepresented in the jury pool from which the jurors were selected for Plain’s trial. Following development of the record on this issue, the district court shall reconsider Plain’s claim that his jury did not represent a fair cross-section of the community.
AFFIRMED ON CONDITION AND REMANDED WITH DIRECTIONS.
Cady, C.J., Wiggins, Appel, and Zager JJ., join this opinion. Cady, C.J., files a special concurrence joined by Wiggins and Appel, JJ. Wiggins, J., files a separate special concurrence joined by Cady, C.J., and Appel, J. Appel, J., files a separate special concurrence joined by Cady, C.J., and Wiggins, J. Waterman, J., files a separate special concurrence joined by Mansfield and Zager, JJ. Mansfield, J., files a separate special concurrence joined by Waterman, J., In 2016, the rule was amended. Under the new iteration of the rule, hearsay is defined as "a statement that: (1) [t]he declarant does not make while testifying at the current trial or hearing, and (2) [a] party offers to prove the truth of the matter asserted in the statement." Iowa R. Evid. 5.801(c)(l)-(2) (2017).
. Plain also attacks the content of the cautionary instruction, contending that it compounds the prejudicial effect of the evidence because it highlighted facts the jury may have otherwise missed or forgotten. We do not address the merits of this argument because this issue was not raised below and thus is not preserved for appeal. See DeWitt, 811 N.W.2d at 467. Plain requested the instruction. He did not object to its content, nor did the State. Thus, error was not preserved for our review. ' See id.
, Similarly, the State’s attorney made a professional statement suggesting he was not aware of objectionable statements on the redacted recording:
Your Honor, I reviewed the 911 call several times before playing it to the jury. I redacted out what I could hear that was inadmissible. I reviewed that with [defense counsel] and actually played the portions so that she could hear where it was redacted. I did not hear any of the times I was reviewing it prior to playing it or.when I was playing it these statements.
. We recently cautioned against conflating prosecutorial error with prosecutorial misconduct in Schlitter, 881 N.W.2d at 393-94. In that case, we distinguished between the two terms as follows:
Prosecutorial misconduct includes those statements "where a prosecutor intentionally violates a clear and unambiguous obligation or standard imposed by law, applicable rule or professional conduct,” as well as "those situations where a prosecutor recklessly disregards a duty to comply with an obligation or standard." Prosecutorial error occurs “where the prosecutor exercises poor judgment” and "where the attorney has made a mistake” based on "excusable *819human error, despite the attorney’s use of reasonable care.”
Id. (citations omitted) (quoting Shawn E. Min-ihan, Measuring Prosecutorial Actions: -An Analysis of Misconduct Versus Error, Prosecutor, Dec. 2014, at 23-25).
. Under Iowa’s jury-selection statutes, a jury “pool” (i.e., venire) consists of all persons who are summoned for jury service and who report. Iowa Code § 607A.3(9). A jury "panel” consists of "those jurors drawn or assigned for service to a courtroom, judge, or trial.” Id. § 607A.3(7). And a "petit” jury (i.e., grand jury) consists of the jurors who are actually called upon to attend court proceedings. See id. § 607A.3(3).
. Article I, section 10 of the Iowa Constitution also entitles a criminal defendant to a jury pool that represents a fair cross-section of the community. See State v. Knutson, 220 N.W.2d 575, 577 (Iowa 1974). The defendant has not raised a challenge under article I, section 10, however, so we do not consider in this case whether the protection offered by that provision is identical to that provided under the Sixth Amendment.
. Hemandez-Estrada also discusses the absolute impact test, which is similar to the absolute disparity test.
Under [the absolute impact] approach, the initial calculation is the same as for the absolute disparity test. However, the resulting number is then multiplied by the number of persons on the particular panel.. The advantage of the test is that it applies the disparity analysis to the actual jury pool to determine the impact. However, because it is based on the absolute disparity test, the absolute impact tdst bears many of the same flaws. • ' '
Hernandez-Estrada, 749 F.3d at 1162.
. Courts set different standards for .what constitutes an absolute disparity that is not a fair and reasonable representation. See Jones, 490 *823N.W.2d at 793 (noting other courts have found absolute disparities of anywhere from 2.9% to 10% insufficient to establish a prima facie case).
. The Sixth and Fourteenth Amendments both protect the impartiality of a jury. See Nina W. Chemoff, Wrong About the Right: How Courts Undermine the Fair Cross-Section Guarantee by Confusing It with Equal Protection, 64 Hastings L.J. 141, 151-52 (2012) [hereinafter Chemoff I], While the Fourteenth Amend-merit’s Equal Protection Clause bars the intentional exclusion of protected minority groups, the Sixth Amendment guarantees that minority groups will not be systematically excluded, even where there is no evidence of intentional exclusion. Coriell, 100 Cornell L. Rev. at 472 (“[W]hile equal protection is a powerful tool to combat intentional discrimination in drafting the jury venire, it is less effective in addressing situations where there is obvious underrepresentation of certain seg*824ments of the community on jury venires but no evidence of intentional exclusion on the part of state actors.”).
In this case, the State asserts the defendant did not establish evidence of discriminatory intent. The State conflates the test for a violation of the Equal Protection Clause (which requires a showing of intent) with the test for a violation of the Sixth Amendment (which does not). See Chernoff I, 64 Hastings L.J. at 151-52.
For an example of how there can be systematic exclusion without discriminatory intent, consider an example from Hartford, Connecticut, where
[ujnbeknownst to the jury office, an inadvertent error caused the computer program to erroneously read the "d” in the capital city of Hartford to mean ''deceased.” The computer program accordingly never selected anyone from Hartford to receive a juiy summons....
The computer program also failed to send jury summons to anyone in the neighboring city of New Britain, the second largest city in the jury district. The New Britain citizens were left out of the jury pool because the source list of names from New Britain was accidentally misplaced and was never entered into the computer program.
... Those two cities alone accounted for 63% of the voting-age African-American population and 68% of the voting-age Hispanic population in the jury district.
Nina W. Chernoff, No Records, No Right: Discovery & the Fair Cross-Section Guarantee, 101 Iowa L. Rev. 1719, 1723-24 (2016) (footnotes omitted).
. The high arrest and incarceration rates have secondary effects, too, presenting barriers for education and employment and contribute to high percentages of families living in poverty (39% for black families versus 13.9% statewide), receiving food stamps (42% for black households versus 11% for white households), rental housing (73.4% for black Iowans versus 29.2% statewide), median income ($27,406 for black households versus $52,652 for white households), unemployment (11.6% for black Iowans versus 4.8% overall), health insurance pre-Affordable Care Act (44% of black Iowans versus 77% of white Iowans), dropout rates (14% for black children who make up just 5% of.the school population), and bachelor degrees (15% for black Iowans versus 26% overall). Kyle Mun-son, Black-White Disparities Persist in Iowa, Des Moines Register, http://www.desmoinesre gister.com/story/news/local/kyle-munson/2015 /07/12/black-iowa-statistics-economics-incarce ration/30059517 (last updated July 12, 2015).
. See, e.g., Hernandez-Estrada, 749 F.3d at 1164-65; United States v. Orange, 447 F.3d 792, 798-99 (10th Cir. 2006); Mosley v. Dretke, 370 F.3d 467, 479 n.5 (5th Cir. 2004); Ramseur v. Beyer, 983 F.2d 1215, 1231 (3d Cir. 1992); State v. Sanderson, 182 Ariz. 534, 898 P.2d 483, 487 n.2 (App. 1995); State v. Lopez, 107 Idaho 726, 692 P.2d 370, 375-76 (App. 1984); State v. Kennedy, 823 So.2d 411, 418-19 (La. App. 5th Cir. 2002); State v. Williams, 525 N.W.2d 538, 543 (Minn. 1994); Evans v. State, 112 Nev. 1172, 926 P.2d 265, 275 (1996); State v. Dixon, 125 N.J. 223, 593 A.2d 266, 271-72 (1991); Commonwealth v. Rosado, No. 2467, 1993 WL 1156055, at *28 (Pa, Com. PL Apr. 15, 1993); State v. Hester, 324 S.W.3d 1, 43 (Tenn. 2010); Ovalle v. *827State, 13 S.W.3d 774, 778 (Tex. Crim. App. 2000); State v. Tillman, 750 P.2d 546, 577 (Utah 1987).