Snap Shot
- The question: How often are sexual assault reports shown to be false, and what do terms such as “unfounded,” “unsubstantiated,” and “not prosecuted” actually mean?
- What the research shows: Where researchers require evidence that the assault did not occur, confirmed false reports make up a small share of reports to authorities, in the commonly cited range of about 2% to 10%; a false-report rate counts only reports shown to be false, and classifications can err in either direction. Reports not substantiated far outnumber them: in one university study of 136 reports, 5.9% were coded false and 44.9% did not proceed to any prosecution or disciplinary action.
- What it does not show: The 2% to 10% range is not evidence that 90% to 98% of reports are accurate, and the large unresolved middle cannot be counted as true or false. Non-prosecution, acquittal, DNA exclusion, and recantation-based figures such as Kanin’s do not measure falsity. No base rate converts into a probability for a particular complainant.
- The law: Statistics quantifying a complainant’s likely truthfulness are the functional equivalent of vouching (Brooks, Mullins, Lindsey), and the rule applies equally to statistics suggesting falsity. A prior accusation is admissible only if shown to be false, and an acquittal is not dispositive (Erikson); Arizona requires clear and convincing evidence (A.R.S. § 13-1421; Gilfillan).
- Issues for each side: Prosecutors should correct arguments equating unfounded or unprosecuted cases with false ones, without using false-report rates to imply this complainant is truthful; defense counsel should require that any cited rate come with its definition and non-substantiation rate, without treating unresolved cases as false.
Sexual assault is common and substantially underreported; in the National Crime Victimization Survey, about one in four rape and sexual assault victimizations were reported to police in the most recent survey year (23.6% in 2024); the estimate has varied considerably from year to year (46.0% in 2023), so no single figure should be treated as fixed (Tapp & Coen, 2025). Most reports that are made do not end in a conviction, and a small number are shown to be false. All three facts reach the courtroom. The question of false reports arises when the defense contends that an allegation is untrue or mistaken, when either party seeks to introduce statistics about how often reports are false, when an expert is asked about the subject on cross-examination, and when the defense seeks to introduce evidence that a complainant has made a false allegation before.
One feature of the system shapes every statistic in this area. Investigations do not ordinarily set out to prove that a report was false. They ask whether the evidence supports charges or discipline, and when it does not, the case is usually closed for insufficient evidence without any finding about what happened. A formal determination that a report was false is uncommon. The proportion of reports that are not substantiated, that is, closed without referral for prosecution or discipline, is therefore the figure that every agency actually produces, while a false-report rate counts only the smaller set of cases in which someone went further and documented that the report was untrue. A false-report rate quoted without the non-substantiation rate from the same study leaves out most of the cases that did not result in a proven offense.
The research on this question is frequently misdescribed by both sides, largely because the terms used to describe reports that do not result in prosecution have different meanings. A report may be false, unfounded, unsubstantiated, not prosecuted, dismissed, or ended by acquittal, and each of these categories reflects a different decision by a different person on a different standard. Only the first describes a report shown to be untrue, and even that term is used inconsistently.
This paper defines the categories, explains why the non-substantiation rate is the more complete measure, reviews the principal studies of false reporting and the methods behind them, discusses research on confirmed false allegations and on the pathways by which inaccurate reports can arise, explains why DNA exclusion data do not measure false reporting, and describes the large unresolved middle category that research cannot assign to either truth or falsity. It closes with the law governing statistical credibility testimony and evidence of prior false allegations.
Definitions
A false report, in the research literature, ordinarily means a report of a sexual assault that did not occur, established by evidence rather than by the absence of proof. Some researchers limit the term to deliberate fabrications; others include sincere but inaccurate reports, such as those arising from a mistaken memory or a misidentification of the assailant. Which definition a study uses affects its results and must be stated whenever its figures are cited.
Unfounded is a law-enforcement classification. Under the Federal Bureau of Investigation’s Uniform Crime Reporting handbook, an agency unfounds a complaint that “is determined through investigation to be false or baseless. In other words, no crime occurred” (Federal Bureau of Investigation, 2004, p. 77). The handbook adds that a victim’s refusal to cooperate with prosecution, the failure to make an arrest, and the findings of a court, jury, or prosecutor do not unfound an offense that law enforcement investigation has established to be legitimate (p. 78). The handbook does not define “baseless” separately; the term is generally understood to cover a report in which the reported conduct, even if it occurred, does not meet the legal elements of the offense. Unfounded therefore includes, but is broader than, false. The term means something different in child protective services, where an unfounded report is simply one that was not substantiated. Because the category covers reports determined false or baseless, and a baseless finding may involve conduct that occurred but did not meet the legal definition, an unfounding rate should not be read as a false-report rate; police also often unfound cases that do not meet false-report criteria (Kelly et al., 2005; Spohn et al., 2014). A figure broader than the confirmed-false rate must be described in this context as the share of reports not substantiated, closed for insufficient evidence, or not referred for prosecution. The Department of Defense uses a similar two-part definition in its disposition data, in which commanders or special trial counsel, after legal review, may find a case false or baseless (U.S. Department of War, 2026, p. 23).
Unsubstantiated, insufficient evidence, and similar terms describe cases in which the investigation did not produce enough evidence to support charges or discipline. They are statements about the available proof. Decisions not to prosecute, dismissals, and acquittals are likewise determinations about whether a legal standard was met, not findings about what happened. Interviews with British police officers and prosecutors have shown that practitioners themselves apply the label “false” to very different kinds of cases, which helps explain why practitioners and researchers often reach different estimates (Saunders, 2012). The written record behind a classification also varies. In a sentiment analysis of 5,638 police narratives tied to previously untested sexual assault kits in one Midwestern U.S. jurisdiction, narratives of reports classified as unfounded were shorter and more negative in tone than other narratives, while narratives in which officers recorded that the victim lied or was doubted did not differ from the rest (Lovell et al., 2026). The study measured how reports were written, not whether the classifications were correct.
The Non-Substantiation Rate and the False-Report Rate
There are two ways to ask how often reports fail. One is to ask what proportion were not substantiated: investigated and closed without referral for prosecution or discipline. Every agency classifies every case it closes, so this figure is available in virtually every study. The other is to ask what proportion were false. That figure exists only when an investigator made an affirmative determination of falsity, for example on a credible admission by the complainant or on strong evidence that the assault did not occur.
Because investigations seldom take that additional step, a false-report rate counts only the cases in which falsity was documented. Reports that were in fact false but were closed for insufficient evidence are not counted, because no one examined them for that purpose. A false-report rate therefore counts only the reports shown to be false; it is not an estimate of how many reports were inaccurate, and classifications can err in either direction, as the police designations reviewed by Kelly et al. (2005) illustrate. Quoting it alone, as though it were the complement of the rate of true reports, is misleading. The non-substantiation rate has the opposite limitation: it includes true reports that could not be proven, sincere but mistaken reports, and the occasional fabrication, so it cannot be read as a false-report rate either. Its value is completeness. It accounts for every report and shows how large the zone of uncertainty actually is.
The studies that report both figures show the gap. Of the 136 reports in the university study discussed below, 8 (5.9%) were coded as false, 61 (44.9%) did not proceed to any prosecution or disciplinary action, 48 (35.3%) were referred for prosecution or discipline, and 19 (13.9%) contained too little information to classify (Lisak et al., 2010). In the Department of Defense’s fiscal year 2025 data, 1,121 cases considered for disciplinary action were closed for insufficient evidence, compared with 27 found false or baseless, a ratio of roughly 40 to 1 (U.S. Department of War, 2026, pp. 20–21, Table 4). Over a longer span, court-martial convictions fell from about 9 per 100 unrestricted reports in fiscal year 2009 to about 3 per 100 in fiscal year 2025, while the share of trials ending in conviction stayed between about 61% and 80%; the companion paper What Department of Defense Sexual Assault Data Can and Cannot Show sets out the year-by-year figures and their limits. In each instance the cases that were not substantiated far outnumber those shown to be false, and neither figure establishes how many reports were true.
Research Methods and Estimates
Estimates of false reporting differ mainly because studies differ in who decides whether a report is false and by what criteria. Studies that rely on police classifications inherit whatever standards the officers used. In a large Home Office study of reported rapes in England and Wales, police designated 216 of 2,643 reported cases (8%) as false, but the researchers found that many of those designations rested on uncertain grounds, including reliance on the complainant’s mental health problems or previous allegations; when they reassessed the cases with enough information against the standard in the police’s own rules, which limit the category to a clear and credible admission or strong evidence, only about 3% of reported cases (67 of 2,643) were probable or possible false reports (Kelly et al., 2005). The report’s summary states the police-designated figure as 9% rather than the 8% computed from the case counts, and describes the 3% figure as applying Home Office counting rules; the 3% rests on the information available in the police case records and is the researchers’ reassessment, not a finding that the remaining designations were correct or incorrect. A legal review of the field concluded that published estimates vary widely and that much of the variation reflects differences in definition, research quality, and reliance on police judgment (Rumney, 2006).
Studies that apply explicit criteria to case files produce more consistent figures. Lisak and colleagues (2010) reviewed all 136 sexual assault reports made to a university’s police over ten years and coded a report as false only where a thorough investigation produced evidence that the assault did not occur, following guidance consistent with the International Association of Chiefs of Police; 8 reports (5.9%) met that standard. Spohn, White, and Tellis (2014) reviewed sexual assault reports made to the Los Angeles Police Department in 2008 and interviewed the detectives; they estimated that about 4.5% of reports were false, found that the department sometimes cleared cases as unfounded that did not meet the false-report criteria, and found that a complainant’s recantation was the strongest predictor of an unfounding decision. A meta-analysis of seven studies in which researchers evaluated case files found a pooled rate of confirmed false reports of 5.2%, with a 95% confidence interval of 3.0% to 8.9% and considerable variation among studies (Ferguson & Malouff, 2016).
These studies support a cautious summary. Where researchers require evidence that an assault did not occur, confirmed false reports make up a small share of reports to authorities, in the commonly cited range of about 2% to 10% (Lisak et al., 2010; Spohn et al., 2014; Ferguson & Malouff, 2016, 5.2%, 95% CI 3.0% to 8.9%). These figures describe police and campus case classifications, leave a large share of cases unresolved, and cannot be applied to any one complainant. That range is an estimate of reports shown to be false. Ferguson and Malouff (2016) noted that including cases classified as equivocal would produce a higher figure, and none of these studies found that the remaining reports were shown to be true. The range is therefore neither an upper limit on inaccurate reports nor evidence that 90% to 98% of reports are accurate. Any figure from this range should be presented together with the non-substantiation rate from the same study.
Kanin (1994) and Its Limitations
Kanin (1994) is frequently cited for a much higher figure. He reviewed forcible rape complaints to a single police agency in a small Midwestern city from 1978 to 1987 and reported that 45 of 109 (41%) were declared false because the complainant recanted. The study’s design limits what it can show. The sole criterion for falsity was recantation; the department routinely offered polygraph examinations to complainants, some recantations occurred in connection with those examinations, and complainants who recanted were told they could be charged with filing a false report. The article does not describe a systematic procedure for evaluating whether recantations were themselves accurate. Offering polygraph examinations to complainants may pressure some of them to withdraw, a concern raised by later commentators rather than a finding of the study, and the study offered no independent check on the police determinations. Kanin himself cautioned that the 41% figure should not be extrapolated to other populations. Because recantation can occur for many reasons in true cases, a figure based on recantation alone cannot be treated as a measure of false reporting, and the study is not representative of the better-controlled research described above.
Research on Confirmed False Allegations
A small body of research examines cases in which falsity was firmly established, in order to learn how such cases arise. De Zutter, Horselenberg, and van Koppen (2018) studied 57 allegations in Dutch national police files from 1997 to 2011 that met three conditions: the complainant retracted and stated that no rape occurred, investigation showed the allegation was false, and other evidence such as a confession or forensic results corroborated what had actually happened. The most commonly coded motive was emotional gain, and about a quarter involved an alibi, such as concealing a consensual encounter; a substantial minority of complainants said they did not know why they had made the report, and motives were undocumented in others. The authors noted that the sample was small and drawn from one country, that the data were coded by one person from case files, and that their strict criteria excluded many cases.
This research describes a small set of confirmed cases. It does not estimate how often false allegations occur, and it does not suggest that complainants generally have such motives. Many of the circumstances that appear in confirmed false cases, such as relationship conflict, a concealed encounter, or intoxication, may also be present in true reports, and I located no research comparing how often they occur in true and false reports. The presence of a possible motive is therefore subject to the same non-specificity limitation that applies to post-assault behavior: it may be relevant to a factfinder’s assessment of the evidence, but research does not show that it distinguishes true reports from false ones.
Pathways to Inaccurate Reports
Not every inaccurate report is a deliberate fabrication. Engle and O’Donohue (2012) proposed eleven pathways by which a report of sexual assault could be inaccurate, including deliberate lying, misinterpretation of whether consent was implied, false memories, intoxication, delirium, psychotic and dissociative conditions, certain personality disorders, and intellectual disability. Their article is a conceptual framework rather than an empirical study; it does not estimate how often any pathway occurs, and the authors called for research on frequency and detection. A separate and well-documented pathway in cases involving strangers is mistaken identification of the assailant, in which the assault occurred but the wrong person was accused. These distinctions matter in court because an honest but mistaken account calls for different evidence and a different argument than a deliberate fabrication, and because research on memory, alcohol, and identification procedures addresses the former. Those topics are covered in this hub’s papers on memory and in the Memory in the Courtroom resource.
DNA Exclusions
DNA exclusion data are sometimes cited as evidence of false reporting. The principal source is a 1996 National Institute of Justice report that reviewed 28 cases in which men convicted of crimes, nearly all involving sexual assault, were exonerated by post-conviction DNA testing; in most of those cases eyewitness identification was the most compelling evidence at trial (Connors et al., 1996). The report also surveyed crime laboratories and found that, in about 21,600 cases submitted for testing, DNA results excluded the suspect in about 23% and were inconclusive in about 16%.
An exclusion shows that the suspect was not the source of the tested biological material. In many such cases the exclusion may reflect a question of identity (the police suspected the wrong person) or material from a consensual partner rather than the assailant, but the report did not determine why suspects were excluded. The report itself noted that the laboratories lacked information about what happened to excluded suspects. The exclusion figures therefore speak to the risk of misidentification and the value of forensic testing, not to how often complainants report assaults that did not occur. The converse also holds: in a case where identity is not disputed and the issue is consent, a DNA match does not show that an assault occurred.
The Unresolved Middle
In every careful study, the largest group of reports is neither confirmed false nor established through prosecution, as the Lisak and Department of Defense figures above illustrate. These cases are unresolved: the investigation did not establish what happened. Some may involve assaults that could not be proven, some may involve sincere but mistaken accounts, and some may involve false reports that were not detected. Research provides no method for dividing them among those possibilities, and no case-level inference can be drawn from the group.
Attrition research shows where most cases leave the system. A systematic review of 36 U.S. studies covering 288,066 cases from 2000 to 2020 found average rates across studies of about 89% of sampled reports founded by police (that is, not unfounded), 27% resulting in arrest, 31% in referral for prosecution, 20% in charges, and 8% in conviction, with wide variation that reflected which cases each study included, such as only cases with sexual assault kits (Hoffman et al., 2025). In most samples, then, the gap between reports and convictions is not explained by unfounding. In U.S. Army data obtained under the Freedom of Information Act, the odds that law enforcement would not find probable cause were 5.30 times greater for sexual assault than for homicide, robbery, and assault from 2008 to 2014, and 4.39 times greater from July 2015 to 2017 (Carpenter, 2023). Investigator judgment also plays a part: in a random sample of reports from one police department in the western United States, officers’ expressions of doubt were associated with lower odds of arrest and referral for prosecution, independent of the legal and extralegal factors measured (King & Bostaph, 2024). Single-jurisdiction and journalistic counts show a similar pattern. Among 5,031 rapes and attempted rapes reported to the Los Angeles Police Department from 2005 through 2009, 12.2% were cleared by arrest and 80.2% of the cases in which charges were filed ended in conviction; the report cautioned that adding clearances by exceptional means to arrest clearances substantially inflates clearance rates, and 43.4% of reports remained open and 10.9% were unfounded (Spohn & Tellis, 2012). A news investigation of agency and court records from eight U.S. cities, covering about ten years where data allowed, reported that less than 4% of reported rapes, sexual assaults, and child sex abuse allegations in those cities ended in a sex crime conviction, with wide variation by city, differing periods and offenses counted, and some counts still subject to change as cases were pending (Walinchus et al., 2025). That investigation is journalistic and not peer reviewed. These studies are observational. They cannot say whether particular decisions were right, whether the doubt was warranted, or how much of the difference reflects the evidence available rather than investigative practice. They do show that a decision not to arrest, refer, or find probable cause is shaped by the investigation and is not a finding of falsity.
Both parties are tempted to resolve the middle category in their favor. Treating unresolved or unprosecuted cases as false inflates false-report estimates and is inconsistent with the definitions used in the research; the Court of Appeals for the Armed Forces has held that an acquittal on a prior allegation does not itself establish that the allegation was false (United States v. Erikson, 2017). Treating confirmed-false estimates as proof that the remaining reports are true makes the opposite error. The accurate statement is that a small proportion of reports are shown to be false, a proportion are referred for prosecution or discipline and some end in conviction, and a large proportion remain unresolved.
Limits of the Evidence
The false-report studies are limited to reports made to police or institutions and cannot describe unreported assaults or unreported false accusations. They depend on case files prepared by investigators whose practices vary, and their classifications are only as good as the investigations behind them. Samples are often small, several key studies are a decade or more old, and they come from a few countries and settings, including university campuses, which may not resemble military or other populations. Definitions differ across studies. Most importantly, a base rate for a population of reports cannot be converted into a probability that a particular complainant is accurate or inaccurate; that question depends on the evidence in the case, and military and Arizona authority forbid experts from offering such probabilities.
Questions to Ask About This Research
When a witness cites the 5.9% false-report figure from Lisak et al. (2010), ask what exactly was counted and out of what total: 8 of 136 reports met a strict evidentiary standard, while 61 (44.9%) did not proceed to prosecution or disciplinary action, so the figure counts only reports shown to be false and does not show that the remaining reports were accurate. When fiscal year 2025 Department of Defense data are offered, ask about both numbers: 27 cases found false or baseless and 1,121 closed for insufficient evidence (U.S. Department of War, 2026, pp. 20–21, Table 4); insufficient evidence describes proof, not falsity. When Kanin (1994) is cited for a higher figure, ask whether the conclusion follows from the design, since the sole criterion was recantation at a single agency that offered complainants polygraph examinations. When the confirmed false cases studied by De Zutter, Horselenberg, and van Koppen (2018) are used to suggest that a complainant’s possible motive points to fabrication, ask who was studied and compared with whom: 57 cases from one country, with no comparison of how often such circumstances occur in true reports. These questions are explained in “Reading Research Claims: A Guide for Attorneys” on this site.
False Reports and the Law
Federal and Military Standards
The Court of Appeals for the Armed Forces treats testimony quantifying the likelihood that a complainant is truthful as the functional equivalent of vouching. In United States v. Brooks (2007) (a child-complainant case), expert testimony about the rarity of false allegations that implied a better than 98 percent likelihood of truthfulness was plain error, and in United States v. Mullins (2010) (also a child-complainant case), testimony implying a 1-in-200 chance of a false allegation was error. The rule applies equally to statistics offered to suggest falsity, such as non-prosecution or acquittal rates. Fed. R. Evid. 702 and Mil. R. Evid. 702, as interpreted in Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993) and United States v. Houser (1993), govern the reliability and fit of any expert testimony on the research itself.
Evidence that a complainant previously made a false allegation is governed by several rules. In Erikson (2017), the court questioned whether Mil. R. Evid. 412 applies to a prior accusation of assault, which does not obviously concern the complainant’s sexual behavior or predisposition, but held that, under any theory of admissibility, evidence of a prior accusation is admissible only if the accused shows that the prior accusation was false; an acquittal at the earlier proceeding was not dispositive, and the military judge did not abuse his discretion in excluding the evidence. Where Mil. R. Evid. 412 applies, it requires a written motion and a closed hearing, and its exception for constitutionally required evidence applies when the evidence is relevant, material, and favorable to the defense. Specific instances of untruthfulness may also be inquired into on cross-examination under Mil. R. Evid. 608(b), in the military judge’s discretion and without extrinsic evidence.
Arizona
Arizona’s rape shield statute permits evidence of false allegations of sexual misconduct made by the victim against others, if the court finds it relevant and material and not outweighed by its prejudicial nature, after a hearing on written motions; the proponent must establish admissibility by clear and convincing evidence (A.R.S. § 13-1421(A)(5), (B)). In State v. Gilfillan (2000), the Court of Appeals upheld exclusion of alleged prior false accusations where the defendant had not met that standard, and noted that Arizona Rule of Evidence 608(b) gives trial courts discretion to limit inquiry into unrelated acts of untruthfulness. State v. Lindsey (1986) bars expert testimony quantifying the proportion of complainants who tell the truth (149 Ariz. at 475–76).
Special Issues in Legal Practice
The principles described above apply with equal force to every party in a legal proceeding. The science operates the same way regardless of who is asking the questions, and the research summarized here does not by itself favor either side. What differs between the parties is not the science but the vantage point — the particular places where these issues tend to surface, and the particular errors each party may need to guard against. The two summaries below identify those recurring issues. They are offered for orientation, not as litigation strategy; in every case the controlling question is what the evidence in that case will, and will not, support, and the expert’s role is to inform the trier of fact rather than to advocate for either side.
Special Issues for Prosecutors
The research supports the statements that confirmed false reports make up a small share of sexual assault reports, that non-prosecution and acquittal do not indicate falsity, and that recantation-based estimates such as Kanin (1994) do not measure false reporting. Prosecutors can explain that the non-substantiation rate is not a false-report rate, to correct a witness or argument that equates unfounded, unprosecuted, or unresolved cases with false ones, and to insist that the definitions used in any cited study be stated. Prior-false-allegation motions are properly resolved under the burden that Erikson and § 13-1421 place on the proponent.
The research also counsels caution. The 2% to 10% range describes reports shown to be false, not the proportion of reports that are true; presenting it without the non-substantiation rate from the same studies understates how often reports go unconfirmed, and the large unresolved category cannot be counted as accurate. More fundamentally, any use of false-report statistics to suggest that this complainant is probably truthful runs directly into Brooks, Mullins, and Lindsey, and creates appellate risk even if the figures are accurate.
Special Issues for Defense
The research supports the statements that false and mistaken reports do occur, that inaccurate reports can arise through several pathways other than deliberate fabrication, including memory error, intoxication, and misidentification, and that confirmed-false estimates count only reports shown to be false and do not establish the accuracy of the remaining reports. Defense counsel can ensure that any witness who cites a false-report rate states its definition and its limits and also gives the non-substantiation rate from the same study, and to develop case-specific evidence of mistake or fabrication under the rules governing prior allegations and impeachment.
The research also counsels caution about the defense’s own arguments. The non-substantiation rate measures uncertainty, not falsity. Unfounded, unsubstantiated, unprosecuted, acquitted, dismissed, and DNA-exclusion cases are not false reports, and Kanin (1994) cannot support a general estimate. Research on motives in confirmed false cases describes those cases only; a possible motive is common in true reports as well and does not by itself indicate falsity. Statistical arguments about falsity are subject to the same credibility rules that bind the prosecution. Arguments grounded in the evidence of the particular case present the most defensible use of the research.
This paper is for educational purposes only and is not legal advice. The author is a psychologist, not an attorney; summaries of law are general and must be verified against current authority or with a licensed attorney.
References
Carpenter, E. R. (2023). Differential treatment of sexual assault cases by US Army law enforcement personnel. Violence Against Women, 29(5), 882–900. https://doi.org/10.1177/10778012221101915
Connors, E., Lundregan, T., Miller, N., & McEwen, T. (1996). Convicted by juries, exonerated by science: Case studies in the use of DNA evidence to establish innocence after trial (NCJ 161258). U.S. Department of Justice, National Institute of Justice.
De Zutter, A. W. E. A., Horselenberg, R., & van Koppen, P. J. (2018). Motives for filing a false allegation of rape. Archives of Sexual Behavior, 47(2), 457–464. https://doi.org/10.1007/s10508-017-0951-3
Engle, J., & O’Donohue, W. (2012). Pathways to false allegations of sexual assault. Journal of Forensic Psychology Practice, 12(2), 97–123.
Federal Bureau of Investigation. (2004). Uniform crime reporting handbook (Rev. ed.). U.S. Department of Justice.
Ferguson, C. E., & Malouff, J. M. (2016). Assessing police classifications of sexual assault reports: A meta-analysis of false reporting rates. Archives of Sexual Behavior, 45(5), 1185–1193. https://doi.org/10.1007/s10508-015-0666-2
Hoffman, E. E., Patton, E., & Greeson, M. R. (2025). A systematic review of sexual assault case attrition in the United States from 2000 to 2020. Trauma, Violence, & Abuse, 26(4), 850–866. https://doi.org/10.1177/15248380241293803
Kanin, E. J. (1994). False rape allegations. Archives of Sexual Behavior, 23(1), 81–92. https://doi.org/10.1007/BF01541619
Kelly, L., Lovett, J., & Regan, L. (2005). A gap or a chasm? Attrition in reported rape cases (Home Office Research Study 293). Home Office Research, Development and Statistics Directorate.
King, L. L., & Bostaph, L. M. G. (2024). “That is not behavior consistent with a rape victim”: The effects of officer displays of doubt on sexual assault case processing and victim participation. Journal of Interpersonal Violence, 39(5–6), 973–995. https://doi.org/10.1177/08862605231200252
Lisak, D., Gardinier, L., Nicksa, S. C., & Cote, A. M. (2010). False allegations of sexual assault: An analysis of ten years of reported cases. Violence Against Women, 16(12), 1318–1334. https://doi.org/10.1177/1077801210387747
Lovell, R. E., Caporale, L., & Du, J. (2026). Decoding disbelief: Using natural language processing’s sentiment analysis to assess 24 years of unfounded rape reports narratives. Behavioral Sciences & the Law, 44(1), 47–62. https://doi.org/10.1002/bsl.70020
Rumney, P. N. S. (2006). False allegations of rape. Cambridge Law Journal, 65(1), 128–158. https://doi.org/10.1017/S0008197306007069
Saunders, C. L. (2012). The truth, the half-truth, and nothing like the truth: Reconceptualizing false allegations of rape. British Journal of Criminology, 52(6), 1152–1171. https://doi.org/10.1093/bjc/azs036
Spohn, C., & Tellis, K. (2012). Policing and prosecuting sexual assault in Los Angeles City and County: A collaborative study in partnership with the Los Angeles Police Department, the Los Angeles County Sheriff’s Department, and the Los Angeles County District Attorney’s Office (NCJ 237582). National Institute of Justice. https://www.ojp.gov/pdffiles1/nij/grants/237582.pdf
Spohn, C., White, C., & Tellis, K. (2014). Unfounding sexual assault: Examining the decision to unfound and identifying false reports. Law & Society Review, 48(1), 161–192. https://doi.org/10.1111/lasr.12060
Tapp, S. N., & Coen, E. J. (2025). Criminal victimization, 2024 (NCJ 310547). U.S. Department of Justice, Bureau of Justice Statistics.
U.S. Department of War. (2026). Appendix B: Statistical data on sexual assault [Appendix to the Department of War annual report on sexual assault in the military, fiscal year 2025]. Sexual Assault Prevention and Response Office. https://www.sapr.mil/Portals/156/FY25_Appendix_B.pdf
Walinchus, L., Smyser, K., & Murphy, J. (2025, January 14). A vanishingly small number of violent sex crimes end in conviction, NBC News investigation shows. NBC News. https://www.nbcnews.com/specials/sex-assault-convictions
Legal Authorities
Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579 (1993).
State v. Gilfillan, 196 Ariz. 396, 998 P.2d 1069 (App. 2000).
State v. Lindsey, 149 Ariz. 472, 720 P.2d 73 (1986).
United States v. Brooks, 64 M.J. 325 (C.A.A.F. 2007).
United States v. Erikson, 76 M.J. 231 (C.A.A.F. 2017).
United States v. Houser, 36 M.J. 392 (C.M.A. 1993).
United States v. Mullins, 69 M.J. 113 (C.A.A.F. 2010).
Ariz. Rev. Stat. Ann. § 13-1421.
Fed. R. Evid. 412.
Fed. R. Evid. 702.
Mil. R. Evid. 412.
Mil. R. Evid. 608(b).
Mil. R. Evid. 702.
Ariz. R. Evid. 608(b).
Related topic papers: What Department of Defense Sexual Assault Data Can and Cannot Show · Behavioral Indicators, Base Rates, and the Limits of Inference · Recantation in Adult Sexual Assault Cases · Memory Contamination: Discussion, Interviews, and Post-Event Information · Vouching, Profile Evidence, and the Human Lie Detector Rule
Related topics from Arizona Forensic Services: Memory in the Courtroom · Alcohol Blackouts · Courts-Martial