by Collin Derrig, Associate Member, University of Cincinnati Law Review Vol. 93
I. Introduction
Usually, when a person calls 911, they are dealing with an emergency ranging from a fire to a sick child to a violent crime. When faced with such traumatic circumstances, no one person deals with the situation the same as another. However, this has not stopped a small school of criminal analysts from attempting to scientifically analyze 911 calls to use as evidence against the caller during criminal investigations.[1] These criminal analysts claim that 911 call analysis is a useful tool for the investigation and prosecution of criminal defendants.[2] Based on Detective Tracy Harpster’s research and book, analysts believe there are over thirty indicators that suggest the person who called 911 is guilty of the crime they are reporting.[3] Yet, despite being the creator of 911 call analysis, Harpster has not taken the stand as an expert in a criminal trial.[4] Instead, Harpster spread his work through publications, teaching classes all over the country, and consulting on over 2,500 cases.[5] Harpster’s work has been used to identify suspects, interrogate them, and, most insidiously, has been mixed into in-court statements and lay testimonies to sneak it past the standards for expert testimony.[6]
This blog will explain why 911 call analysis cannot be used as a reliable and verifiable method to adjudicate crimes. Part II will first review the difference between lay and expert testimony and outline the standards for admitting expert testimony from Daubert v. Merrell Dow Pharmaceuticals, Inc. and Frye v. United States. Part II will then provide a brief introduction to 911 call analysis, how it is used, and how independent research puts its validity into question. Part III will demonstrate that 911 call analysis is a form of expert testimony that falls short of admissibility under the tests derived from Daubert and Frye.[7] Then, Part III will discuss potential solutions to address the improper use of 911 call analysis in investigations and court. Part IV will call defense attorneys, judges, police, and prosecutors to take the appropriate steps to immediately end the use of 911 call analysis in investigation and prosecution.
II. Background
The Background will first lay out the differences between expert and lay testimony and explain the different standards for admissibility. The Background will then explore the origins of 911 call analysis, how it works, and how it has been used in real investigations and prosecutions. Finally, the Background will overview peer-reviewed studies attempts to replicate 911 call analysis and criticisms against 911 call analysis.
A. Expert vs. Lay Witness Testimony: The Daubert and Frye Standards
Rules 701 and 702 of the Federal Rules of Evidence (“FRE”) outline the differences between lay and expert witness testimony.[8] FRE 701 provides that lay witness testimony is limited to an opinion that is: (a) rationally based on the witness’s perception; (b) helpful to clearly understanding the witness’s testimony or to determining a fact in issue; and (c) not based on scientific, technical, or other specialized knowledge within the scope of FRE 702.[9] FRE 702 provides that, for a qualified expert’s testimony to be admissible: (a) the expert’s scientific, technical, or other specialized knowledge must help the trier of fact to understand the evidence or to determine a fact in issue; (b) the testimony must be based on sufficient facts or data; (c) the testimony must be the product of reliable principles or methods; and (d) the expert’s opinion must reflect a reliable application of the principles and methods to the facts of the case.[10] Most states use similar standards in their own state rules of evidence.[11] FRE 701 and 702, and their state equivalents, establish the clear principle that if a witness wishes to testify using information acquired through specialized knowledge, skill, experience, training, or education the witness must first be qualified as an expert.[12] Depending on the jurisdiction, an expert witness must be qualified under the Daubert or Frye standard.[13]
The most widely used standard for admitting expert testimony is the Daubert standard.[14] This standard was first established in Daubert v. Merrell Dow Pharmaceuticals Inc.[15] In Daubert, the Supreme Court worried about the admission of pseudoscientific or unreliable expert testimony at trial and made judges responsible for determining if expert testimony was too unreliable or unverifiable to be admissible.[16] When assessing the reliability of an expert, Daubert requires judges to assess the reasoning and methodology employed by an expert as opposed to merely relying on an expert’s credentials.[17] Under Daubert, judges must use a five-factor test to determine if expert testimony is admissible.[18] These five factors are: 1) whether the technique or theory in question can be, and has been, tested; 2) whether it has been subjected to publication and peer review; 3) its known or potential error rate; 4) the existence and maintenance of standards controlling its operation; and 5) whether it has attracted widespread acceptance within a relevant scientific community.[19] When determining whether to admit expert testimony, federal courts must consider and weigh all five Daubert factors to determine validity.[20]
All federal courts and most state courts employ the Daubert standard.[21] However, some state courts still use an older method of admitting expert testimony.[22] In Frye v. United States, a case decided seventy years before Daubert, the United States Court of Appeals for the District of Columbia held that expert testimony may only be introduced if it is widely accepted and used in a field.[23] As such, the Frye standard is more limiting than Daubert and only allows testimony by experts using methodologies that generally follow accepted practices in a given field.[24]
B. The Creation of 911 Call Homicide Analysis
In 2006, Tracy Harpster, then a police lieutenant in Moraine, Ohio, wrote his master’s thesis titled, The Nature of 911 Homicide Calls: Using 911 Homicide Calls to Identify Indicators of Innocence and Guilt.[25] In this thesis, Harpster used a sample of one hundred 911 calls from homicide cases which resulted in either a conviction or a plea.[26] Harpster collected this sample by directly contacting the departments and officers who investigated these cases.[27] In half of these cases, the 911 caller was convicted or confessed to the crime that they called about.[28] In the other half, the 911 caller was either never a suspect or was eliminated as a suspect at some point in the investigation.[29] Based on this original sample, Harpster identified sixteen factors, the presence or absence of which are indicative of either guilt or innocence.[30]
Harpster’s thesis was inspired by a 2004 training conducted by FBI Analyst Dr. Susan Adams.[31] Two years after publishing his thesis, Harpster partnered with Dr. Adams to release an article in the FBI Law Enforcement Bulletin publicly introducing the world to 911 call analysis.[32] This article breaks down several of the most important indicators first referenced in Harpster’s thesis.[33] In 2009, shortly after the publication of this article, Harpster, Adams, and a third researcher, John Jarvis, released an exploratory paper.[34] This paper reinforced the results of Harpster’s previous work, this time identifying eighteen guilt indicators on 911 calls.[35] Most recently, in 2017, Harpster and Dr. Adams published a book providing a comprehensive breakdown of an updated version of 911 call analysis and instructions on how to conduct it.[36]
Harpster and Dr. Adam’s book, Analyzing 911 Homicide Calls: Practical Aspects and Applications, contains the most up-to-date version of the duo’s research.[37] Harpster and Adams used an expanded sample of two hundred 911 homicide calls and have now identified thirty-eight indicators that can be used to identify whether a 911 caller is guilty or innocent of the crime they are reporting.[38] These indicators are organized into a proprietary scale called the Considering Offender Probability in Statements (“COPS”) Scale.[39] The COPS scale indicators can be broken down into six basic categories: 1) who the call is about; 2) what the call is about; 3) how the call is made; 4) passive defenses; 5) resistance to dispatcher; and 6) additional guilty indicators.[40] The first three categories of indicators contrast the presence or absence of a behavior or statement to indicate guilt or innocence.[41] For example, an immediate plea for help for the victim indicates innocence, while no immediate plea for help for the victim indicates guilt.[42] The second three categories of indicators, however, tend to indicate guilt if they are present in a 911 call.[43] For example, having a “Defendant Mentality,” making unintelligible comments, using awkward phrases, saying “Huh?” or “I don’t know” are all indicators of guilt.[44]
Despite the publication of Harpster and Adams’s book, much is still unknown about 911 call analysis. Harpster and Adams have not released how they obtained the calls used to create the current COPS scale or how they conducted their research.[45] They have also never made their data publicly available.[46] This is contrary to standard practice for this sort of research.[47]
C. Use of 911 Call Analysis in Investigation and Prosecution
After the publication of the 2008 FBI Bulletin, Harpster’s work quickly became popular.[48] Harpster and Adams were approached by departments interested in using their work as part of criminal investigations.[49] As of March 2024, Harpster has consulted on at least 2,500 homicide investigations.[50] In addition to direct consultation, Harpster also conducts training for dispatchers, law enforcement officers, and attorneys all over the country.[51] According to ProPublica, prosecutors and investigators in at least twenty-six states have used 911 call analysis in investigations and prosecutions.[52] Harpster often promotes his work and training by using testimonials provided by investigators and prosecutors who have used 911 call analysis successfully.[53]
Harpster encourages dispatchers, patrol officers, investigators, and, most importantly, prosecutors to use 911 call analysis.[54] He claims that 911 call analysis can be used by law enforcement to inform the investigation of the crime scene, by dispatchers to flag suspicious indicators so the dispatcher can keep the caller talking to implicate themselves further, and by prosecutors when trying the case.[55]
1. 911 Call analysis and Law Enforcement Investigation
In principle, conducting 911 call analysis is exceedingly simple. An analyst trained in Harpster’s method listens to the audio of a 911 call or reads through a transcript of a 911 call.[56] As they listen or read, the analyst analyzes the call using the COPS scale.[57] The analyst checks a box for every indicator of guilt or innocence present.[58] If an indicator is not present, the analyst circles “n/a.”[59] The more indicators leaning towards guilt or innocence, the higher the likelihood that the caller is either innocent or guilty.[60] However, Harpster’s work states that some indicators have stronger relationships to guilt or innocence than others, meaning certain indicators may be weighed more heavily by analysts when determining guilt or innocence.[61]
911 call analysis is being used by police to identify and convict people who call 911 for help. In a Michigan case, police used 911 call analysis to identify the caller as a suspect because the caller made the mistake of saying “I need help.”[62] In a 2012 case from Fremont, Ohio, Detective Jason Kiddey, who had previously seen Harpster speak, used 911 call analysis to elicit a confession from a man accused of murdering his wife.[63] Kiddey later wrote an email to Harpster stating he did as Harpster instructed and that the man “cracked” after Kiddey threw down a 911 call analysis handout.[64] The man was charged with murder, and pled to manslaughter and an 11-year prison sentence.[65]
911 call analysis has also transformed otherwise undetermined deaths into homicides. In 2018, a young mother in Washington called 911 to report that her baby had stopped breathing in his sleep.[66] Three different pathologists analyzed the physical evidence and were unable to determine if the death was a homicide.[67] However, the death was deemed suspicious after Detective Marty Garland performed a 911 call analysis of the mother’s call.[68] Garland told Harpster in an email that he noticed problems with the mother’s first word—“hi”—because being too polite indicates that the caller may be hiding something.[69] After Garland reported his findings to a supervisor and shared the mother’s call analysis with the pathologist, the baby’s death was re-classified as a homicide.[70] The mother took an Alford plea to manslaughter rather than face a second-degree murder charge and a potential life sentence.[71] Garland told Harpster, “we would never have known the truth if it hadn’t been for your book and excellent training.”[72]
2. 911 Call Analysis and Homicide Prosecutions
In addition to the role 911 call analysis plays in homicide investigations, it is also being utilized by prosecutors at trial.[73] However, it is rarely used and admitted as expert evidence at trial.[74] In 2009, Harpster was rejected as an expert witness at a trial in Alabama.[75] Similarly, a Nevada judge rejected a detective trained in 911 call analysis.[76] In one Michigan case, a judge admitted a police detective trained in 911 call analysis as an expert on the basis that Harpster’s training class was recognized by the Michigan Commission on Law Enforcement Standards.[77] The defendant, 16-year-old Riley Spitler, was convicted of murder, but his conviction was overturned because 911 call analysis and other statement analysis techniques should not have been admitted.[78]
The difficulties in getting 911 call analysis admitted as expert testimony have led Harpster and prosecutors to take a different approach.[79] Because 911 calls are already frequently admitted into evidence and played in court to provide juries with context and insight, Harpster suggests that prosecutors trained in 911 call analysis can “guide juries through both the audio and transcript of the calls,” which allows prosecutors to “create lasting impressions of this critical evidence.”[80] Harpster highlights that 911 call analysis is “logical and straightforward” and has basic principles that jurors can easily understand.[81]
Harpster provides ten tips for prosecutors to use 911 call analysis as part of their prosecution strategy without needing to seek the required qualification for expert testimony.[82] Harpster recommends that prosecutors provide the 911 call audio and transcript to all witnesses—including expert witnesses like medical examiners, blood splatter experts, and DNA analysts.[83] Harpster also recommends that prosecutors prepare questions about the 911 call for witnesses and make sure they know 911 call analysis.[84] This allows witnesses to reference relevant COPS scale indicators when testifying and, in the case of expert witnesses, to use 911 call analysis to inform how they conduct their expert analysis.[85] Harpster suggests that prosecutors should prime the jury for 911 call analysis by “weaving the 911 research throughout the case,” starting in voir dire, continuing in their opening statement, then through witness testimony, followed by cross-examination, and finally bringing it all together in their closing argument.[86] Harpster states that, if done properly, this can lead the jury members to an “aha moment” where they think “I wouldn’t act or talk how the defendant did if I called 911 in that situation.”[87]
Prosecutors have employed Harpster’s strategies to skirt evidentiary rules and bring 911 call analysis to trial. A former Macomb County, Michigan prosecutor highlighted how she used 911 call analysis to get convictions in two child death cases.[88] In the first case, a dispatcher trained in 911 call analysis took the stand to testify, making a “HUGE impact” according to the prosecutor.[89] In the second case, the prosecutor highlighted 911 call guilty indicators in her closing argument to help obtain a guilty verdict.[90] In Ohio, a prosecutor who had taken Harpster’s course contacted Harpster for help preparing questions for witnesses.[91] The prosecutor later sent an email thanking Harpster, saying his assistance helped them to secure a guilty verdict by highlighting the parts of a 911 call that indicated guilt.[92]
D. Review of Independent Research and Criticism of 911 Call Analysis
Despite the extensive use of 911 call analysis in practice, by both Harpster and those he has trained, no peer-reviewed study has been able to replicate or authenticate Harpster’s work.[93] Since the publication of Harpster’s book in 2017, seven peer-reviewed studies have looked at Harpster’s methods and failed to validate their reliability.[94] Some have found inverse relationships from Harpster’s work between innocence and guilt regarding certain indicators.[95] First, a 2019 study at James Madison University found that seven of nine COPS scale indicators they tested did indicate innocence or guilt.[96] Then, a 2021 study by an FBI working group was unable to replicate Harpster’s results when looking at one-hundred-seventy-five 911 homicide calls.[97] Two years later, a different FBI working group again failed to replicate Harpster’s results when attempting to apply the COPS scale to missing child calls, warning that 911 call analysis may reinforce biases.[98] Around this same time, two separate studies conducted by a research team at Villanova also failed to replicate Harpster’s results.[99] However, none of these early studies applied the full COPS scale to homicide cases as Harpster did in his research.[100]
Most recently in February 2025, two studies applied the full COPS Scale and found that when applied, the COPS scale had no noticeable ability to differentiate between innocent and guilty callers.[101] The 2025 FBI study found that twenty-three of the indicators bore little to no relationship to either guilt or innocence and that some guilt indicators indicated innocence.[102] Both studies concluded that in light of the high stakes of homicide investigations and the limitations of the COPs scale in differentiating between innocent and guilty callers, 911 call analysis has limited utility in real cases.[103] Following the publication of the 2025 studies the American Psychological Association released a statement echoing their results and stating the COPS scale is “not a valid tool” for guilt detection.[104]
Additionally, the authors of these studies and other experts raise concerns with Harpster’s original methodology.[105] Experts worry that his research may suffer from confirmation bias, as he sourced and coded the initial one hundred calls himself.[106] Experts share this worry about the 2017 book’s data as Harpster never released any information on how calls were sourced or who did the coding.[107] Such a limited sample does not account for race, upbringing, dialect, education, or other demographic factors that may affect how a caller acts or responds to a dispatcher.[108] Even John Jarvis, one of the co-authors of Harpster’s 2009 paper, criticized the use of 911 call analysis in real cases and stated that he told Harpster and Adams that 911 call analysis was not ready for use and more research was necessary.[109]
III. Discussion
Despite its shaky methodological foundations and the failure of other researchers to replicate its results, 911 call analysis has proliferated as a tool for both homicide investigations and prosecutions since its promulgation in 2008.[110] Most troubling is its use in court by prosecutors who disguise Harpster’s methods as lay evidence and work in 911 call analysis as part of their opening and closing statements.[111] This deceptive deployment of an unproven and unreliable science denies defendants their right to properly confront witnesses against them and undermines the integrity and legitimacy of the justice system.
Courts must reject the use of 911 call analysis because, properly characterized as expert testimony, it cannot pass the Daubert or Frye standards. Considering the severe implications presented by unproven and unreliable scientific evidence, judges, defense attorneys, and prosecutors must all take the necessary steps to reign in the use of 911 call analysis. Judges and defense attorneys should educate themselves so they can recognize 911 call analysis in disguise, while police and prosecutors should immediately stop using 911 call analysis in investigations and prosecutions.
A. 911 Call Analysis Should be Considered Expert Testimony
Under FRE 701 and 702, courts must characterize 911 call analysis as expert testimony because it relies on specialized knowledge that can only be learned through Harpster’s teachings.[112] Harpster himself states that not just anyone can perform 911 call analysis—it can only be properly performed by someone who has gained knowledge and experience through reading his work and attending his training.[113]
The back cover of Harpster’s book states that the book will provide those who read it with the skills to conduct 911 call analysis.[114] This fact alone indicates the specialized nature of 911 call analysis. However, even more indicative of this specialized nature is Harpster’s assertion that scientists and researchers have failed to replicate his results because they have only read his book but have not received his training, which Harpster claims is necessary to properly perform 911 call analysis.[115] Thus, because specialized knowledge and training are required to perform 911 call analysis, it is appropriately categorized as expert testimony and should not be admissible through lay opinion testimony.
B. 911 Call Analysis Clearly Fails Both Daubert and Frye
Since 911 call analysis is expert testimony, courts must apply the Daubert or Frye standards to determine whether it is admissible. Given that 911 call analysis is neither reliable nor verifiable, it should not pass either standard.
911 call analysis fails all five of the Daubert factors. First, 911 call analysis cannot be properly tested because Harpster has never shared his data or exact methodology.[116] Second, all peer-reviewed studies that have applied the COPS scale have been unable to replicate Harpster’s results.[117] Third, there is no known error rate because Harpster’s research does not provide a potential error rate.[118] Fourth, the only published standard for 911 call analysis comes from Harpster himself.[119] Finally, Harpster’s methods have not been accepted by either the scientific or the legal community as it has been expressly rejected by both law enforcement and psychological authorities.[120] As such, when Daubert is properly applied, 911 call analysis must be rejected as all five Daubert factors weigh against its admissibility.
911 call analysis is even more easily rejected in jurisdictions that still employ the Frye standard, as Harpster and his coauthor Adams are the only researchers who have accepted its use in real cases.[121] Their co-author John Jarvis, the authors of the studies attempting to replicate their work, and the APA have all cautioned that 911 call analysis should not be used in real cases.[122] As such, 911 call analysis fails to meet the broad acceptance required by Frye.
C. Potential Solutions
911 call analysis is too unreliable to be counted on in court or homicide investigations and its continued use will undermine the integrity of the criminal justice system. Concrete actions must be taken to prevent its use and the miscarriages of justice that such unreliable evidence may lead to. All actors in the criminal system, including defense attorneys, judges, police, and prosecutors, have important roles to play in this process. Defense attorneys and judges must make efforts to learn about 911 call analysis so they can recognize when it is being used inappropriately. However, the most important players are the police and prosecutors. Whether or not 911 call analysis is used and how it is used is ultimately the decision of police and prosecutors, and both groups should make concerted efforts to prevent further proliferation.
Because 911 call analysis is more often used as a part of a clever prosecutor’s broader trial strategy, as opposed to being introduced as expert testimony, defense attorney associations should offer classes on how to recognize when COPS scale indicators are being employed.[123] During discovery defense attorneys should examine the credentials of testifying witnesses to see if they have attended Harpster classes in the past. Similarly, defense attorneys should also attempt to find out if the prosecutor has attended Harpster training in the past. If defense attorneys know what 911 call analysis looks like they will be more readily able to recognize it if and when prosecutors or testifying witnesses attempt to employ it improperly.
Of course, the arbiter of admissibility in any criminal case is the judge.[124] Similar to defense attorneys, judges should educate themselves on 911 call analysis, so if an objection is raised or a testifying witness veers into expert testimony employing 911 call analysis, the presiding judge can recognize the issue and take appropriate remedial action. By rejecting the use of 911 call analysis, judges can send a clear message that they will not tolerate the use of unverified and unreliable expert testimony in their courts no matter what guise it is cloaked in.
Ultimately, police and prosecutors are in control of whether 911 call analysis is used in investigation or prosecution. Considering the failure of researchers to replicate Harpster’s results and the rejection of the COPs scale by scientific and legal authorities, police and prosecutors must recognize the potential that 911 call analysis may prejudice or undermine their efforts to seek justice. Some prosecutors have recognized this fact.[125] Following this, police and prosecutors should immediately halt the use of 911 call analysis in real cases and consider ways to prevent the premature adoption of unverified forensic sciences in the future.
IV. Conclusion
The use of 911 call analysis both in investigation and in prosecution undermines reliability and fairness in criminal trials. This form of forensic analysis is untested, unproven, and relies on confirmation bias. Every study that has independently tested 911 call analysis has failed to replicate Harpster and Adam’s results and has cautioned against using this method in investigations and trial. 911 call analysis is not science and is not reliable. It should not be used in homicide investigations, and it certainly should not be allowed in court. Defense attorneys and judges must educate themselves on the problems with 911 call analysis and how it is used so they can prevent it from improperly influencing juries. The most important duty lies with police and prosecutors to recognize that the use of such unreliable scientific evidence undermines the integrity of their investigations. Responsible police and prosecutors must and will reject the usage of 911 call analysis.
[1] Brett Murphy, They Called 911 for Help. Police and Prosecutors Used a New Junk Science to Decide They Were Liars., ProPublica (Dec. 28, 2022), https://www.propublica.org/article/911-call-analysis-fbi-police-courts [https://perma.cc/T4HM-8EXF].
[2] See generally Tracy Harpster & Susan H. Adams, Analyzing 911 Homicide Calls: Practical Aspects and Applications (CRC Press, 1st ed. 2017).
[3] Id.; See also Murphy, supra note 1.
[4] Murphy, supra note 1; Additionally, in this author’s research he found no public records indicating Tracy Harpster taking the stand as an expert witness in a criminal trial.
[5] Patrick M. Markey et al., Validity Concerns of the Considering Offender Probability Statements Scale in 911 Homicide Call Analysis: An Empirical Study, 31 Psych, Pub. Pol’Y, & L. 1, 23 (2025).
[6] Murphy, supra note 1.
[7] Daubert v. Merrell Dow Pharms. Inc., 509 U.S. 579 (1993); Frye v. United States, 293 F. 1013 (D.C. Cir. 1923).
[8] Fed. R. Evid. 701-702.
[9] Fed. R. Evid. 701(a)-(c).
[10] Fed. R. Evid. 702(a)-(d).
[11] Christine Funk, Daubert vs. Frye: A State-by-State Comparison, Expert inst. (July 10, 2024), https://www.expertinstitute.com/resources/insights/daubert-versus-frye-a-national-look-at-expert-evidentiary-standards/ [https://perma.cc/T25J-YNQ8].
[12] Fed. R. Evid. 701-702.
[13] Funk, supra note 11.
[14] Id.
[15] Daubert v. Merrell Dow Pharms. Inc., 509 U.S. 579 (1993).
[16] Id. at 592-593.
[17] Id.
[18] Id. at 593-595.
[19] Id.
[20] Id. at 595.
[21] See Jim Robinson, Daubert Standard, Legal Info. Inst. https://www.law.cornell.edu/wex/daubert_standard [https://perma.cc/L25W-XJSR]; Funk, supra note 11.
[22] See Funk, supra note 11.
[23] Frye v. United States, 293 F. 1013 (D.C. Cir. 1923).
[24] See Daubert v. Merrell Dow Pharms. Inc., 509 U.S. 579 (1993).
[25] Tracy K. Harpster, The Nature of 911 Homicide Calls: Using 911 Homicide Calls to Identify Indicators of Innocence and Guilt (Oct. 25, 2006) (M.S. thesis, University of Cincinnati) (on file with the University of Cincinnati Library system).
[26] Id. at 32-33.
[27] Id.
[28] Id.
[29] Id.
[30] Id. at 55-58.
[31] Murphy, supra note 1.
[32] Susan H. Adams & Tracy Harpster, 911 Homicide Calls and Statement Analysis: Is the Caller the Killer, 77 FBI L. Enf’t Bull. 22 (June 2008).
[33] See Id.
[34] Harpster et al., Analyzing 911 Homicide Call for Indicators of Guilt or Innocence: An Exploratory Analysis, 13(1) homicide stud. 69 (2009).
[35] Id. at 69.
[36] Harpster & Adams, supra note 2.
[37] Id.
[38] Id. at 6 (The following list is formatted as innocent v. guilty; Full list of COPS scale indicators 1) Who the call is about – Immediate Plea for Help for Victim v. No Immediate Plea, Immediate Assessment of Victim v. No Immediate Assessment, Focus on Victim v. Focus on Caller, No Acceptance of Victim’s Death v. Acceptance of Victim’s Death, Aid Provided for Victim v. No Aid; 2) What the call is about – Relevant Information v. Extraneous Information, Sensory Details v. Lack of Sensory Details, Prioritized Order v. Inappropriate Order, Bleeding Comments v. Blood/Brain Comments; 3) How is the call made – Urgency v. No Urgency, Fear for Caller’s Safety v. No Fear, Proximity to Victim v. No Proximity, Initial Sounds or Comments v. Initial Delays; 4) Aggressive Demands v. Passive Defenses (Defendant Mentality, Ingratiating Remarks, Insulting or Blaming Victim, Mental Miscues, Minimization); 5) Cooperation with Dispatcher v. Resistance to Dispatcher (Dispatcher Confusion, Diversion, Equivocation, Evasion, Hangs Up, Only Answers What’s Asked, Pauses, Repetition, Self-Interruption, Short Answers, Unintelligible Comments); 6) Additional Guilty Indicators – Attempts to Convince, Awkward Phrases, Conflicting Facts, “Huh?” Factor, I Don’t Know, Isolated “Please”, Lack of Contractions, No Modulation, and Unexplained Knowledge).
[39] Id.
[40] Id.; See also Daniel E. O’Donnel et al., Evaluation of the 911 Considering Offender Probability in Statements Scale as a Deception Detection Method for 911 Calls, 31 Psych. Pub. Pol’y & L. 1, 16 (Feb. 27, 2025) (Advance online publication); Markey, supra note 5.
[41] Harpster & Adams, supra note 2 at 6.
[42] Id.
[43] Id.
[44] Id.
[45] Murphy, supra note 1.
[46] Id.
[47] See Murphy, supra note 1; Jonathan Jarry, Is the Caller the Killer? 911 Call analysis Can’t Give You the Right Answer, Mcgill Univ. off. for Sci. & Soc’y (Mar. 24, 2023) https://www.mcgill.ca/oss/article/critical-thinking/caller-killer-911-call-analysis-cant-give-you-right-answer [https://perma.cc/SUT6-GH67].
[48] Murphy, supra note 1.
[49] Id.
[50] Markey, supra note 5 at 2.
[51] Murphy, supra note 1.
[52] Id.
[53] Id.
[54] Harpster & Adams, supra note 2 at 229, 239, 245, 249.
[55] Id.
[56] Id. at 14.
[57] Id. at 5.
[58] Id.
[59] Id.
[60] Id.
[61] See generally Harpster & Adams, supra note 2 (Each indicator is covered by a chapter in the book and Harpster discusses the prevalence of indicators and how strongly they correlate to guilt or innocence in each in depth).
[62] Murphy, supra note 1.
[63] Id.
[64] Id.
[65] Id.
[66] Id.
[67] Id.
[68] Id.
[69]Id.; See also Harpster & Adams, supra note 2 at 95-98.
[70] Murphy, supra note 1.
[71] Id.
[72] Id.
[73] See Id.
[74] Id.
[75] Id.
[76] Id.
[77] Id.
[78] Id.
[79] See Id.; Harpster & Adams, supra note 2 at 249-254.
[80] Harpster & Adams, supra note 2 at 249.
[81] Id.
[82] Id.
[83] Id. at 249-250.
[84] Id. at 250-251.
[85] Id. at 249, 251.
[86] Id. at 250-252.
[87] Id. at 252.
[88] Murphy, supra note 1.
[89] Id.
[90] Id.
[91] Id.
[92] Id.
[93] Id.
[94] See Id.; Guilty caller? Real-world implications of a flawed 911 tool, Am. Psych. Ass’n (Feb. 28, 2025) https://www.apa.org/pubs/highlights/spotlight/considering-offender-probability-statements-scale [https://perma.cc/7NPX-PVFC].
[95] See Michelle L. Miller et al., “911 What’s Your Emergency?”: Deception in 911 Homicide and Suicide Staged as Homicide Calls, 25 Homicide Stud. 3 (2020); Daniel E. O’Donnel et al., “My child is missing”: 911 calls in mysterious disappearances of children, 67 Aggression & Violent Behav. 101795 (2022); O’Donnel, supra note 40.
[96] See Jon D. Cromer et al., 911 Calls in Homicide Case: What Does the Verbal Behavior of the Caller Reveal?, 34 J. Police Crim. Psych. 156 (2019).
[97] Miller, supra note 95 at 7, 13-15.
[98] O’Donnel, supra note 95 at 7, 9.
[99] See Patrick Markey et al., The (Mis)utilization of Cues During Deception Detection in 911 Homicide Calls, 29 Homicide Stud. 2 (2022); Patrick M. Markey et al., Deception Cues During High-Risk Situation: 911 Homicide Calls, 33(7) Psych. Sci. 1040 (2022).
[100] See generally Miller, supra note 95; O’Donnel, supra note 95; Cromer, supra note 96; Homicide Stud., supra note 99; Psych. Sci., supra note 99; See also Markey, supra note 5; O’Donnel, supra note 40 (Both 2025 studies explicitly discuss that no previous research has applied the entire COPS scale to homicide 911 calls).
[101] O’Donnel, supra note 40 at 6; Markey, supra note 5 at 5-6.
[102] O’Donnel, supra note 40 at 10-11.
[103] Id. at 13; Markey, supra note 5 at 7-8.
[104] American Psych. Assoc., supra note 94.
[105] See O’Donnel, supra note 40, Markey, supra note 5.
[106] O’Donnel, supra note 40 at 2; See also Jarry, supra note 47;Murphy, supra note 1.
[107] Markey, supra note 5 at 2; See also Jarry, supra note 47; Murphy, supra note 1.
[108] Murphy, supra note 1.
[109] Id.
[110] Id.
[111] Id.
[112] See Id.; Markey, supra note 5 at 8.
[113] Markey, supra note 5 at 8.
[114] Harpster & Adams, supra note 2.
[115] Markey, supra note 5 at 8; See also Murphy, supra note 1.
[116] In this author’s extensive research, he found no place where Harpster’s dataset was publicly available; See also Murphy supra note 1; O’Donnel supra note 40; Markey supra note 5 (All three authors also make this assertion in their own work).
[117] See generally Cromer, supra note 96; Miller, supra note 95; Homicide Stud., supra note 99; Psych. Sci., supra note 99; O’Donnel, supra note 40; Markey supra note 5.
[118] See Harpster and Adams, supra note 2.
[119] See Id.
[120] See American Psych. Assoc., supra note 94; Miriam Krinsky and Isabelle Cohn, 911 Call Scrutiny Should Not Be Used To Identify Suspects, Law360 (Feb. 2, 2024) https://www.law360.com/articles/1793107/911-call-scrutiny-should-not-be-used-to-identify-suspects [https://perma.cc/ZJQ7-USTY]; Brett Murphy, Prosecutors and Judges Push for Conviction Reviews, Ban on Junk Science of 911 Call Analysis, ProPublica (Feb. 24, 2023) https://www.propublica.org/article/911-call-analysis-ban-review-convictions [https://perma.cc/AT2Z-3WZX]; See also Murphy, supra note 1 (Murphy discusses several rejections of 911 call analysis by prosecutors, defense attorneys, judges, and law professors).
[121] See generally Cromer, supra note 96; Miller, supra note 95; Homicide Stud., supra note 99; Psych. Sci., supra note 99; O’Donnel, supra note 40; Markey supra note 5.
[122] Murphy, supra note 1.
[123] Id.
[124] Daubert v. Merrell Dow Pharms. Inc., 509 U.S. 579, 592-593 (1993).
[125] See Krinsky, supra note 120; Murphy, supra note 120.
Cover Photo by Michael Förtsch on Unsplash.
