by David Kim

The Supreme Court of Iowa held that generalized expert testimony concerning psychological factors affecting eyewitness identification accuracy and witness confidence should “ordinarily” be admitted when a criminal prosecution relies in part or whole on eyewitness testimony. Providing the first guidance on the boundaries of trial court discretion in nearly three decades since State v. Schutz, 579 N.W.2d 317 (Iowa 1998), the Court established a new analytical framework by using principles from its child sexual abuse expert testimony jurisprudence. The Court further announced that the presence or absence of corroborating evidence is “irrelevant to the admissibility analysis.” Applying these principles, the Court concluded that the district court abused its discretion by categorically excluding defense expert testimony that would have addressed only general psychological principles without commenting on any specific witness or identification. Because the exclusion was prejudicial where eyewitness identification constituted the central dispute at trial, the Court reversed the conviction and remanded for a new trial.

Background

In spring 2022, two women in Boone, Iowa, reported separate incidents of indecent exposure. On March 30, K.W. was sitting in her vehicle at a grocery store parking lot when a man in an adjacent car exposed himself. She reported the incident that evening and described the perpetrator as a white male, possibly in his forties, who may have been balding, wearing light-colored gym shorts and driving a silver sedan. Law enforcement obtained surveillance footage but could not capture the incident itself due to camera blind spots. An officer did observe a silver vehicle in the footage with distinctive features: a bent license plate, front bumper discoloration, and an unusual stance with the rear sitting lower than the front.

Approximately one week later, on April 5, E.P. experienced a similar incident while leaving a gym. She observed a man masturbating in his vehicle. With the gym owner’s assistance, she attempted to record the license plate number as the vehicle departed. Their attempts yielded two similar but inexact numbers. Police investigation revealed that a closely matching plate, IAP625, was registered to a silver Camry owned by the defendant, Pat Kepner.

Detective John Mayse assembled a six-photo array using Kepner’s photo alongside five similar individuals from the county jail, with backgrounds altered to appear uniform. When shown the array, K.W. could not identify anyone. However, E.P. selected Kepner’s photograph. Mayse subsequently visited Kepner’s home, and the resulting bodycam footage was later shown to both witnesses. Each woman stated that the man in the video was the person who had exposed himself to her.

The State charged Kepner with two counts of indecent exposure under Iowa Code § 709.9(1). At trial, both K.W. and E.P. testified about their experiences and identified Kepner in the courtroom. On redirect examination, each expressed absolute certainty, testifying there was “no doubt” Kepner was the perpetrator.

Prior to trial, the defense retained Dr. Kim MacLin, a psychologist specializing in sensation, perception, and memory in legal contexts. MacLin was prepared to testify about research-based factors affecting eyewitness identification reliability. The State moved to exclude or limit this testimony. The district court deferred ruling until trial, at which point the defense made an offer of proof.

During the offer of proof, MacLin explained that lineups, photo arrays, and investigative interviews all constitute methods of collecting memory evidence. Scientific research has produced protocols for conducting these procedures to yield more trustworthy results. She testified that flawed photo selection – such as when filler photographs do not match witness descriptions as closely as the suspect’s photo or differ in lighting, positioning, or source – can draw witnesses toward particular images in ways that feel like genuine recognition but actually reflect problems with the array itself.

MacLin also emphasized the importance of preliminary instructions to witnesses. Before presenting photographs, investigators should inform witnesses that the investigation will continue regardless of their response and that excluding innocent individuals is equally important as identifying perpetrators. Without such guidance, witnesses may feel compelled to select someone, which research shows increases false identifications.

Additionally, MacLin testified that array administrators should ideally be unaware of which photograph depicts the suspect to avoid inadvertently communicating that information through subtle cues. She explained that administrator feedback following a selection can artificially inflate witness confidence in subsequent identifications. Regarding repeated identification procedures, MacLin observed that memory changes over time and that witness confidence often increases through successive questioning and identification processes, but research demonstrates that confidence does not reliably indicate accuracy.

The district court excluded MacLin’s testimony, determining that it indirectly addressed witness credibility, a matter within the exclusive province of the jury. The court ruled that testimony regarding memory and confidence would confuse jurors and constitute improper commentary on witness credibility.

The jury convicted Kepner on both counts. The Court of Appeals affirmed, and the Iowa Supreme Court granted further review.

Analysis

The Court began its analysis by discussing Iowa Rule of Evidence 5.702, which permits experts to testify from specialized knowledge when their testimony will assist the trier of fact in understanding evidence or determining factual issues. The Court stated that Iowa courts hold a “liberal view” toward admitting expert testimony that satisfies this helpfulness requirement, while simultaneously excluding testimony on matters fully comprehensible to ordinary jurors without expert assistance. State v. Stendrup, 983 N.W.2d 231 (Iowa 2022); State v. Krogmann, 998 N.W.2d 141 (Iowa 2023).

The Court traced Iowa’s evolving approach to eyewitness identification expert testimony, ultimately providing the first substantive guidance since Schutz on when trial courts abuse their discretion in excluding such evidence. In State v. Galloway, 275 N.W.2d 736 (Iowa 1979), a seven-justice special concurrence concluded that such testimony could properly be excluded because the subject matter was not beyond ordinary juror knowledge. For nearly two decades, Iowa maintained this effective bar on eyewitness identification experts.

The Court reconsidered this approach in Schutz, observing that numerous jurisdictions had permitted such testimony following California’s influential decision in People v. McDonald, 690 P.2d 709 (Cal. 1984). Schutz noted California’s conclusion that available psychological research on eyewitness identification was “sufficiently beyond common experience” that expert testimony “could assist the trier of fact” in appropriate cases. Finding no other state with a categorical exclusion rule, Schutz abandoned Galloway’s per se approach and committed admissibility decisions to trial court discretion.

Importantly, the Court acknowledged that “in the almost three decades since Schutz, we have not had an opportunity to explore the boundaries of that discretion.” Two post-Schutz developments informed the Court’s current analysis. First, pro-admission trends have continued in other jurisdictions. Citing Commonwealth v. Walker, 92 A.3d 766 (Pa. 2014) (collecting cases), the Court observed that 44 states plus the District of Columbia now permit such testimony at trial court discretion, as do virtually all federal circuits. State v. Doolin, 942 N.W.2d 500 (Iowa 2020). Second, Iowa has developed substantial jurisprudence governing expert testimony in child sexual abuse cases, jurisprudence the Court determined was instructive for eyewitness identification testimony.

New Framework: Importing
Child Sexual Abuse Expert Testimony Principles

The Court declared that “the same approach should govern expert testimony about eyewitness identifications” as governs expert testimony in child sexual abuse prosecutions. This framework distinguishes permissible generalized testimony from impermissible credibility vouching.

In State v. Dudley, 856 N.W.2d 668 (Iowa 2014), the Court prohibited experts from vouching for a child witness’ truthfulness or offering “consistent with” testimony connecting the child’s symptoms to abuse generally. Such testimony impermissibly infringes on the jury’s exclusive credibility-determination function and may mislead jurors by providing a “scientific certainty stamp of approval” on testimony that experts cannot reliably validate.

Nevertheless, Dudley recognized that generalized expert testimony about child victim behavior “can be very beneficial to assist the jury in understanding some of the seemingly unusual behavior child victims tend to display.” Juries may harbor misconceptions about how abused children behave, such as why victims delay reporting or give inconsistent accounts, and experts can help jurors evaluate testimony in light of psychological research.

In State v. Leedom, 938 N.W.2d 177 (Iowa 2020), the Supreme Court approved generalized expert testimony about child abuse reporting patterns where the expert had never met the victim, never used the victim’s name, offered no opinion regarding the victim’s truthfulness, and drew no connection between the victim’s experience and the research she described. Such generalized testimony was permissible under established precedent, according to the Leedom Court.

The Court concluded that identical principles govern eyewitness identification expert testimony. Just as jurors may misunderstand child abuse victim behavior, they may also harbor misconceptions about memory function during investigative processes and how those processes can alter witnesses’ perceptions of their own recollections, including artificially increasing confidence through repeated identifications and investigator feedback.

Building on this newly adopted framework, the Court announced an important presumption: “Ordinarily, this sort of generalized testimony should be admitted where, as here, a criminal prosecution relies in part or whole on a witness’s eyewitness identification.” This language establishes that when prosecutions depend in part or whole on eyewitness testimony, trial courts should generally admit properly limited expert testimony addressing the psychological factors that can affect identification reliability and witness confidence, the Court instructed.

The Court rejected the Court of Appeals’ suggestion that admissibility should depend on whether eyewitness testimony was corroborated by other evidence, an approach some jurisdictions have adopted. The Court acknowledged cases from Georgia and New York that had conditioned admissibility on the absence of corroboration. Johnson v. State, 526 S.E.2d 549 (Ga. 2000); People v. Santiago, 958 N.E.2d 874 (N.Y. 2011), abrogated by People v. Vaughn, 258 N.E.3d 1159 (N.Y. 2024).

Expressly declining to follow this approach, the Court reasoned that expert testimony satisfies Rule 5.702’s helpfulness requirement whenever it assists jurors in evaluating presented evidence, including eyewitness testimony, regardless of what other evidence exists. The Court explained that corroborative evidence is “irrelevant to the admissibility analysis.” State v. Guilbert, 49 A.3d 705 (Conn. 2012); In re L.C., 92 A.3d 290 (D.C. 2014).

Application to
MacLin’s Testimony

Applying these principles, the Court determined that there was no proper basis for excluding MacLin’s testimony. Her proposed testimony would have provided scientific insights about how investigative methods can affect identification reliability and witness confidence, matters “sufficiently beyond common experience” to benefit jurors, the Court stated. Schutz. Critically, the Court explained that like the approved testimony in Leedom, MacLin’s testimony was wholly generalized. She did not claim familiarity with the specific identifications at issue, never mentioned any witness by name, drew no connections between general principles and case-specific facts, and offered no opinion on the validity of any identification or the propriety of procedures actually used.

The Court held that the district court erred by concluding that expert witness testimony concerning psychological factors affecting identification accuracy and witness confidence categorically constitutes impermissible credibility commentary. While experts may not vouch for witnesses directly or indirectly, this prohibition does not bar generalized psychological education that may inform how jurors evaluate testimony. The district court’s categorical approach misapplied governing law and therefore constituted an abuse of discretion, according to the Court.

Prejudice Analysis

Turning to whether the error affected Kepner’s substantial rights, the Court applied the presumed-prejudice standard governing nonconstitutional evidentiary errors. Under this framework, reviewing courts presume prejudice and reverse unless the record affirmatively establishes otherwise. State v. Sullivan, 679 N.W.2d 19 (Iowa 2004); McGrew v. Otoadese, 969 N.W.2d 311 (Iowa 2022).

The Court determined that the record did not affirmatively establish harmlessness. The perpetrator’s identity was the sole contested issue, and MacLin’s testimony directly addressed that dispute. Errors touching “a hotly contested central dispute” are difficult to characterize as harmless. Tarbox ex rel. S.K. v. Obstetric & Gynecologic Assocs. of Iowa City & Coralville, P.C., 13 N.W.3d 546 (Iowa 2024).

Additionally, the State repeatedly emphasized the eyewitness identifications’ centrality. In opening statement, the prosecutor told jurors that while various evidence would be presented, the case would ultimately depend on the eyewitnesses pointing to the defendant. Closing argument reinforced this theme, declaring the case “boils strictly down to identification.” When prosecutors emphasize particular evidence as especially important, courts are reluctant to dismiss related errors as inconsequential. State v. Skahill, 966 N.W.2d 1 (Iowa 2021).

The jury instruction on eyewitness identification reliability, while valuable, did not substitute for MacLin’s testimony. The instruction did not explain that investigators should inform witnesses that investigations continue regardless of their responses and that excluding innocent persons is equally important as identifying perpetrators. Nor did it warn jurors that witness confidence may have been artificially enhanced through investigative processes and may not reliably indicate accuracy, information particularly relevant given K.W.’s and E.P.’s emphatic expressions of certainty.

Conclusion

Accordingly, the Court vacated the Court of Appeals decision, reversed the district court judgment, and remanded for a new trial. See: State v. Kepner, 27 N.W.3d 545 (Iowa 2025)

Editor’s Note: Anyone interested in the admissibility of expert witness testimony on eyewitness identification reliability is encouraged to read the Court’s full opinion.   

Subscribe today

Already a subscriber? Login