THE EXPERT WITNESS: Michigan Forensic Interviewing Protocol, 4th edition review


By Michael G. Brock

There is an update of the Michigan Forensic Interviewing Protocol,1 (Protocol) which I recently ran across and printed off while doing preparation for an upcoming case. This article is a review of the changes in the new Protocol, and some thoughts about what this will mean to attorneys, experts, and clients involved in criminal sexual conduct cases. The preface states that the current (4th) edition began revision in 2016 and was posted in October 2017. Presumably, work on the current Protocol commenced subsequent to publication of Debra Poole’s new book on forensic interviewing2, which I reviewed on my blog and in the Legal News (Is Forensic Science an Oxymoron?)  You may also read this and related articles on my website,

The importance of the Protocol cannot be overstated, since it is the legally mandated format for the child’s interview in an abuse case.  The documented transcript and/or recording of the child’s forensic interview should tell attorneys and expert witnesses whether the interview was conducted properly, and whether the child’s statements provide a consistent and credible narrative.  If the child was reporting an actual event or events, was coached, or is making up a story, the forensic interview should provide evidence of it and suggest possible effective direct and cross examination.  Moreover, your expert will be expected to have knowledge of the Protocol and its use, and, though I am not a lawyer, I have no idea how counsel can prepare a defense without knowing how the evidence was obtained.

Forensic interviews are similar to what one might experience in other kinds of criminal prosecutions. In a murder trial, for example, the defense would be very interested in how a confession was obtained. If it was coerced from his client under threats, bullying, physical abuse or hours-long interrogation during which the defendant was denied access to counsel, the defense attorney could be expected to argue that the confession was obtained illegally and should be kept out of evidence, and the prosecution would argue that it should be admissible. In a child sex abuse case, there is typically no physical evidence and the decision to indict is based on the child’s forensic interview. If the child makes an allegation, regardless whether the interview was done properly, or the story has any continuity or credible detail, the prosecution is going to indict, but unlike the murder trial, they will not want the jury to hear contradictions in the child’s story, and will try to keep details of the interview out.

If defense counsel cannot keep the child’s evidence out completely due to the child’s statements being irreversibly tainted though bad interviewing techniques,3 they will need to be able to impeach the child’s testimony by showing inconsistencies with previous statements, especially the first recorded interview, which the forensically literature reports to be the most reliable evidence obtained from the child.4 They may also explore this discrepancy at trial by allowing the child to expand on questionable allegations with details that are obviously untrue. The defense will also want to make sure the jury understands what constitutes a credible narrative, and what may be evidence of coaching or a motive to lie.

Multiple interviews, when employed, are more likely to produce disclosure of abuse, but they are also more likely to point out discrepancies between what was said in one interview and what was said in another, the preliminary hearing, the trial itself, and what was disclosed to the adult who presented with the child and made the initial report of abuse. As Dr. Poole states, “A second or third interview can benefit the prosecution when children do not contradict central details and new, valuable information emerges, but subsequent interviews are invaluable to the defense when these conversations document significantly changed stories, a pattern of adult influence (e.g., the infiltration of changing adult beliefs into children’s reports), and expanding stories with obvious confabulations.”5

Significantly, alterations to the Protocol6 in the three editions issued since the original have not made any changes to the basic format of the interview, or “The Phased Interview,” as it is known (These phases consist of the following steps: • Prepare for the Interview. • Introduce Yourself and Start Building Rapport. • Establish the Ground Rules. • Conduct a Practice Narrative. • Introduce the Topic. • Elicit a Free Narrative. • Question, Clarify, and Test Hypotheses7. • Close the Interview (P. 7)), but tend to expand on it.  One notable deletion from the original preface is the following statement. “The purpose of this Protocol and training is to prepare local investigators to conduct competent child interviews which will reduce trauma to children, make the information gained more credible in the court process, and protect the rights of the accused.”8 Considering that all other changes through four editions (1998, 2005, 2012, and 2017) have expanded on the original concepts, it is curious that the reference to making sure the process is fair to the accused has been deleted.

The new edition reminds us that this Protocol, or a reasonable facsimile, must be followed when interviewing children suspected of being abused as a matter of law: “In 1998, the Child Protection Law was amended to require each county to implement a standard child abuse and neglect investigation and interview Protocol using as a model the Protocols developed by the Task Force as published in DHS Publication 794, A Model Child Abuse Protocol—Coordinated Investigative Team Approach and DHS Publication 779, Forensic Interviewing Protocol, or an updated version of those publications.”9

In each edition subsequent to the first: “...the Committee edited sections for clarity, improved the examples, added Quick Guides, and provided some additional reference materials, including relevant statutes. Recent research continues to support the methodology used in Michigan’s Protocol.”10  Training of personnel employed by the State is required before they are considered qualified to conduct forensic interviews. “Training is to be provided only by the current holder of the MDHHS service contract that provides forensic interviewing training.”11 It is noteworthy that this training is not available to those who are not prosecution team members, and one has to wonder, why not. Law enforcement conducted in secrecy may not be in the public’s best interest—indeed, the term “secret police” has an ominous tone. Since prosecutors represent “the people” of the State of Michigan, and since transparency is essential for preserving democracy, why are their website and trainings closed to non-prosecutors? However, defense attorneys need to be prepared for their experts to be attacked by prosecutors for not having had this training. It would seem that a major strategy of contemporary prosecution is to be sure that the defense cannot put on a case.

Training for forensic interviewers is, of course, a good idea for obvious reasons: many of the people performing forensic interviews for the State are not licensed mental health professionals (MHPs), and have no experience in the practice of mental health. Consequently, a police officer or bachelors level MHP would not be qualified to practice mental health treatment independently or as part of a freestanding outpatient mental health clinic, licensed by the State, authorized by insurance companies to treat their clients, and accreditation by an appropriate agency.12 Therefore, if they are going to be practicing mental health in a setting where they will participate in the process of deciding who remains free and who spends the rest of their life in prison, a weekend or even a weeklong training session is probably a good idea.

However, it should be noted that this Protocol is written as a requirement for State employees and not as a mandate for professional MHPs, who have their own forensic guidelines and standards of practice13, and who may choose between using Dr. Poole’s Protocol, one developed by Katherine Kuehnle, or that authored by Michael Lamb (known as the NICHD Protocol), to name some that are commonly used, or they may choose to develop their own. When I did forensic interviewing as part of child custody evaluations for many years, I used Dr. Poole’s Protocol, since I did not see any reason to reinvent the wheel, nor did I think I could improve on what she had done. All viable protocols contain the same core elements in any case, as she noted in her most recent (2016) book: “The Protocols described later in this chapter [7] shared core recommendations ...such as the benefits of eliciting freely recalled narratives, following witnesses’ trains of thought, and avoiding suggestive questions.”14

On page one, the Protocol introduction states that: “The goal of a forensic interview is to obtain a statement from a child—in a developmentally-sensitive, unbiased, and truth-seeking manner—that will support accurate and fair decision-making in the criminal justice and child welfare systems.” This Protocol was developed subsequent to the daycare hysteria of the 1980s, when hundreds of people were falsely changed—and many convicted—of ridiculous and far-fetched allegations of child sexual abuse, for which there was literally no evidence except statements obtained from children using heavy-handed and coercive interviewing techniques.

The underlying assumptions to the development of these protocols (Michigan’s and others like it) is that a child’s statements can constitute valid evidence in a court of law if they are obtained in a scientific and systematic manner, but that if they are not obtained in this manner, they are not reliable.

Hence, after a warning about mixing the dual and conflicting roles of treatment provider and forensic interviewer, previous versions of the Protocol continued by emphasizing the scientific nature of inquiry: “There are two overriding features of a forensic interview: Hypothesis testing [and] a child-centered approach. First, forensic interviews are hypothesis-testing rather than hypothesis-confirming (Ceci & Bruck, 1995).” What does this mean exactly? It means that if you start out prepared to find something (this child was molested by the accused), then you will find that. If you take a position of skeptical neutrality, including that the child could have been coached of may be lying for any number of reasons, then you are more likely to find the truth.

This emphasis on the science is extremely important, but experience has been the typical forensic interview is not well done. One main reason for this is apparent to anyone looking at one of these interviews with a degree of objectivity. The interviewers typically ignore exculpatory statements, absurd scenarios, and statements that provide opportunities for inquiry into alternative hypotheses. They are also anything but child-centered, and resemble interrogations more than interviews. My theory is that this is because interviewers see themselves as part of law enforcement team, allied with police and prosecutors. Moreover, this mindset is reinforced by the fact that the person evaluating their work is not an objective party, but the very prosecutor who will be arguing for conviction if the case goes to trial. The defense is completely unrepresented in the decision of whether to indict a suspect.

The law enforcement mindset is to get convictions, not to exonerate the innocent. This sounds callous and cold, as indeed it is, and it is beyond the capacity of the average juror to understand. However, though there are undoubtedly prosecutors who are not convinced that everyone accused is guilty, I do not run across those cases; they get tossed before I would become involved. But the fact that there is so much very bad evidence based on hypothesis-confirming techniques tells me that the mindset of the prosecution team is generally to prosecute any time a child makes an allegation during an interview, regardless of how coercive the interviewing technique, or how ludicrous and improbable the story.

As an example of this mindset, let’s take a look at a matter in the headlines. Special Prosecutor Robert Mueller was appointed to investigation collusion by the Whitehouse Transition Team of President Trump with Russia. Collusion is not a crime, but evidence of collusion between the Democratic Party, British spy Christopher Steele, the FBI and Russia has been made public.15 If the real goal was to investigate Russian collusion, or crimes connected with it, then there would seem to be sufficient smoke to look for fire where there is evidence of collusion.

But the obvious goal of this investigation is to bring down the Trump Administration—a goal that seems to be shared by both Democrats and a significant number of Republicans. Moreover, former FBI Director Comey stated that the Steele dossier was “salacious and unverified,” but AG Rod Rosenstein would not say whether it was used to get a FISA warrant to spy on Trump’s team and acquire information that would lead to a guilty plea by Michael Flynn—which means that it probably was. To use bogus evidence to obtain an illegal warrant ought to be a crime, but not if you’re the designated good guys. The people involved in the collusion were “demoted,” but Michael Flynn pleaded guilty to a felony for lying to the FBI about information they obtained illegally. The secret police don’t have to say where they got their evidence. What’s the message? Only the targets of the investigation have to play by the rules, not the investigators.

How does this apply to what we’re talking about? Only this: On 11/30/2017, a Michigan Court of Appeals upheld the conviction of Shae Lynn Mullins16 for convincing her daughter to tell a school teacher that her father had molested her. Since this was the fourth such false allegation of sexual abuse by Ms. Mullins the prosecution finally got around to prosecuting the designated victim17—though to do so the prosecutor had to appeal a lower court decision not to allow a criminal prosecution—and won a conviction. Mullins appealed the conviction on the basis that it wasn’t she who made the false allegation, it was her daughter. The appeals court didn’t buy it, and she is stuck with a felony.

Sounds like justice, right? Well, it might, unless you consider the fact that the accused would have spent the rest of his life behind bars and on the sex offenders list if he were convicted on any one of those four false allegations. What did she receive for a sentence? One week behind bars. Yes, one week in jail! As long as this kind of discrepancy exists between sentences for designated perpetrators18 (about whom we are allowed to assume the worst) and designated victims (whom we are supposed to give the benefit of the doubt), then crime will definitely pay for the designated victims. What genuinely surprises me is that anyone ever gets caught making a false allegation. If you have the majority of time with the child, you should be able to brainwash her sufficiently to make a false allegation stick.

Usually, no matter how stupid the allegations or how obvious the motive, the person will be convicted and sent away, but the fact that even if you get caught the consequences are minimal makes it worth the risk if you’re a designated victim. I was an expert on a case where abuse allegations emerged after the accused testified against his ex-girlfriend in an attempted murder case. She got two years for trying to shoot someone and is back on the street; he’s gone for a long time.
The center piece of the forensic interview remains the free narrative, and follow up questions should ask for more narrative regarding topics that were mentioned by the child in the free narrative.  This is considered the most reliable evidence.  However, as the Protocol notes on P. 17-18, “The most common interviewer errors are to omit the free narrative or to shift prematurely to focused questions. Instead of asking the child to talk about the event and then shifting to the Question, Clarify, and Test Hypotheses Phase, the interviewer should prolong the Elicit a Free Narrative Phase with numerous open-ended prompts, such as ‘And then what happened?’ and ‘Tell me more about [child’s words for an event].’ To elicit a free narrative, the interviewer simply tacks on an open-ended broad prompt (also called an invitation) after the topic is raised: • ‘What happened?’ • ‘Tell me everything you can about [refer back to child’s statement].’ • ‘Tell me all about [refer back to child’s statement].’” But these techniques are almost never used!

Why is it still the case, almost two decades after the first Michigan Protocol was developed, that the narrative is being omitted? And, more importantly, what does it say about the quality of evidence being produced? The answers are: 1) Because there is no objectivity on the part of those evaluating the validity of the interview; and, 2) improperly done forensic interviews are producing poor quality evidence and innocent people are going to prison. Prosecutors control the process and they see their job as getting convictions; if they destroy innocent lives and ultimately the country in the process, well, so be it.

The evidence of this is present in almost every forensic interview. Despite the fact that interviewers are warned, “...the interviewer should not interrupt until it is clear that the child has finished giving a free narrative,”19 and, “Be tolerant of pauses in the conversation. It is appropriate to look away and give the child time to continue talking.”20 Interviewers continue to interrupt because they want specifics that will provide them with the elements of a crime, and multiple counts. Their typical response, once the child has made a “disclosure” is, “Did that happen one day or more than one day?” “What is your mom’s boyfriend’s name?” “What were your wearing?” “Were you lying down or standing up?”

Typically, the rest of the interview is made up of questions that require short answers and make it seem like the child is telling the story, when, in fact, it is the forensic interviewer who is feeding suggestions in the form of questions. It is, therefore, the interviewer who is actually telling the story. If the child says something ludicrous, like, “[The suspect] made me perform oral sex on my knees when I was four, but my dad was there in the room to protect me,” the interviewer will typically ignore the obvious exculpatory evidence/alternative hypothesis suggested by the phrase “my dad was in the room to protect me,” and ask a stupid question like, “He made you perform oral sex. Where were his hands?” What does not happen is what the Protocol refers to as Resolving Inconsistent Information: “You said [child’s first words on the issue], but then you said [child’s second words on the issue]. I’m confused about that. Tell me again how that happened...”21

Of course, prosecutors will do everything they can to keep the jury from hearing this evidence, but defense counsel has to get it in or the jury will assume the prosecutor is a reasonable person and had a good reason for charging the defendant. So what if the jury is going to hear the kid say again that it happened? If they don’t know the context they are going to assume that it happened if they hear the kid say it only once. You can’t pretend the burden of proof is on the prosecution. The only way a jury is not going to believe it happened is if they are convinced the story sounds so coerced, coached or made up it can’t possibly be true. If the defense can impeach the child without introducing the forensic interview or putting an expert on the stand, that may be desirable in some situations, but it is certainly not possible in all.

At the very least, major contradictions—like the one mentioned above—in a forensic interview offer an opportunity to impeach a child’s testimony without attacking the child. How does she explain that her protector was in the room, yet did not intervene when she was being abused? She can’t. What should be obvious—if she is allowed at trial to tell the narrative that the forensic interviewer did not ask for, or, asked for but did not allow— is that she was coached to make the allegations against one person, and to also say who the good guys are, and she mixed them up in one improbable scenario. Defense counsel needs to have the jury hear this narrative.

Presumed innocent until proven guilty is a fiction that is only used these days to comfort the general population and keep them from thinking about the fact that more and more of them are going to prison all the time and that a substantial portion of them are innocent. They will only find out the fallacy of the presumption of innocence after it is too late to do them any good. If there is one person that can’t afford to buy this propaganda, it is defense counsel. After all, it is you against the police, the prosecutor, the forensic interviewer, the jury, most judges, the general public, and sometimes your client. Let them sell stupid to everyone else, you can’t afford to buy it. Your job is to get the jury to recognize stupid.

But the new Protocol, makes the job harder in significant ways with the addition of this statement in the first paragraph on page 1: “Forensic interviews are part of investigations that sometimes involve retrieval of physical evidence, conversations with collateral contacts, and other fact-finding efforts. Therefore, interviewers should explore topics that might lead to other evidence keeping in mind that a forensic interview is only part of an investigation.”

What is particularly important about this statement is that it communicates to the interviewer that her job is more than just obtaining a scientifically objective statement from a child. She is, in fact, not just in the way she views herself and is viewed by the prosecutor, part of the crime-fighting team, and needs to keep that in the forefront of her mind when she is doing the interview. She needs to find out if there is physical evidence because police and hard science forensic experts might not be able to do their job so she has to back them up. She needs to find out if there are other people who need to be interviewed, or other facts that police and prosecutors might have overlooked.

The problem with these additional assignments is that it is very difficult to do the job of forensic interviewing properly as it is. What makes it most difficult is that it is a very unnatural way to converse with children, as we have explored in previous articles22. In brief, forensic interviews are an effort to obtain information that the child knows, but the interviewer doesn’t. Ordinarily, when children are being questioned by adults, the adults know the answer and they are trying to find out if the child knows what they know. Consequently, it is natural to lead the child, and the child expects to be led. To hold back and not jump in at every opportunity is difficult, even for someone who has no agenda, and who has been counseling children and adults for decades.

But forensic interviewers have an agenda; they are expected to obtain evidence, as discussed earlier, the new Protocol makes it clear that the evidence they are expected to find is inculpatory, not exculpatory. Additionally, they are relatively inexperienced bachelors level MHPs and police officers, not experienced masters and doctorate level people, so the job of being the first step in the process of deciding whether a child has been the victim of a crime has just been made more difficult. Perhaps that is inevitable, for while there were seven members of prosecutorial teams consulted in the revision of this document by Dr. Poole, there was only one defense attorney consulted, and no defense experts.
2Interviewing Children: The Science of Conversation in Forensic Contexts, Debra Ann Poole (2016)
3 Though taint hearings have been used in high profile cases (i.e., NJ v. Margaret Kelly Michaels, Suggestibility, Reliability, and the Legal Process, Robert Rosenthal, J. D., Developmental Review 22, (2002) 334-369 Sep 2002), experience has been that this legal maneuver is rarely used and even more rarely successful in lower courts. At the time of the Michaels trial, forensic interviewing had not yet been developed as a mechanism to avoid taint and it was easier to argue that science had not caught up with the law. Now, though the same techniques and biases the convicted Michaels are still evident, the courts will presume that past errors have been corrected, and will err on the side of “protecting the child,” which also means protecting those would maliciously (or erroneously) use a child to make a false allegation.
4 “Whether repeated interviews are productive or problematic depends on the features of the case, including the length of time between interviews, the quality of the interviews, and the motivations of people in children’s lives to influence their testimony through threats, intentional coaching, or inadvertent influences (e.g., overheard conversations). When external influence is not an issue and interviews are nonsuggestive, children’s reports across repeated interviews generally contain few explicit contradictions.” Poole, Debra A., Interviewing Children: The Science of Conservation in Forensic Contexts (Kindle Locations 3195-3198). Kindle Edition.
5 Poole, Debra A., Interviewing Children: The Science of Conservation in Forensic Contexts (Kindle Locations 3205-3209). Kindle Edition.
6 “Debra Poole, Ph.D., of Central Michigan University was contracted by DHS to develop a forensic interviewing Protocol...The first edition of the Forensic Interviewing Protocol was published in 1998.” 4th Edition Protocol Preface, P. iv
7 Hypothesis testing was described and recommended in previous versions of the Protocol, but was not explicitly contained in the “Question and Clarify” section of the Phased Interview.
8 Protocol Introduction (Preface) 1st edition, last sentence.
9 4th Edition Protocol Preface, P. iv
10 4th Edition Protocol Preface, P. iv
11 4th Edition Protocol Preface, P. iv
12 “To qualify for this position, you will need the following: Bachelor’s Degree in a Human Services Related Field is work, psychology, counseling or criminal justice...” The Guidance Center...Forensic Interviewer in the Kids-TALK Program
13 See for example, American Psychological Association Specialty Guidelines for Forensic Practice, , or APA Guidelines for Child Custody Evaluations in Family Law Proceedings
14 Poole, Debra A. Interviewing Children: The Science of Conservation in Forensic Contexts (Kindle Locations 3357-3359). Kindle Edition. Chapter 7
15 , Molly Hemmingway
16 Court of Appeals of Michigan. PEOPLE OF THE STATE OF MICHIGAN, Plaintiff-Appellee, v SHAE LYNN MULLINS, Defendant-Appellant. No. 334098, Decided: November 30, 2017 Before: SWARTZLE, P.J., and SAWYER and MARKEY, JJ.
17 Persons currently considered a protected species, deemed to be deserving of protection under the law
18 Persons not currently considered a protected species, therefore, not deserving protection under the law
19 4th Edition Protocol, P. 8
20 4th Edition Protocol, P. 6
21 4th Edition Protocol, P. 44
Michael G. Brock, MA, LMSW, is a forensic mental health professional in private practice at Counseling and Evaluation Services in Wyandotte, Michigan. He has worked in the mental health field since 1974, and has been in full-time private practice since 1985. Much of his practice in recent years relates to driver license restoration and substance abuse evaluation, but he also consults and serves as an expert witness regarding forensic interviewing and the use of forensic interviewing protocols in cases of child sexual abuse allegations. He may be contacted at Michael G. Brock, Counseling and Evaluation Services, 2514 Biddle, Wyandotte, 48192; 313-802-0863, fax/phone 734-692-1082; e-mail:, website,