Amended August 18, 2015 State of Iowa v. Tyler James Webster

IN THE SUPREME COURT OF IOWA No. 13–1095 Filed June 19, 2015 Amended August 18, 2015 STATE OF IOWA, Appellee, vs. TYLER JAMES WEBSTER, Appellant. On review from the Iowa Court of Appeals. Appeal from the Iowa District Court for Jefferson County, Myron L. Gookin, Judge. The State seeks further review of a court of appeals decision reversing the defendant’s conviction of second-degree murder and remanding for a new trial. DECISION OF COURT OF APPEALS VACATED; DISTRICT COURT JUDGMENT AFFIRMED. Mark C. Smith, State Appellate Defender, and Rachel C. Regenold, Assistant Appellate Defender, for appellant. Thomas J. Miller, Attorney General, Kevin Cmelik, Denise A. Timmins, and Heather Ann Mapes, Assistant Attorneys General, and Timothy W. Dille, County Attorney, for appellee. Alan R. Ostergren, Muscatine, for amicus curiae Iowa County Attorneys Association. 2 APPEL, Justice. Tyler Webster shot and killed Buddy Frisbie. The State charged Webster with first-degree murder. A jury returned a guilty verdict of second-degree murder. Webster filed a posttrial motion seeking to vacate the conviction on grounds of juror misconduct. Webster also challenged several evidentiary rulings in which the district court refused to admit evidence that reflected poorly on Frisbie. The district court denied Webster’s posttrial motion and entered judgment. Webster appealed. On appeal, Webster claimed his conviction should be vacated because of juror misconduct and juror bias. Specifically, Webster claimed a juror failed to disclose that her daughter was a good friend of Frisbie’s stepsister. Webster further claimed the same juror engaged in discussions about the case with third parties, posted comments on Facebook, and “liked” a comment posted by Frisbie’s stepmother on Facebook related to the trial. Webster also appealed the judgment based upon assorted errors in the district court’s evidentiary rulings. We transferred the case to the court of appeals. The court of appeals held there was no reversible juror misconduct, but reversed Webster’s conviction on the issue of juror bias. We granted further review. For the reasons expressed below, we vacate the decision of the court of appeals and affirm the judgment of the district court. I. Factual Background and Proceedings. A. Factual Overview. Webster and Frisbie were long-time friends. On August 25, 2012, Webster, Frisbie, a mutual friend Doug Knight, and Frisbie’s girlfriend Shelby Hall attended a party together. They decided to go fishing. Frisbie and Hall went to Frisbie’s trailer to retrieve fishing gear, and Webster joined them in the trailer as it began to rain. Knight went to his own trailer nearby. 3 In Frisbie’s trailer, Webster claimed he believed Frisbie was sexually assaulting Hall. He left the trailer, went to his truck, grabbed a gun, came back to the trailer, and shot Frisbie in the face at close range to avoid hitting Hall, whose body partially covered Frisbie. Hall ran to Knight’s trailer and explained what had just happened. Knight retrieved a shotgun, and when Webster approached the trailer with a gun in his hand, Knight told him to put the gun down. Webster complied and Knight called 911. Webster admitted to the dispatcher that he had just shot Frisbie. Dispatch instructed Webster to walk to the end of the driveway, lie on his stomach, and wait for law enforcement to arrive. Law enforcement arrived and arrested Webster without incident. The State charged Webster with first-degree murder. Prior to the start of trial, the court ruled on motions in limine. Jury selection was not reported. During the week-long trial, the court reminded the jury of its “long admonition” previously given. The long admonition, however, is not part of the record. The first admonition found in the record occurred before the noon recess during the first day of trial. This admonition stated: And so at this time I will again remind you of that long admonition that I previously read to you about you don’t talk between yourselves, you don’t talk with anyone else, you don’t listen or read any news reports. This matter is not yet submitted so you don’t communicate with anyone or each other about what you have heard so far, and you keep an open mind and you don’t come to any conclusions. The court gave a similar rendition of this admonition numerous times throughout the trial. B. In Camera Examination of the Juror. After the defense rested, and outside the presence of the jury, the district court alerted the parties to an issue that had arisen in the case. The district court told the 4 parties that the clerk of court and the court attendant had advised the court they had received information that one of the jurors had stated prior to being seated that she would probably never be picked for the jury because “she knew the family.” The district court stated the court attendant was concerned because she had sat through voir dire and did not recall the juror saying anything about her connection to either family. The court then conducted an in camera hearing and the juror was questioned. The juror told the court that her twenty-seven-year-old daughter was friends with Frisbie’s half-sister or stepsister as they had attended high school together. The juror stated she did not know Frisbie, and other than telling her daughter she had jury duty, she did not discuss the case with her. She also stated she was friendly with Frisbie’s parents, as they worked in the courthouse and she also worked in the courthouse. She also noted she thought she knew a family member of Webster’s wife. Webster’s attorney noted the juror’s familiarity with these individuals commenting, “I understand this is a small town.” When asked if the relationships would cause her to be biased, the juror stated she would not be biased and would rely upon her notes in making her decision. The juror further stated she was a Facebook 1 user and knew about the shooting the night of the incident through Facebook. She stated that while she had been on Facebook during the trial playing games, she had “not read anybody else’s postings, because [she] kn[ew] if they posted something [she] didn’t want to know about it.” At the conclusion of the 1For an overview of Facebook terminology, see Facebook’s Glossary of Terms, available at http://www.facebook.com/help/219443701509174 (last visited 5/21/2015). 5 in camera examination of the juror, the defense declined to challenge the juror for cause. C. Submission of the Case and the Verdict. The next day, the parties completed closing arguments and the case was submitted to the jury. In its jury instructions, the district court stated, in relevant part: You may not communicate about this case before reaching your verdict. This includes cell phones, and electronic media such as text messages, Facebook, MySpace, LinkedIn, YouTube, Twitter, email, etc. Do not do any research or make any investigation about this case on your own. Also, do not research any information about this case, the law, or the people involved, including the parties, the witnesses, the lawyers, or the judge. This includes using the Internet to research events or people referenced at trial. After being instructed, the jury deliberated and returned a verdict finding Webster guilty of murder in the second degree in violation of Iowa Code sections 707.1 and 707.3 (2013). D. Posttrial Motion. After the verdict, Webster filed a combined motion for new trial and arrest of judgment. Webster asserted that subsequent to the verdict Karen Taylor, a convenience store employee, told the defense that during the course of trial she observed the same juror in the store talking about the trial with two or three other customers. In addition, Webster also asserted the defense received information from Sheila Ross, who employed the defendant’s mother as a housekeeper. Ross recalled a conversation she had with the juror a few days after the verdict in which the juror told her that she looked up Knight’s age and that he was not as “old” as defense counsel had contended. Ross further indicated the juror informed her that the juror’s daughter regarded Frisbie as “the sweetest nicest most soft spoken good guy” and regarded Webster as “loud mouth verbally aggressive and generally not a good person.” Ross also stated the juror told her that “no 6 one asked [the juror] to recuse herself for knowing the Frisbie family so she just kept her mouth shut and did not offer that up.” In addition, Webster contended the district court denied him a fair trial by prohibiting him from introducing evidence that Frisbie’s ex-wife was nine months pregnant when Frisbie punched her in the stomach and by prohibiting Webster from introducing evidence of Frisbie’s felon status and “prison mentality.” Webster claimed this information was essential to his defense of justification because it would put his actions in context. The district court held a hearing on Webster’s combined motion. Taylor was the first witness to testify. She stated during the course of the trial she heard the juror discuss the case at a convenience store where she worked. Taylor testified she heard a customer state, “[E]veryone knows he’s guilty,” and the juror responded “Yeah.” Taylor also testified the juror told her that Webster “had plead not guilty . . . and [the jury had] to decide guilt or innocence.” Ross testified next. She did not know the juror personally but became familiar with her when she attended jury selection and the trial. Webster’s mother had been Ross’s housekeeper and Ross considered Webster’s mother a friend. Ross testified that after trial she heard the juror in a Hy-Vee store “loudly proclaiming about the trial and the results of the trial . . . and [that] he deserved what he got.” Ross further testified the juror indicated that Webster could have gotten help from Knight because Knight was not an old man, as she had “looked up his age.” Ross further testified the juror told her that the juror’s daughter “knew both [Webster and Frisbie] when they were young, and that [Frisbie] was just this kind, sweet, gentle, polite person” and Webster “was a mouthy, aggressive . . . verbally—you know, aggressive person.” 7 When Ross asked the juror why she did not recuse herself like another potential juror did when that juror told the lawyers that the juror knew the Webster and Frisbie families, the juror responded “[T]hey did not ask me, so I didn’t tell them.” When Ross inquired if the lawyers had asked whether anyone knew the Frisbie family, the juror replied “nope,” giving Ross the impression that the juror was proud that she had “dodged that one.” Webster’s wife, Ann Webster, also testified at the hearing. She stated she had heard rumors there was a juror who knew the Frisbie family and Ann began looking at Facebook. She testified that about a month before trial, Frisbie’s stepmother wrote on the juror’s wall, “Have a wonderful day.” Ann also testified that on April 11, during the trial, Frisbie’s stepmother posted on her own Facebook wall, “Give me strength,” and at some point 2 the juror’s daughter “[l]ike[d]” the comment. 3 After trial, Ann stated the juror posted a summary of her jury service on the juror’s Facebook wall. Also after trial, Ann noted the juror responded to a Facebook posting by Frisbie’s stepmother by stating, “I wish you could have gotten murder in lst degree. I can safely say that this was a very hard decision. I could talk to you more about it if you wanted at some point—just message me.” The juror also testified at the posttrial hearing. First, regarding the convenience store communication, the juror admitted there were a few 2It is not entirely clear from the record, including the exhibit of the Facebook page admitted at the posttrial hearing on the combined motion, when the comment was liked. However, the parties appear to agree that it was liked “during trial.” 3Later, when shown a copy of the Facebook printout admitted at the hearing, Ann was asked to “read that line right up there,” in reference to the “Give me strength” comment. Ann answered, “This one says, ‘[the juror], [a second individual] and 12 others like this.’ ” After reviewing the Facebook printout, it is unclear whether the line Ann was reading refers to the “Give me strength” comment or a different comment. The juror later testified that she “probably” had “liked” Frisbie’s stepmother’s comment. 8 words spoken related to the trial by the customers but stated she wanted to avoid the discussion. She admitted telling the clerk at the store that Webster “pled not guilty [and that y]ou have to prove guilt or innocence.” The juror testified that although she may have used the term “Yeah” at one point, she did not agree with a comment from one of the customers that “[e]veryone knows he’s guilty.” The juror denied being influenced by any of the conversations at the convenience store stating, “In no way . . . I was not convinced by anything that had come up previously yet at all. Nothing had convinced me.” With respect to the Hy-Vee incident that occurred after the trial, the juror admitted telling a person at Hy-Vee that she thought her relationship with the Frisbie family would keep her off the jury but that she was not asked about this relationship. At the hearing the juror stated, “I didn’t know how to tell. I guess I’m dumb to the rules.” The juror claimed Ross then approached her and stated, “You put an innocent man in jail.” According to the juror she told Ross, Who do you believe? You say [Webster is] a good man. The Frisbies are good people. Who do you believe? You go with the evidence that is in front of you, and you go with that. That’s what you have to go with. The juror acknowledged, as Ross had testified, that she looked up Knight’s age, but had done so the morning after the trial ended and determined that Knight was about the same age as she was. Regarding the question of her relationship generally with the Frisbie family, the juror testified she “never kept any of that a secret, that [she] was friends with [the Frisbie family].” When the juror was asked, “Now, you didn’t tell us that you knew the Frisbie family until we went back into the judge’s chambers in private, correct?” the juror responded: 9 I did not tell you, but I didn’t know how to tell you any other way. I was asked in jury selection up here, I said—you know, we were asked if we knew these people. I didn’t know the people I was asked. I honestly thought that, okay, they’re not asking for the prosecution, and here’s going to come defense and you were gonna ask. And I was—I’m like okay, how do I do this now? I mean, because I was going to say something then, but I thought we had to wait to be asked specific questions. The juror also testified about her Facebook activity. She testified she was “friends” with Frisbie’s stepmother on Facebook. At the same time, the juror testified she was not “close friends” with the Frisbies. She admitted she “probably” clicked “Like” on Frisbie’s stepmother’s comment, “Give me strength,” but denied that she communicated with Frisbie’s stepmother, stating she simply “clicked a button that said, ‘like.’ ” Additionally, although she claimed not to be close to the Frisbie family, she knew it was a good family because her daughter had said so and her daughter would never allow the Frisbie family to babysit for her child if she did not feel that way. In his combined motion for new trial and arrest of judgment, Webster argued he was denied a fair trial, but cited no constitutional or other provision in support of his motion. At the hearing, Webster’s attorney asked the court to set aside the verdict and order a new trial, contending it was “the only way [Webster’s] due process rights to a fair trial [could] be protected under both the Iowa Constitution as well as the Federal Constitution.” Citing Iowa Rule of Criminal Procedure 2.24(2)(b)(3)–(4), and (3)(a), the district court denied Webster’s posttrial motion and held that a new trial was not warranted and that judgment should not be arrested “on the basis of juror misconduct,” as there was “insufficient evidence of misconduct to set aside the verdict or grant a new trial.” The district 10 court also noted that “if there was any undisclosed bias” on the part of the juror, “it was not reflected in the verdict in which she participated,” nor was there “any indication that any such ostensible bias influenced or infected any discussions or deliberations of the jury as a whole.” Webster appealed and we transferred the case to the court of appeals. The court of appeals determined that while there was no basis to upset the verdict based upon juror misconduct, the district court ruling was reversed on the question of juror bias. We granted further review. II. Standard of Review. We review a denial of a motion for a new trial based upon juror misconduct or juror bias for an abuse of discretion. Fry v. Blauvelt, 818 N.W.2d 123, 128 (Iowa 2012) (juror misconduct); State v. Hendrickson, 444 N.W.2d 468, 472 (Iowa 1989) (juror misconduct and bias); see also State v. Johnson, 445 N.W.2d 337, 340 (Iowa 1989) (same). Additionally, we review evidentiary rulings for an abuse of discretion. See State v. Nelson, 791 N.W.2d 414, 419 (Iowa 2010). “An abuse of discretion occurs when the trial court exercises its discretion ‘on grounds or for reasons clearly untenable or to an extent clearly unreasonable.’ ” State v. Rodriquez, 636 N.W.2d 234, 239 (Iowa 2001) (quoting State v. Maghee, 573 N.W.2d 1, 5 (Iowa 1997)). “ ‘A ground or reason is untenable when it is not supported by substantial evidence or when it is based on an erroneous application of the law.’ ” Id. (quoting Graber v. City of Ankeny, 616 N.W.2d 633, 638 (Iowa 2000)). The burden is on the party seeking to overturn the verdict. See State v. Henning, 545 N.W.2d 322, 324–25 (Iowa 1996). 4 4There is a question of the proper standard of review regarding fact-finding performed by the district court in the context of a motion for a new trial. There is 11 We review ineffective-assistance-of-counsel claims de novo. State v. Halverson, 857 N.W.2d 632, 634 (Iowa 2015). In order to succeed on an ineffective-assistance-of-counsel claim, a defendant must prove each of the following two elements by a preponderance of the evidence: (1) trial counsel failed to perform an essential duty, and (2) this failure resulted in prejudice. State v. Dalton, 674 N.W.2d 111, 119 (Iowa 2004); see also Strickland v. Washington, 466 U.S. 668, 687, 104 S. Ct. 2052, 2064, 80 L. Ed. 2d 674, 693 (1984). Therefore, if the claim lacks prejudice and can be decided on that ground alone, we need not address whether the attorney failed to perform an essential duty. See Dalton, 674 N.W.2d at 119. III. Preservation Issues Related to Juror Misconduct and Juror Bias. Our review in this case is impacted by the limited nature of the advocacy and the limited record developed below. Turning first to the advocacy of the parties, Webster did not indicate in his combined motion for new trial and arrest of judgment the basis for his claim. He did not cite a statute, rule, or constitutional provision. At oral argument, Webster generally claimed a right to a fair trial under the Iowa and United States Constitutions, but did not identify any particular provision. In its order, the district court relied upon Iowa Rule of Criminal Procedure 2.24(2) and (3), but did not make a ruling on any Iowa or United States constitutional claim. Webster did not seek an ______________________ authority in other jurisdictions that fact-finding made by the district court in considering a motion for a new trial is subject to review under a clearly erroneous standard. See, e.g., State v. Dellinger, 696 S.E.2d 38, 42 (W. Va. 2010) (per curiam). On the other hand, we have held, in other contexts, that when constitutional issues are involved, our review of fact-finding is de novo. See, e.g., State v. Hoskins, 711 N.W.2d 720, 725 (Iowa 2006) (noting in search-and-seizure context that our review of the facts is de novo, however, we give deference to the district court’s findings as it “had the opportunity to evaluate the credibility of the witnesses”). In this case, we do not resolve the issue because we generally agree with the fact-finding of the district court. 12 expanded ruling. See Lamasters v. State, 821 N.W.2d 856, 863–64 (Iowa 2012) (noting rule 1.904(2) “is one means, but not the only means, for requesting” a ruling on a matter in order to preserve error); State v. Krogmann, 804 N.W.2d 518, 524 (Iowa 2011) (stating that “when a court fails to rule on a matter, a party must request a ruling by some means”); see also Meier v. Senecaut, 641 N.W.2d 532, 539 (Iowa 2002). As a result, we find the question of whether the refusal of the district court to grant a new trial or arrest judgment violated Iowa Rule of Criminal Procedure 2.24(2) or (3) was preserved. Any claim that the district court action violated a constitutional provision, state or federal, however, is waived. In addition, Webster explicitly claimed in his combined motion that he was denied a fair trial based on juror misconduct, but did not advance a separate argument based on juror bias. Juror misconduct and juror bias are related, overlapping, but analytically distinct concepts. Juror misconduct ordinarily relates to actions of a juror, often contrary to the court’s instructions or admonitions, which impair the integrity of the fact-finding process at trial. See generally Jimmie E. Tinsley, Jury Misconduct Warranting New Trial, 24 Am. Jur. Proof of Facts §§ 35–38, at 697–704 and §§ 4–11, at 255–308 (2d ed. 1980 & Supp. 2014) (citing examples). Typical acts of misconduct include communication with others outside the jury about the case, independently investigating the crime or accident scenes outside of judicial oversight, or engaging in independent research about questions of law or fact. See id. Juror bias, on the other hand, focuses on the ability of a juror to impartially consider questions raised at trial. See 50A C.J.S. Juries § 369, at 495– 97 (2008). A biased juror is simply unable to come to a fair decision in a case based upon the facts and law presented at trial. See id. A juror 13 may be biased without engaging in any kind of misconduct. Conversely, an impartial and fair-minded juror may nonetheless engage in juror misconduct. Jurisdictions have developed a variety of approaches to deal with questions of misconduct and bias. See generally Robert G. Loewy, Note, When Jurors Lie: Differing Standards for New Trials, 22 Am. J. Crim. L. 733 (1995) [hereinafter Loewy] (surveying different state and federal approaches). While Webster emphasized the label “juror misconduct” in his combined motion, he also stated that the juror was biased. In its ruling, the district court followed Webster’s word usage, generally referring to “juror misconduct” but seemingly including juror bias within this larger concept. The question of a challenge for juror bias may be considered preserved based on the theory that the substance of the claim, rather than its label, controls. See Lee v. State, 815 N.W.2d 731, 739 (Iowa 2012) (“We will not exalt form over substance when the objectives of our error preservation rules have been met.”); Griffin Pipe Prods. Co. v. Bd. of Review, 789 N.W.2d 769, 772 (Iowa 2010) (“Our issue preservation rules are not designed to be hypertechnical.”). For the purposes of this appeal, we address the underlying merits of the bias claim without deciding the question of preservation. See State v. Hochmuth, 585 N.W.2d 234, 236 (Iowa 1998) (per curiam) (“Assuming without deciding that Hochmuth has preserved error, we find her challenge . . . is without merit.”); see also Ostergren v. Iowa Dist. Ct., 863 N.W.2d 294, 297 (Iowa 2015) (noting same). There are further problems, however, to the claims of juror misconduct and juror bias. Although this was a first-degree murder case, no record was made of voir dire or of the district court’s preliminary admonition to the jury. These record shortcomings, however, do not 14 raise a question of preservation of the claims so much as affect the viability of Webster’s claims on the merits, to which we now turn. IV. Merits of Juror Misconduct and Juror Bias. A. Introduction. This case requires us to explore a delicate area of the law. It is a bedrock component of our system of justice that an accused charged with a criminal offense receives a fair trial before an unbiased decision-maker. See Iowa Const. art. I, § 9. We have also established a criminal trial process with a right to counsel at every critical stage of the proceeding and the right to confront witnesses. See id. § 10. A jury that considers evidence produced outside the trial process deprives the defendant of the right to counsel and the right to confront, as well as defeats the policies advanced by our rules of evidence and rules of procedure that help ensure just results. Further, the judge’s instructions in a case are designed to channel the jury deliberations according to the rule of law. See 75A Am. Jur. 2d Trial § 920, at 558–59 (2007). These basic concepts—unbiased juries, structured trial process with the assistance of counsel for the accused, the right to confront witnesses, and judicially crafted instructions provided to the jury to channel decision-making according to the rule of law—are universally admired norms. Ensuring that these celebrated norms are the reality in each criminal case is a crucial responsibility of the judicial branch. As with so many things, enforcement of these norms is sometimes easier said than done. The line between permissible and impermissible is often difficult to discern. For instance, while we do not want jurors to be biased, we do want them to draw upon their common experience that may cause them to perceive evidence in a distinctive way. See, e.g., State v. Smith, 196 Iowa 1003, 1012, 193 N.W. 418, 422–23 (1923). Further, here, a jury 15 verdict has been rendered after a lengthy trial, and we have no desire to start again for trifles. As has been often said, the accused is not entitled to a perfect trial, but only a fair trial. See, e.g., State v. Gansz, 376 N.W.2d 887, 891 (Iowa 1985). B. Positions of the Parties. In this appeal, Webster makes a number of charges of juror misconduct. Although packaged as a claim of juror misconduct, Webster also asserts he was denied a fair trial due to impermissible jury bias. To the extent the claims presented on appeal related to his combined motion for a new trial and arrest of judgment were not preserved, Webster asserts ineffective assistance of counsel. With respect to all his misconduct and bias claims, Webster asserts the district court’s refusal to grant a new trial is subject to review for an abuse of discretion. Johnson, 445 N.W.2d at 340. With respect to his claim that counsel provided ineffective assistance to the extent counsel waived his claims, Webster asserts that such constitutional claims are reviewed de novo. State v. Risdal, 404 N.W.2d 130, 131 (Iowa 1987). 5 On his new trial claims based upon both juror misconduct and juror bias, Webster seems to apply the three-part substantive standard for granting a new trial based on juror misconduct articulated in State v. Cullen, 357 N.W.2d 24, 27 (Iowa 1984). In Cullen, we stated that in order to be entitled to a new trial based upon juror misconduct, the (1) evidence from the jurors must consist only of objective facts as to what actually occurred in or out of the jury room bearing on misconduct; (2) the acts or statements complained of must exceed tolerable bounds of jury deliberation; and (3) it must appear the misconduct was 5Webster, however, does not assert his claim that he was entitled to a new trial is subject to de novo review, thereby suggesting he does not assert a constitutional basis for the motion. 16 calculated to, and with reasonable probability did, influence the verdict. Id.; see also Ryan v. Arneson, 422 N.W.2d 491, 495 (Iowa 1988) (clarifying the first prong of the Cullen test by interpreting Iowa Rule of Evidence (5).606(b) to allow statements regarding extraneous prejudicial information or outside influence that was brought to bear on the jury, but excluding evidence of internal deliberations of the jury). Webster claims the juror committed misconduct by failing to fully disclose her relationship to the Frisbie family during voir dire and at the in camera examination during trial, by violating the admonitions of the district court by “liking” a comment of the victim’s stepmother on Facebook, by engaging in improper conversations about the trial with third parties at a convenience store, and by improperly conducting extra- record research on the age of a witness. Although packaged as a claim of juror misconduct, Webster also asserts he was denied a fair trial due to impermissible juror bias. Webster recognizes that no challenge for bias was raised at trial and therefore the issue is ordinarily waived. See State v. Cuevas, 288 N.W.2d 525, 534 (Iowa 1980). However, Webster claims waiver does not apply here because the juror was less than honest in voir dire and at the in camera hearing. The waiver rule, he asserts, does not apply to instances of concealed bias. See id. at 535. Based on the record, Webster claims the juror was impermissibly biased. In support of the bias theory, Webster contrasts the disclosures made by the juror at the in camera hearing during trial with her more robust explanations of her relationship with the Frisbie family in her after-trial statements and at the posttrial hearing. According to Webster, the juror saw the Frisbie family as a “good family,” recognized that she 17 would not be seated as a juror because she knew the family, was not forthcoming during voir dire about her relationship with the Frisbies, and was evasive about her relationship with the Frisbies when questioned in camera in order to remain on the jury. Further, Webster contends, the juror showed her true colors after the trial when she smugly told a friend of Frisbie’s stepmother that she was never questioned about her relationship with the Frisbies and commented on Frisbie’s stepmother’s Facebook post shortly after the verdict that the juror wished there could have been a first-degree murder verdict. The State accepts the same legal framework for reviewing a denial of a motion for a new trial and a motion in arrest of judgment as Webster. On the misconduct claims, the State asserts the record shows the juror did not deliberately lie about her relationship with the Frisbies. While it might have been desirable for her to have volunteered more information for the attorneys to explore on voir dire, the State notes there is no evidence she was specifically asked about her relationship with the Frisbies in voir dire. Further, during the in camera hearing, the juror indicated she knew the Frisbie family because her daughter was friends with the victim’s stepsister, knew the victim’s parents well enough to engage in small talk with them, and was Facebook friends with the victim’s stepmother. The State notes the juror truthfully indicated she did not know the victim and would not have recognized him on the street. According to the State, counsel for Webster had a full opportunity to explore any potential bias issues at the in camera hearing and any claim for juror bias was waived by Webster’s failure to develop the record to dismiss the juror for cause. On the question of the asserted misconduct arising from the juror’s liking a comment of the victim’s stepmother on Facebook during trial, the 18 State contends the juror’s conduct was of little importance. The comment which she liked on Facebook was a declaration from the stepmother in the midst of trial, “Give me strength.” According to the State, the juror’s statement that she liked the “Give me strength” statement does not relate to the merits of the case and had no relationship to the guilt or innocence of the defendant. According to the State, the juror did not think the inconsequential act of clicking a computer button liking the stepmother’s status amounted to a communication in violation of the district court’s admonition. With respect to the claim of misconduct at the convenience store, the State indicates the juror simply wanted to get out of the store and did not engage in substantive conversation with other customers about the trial itself. The State asserts there was no evidence the juror was in any way influenced by the brief interaction. On the issue of researching the age of one of the witnesses, the State responds the evidence showed this happened after the verdict had been rendered. As a result, the information obtained by the juror could not have had any influence on the verdict in this case. Finally, the State asserts there is no evidence of impermissible juror bias. The State emphasizes the juror repeatedly stated she could decide the case based upon the evidence alone. Further, the State stresses the decision not to grant a new trial rests in the sound discretion of the district court. See Johnson, 445 N.W.2d at 340. The State emphasizes Webster failed to show that any alleged misconduct “was calculated to, and with reasonable probability did, influence the verdict.” Cullen, 357 N.W.2d at 27. C. Discussion. We first dispose of the less difficult issues in this case. With respect to the juror’s conversations at the convenience store, 19 we find Webster’s challenge lacks merit. There can be no question that communications with third parties about the merits of a case outside the confines of jury deliberations is a species of misconduct. See Iowa R. of Crim. P. 2.24(2)(b)(2)–(3). Here, however, the juror did not initiate impermissible conversations, but was confronted with brief conclusory statements by third parties when picking up pizza at a convenience store. She apparently rolled her eyes and uttered an ambiguous “Yeah” while seeking to disengage. Our review of the record leads us to the same factual conclusion as that of the district court, namely, that the juror did not engage in an extended conversation on the merits of the case with these third-party intermeddlers but simply sought to end the nettlesome interaction and be on her way. We find no juror misconduct here, and even if we did, we would not find that the misconduct “was calculated to, and with reasonable probability did, influence the verdict.” Cullen, 357 N.W.2d at 27; see also State v. Anderson, 448 N.W.2d 32, 34– 35 (Iowa 1989) (fleeting conversation between juror and reserve special deputy, who declared, “what’s this not guilty shit,” was not basis for new trial). With respect to Webster’s claim that the juror conducted impermissible research on the age of one of the witnesses, we also find this claim lacking in merit. Iowa Rule of Criminal Procedure 2.24(2)(b)(2) authorizes a district court to grant a new trial when the jury has received out-of-court evidence. Of course, the introduction of material evidence outside the rigors of the trial process raises serious problems that may require a new trial. See State v. Folck, 325 N.W.2d 368, 372 (Iowa 1982) (noting that “[c]onsideration of matters outside the record may, under certain circumstances, require a new trial”); State v. Kirk, 168 Iowa 244, 256–62, 150 N.W. 91, 94–96 (1914) (remanding for new trial when jury 20 secured copy of a Code book and rendered a verdict of guilty of manslaughter with a recommendation of mercy, believing the punishment was fixed by the Code, when, in reality, it was fixed by the indeterminate sentence act). The record here, however, reveals the juror engaged in the research after the jury rendered its verdict to confirm her views based upon the evidence offered at trial. Once the jury renders its verdict, jurors are free to research factual and legal questions as much, or as little, as they desire. See, e.g., Wilgus v. F/V Sirius, Inc., 665 F. Supp. 2d 23, 27–28 (D. Me. 2009) (finding record failed to show that during trial juror discovered extraneous information and therefore finding no evidence of juror misconduct). The most serious issues in this case relate to the question of juror bias. Iowa Rule of Criminal Procedure 2.24(2)(b)(9) provides that the district court may grant a new trial when “the defendant has not received a fair and impartial trial.” Juror bias may be actual or implied. Actual juror bias occurs when the evidence shows that a juror, in fact, is unable to lay aside prejudices and judge a case fairly on the merits. See United States v. Wood, 299 U.S. 123, 133, 57 S. Ct. 177, 179, 81 L. Ed. 78, 82 (1936). Implied bias arises when the relationship of a prospective juror to a case is so troublesome that the law presumes a juror would not be impartial. See id.; see also McDonough Power Equip., Inc. v. Greenwood, 464 U.S. 548, 558, 104 S. Ct. 845, 851, 78 L. Ed. 2d 663, 673 (1984) (Brennan, J., concurring in judgment); Philip Staten, Clarifying the Implied Bias Doctrine: Bringing Greater Certainty to the Voir Dire Process in the Military Justice System, 2011-Mar. Army Law. 17, 17–21 (2011) [hereinafter Staten] (canvassing Supreme Court precedent regarding 21 implied juror bias). 6 Implied bias has been found to arise, for instance, when a juror is employed by a party or is closely related to a party or witness. See, e.g., McHugh v. Proctor & Gamble Paper Prods. Co., 776 A.2d 266, 270 (Pa. Super. Ct. 2001) (noting that “close relationship, familial, financial, or situational, with the parties, counsel, victims, or witnesses” can give rise to a presumption of bias); Staten, 2011-Mar. Army Law. 17, at 20. Here, our focus is on the question of whether the district court abused its discretion in finding that the prospective juror was not actually biased. 7 Showing that a juror is actually biased poses difficult problems of proof. Ordinarily, however, questioning of prospective jurors in voir dire is the method to smoke out actual juror bias. Voir dire, which literally means, “ ‘to speak the truth,’ ” allows attorneys to determine whether there is a case for dismissing a juror and to form an intelligent basis for the exercise of preemptory challenges. See State v. Dellinger, 696 S.E.2d 38, 43 (W. Va. 2010) (per curiam) (quoting Michael ex rel. Estate of 6Chief Justice Marshall recognized implied bias in the famous treason case of Aaron Burr, stating that a prospective juror “may declare that notwithstanding these prejudices he is determined to listen to the evidence, and be governed by it; but the law will not trust him.” United States v. Burr, 25 F. Cas. 49, 50 (D. Va. 1807). 7The United States Supreme Court has held that when a juror is seated who deliberately concealed bias that would have required he or she be dismissed for cause, reversal of any subsequent conviction is required. United States v. Martinez-Salazar, 528 U.S. 304, 316, 120 S. Ct. 774, 782, 145 L. Ed. 2d 792, 803–04 (2000). The theory is that a jury consisting of eleven impartial jurors and one actually biased juror is constitutionally infirm without any showing that there was juror misconduct which was “calculated to, and with reasonable probability did, influence the verdict.” Cullen, 357 N.W.2d at 27; see also Martinez-Salazar, 528 U.S. at 316, 120 S. Ct. at 782, 145 L. Ed. 2d at 803–04; Parker v. Gladden, 385 U.S. 363, 366, 87 S. Ct. 468, 471, 17 L. Ed. 2d 420, 423 (1966) (per curiam) (noting a defendant is “entitled to be tried by 12, not 9 or even 10, impartial and unprejudiced jurors”). In two relatively recent cases involving both juror misconduct and juror bias, we applied the three-pronged Cullen test to juror misconduct but not to juror bias. See Johnson, 445 N.W.2d at 340–42; Cuevas, 288 N.W.2d at 534–35. We need not reach the issue in this case, however, as we do not find the district court abused its discretion in determining a lack of actual bias in this case. 22 Michael v. Sabado, 453 S.E.2d 419, 426 (W. Va. 1994)). We have held that a party who fails to avail himself or herself of procedures for identifying bias waives later challenges for juror impartiality. See State v. Coffee, 182 N.W.2d 390, 395–96 (Iowa 1970) (citing cases). A troublesome scenario emerges, however, when a potential juror fails to tell the truth in response to questions in voir dire designed to probe potential juror bias. Of course, a nervous prospective juror may simply forget a fact, make an unintentional misstatement, or misunderstand the question. Or, a juror may decline to answer a question to avoid personal embarrassment. See, e.g., United States v. Stewart, 317 F. Supp. 2d 432, 438–39 (S.D.N.Y. 2004) (noting failure of prospective jurors to disclose criminal justice records does not show bias under the circumstances). Of particular concern, however, is a juror who declines to truthfully answer a voir dire question in order to avoid being removed from the jury panel. When a juror conceals information in voir dire in order to avoid either a strike for cause or a preemptory strike, it deprives the accused “of the ability to determine whether [the juror] harbored any prejudices or biases against [the accused] or in favor of the State.” Dellinger, 696 S.E.2d at 43. In addition, deliberate lying by a juror in voir dire may strongly suggest the kind of actual bias that may require disqualification of that juror and, if the juror participates in jury deliberation, may require a new trial. See, e.g., United States v. Colombo, 869 F.2d 149, 150–52 (2d Cir. 1989) (remanding for findings of fact when juror intentionally failed to disclose brother-in-law was government attorney in order to sit on case); United States v. Scott, 854 F.2d 697, 698–700 (5th Cir. 1988) (reversing and remanding for new trial when prospective juror did not disclose that his brother was a deputy sheriff 23 investigating the case and court found this as evidence that the juror wanted to serve but did not disclose the relationship because it would have prevented his service). Here, however, there is no evidence the juror provided false testimony during voir dire. Of course, the fact that voir dire was not reported poses a difficult evidentiary problem for Webster. The parties seem to agree, however, that during voir dire, one prospective juror volunteered she had relationships with the parties. The parties further agree that the juror was not specifically asked about her relationship with the Frisbies. We do not think the failure to volunteer an answer to an unasked question amounts to juror misconduct. See, e.g., McGaha v. Commonwealth, 414 S.W.3d 1, 4–7 (Ky. 2013) (finding juror who did not disclose in voir dire that she was friends with victim’s wife on Facebook did not give false answer when juror stated they were casual friends but did not volunteer the Facebook connection). Further, in the in camera examination, the juror stated that she knew the Frisbies in passing, that her daughter was a friend of Frisbie’s stepsister, that she knew Frisbie’s parents well enough to say “Hi” to them, and that she was Facebook friends with Frisbie’s stepmother. 8 At this point, the door was open to further explore these issues. See Johnson, 445 N.W.2d at 340 (noting that party must pursue issue and juror for cause when party has adequate notice of possible prejudice 8In Sluss v. Commonwealth, two jurors may have been Facebook friends with the mother of the victim, despite statements in voir dire that they did not know the victim or her family. 381 S.W.3d 215, 221, 229 (Ky. 2012) (remanding for hearing to investigate potential juror misconduct). During voir dire, the jurors were not asked directly whether they were Facebook friends with the victim’s mother. Id. at 221. The court noted the mere fact that the jurors were Facebook friends was not determinative, but it was the underlying nature of the relationship and the information that a juror knows that frames whether that juror could reasonably be viewed as biased. Id. at 222– 23. The court further noted that false answers by jurors on voir dire could give rise to serious constitutional issues. Id. at 225. 24 toward defendant). However, Webster’s lawyer elected not to thoroughly explore the nature of the relationship, including the intriguing mention that she and Frisbie’s stepmother were friends on Facebook. Instead, Webster’s counsel asked a series of questions that seemed more designed to rehabilitate the juror than challenge her. At the conclusion of the hearing, Webster did not challenge the juror for cause. Based on the record before us, we cannot conclude the juror engaged in misconduct by lying during the in camera hearing. 9 In short, we are not faced with a juror who lied during voir dire or during an in camera hearing in order to avoid the risk of being disqualified. Thus, an important feature present in many actual disqualification cases is lacking here. Further, both at the in camera hearing and in testimony related to the combined motion for new trial and arrest of judgment, the juror emphatically emphasized that she was capable of, and did in fact, base her verdict solely on the evidence. See, e.g., State v. Walters, 426 N.W.2d 136, 139 (Iowa 1988) (“ ‘It is sufficient if the juror can lay aside his impressions or opinion and render a verdict based on the evidence presented in court.’ ” (quoting Irvin v. Dowd, 366 U.S. 717, 723, 81 S. Ct. 1639, 1643, 6 L. Ed. 2d 751, 756 (1961))). Of course, the juror’s testimony may be self-serving, but the district court found her testimony credible. The mere fact a juror has knowledge of parties or witnesses does not indicate actual bias or require juror disqualification. See, e.g., Green v. State, 757 S.E.2d 856, 858–59 (Ga. 2014) (noting “juror’s knowledge of, or non-familial relationship with, a 9On the issue of juror’s lying during voir dire, there is some division in the caselaw as to whether the standard is objective or subjective. See Loewy, 22 Am. J. Crim. L. at 736–37 & n.21 (noting “critical distinction” between standards and surveying state approaches). We need not determine the issue here because the claim fails under either test. 25 witness, attorney, or party provides a basis for disqualification only if it is shown that it has resulted in the juror having a fixed opinion of the accused’s guilt or innocence or a bias for or against the accused”); see also Ex parte Killingsworth, 82 So. 3d 761, 764–65 (Ala. 2010) (per curiam) (noting same); Marsillett v. State, 495 N.E.2d 699, 707 (Ind. 1986) (noting juror’s mere knowledge of or acquaintance with a victim is not enough to disqualify juror). The district court explained to the parties prior to the in camera hearing that court personnel had advised him the juror had stated the fact that she knew the victim’s family would disqualify her from jury duty. As the above cases demonstrate, if the juror in fact had that belief, it would have been a mistake of law. That brings us to the most troublesome point in the case. There is some suggestion the juror, after the in camera inspection but prior to the verdict, clicked “like” on a Facebook comment by the victim’s stepmother which stated, “Give me strength.” A juror who directly violates the admonitions of the court and communicates with the mother of a crime victim about a case certainly raises questions about her ability to be an impartial juror. This action occurred after voir dire and apparently after the in camera hearing. Thus, Webster has not waived his bias challenge based upon this event, which would not have been uncovered through diligent use of ordinary trial processes. However, the record here does not disclose the court’s initial admonition or when the juror clicked “like.” In any event, while the short form admonition to the jury in the record indicated that the juror should not communicate with parties and witnesses about the case, the juror apparently thought (erroneously) that merely clicking “like” on Facebook was not a “communication.” Moreover, the communication did not relate to the guilt or innocence of the accused, but only showed a degree of 26 empathy for a grieving stepmother who lost her son. A juror who does not have empathy for a grieving mother whose son was a homicide victim would be awfully cold hearted. If we disqualified jurors because they empathized with the family of crime victims, we would have no jurors. See, e.g., Oro-Jimenez v. Commonwealth, 412 S.W.3d 174, 180–81 (Ky. 2013) (finding juror conversation with a victim in a robbery case regarding “how [the victim] was doing” and stating that the juror was sorry the victim had to go through all that, was innocent and did not give rise to a basis for a new trial). We find no abuse of discretion on this point. Notwithstanding our resolution of the issues in this appeal, we do not approve of the juror’s conduct in this case. While the click of the mouse does not require reversal of Webster’s criminal conviction, it is troublesome nevertheless. While it did not occur in this case, a single click of the mouse on Facebook can trigger cascading responses. Further, messages posted on Facebook may be viewed by many persons, generating a perception of a miscarriage of justice. In the future our district courts would do well to recognize that in this day and age, our jurors are part of the new electronic world. This can pose a problem in our jury trials. We have held that the click of the mouse in this case was not misconduct sufficient to require a new trial under the three-part Cullen test. We have also held that the click of the mouse was insufficient to establish that the juror was actually biased in light of the record developed at the posttrial hearing. The click of the mouse did, however, show poor judgment. Indeed, there is a growing body of highly publicized cases showing the risk posed by jurors engaged in electronic and social media activity. In one case, a juror conducted a Facebook poll regarding how she should 27 vote in a case during jury deliberation. See Urmee Khan, Juror Dismissed From a Trial After Using Facebook to Help Make a Decision, The Telegraph, Nov. 24, 2008, http://www.telegraph.co.uk/news/news topics/lawreports/3510926/Juror-dismissed-from-a-trial-after-using-Fa cebook-to-help-make-a-decision.html. In an Arkansas death penalty case, a juror tweeted 10 throughout the trial and continued to do so even after being told to stop. Dimas-Martinez v. State, 385 S.W.3d 238, 247– 48 (Ark. 2011). In another case, a tweeting juror rambled about trial proceedings. United States v. Fumo, 639 F. Supp. 2d 544, 555 (E.D. Pa. 2009), aff’d in part and rev’d in part on other grounds, 655 F.3d 288 (3d Cir. 2011). In a Tennessee case, a juror contacted an expert witness whom she had known in the past to opine, “[Y]ou did a great job . . . . You really explained things so great!!” State v. Smith, 418 S.W.3d 38, 43 (Tenn. 2013). In a Georgia case, a juror found definitions on Google relating to an affirmative defense of habitation as it related to motor vehicles. Chambers v. State, 739 S.E.2d 513, 517–18 (Ga. Ct. App. 2013). As noted by one commentator, “many jurors do not see blogging, tweeting or posting as communication, or at least they don’t consider it to fall within the rubric of traditional admonitions.” Rosalind R. Greene & Jan Mills Spaeth, Are Tweeters or Googlers in Your Jury Box?, 46-Feb. Ariz. Atty. 38, 39 (2010); see also Robert P. MacKenzie III & C. Clayton Bromberg Jr., Jury Misconduct What Happens Behind Closed Doors, 62 Ala. L. Rev. 623, 638 (2011) (“The fastest developing area in the realm of juror misconduct involves the use of e-mail, social networking sites such as Facebook, and microblogging sites such as Twitter during trial.”). 10For an overview of Twitter terminology, see The Story of a Tweet, available at https://about.twitter.com/what-is-twitter/story-of-a-tweet. 28 In order to address the new risks, authorities suggest that courts should frequently, as a matter of course, instruct jurors not to use social media to communicate about the trial and clearly explain what constitutes communication. See, e.g., Steiner v. Superior Ct., 164 Cal. Rptr. 3d 155, 163 (Ct. App. 2013); Kervick v. Silver Hill Hosp., 72 A.3d 1044, 1059 & n.11 (Conn. 2013). The admonition should be given early and often, beginning at the time jurors are sworn and repeated periodically as the trial progresses. While there are many sources of potential admonitions, the United States Judicial Conference Committee on Court Administration and Case Management has recommended that federal district courts use an admonition related to electronic media that may provide a guide to Iowa judges. See Judicial Conference Comm. on Ct. Admin. & Case Mgmt., U.S. Cts., Proposed Model Jury Instructions: The Use of Electronic Technology to Conduct Research on or Communicate about a Case (2012), available at www.uscourts.gov/ file/3159/download?token=3s0ovosm [hereinafter Proposed Model Jury Instructions]; see also U.S. Cts., Revised Jury Instructions Hope to Deter Juror Use of Social Media During Trial (2012), available at www.uscourts.gov/news/2012/08/21/revised-jury-instructions-hope- deter-juror-use-social-media-during-trial. The instruction states, in part: I know that many of you use cell phones, Blackberries, the internet and other tools of technology. You also must not talk to anyone at any time about this case or use these tools to communicate electronically with anyone about the case. This includes your family and friends. You may not communicate with anyone about the case on your cell phone, through e-mail, Blackberry, iPhone, text messaging, or on Twitter, through any blog or website, including Facebook, Google+, My Space, LinkedIn, or YouTube. You may not use any similar technology of social media, even if I have not specifically mentioned it here. I expect you will inform me as soon as you become aware of another juror’s violation of these instructions. 29 Proposed Model Jury Instructions. The instructions also state this admonition “should be provided to jurors before trial, at the close of a case, at the end of each day before jurors return home, and other times, as appropriate.” Id. A trial court providing jurors with admonitions such as those in the federal model will minimize the risk of unnecessary and costly mistrials due to the failure of jurors to clearly understand their obligations in the electronic world. V. Evidentiary Issues. A. Positions of the Parties. Webster makes two evidentiary arguments on appeal. First, he contends the district court erred when ruling Webster could elicit testimony that Frisbie punched his ex-wife, but not testimony that she was pregnant at the time. Second, Webster argues he should have been able to introduce evidence regarding Frisbie’s “prison mentality,” specifically his violent aggression towards authority figures, and felon status. Webster emphasizes this evidence was essential to show Frisbie’s violent character and support his assertion of self-defense and defense of others. Prior to trial, the trial court made several preliminary rulings regarding admissibility of evidence, including disallowing testimony that Frisbie punched his ex-wife while she was pregnant. During trial, Webster made an offer of proof on this matter. Regarding Frisbie’s prison mentality, although Webster resisted the State’s motion in limine regarding Frisbie’s prior nonviolent criminal history, the trial court sustained the State’s motion in limine on this point, again indicating it was not a final ruling. Webster’s trial counsel did not raise the issue of the admissibility of Frisbie’s felonies at trial, Webster therefore argues his attorney provided ineffective assistance for failing to request to present this information. 30 At the outset, the State concedes that because Webster raised the issue of Frisbie punching his pregnant ex-wife during trial, this evidentiary claim was preserved; however, evidence relating to Frisbie’s prison mentality and felon status was not raised during trial and therefore this issue must be presented under an ineffective-assistance-of- counsel framework. On the merits, the State argues the district court properly excluded evidence showing Frisbie’s ex-wife was pregnant when he punched her in the stomach because the probative value was substantially outweighed by its prejudicial impact. The State asserts this additional evidence would not have added significant probative value not otherwise shown by Frisbie’s ex-wife’s testimony that Frisbie punched her, and further, the risk of unfair prejudice was high, in that a jury would be especially sensitive to crimes against a pregnant woman. Regarding Webster’s second contention, under an ineffective- assistance framework, the State asserts that a request to admit the prior felonies would have been meritless, as references to Frisbie’s violent reactions to authority were shown by other witnesses and the probative value of references to prior felonies would have been substantially outweighed by their prejudicial impact. Finally, the State argues Webster has not proven by a reasonable probability that the result would have been different had the evidence been admitted. B. Preservation. It is undisputed that at the pretrial hearing on the motions in limine the trial court made no final evidentiary rulings related to the claims at issue. Webster’s first evidentiary claim, that Frisbie punched his ex-wife in the stomach when she was pregnant, was preserved, as an offer of proof was made at trial regarding this incident. 31 Webster’s second evidentiary claim, regarding Frisbie’s felon status and prison mentality, however, was not preserved. It is undisputed that Webster never attempted to enter into evidence the fact that Frisbie had been to prison or that he was consumed by a prison mentality, which he argued was needed to give context to specific instances of Frisbie’s conduct and his reaction to authority figures (i.e., Frisbie would have seen Webster as an authority figure in trying to intervene and protect Hall and therefore Webster’s actions were reasonable). After Webster attempted to address these issues in the pretrial hearing and was preliminarily denied, he did not attempt to introduce this evidence at trial. He raised the issue again for the first time in his combined motion for new trial and arrest of judgment. The district court denied Webster’s combined motion. Therefore, these claims were not preserved and must be addressed under an ineffective-assistance-of-counsel framework. See Twyford v. Weber, 220 N.W.2d 919, 923 (Iowa 1974) (noting a ruling on a motion in limine “is not a ruling on evidence and should not, except on a clear showing, be used to reject evidence”); see also Quad City Bank & Trust v. Jim Kircher & Assocs., P.C., 804 N.W.2d 83, 90–92 (Iowa 2011) (holding that failure to renew request to present expert testimony about accountant’s work papers or to elicit further ruling from trial court as to its admissibility failed to preserve error); Johnson v. Interstate Power Co., 481 N.W.2d 310, 317 (Iowa 1992) (holding that ruling sustaining motion in limine only prohibited mentioning matter to jury without first securing court permission and did not specifically rule evidence in issue was inadmissible and therefore alleged error was not preserved when the defendant failed to raise issue at time of cross-examination, make offer of proof, or secure ruling on admissibility). 32 C. Merits. Although relevant evidence is generally admissible, Iowa R. Evid. 5.402, the trial court may exclude relevant evidence “if its probative value is substantially outweighed by the danger of unfair prejudice . . . or [amounts to a] needless presentation of cumulative evidence,” id. r. 5.403. We utilize a two-part test to decide whether evidence should be excluded under rule 5.403. State v. Huston, 825 N.W.2d 531, 537 (Iowa 2013). “First, we consider the probative value of the evidence. Second, we balance the probative value against the danger of its prejudicial or wrongful effect upon the triers of fact.” Id. (citations omitted) (internal quotation marks omitted). Probative value refers to “ ‘the strength and force of the evidence to make a consequential fact more or less probable.’ ” State v. Martin, 704 N.W.2d 665, 671 (Iowa 2005) (quoting Rodriquez, 636 N.W.2d at 240). Unfairly prejudicial evidence, on the other hand, appeals to the jury’s sympathies, arouses its sense of horror, provokes its instinct to punish, or triggers other mainsprings of human action that may cause a jury to base its decision on something other than the established propositions in the case. Id. at 672 (internal quotation marks omitted); see also State v. Brown, 569 N.W.2d 113, 117 (Iowa 1997). In Martin, we noted that in weighing probative value and unfair prejudice, the court considers: (1) the need for the proffered evidence “in view of the issues and other available evidence,” (2) whether there is clear proof it occurred, (3) the “strength or weakness of the prior-acts evidence in supporting the issue sought to be prove[d],” and (4) the degree to which the evidence would improperly influence the jury. 704 N.W.2d at 672 (quoting State v. Henderson, 696 N.W.2d 5, 11 (Iowa 2005)); see also State v. Putman, 848 N.W.2d 1, 10 (Iowa 2014) (“Weighing probative value against prejudicial effect ‘is not an exact science,’ so ‘we give a great deal of leeway to the trial judge who must 33 make this judgment call.’ ” (quoting State v. Newell, 710 N.W.2d 6, 20–21 (Iowa 2006))). “Evidence of a homicide victim’s prior violent or turbulent character is ordinarily immaterial and not admissible at trial.” State v. Shearon, 449 N.W.2d 86, 88 (Iowa Ct. App. 1989) (citing State v. Jacoby, 260 N.W.2d 828, 837 (Iowa 1977)). “An exception to this general rule exists where the accused asserts that he or she acted in self-defense and the slightest supporting evidence is introduced.” Id. (noting “[s]pecific instances of conduct may be used to demonstrate character when character is an essential element of a claimed defense”); Jacoby, 260 N.W.2d at 837 (noting that when the defendant asserts self-defense, “the violent, quarrelsome, dangerous or turbulent character of the deceased may be shown, both by evidence of his or her reputation in that respect and by witnesses who can testify from an actual knowledge of the victim’s character”). However, even otherwise relevant admissible evidence is inadmissible “if its probative value is substantially outweighed by the danger of unfair prejudice.” Huston, 825 N.W.2d at 537 (internal quotation marks omitted). 1. Frisbie punching his pregnant ex-wife. During trial, the district court ruled that the “specific instance of the violent act, the striking, is relevant,” but the “probative value [of the fact that Frisbie’s ex-wife was pregnant at the time of the strike] is substantially outweighed by the danger of unfair prejudice.” Here, the trial court correctly found that Frisbie’s act of striking his ex-wife was relevant to show Frisbie’s violent/aggressive character. Punching a woman would support the “violence against women” claim that Webster contended, and the district court found, was relevant to show Frisbie was the first aggressor. However, admitting the additional 34 fact of Frisbie’s ex-wife’s pregnancy was not essential for Webster to prove his defense. Therefore, although this evidence is minimally relevant, the district court did not abuse its discretion in excluding it by finding its probative value was substantially outweighed by the danger of unfair prejudice. See, e.g., Shearon, 449 N.W.2d at 87–88 (excluding testimony of the victim’s attempted rape of a woman hours before the victim’s death because it presented a risk that the jury might think he “got what he deserved”); see also id. at 88 (“Unfair prejudice exists when minimally relevant evidence could lead a jury to improperly use it to reach a decision based on inflammatory and emotional considerations that are unfavorable to a victim because of his or her conduct or lifestyle.”). In considering the Martin factors, on balance, we cannot say the trial court did not fairly weigh the probative value of the evidence against the probable dangers of unfair prejudice of admitting it. The first (need for evidence) and fourth (improperly influence) factors weigh against admitting the evidence. See Martin, 704 N.W.2d at 672–73 (holding the district court should have excluded testimony about defendant’s prior arrests and violent tendencies, as “the first and fourth factors weigh[ed] heavily against admission of the evidence”). Regarding the first factor, there was evidence presented from numerous sources that Frisbie was sexually violent towards women, including testimony from Frisbie’s ex- wife that Frisbie raped her twice over the course of their relationship and struck her multiple times, including in the stomach. In addition, Webster’s wife Ann testified that in her opinion Frisbie was sexually violent. Webster testified about what Frisbie told him about how he violently sexually assaulted women, including his first wife. Finally, Special Agent Jeff Uhlmeyer testified that he retrieved a computer Frisbie 35 had utilized in the past that contained a commercially produced video relating to sexual violence against women. Therefore, “the need for the proffered evidence was weak in light of the other available evidence.” Martin, 704 N.W.2d at 672; cf. Gregg v. United States, 683 F.3d 941, 945– 46 (8th Cir. 2012) (holding prior specific acts would be unfairly prejudicial when, in proving self-defense, the defendant introduced enough evidence to show the victim’s propensity for violence). Although the second (clear proof of occurrence) and third (strength or weakness of prior-acts evidence) Martin factors tend to lend support for admission of the evidence, the fourth factor also weighs against admission of the evidence. 704 N.W.2d at 672. As the State contends, “the jury would foreseeably be sensitive to crimes against a pregnant woman.” Although this fourth factor is a somewhat close call, based on this factor, we cannot say the trial court abused its discretion in excluding the fact that Frisbie’s ex-wife was pregnant at the time Frisbie punched her. Cf. id. at 672–73 (noting the fact that the defendant was violent “could only serve to inflame the passions of the jury”). We emphasize the narrowness of the trial court’s ruling, as evidence of Frisbie’s other violent behavior towards his ex-wife was allowed. Lastly, we note that even if the probative value of the evidence related to Frisbie’s ex-wife’s pregnancy warranted its admission, its exclusion would fall within the realm of harmless error. See id. at 673 (applying harmless error test to trial court’s evidentiary error); State v. Caples, 857 N.W.2d 641, 648 (Iowa Ct. App. 2014) (concluding that there was “overwhelming evidence of guilt and any evidentiary error was harmless”); Shearon, 449 N.W.2d at 88 (“Evidence of Shearon’s guilt was overwhelming and would sustain a finding of harmless error in this case.”). 36 2. Felon status and prison mentality. Webster next claims his counsel was ineffective for not attempting to introduce evidence of Frisbie’s felon status and prison mentality, arguing that without this information, the jury could not fairly assess the reasonableness of Webster’s actions in the face of Frisbie’s violence against Hall. Beginning with the prejudice prong under an ineffective-assistance claim, we find Webster has not shown there is a “reasonable probability that the outcome would have been different” had this specific evidence been admitted. See Everett v. State, 789 N.W.2d 151, 160 (Iowa 2010). Webster was able to present evidence of Frisbie’s “hatred of authority” through other means and therefore, had testimony regarding Frisbie’s prior felonies been admitted, it would not have assisted Webster’s defense. There was evidence presented that Frisbie was generally known as a violent person who hated “authority figures.” Webster himself testified at length about the importance of not “crossing the line” with Frisbie, because once he exerted any amount of perceived authority, Frisbie could snap. Regarding the suggested threesome, Webster testified that he did not want to participate and when he discussed the topic with Frisbie, he would always try to change the conversation because [t]here was a delicate balance that had to be maintained, and it was important not to upset the balance with [Frisbie], because if you were to appear as what I would describe as an authoritative figure, then you could essentially be looked at differently in his eyes more so as an enemy and not as a friend. Webster clarified that if he were to cross the line he would be viewed by Frisbie like others who had crossed the line and would be subject to Frisbie’s violent outbursts. This testimony helped explain to the jury Webster’s theory of the case, namely that he did not have any other 37 choice but to use deadly force when he saw what he believed was Frisbie attempting to rape Hall, and that if he had tried to intervene, Frisbie could have reacted violently towards him or Hall. The jury, therefore, had ample evidence presented to it regarding Frisbie’s relevant prison mentality to put the contended reasonableness of Webster’s actions in context. The jury also heard testimony from multiple sources related to the unreasonableness of Webster’s use of deadly force. Therefore, Webster has failed to prove by a preponderance of the evidence that his trial counsel’s failure to introduce evidence of Frisbie’s felon status and prison mentality resulted in prejudice. VI. Conclusion. For the above reasons, the decision of the court of appeals is vacated and the judgment of the district court is affirmed. DECISION OF COURT OF APPEALS VACATED; DISTRICT COURT JUDGMENT AFFIRMED. All justices concur except Wiggins and Zager, JJ., who concur specially, and Hecht, J., who concurs in part and dissents in part. 38 #13–1095, State v. Webster WIGGINS, Justice (concurring specially). Although I agree with the majority that, on the record made, Tyler Webster is not entitled to a new trial, I have questions whether the representation given Webster was effective. However, issues regarding ineffective assistance of counsel, if any, will have to be fully explored in a postconviction-relief proceeding. Additionally, I am shocked that in a criminal trial in which the State charges the defendant with a serious crime a record was not made of the court’s admonitions or jury selection. 11 Zager, J., joins this special concurrence. 11Our rules of procedure require the court to report these matters unless waived. Iowa R. Civ. P. 1.903(2)(a)–(b); Iowa R. Crim. P. 2.19(4). Waiver of these matters by the parties should be shown in the record. Iowa R. Civ. P. 1.903(2). The record shows Webster’s attorney waived reporting of jury selection. The record does not indicate Webster’s input in the decision to waive the reporting of jury selection. 39 #13–1095, State v. Webster HECHT, Justice (concurring in part and dissenting in part). I agree the juror’s conversation at a convenience store and her posttrial research into Knight’s age do not constitute juror misconduct warranting a new trial. I also concur with the majority’s affirmance of the district court’s evidentiary rulings. Further, I wholeheartedly join in the majority’s recommendation of careful admonitions regarding jurors’ social media usage during trial. However, I part ways with the majority on one very important point. I believe the evidence establishes the juror and Frisbie’s stepmother were more than casual acquaintances and the juror’s interactions with her on Facebook went beyond mere expressions of empathy. I conclude the record sufficiently demonstrates the juror and the stepmother had a relationship that rendered the juror biased. Because my confidence in the fairness of the trial and the jury’s verdict is shaken by evidence of bias, I would affirm the court of appeals decision granting Webster a new trial. “[J]urors must be—and must be perceived to be—disinterested and impartial.” State v. Smith, 418 S.W.3d 38, 45 (Tenn. 2013) (emphasis added); accord State v. Carey, 165 N.W.2d 27, 29 (Iowa 1969) (“[T]he jury is to be above suspicion . . . . [W]e along with all courts have zealously guarded the utter independence of jurors.”). Whether or not the juror in this case was actually biased, the circumstances presented here are such that an observer could reasonably believe she was. Cf. Carey, 165 N.W.2d at 28, 30 (granting a new trial after the county attorney’s office provided the jury with free coffee because “any member of the public who might become familiar with [the relevant facts]” could reasonably conclude it was “an intentional attempt to secure favor,” even if it genuinely was not). Even the perception of unfairness damages our 40 judicial system. See Lynch v. Kleindolph, 204 Iowa 762, 764, 216 N.W. 2, 3 (1927) (“The question before us is not . . . whether any actual wrong resulted from the association with the juror under the circumstances [presented here], but whether it created a condition from which the opposing litigants and the general public might suspect that wrong resulted from this association.” (Emphasis added.)). I would avoid that perception and prevent that damage by granting Webster a new trial in this case. See State v. Delgado, 588 N.W.2d 1, 5 (Wis. 1999) (“[T]he value of finality and the sanctity of a jury verdict must yield when juror bias undermines confidence in the fairness and impartiality of the trial.”); State v. Gesch, 482 N.W.2d 99, 103 (Wis. 1992) (“Whether [a certain juror]’s presence in the jury room actually hindered [deliberations] we will never know, but what is important is the fact that it could have.”). I. The Juror’s Actions and Omissions. I conclude the available evidence, when considered in its entirety, weighs against the juror’s insistence she could be impartial and establishes her bias. Before trial, the juror told a few of her coworkers— the clerk of court and a court attendant—that she would probably not be selected to serve on the jury because she knew the family. Further, evidence in the posttrial record tends to prove she stated after the trial that she remained on the jury because she was not asked a specific enough question during voir dire that would allow her to disclose her relationship with the Frisbie family. The statement to her coworkers and the posttrial statement are evidence the juror knew she had information that could reasonably be perceived as supporting a finding of bias. “She was intruding into a relation for which she believed herself ineligible.” Clark v. United States, 289 U.S. 1, 10, 53 S. Ct. 465, 468, 77 L. Ed. 993, 998 (1933) (emphasis added). 41 Despite her acquaintance with the Frisbie family and her belief she would be excused from service because of it, the juror did not volunteer any information about her relationship with the Frisbie family during voir dire. She later testified she did not volunteer any information because she did not know how; she thought she had to be asked a direct question to speak on the subject. However, although voir dire was not reported, the parties apparently do not dispute that another prospective juror volunteered during voir dire that he knew both the Webster family and the Frisbie family and was consequently excused from service. Having seen another prospective juror speak up, the challenged juror was on notice that she could (and should) do the same. See Delgado, 588 N.W.2d at 7 (concluding that, when four members of the voir dire panel volunteered information tending to show they would be biased, another juror was “on notice that she should reveal” similar information about herself). Accordingly, the challenged juror’s explanation that she was “dumb to the rules” is unpersuasive at best. In Delgado, the Wisconsin Supreme Court concluded that when a juror withheld material information even after other jurors disclosed similar knowledge, a trial court erred in finding no juror bias. See id. at 7–8. I believe Delgado should inform our analysis here. After a court attendant brought the juror’s pretrial comments to the judge’s attention and expressed surprise the juror had been empaneled, the juror testified in camera during trial that she was a Facebook friend of Frisbie’s stepmother. I agree this connection alone might not support a finding that the juror harbored disqualifying bias. As the Kentucky Supreme Court has explained: “[F]riendships” on Facebook and other similar social networking websites do not necessarily carry the same weight as true friendships or relationships in the community 42 . . . . The degree of relationship between Facebook “friends” varies greatly, from passing acquaintanceships and distant relatives to close friends and family. The mere status of being a “friend” on Facebook does not reflect this nuance and fails to reveal where in the spectrum of acquaintanceship the relationship actually falls. Facebook allows only one binary choice between two individuals where they either are “friends” or are not “friends,” with no status in between. .... Consequently, a juror who is a “Facebook friend” with a family member of the victim, standing alone, is arguably not enough evidence to presume juror bias sufficient to require a new trial. As with every other instance where a juror knows or is acquainted with someone closely tied to a case, it is the extent of the interaction and the scope of the relationship that is the relevant inquiry. Sluss v. Commonwealth, 381 S.W.3d 215, 222–23 (Ky. 2012). However, in this case the extent of the interaction and scope of the relationship— the “relevant inquiry” according to the Kentucky Supreme Court—reveals a closer connection than “passing acquaintanceship.” See id. Although they were friends before trial, the nature of the relationship between the juror and Frisbie’s stepmother is most dramatically evidenced by their Facebook communication during and after trial. The stepmother posted a request for emotional support and the juror responded, through the Facebook act of “liking.” As the majority notes, the juror’s affirmative act of responding to the request does not expressly relate to Webster’s guilt or innocence. Yet, I conclude actions establishing disqualifying bias need not rise to that high level of materiality. I think it evident that a reasonable person observing the stepmother’s request for support would have understood it was directly related to the emotional turmoil arising from the ongoing trial of the murder charge against Webster. More importantly, a reasonable person 43 could understand the juror’s act of liking the request as the juror’s affirmative expression of emotional support for the stepmother. This case presents more troubling facts than Oro-Jimenez v. Commonwealth, 412 S.W.3d 174, 180–81 (Ky. 2013). In Oro-Jimenez, a juror conversed with a victim in a robbery case regarding “how [the victim] was doing” and stated the juror was sorry the victim had “to go through that.” Id. at 180. The Kentucky Supreme Court held this interaction did not provide grounds for a mistrial. Id. at 181. However, the interaction took place during the penalty phase, after the jury had already rendered a guilty verdict. See id. at 180. In other words, while the juror perhaps felt sympathy for the victim during trial, the juror at least waited to express that sympathy directly until after rendering a verdict on guilt. Further, I see nothing in Oro-Jimenez indicating that the juror and the victim had a social relationship that antedated the trial. In this case, the juror’s relationship with Frisbie’s stepmother existed before the trial, and her communication during trial occurred before the jury started deliberating. The communications evidencing a relationship between the juror and Frisbie’s stepmother continued after the trial. It seems quite clear to me that the juror’s perception of the closeness of the relationship demanded she communicate with the stepmother about the jury’s verdict. Indeed, the content of the juror’s posttrial communication in this case—“I wish you could have gotten murder in 1st degree”—carries a more troublesome substantive connotation than the manifestation of generalized empathy at issue in Oro-Jimenez. The juror’s posttrial communication in this case reveals she was motivated to return a verdict she knew her friend wanted and felt obligated to offer an explanation as to why that did not occur. It allows for the possibility that the previous 44 Facebook expression of emotional support to the stepmother during trial really meant “I’ve got your back,” and it evidences a closer relationship than a juror and a victim’s family would or should typically share. See Lynch, 204 Iowa at 764–65, 216 N.W. at 3 (concluding “the district court should have granted a new trial” when the defendant in a civil case had lunch with a juror during trial, even though the two men had never met before and even though “the case on trial was never mentioned at any time” during the meal); cf. Clark, 289 U.S. at 11, 53 S. Ct. at 468, 77 L. Ed. at 998 (“What was sought to be attained was the choice of an impartial arbiter. What happened was the intrusion of a partisan defender.”). II. Assurances of Impartiality. Furthermore, I conclude the district court afforded too much weight to the juror’s professed ability to be impartial. Indeed, I find unsatisfying the notion that the principal consideration in determining bias is the potentially biased juror’s own assurances that he or she can be impartial. See Mary R. Rose & Shari Seidman Diamond, Judging Bias: Juror Confidence and Judicial Rulings on Challenges for Cause, 42 Law & Soc’y Rev. 513, 518 (2008) [hereinafter Rose & Diamond] (“[S]tudies suggest that judges declare jurors to be fair when jurors say they can be fair.”). I recognize that in many instances, there is no other information available, because bias challenges tend to be raised during voir dire and adjudicated on very limited inquiries attributable in part to the press of time. See Dov Fox, Neuro-Voir Dire and the Architecture of Bias, 65 Hastings L.J. 999, 1012 (2014) [hereinafter Fox] (acknowledging trial judges and lawyers conducting voir dire are often “in the precarious position of trying to read [potential jurors’] minds in an effort to ascertain outside influences that might affect the jurors’ decisionmaking”). I also 45 recognize that some prospective jurors claim bias “as a pretext for getting out of service,” and judges therefore often approach bias claims skeptically, aiming to keep a juror eligible if possible to prevent a mass exodus from the venire. Rose & Diamond, 42 Law & Soc’y Rev. at 515– 16. But this case presents neither of those features; it arises from a posttrial motion, when the judge could consider not only the juror’s testimony, but the totality of other available evidence. In this procedural posture, I think the juror’s own assurances deserve less weight and should be treated just as skeptically as those made during voir dire by a potential juror who flippantly says whatever they think will get them excused from service. See State v. Lindell, 629 N.W.2d 223, 233 (Wis. 2001) (“[I]n some cases bias can be detected . . . even though [the juror] pledges impartiality.”); see also Clark, 289 U.S. at 10, 53 S. Ct. at 468, 77 L. Ed. at 998 (affording little weight to a juror’s own testimony, even though she “stated to the court that her mind was free from bias,” because the other available evidence was in conflict). Self-assessments of one’s own impartiality are often overly optimistic. See Rose & Diamond, 42 Law & Soc’y Rev. at 516 (“[P]eople often have difficulty producing accurate self-assessments of bias and find it difficult to estimate whether events or prior experiences are likely to influence them.”); see also Shari Seidman Diamond et al., Realistic Responses to the Limitations of Batson v. Kentucky, 7 Cornell J.L. & Pub. Pol’y 77, 92 (1997) [hereinafter Diamond et al.] (“People are often unable to recognize the extent to which their experiences or attitudes affect their judgments.”). Thus, “simply asking jurors whether they can be impartial is not likely to reveal with any reliability the presence or strength of many of the outside influences that they would in fact bring to bear on the questions at trial.” Fox, 65 Hastings L.J. at 1011. Jurors may say 46 they can be impartial—even if they cannot be—for a variety of reasons, and they might even be “oblivious to the . . . bias they harbor.” Id. at 1020; see also Rose & Diamond, 42 Law & Soc’y Rev. at 516 (recognizing judges faced with a juror “who promises to be fair” must evaluate whether the juror “is perhaps underestimating the potential difficulty of remaining fair and impartial during the trial”). The most pernicious example of overstated impartiality occurs when a juror has a personal or financial interest in the parties or the outcome, and therefore wants to be selected for or remain on the jury to influence it. See Fox, 65 Hastings L.J. at 1023–25. There are, of course, other more benign reasons jurors are so often confident they can be fair. For example, jurors may profess impartiality to boost their self-image because “questions suggest that it is ‘better’ to answer one way than another.” Rose & Diamond, 42 Law & Soc’y Rev. at 516. In other words, “individuals recognize that fairness is a desirable characteristic, and most people want to believe that they possess it.” Id.; see also Diamond et al., 7 Cornell J.L. & Pub. Pol’y at 90 (“[J]urors may be hesitant to reveal experiences or attitudes that would embarrass them . . . .”); Valerie P. Hans & Alayna Jehle, Avoid Bald Men and People with Green Socks? Other Ways to Improve the Voir Dire Process in Jury Selection, 78 Chi.-Kent L. Rev. 1179, 1195 (2003) [hereinafter Hans & Jehle] (“The desire to appear favorably is a main concern of prospective jurors, and that shapes [what] they disclose . . . .”). Additionally, jurors may “identify with the community or feel pressure to conform to its norms in ways that favor a particular side”—a concern that prompted the court to change venue for the trial of Timothy McVeigh because the possibility of jurors feeling “pressure to conform . . . decisionmaking in a way that would serve [the] community’s perceived interests” was too great. Fox, 47 65 Hastings L.J. at 1029, 1031–32. Frisbie’s death was gruesome and the juror in this case acknowledged that Fairfield was abuzz both on the night Frisbie died and during Webster’s trial. I think it is entirely reasonable to conclude the community was very interested in these proceedings, and this general atmosphere of rapt attention may have influenced the juror even though she confidently testified anything she had previously heard or read about the case would not and did not affect her deliberations. But most importantly, jurors often state they can be impartial simply because they believe the judge wants them to be. See Mark W. Bennett, Unraveling the Gordian Knot of Implicit Bias in Jury Selection: The Problems of Judge-Dominated Voir Dire, the Failed Promise of Batson, and Proposed Solutions, 4 Harv. L. & Pol’y Rev. 149, 160 (2010) (“As a [federal] district court judge for over fifteen years, I cannot help but notice that jurors are all too likely to give me the answer that they think I want, and they almost uniformly answer that they can ‘be fair.’ ”); Diamond et al., 7 Cornell J.L. & Pub. Pol’y at 91 (“[J]urors may be hesitant to reveal opinions that they believe might engender disapproval from the judge or others.”); Kurt F. Ellison, Comment, Getting Out of the Funk: How Wisconsin Courts Can Protect Against the Threat to Impartial Jury Trials, 96 Marq. L. Rev. 953, 979 (2013) (“[J]urors’ statements of impartiality are often motivated by pressure from the judge . . . .”). It is easy to capitulate when the approval of an authoritative figure might be at stake: [J]urors may give in to the pressure to comply and say they can be impartial, even though their real feelings have not changed. The judge’s approval is important to a lot of . . . jurors and many will alter their responses or hide certain attitudes in order to be perceived favorably. 48 Hans & Jehle, 78 Chi.-Kent L. Rev. at 1194. Of course, if a juror is just saying what they think the judge wants to hear, their words illuminate very little about whether they are actually biased. Because the juror in this case could have had other reasons for stating she could be impartial, and because the record includes other evidence of bias, I view the juror’s assurances very skeptically. III. The District Court’s Untenable Conclusions. Finally, in addition to the juror’s relationship with Frisbie’s stepmother and the weight afforded her assurances of impartiality, I believe two of the grounds for the district court’s decision to deny Webster’s motion for new trial are clearly untenable. Trial courts have broad discretion to decide whether juror bias warrants a new trial. See State v. Johnson, 445 N.W.2d 337, 340 (Iowa 1989). However, a district court abuses that discretion if it renders a decision on clearly untenable grounds. I think the district court did so here. First, the district court concluded any bias the juror harbored “was not reflected in the verdict in which she participated,” because the jury convicted Webster of second-degree murder rather than first-degree murder. In other words, the court concluded empaneling a juror biased against a criminal defendant is harmless error if the biased person does not succeed in convincing the rest of the jury that it should return a conviction on the most serious charge available. But this principle cannot be right, because it would deny a hypothetical defendant a new trial if a juror concealed racial bias until after a trial convicting the defendant only of a lesser included offense—but brazenly admitted it when questioned after the verdict. As Webster’s trial counsel asserted before the district court, the important consideration is not the ultimate result, but the process of reaching it: 49 [A biased juror] is one mind, one person, that we were deprived of having the opportunity to convince. That could have been the one hung juror, that could have been the one juror that held out until the very end and swayed everyone else. Because of what happened, we are forever denied of knowing what would have been different. I conclude the district court’s reliance on harmless error analysis—due to the jury’s determination that Webster committed a lesser included offense—makes its decision on the bias issue clearly untenable. Second, the district court denied Webster’s motion for new trial because there was no “indication that any . . . ostensible bias influenced or infected any discussions or deliberations of the jury as a whole.” But how would we know if the juror’s bias influenced the deliberations? Our rules of evidence limit challenges to verdicts by precluding presentation of any evidence regarding jury deliberations except for the question of “whether extraneous prejudicial information was improperly brought to the jury’s attention or whether any outside influence was improperly brought to bear upon any juror.” Iowa R. Evid. 5.606(b). Jurors generally may not testify about what was said in the jury room or what did or did not motivate any juror to reach a particular verdict. Id. (“[A] juror may not testify as to any matter or statement occurring during the course of the jury’s deliberations or to the effect of anything upon that or any other juror’s mind or emotions as influencing the juror . . . .”). In this case, any evidence the juror’s bias was manifest during deliberations would fall right in the sweet spot of rule 5.606(b) and be excluded from consideration. It was in my view clearly untenable for the district court to support its decision by citing an absence of evidence on whether bias influenced the jury’s deliberations when evidence on this question is so severely curtailed by the applicable rule. 50 IV. Conclusion. When considered together and in context, the juror’s belief she would be disqualified from service if her relationship with Frisbie’s family was known, her silence during voir dire about her acquaintance with that family, her less than fulsome disclosure of recent Facebook communications with the victim’s stepmother during the in-chambers examination, and her Facebook contacts with Frisbie’s stepmother during and after trial are enough in my view to show a relationship giving rise to an inference of the juror’s actual bias. The juror insisted she could decide the case impartially, but as I have noted, people often overestimate their own capacity for impartiality. See United States v. Burr, 25 F. Cas. 49, 50 (C.C.D. Va. 1807) (No. 14692G) (“[A juror] may declare that notwithstanding [his] prejudices he is determined to listen to the evidence, and be governed by it; but the law will not trust him.”); Lindell, 629 N.W.2d at 235 (“It is not always enough that a . . . juror assures counsel or the court that he or she will be impartial.”). Although I strongly endorse the majority’s recommendation of careful jury admonitions regarding jurors’ use of social media during trial, I would affirm the court of appeals decision granting Webster a new trial in this case.