Chat with us, powered by LiveChat

How To Manage Investigations Post-Regs, Hit the G.A.S.

By: Brett A. Sokolow, Esq.

Some courts and OCR want live hearings, and ATIXA’s position is well-documented: we don’t think overall that live hearings will improve resolution processes or create more accurate outcomes. We also believe that the way OCR is structuring hearing requirements will chill the willingness of students and employees to come forward to report sex offenses. Still, it is ATIXA’s job to teach compliant practices and teach them we will.

OCR has provided detailed information in the Title IX regulations about investigation processes and procedures, but they’ve not really touched on techniques. Fortunately, high-quality professional investigations are our wheelhouse. This guide offers a discussion of how we think substantive investigations should look now that the regulations have fundamentally altered the resolution process. The regulations seek to create three important substantive changes over existing investigation practices:

  • A separation and delineation between the investigation function and decision-making;
  • A full and fair opportunity for all parties to see and know all relevant evidence, and address and comment on it prior to the finalization of the investigation report; and
  • Categorization of the evidence three different “buckets”, both by the investigator, and then again by the hearing chair or decision-maker (see Buckets Graphic below).

If the entire resolution process exists on a sliding scale between the investigation and hearing functions, in terms of the emphasis recipients place on each aspect, it may be helpful to review how the scale has developed to this point. Before OCR’s 2011 Dear Colleague Letter (DCL), most recipients’ resolution processes were about 20% emphasis on the investigation function and 80% emphasis on the hearing/final determination.¹ After the DCL, that balance reversed, with 80% of the resolution relying on the investigation and 20% relying on the hearing. Some recipients went to a single-investigator model (100%/0%) and others used a blended or hybrid approach (80- 100%/0-20%). Regardless, the resolution process was primarily based on a robust investigation. Now that the 2020 regulations are upon us, the balance will change again. Where will it end up?

The tendency for public universities subject to the due process scrutiny of the courts will likely be to limit investigators to a “fact-gatherer-only” role. This, however, is even more restrictive than courts of law, where police officers (who investigate crimes) are often called upon to give opinion testimony in court.² Merely doing so does not offend due process, because their opinion is not binding on judge or jury.

We hope that recipients will task their investigators to do more than fact-gather, and the regulations require more of investigators as it pertains to assessing evidence. Investigators are extremely valuable to the resolution process and recipients should not diminish that value at the expense of an effective final resolution. Yet, we still need to respect the fact that OCR is vesting decision-making within a hearing decision-maker or panel.

The challenge, then, is to optimize the role of the investigator without usurping the role of the hearing decision-maker. That said, the regulations and subsequent guidance are a bit squishy on some key points. While OCR set out to kill the single investigator model (and good riddance), they’ve indicated in the preamble to the regulations that an investigation report can reach a conclusion or recommendation. They’ve further clarified in OPEN Center responses to ATIXA that an investigator can testify to their opinions at a hearing, and that hearing decision-makers and investigators can have discussions about the complaint and their opinions about it off-line, outside the hearing.

OCR just insists that no matter how much of this sharing of investigator conclusions or recommendations takes place, they are not binding on the decision-maker, who must render an independent and objective determination. How do we empower the investigation without usurping the decision-making role? Easier said than done, and we try to lay out below our recommendations for how best to respect this boundary, while maximizing the robustness of the investigation to the extent possible and preserving the role of the hearing officer as decision-maker, as OCR has envisioned.

The quality of the ultimate decision depends, in large part, on the quality of the investigation. A high-quality, robust investigation sets the hearing up for a well-reasoned, complete decision. A low-quality, tepid investigation sets the hearing up for potential failure. Ultimately, we expect the sliding scale of decision-making to land around 40% investigation/60% hearing, and maybe as high as 49%/51% once the role of the investigator is factored into the hearing. The hearing will always bear the lion’s share of the decision-making burden in OCR’s construct, but a robust investigation can comprise 40-49% of the heavy lift. So, if the investigation is more than fact-gathering (20%), what else is it? It already includes some critical procedural elements, such as:

  • Communication protocols, including delivery of the Notice of Investigation/Allegations;
  • Managing advisor involvement;
  • Strategizing acquisition of information, testimony, and evidence;
  • Report writing;
  • Investigation report sharing and incorporation of feedback;

Those aren’t major procedural shifts for investigations from the status quo. The key to this discussion is how much more substantive work investigators can do as they gather evidence and compile the report. ATIXA does not recommend that investigators reach conclusions or make recommendations in the investigation report. We don’t suggest that recipients permit investigators to give opinions on whether policy was violated, when they testify at the hearing. And, we don’t think decision-makers or parties should be allowed to ask about those opinions – they are not relevant. We don’t think offline conversations between the investigator and decision-maker about conclusions or recommendations are wise.

We know each recipient needs to decide these questions for itself, and yes, it looks like OCR would permit these practices, but it almost feels like in doing so OCR is setting recipients up for due process and conflict of interest litigation. Once you blur the line between the investigation and decision-making functions, how do you prove independence? How do you demonstrate in court that a hearing decision was independent if it happens to align with the investigator’s recommendations exactly, or uses the rationale from the investigation report in the letter of hearing outcome? How do you prove a decision-maker wasn’t unduly influenced by an investigator outside the hearing, where the parties did not have a chance to cross-examine with respect to that off-line conversation? Proving a negative is difficult in court, and we think the actions OCR would permit here would also likely be enough for a plaintiff suing a recipient to survive a motion to dismiss in federal court. Instead, given the model that OCR is requiring, we recommend a clear line be drawn between the investigation function and the hearing function. So, where should recipients draw that line?

That’s where the G.A.S. model comes in:

= Gather evidence
= Assess credibility and relevance
= Synthetize areas of dispute/agreement and all questions asked


Collecting the evidence from all sources, organizing it, and summarizing it in the written report is the fact-gathering function. It’s a function all investigators have performed since at least 2011, and it’s not really new or different as a result of OCR’s 2020 regulations. That said, we think investigators need to take care not to gather evidence about past sexual conduct or predisposition of the Complainant, as that information will not be admissible in the hearing unless it is offered to prove that someone other than the respondent committed the conduct alleged by the Complainant, or if the questions and evidence concern specific incidents of the complainant’s prior sexual behavior with respect to the Respondent and are offered to prove consent.


Assessing credibility is also already part of a robust investigation. We think it will continue to be an essential role that investigators can perform well without treading too far into the decision-maker’s territory. Assessment of credibility does not include making conclusions about whether a witness or evidence is credible, or adjudging the comparative credibility of evidence or witnesses. It stops just short of that, pointing to issues of credibility without deciding them.

A solid assessment of credibility sets the decision-maker up for success when making the ultimate judgment call on who is more credible and why, but it’s a difficult call for the deci- sion-maker to make in a vacuum. Credibility assessments begin in the investigation and inform the hearing. Decision-makers will want to compare and contrast live testimony against what the parties and witnesses have told the investigators during the interviews. Investigators assessing credibility should stop short of reaching conclusions about credibility, but will aid decision-makers by highlighting the key issues.

So, as an example, it would cross the line for an investigator to include in their report, “Mark, the Respondent, was less credible than Mariana, the Complainant,” or “The decision-maker should find Mark to be unbelievable in his testimony about having received consent for the following reasons…” That’s not assessment, it’s analysis. It goes too far.

But the investigator can and should include in the report, “Mark’s testimony about X contrasts with Mariana’s testimony about X, and the accounts of Witness 1 and Witness 7 aligned with Mariana’s testimony, not Mark’s, during the investigation.” That’s not a conclusion, but it is an assessment of the evidence.

Similarly, an investigation report could include, “The decision-maker may benefit from looking carefully at Mark’s assertions about having received consent and explore this more deeply with the witnesses during the hearing.” This allows the investigator to point to potential dis- crepancies between testimonial accounts or between testimony and other available evidence, but not come to a conclusion. The investigator highlights disputed accounts or conflicting evidence, but the decision-maker is at liberty to make their own final conclusions.


The investigator assesses evidence for its relevance, as well as credibility. Only relevant evidence should be summarized in the investigation report, though all evidence that is relevant and directly related to the complaint (whether inculpatory or exculpatory, from any source) must be shared with the parties and their advisors in an interactive process between the investigator and parties to ultimately finalize the report.

The process is one where the parties and investigators interact to synthesize all the evidence down to that which will be relied upon in the report, though a similar interactive process with the decision-maker will also allow the recipient to refine which evidence will be relied upon in the hearing. So, there are two interactive vetting processes, the first by the investigator, and the second by the hearing decision-maker or chair. It may be helpful to think of the universe of all gathered evidence as eventually being culled into three different buckets.


Bucket #1 is all evidence that the investigator determines is relevant and is therefore incorporated into the investigation report. Evidence is relevant when it tends to prove or disprove an issue in the complaint.


OCR doesn’t tell us what “directly related” means, or how it differs from relevant evidence. So, we came up with a description that makes sense to us: evidence is directly related when it is connected to the complaint, but is neither inculpatory (tending to prove a violation) nor exculpatory (tending to disprove a violation) and will not be relied upon by the investigation report. For example, the mental health of a party may be introduced into evidence in a sexual assault complaint, but if the investigator determines the mental health issue is not relevant to consent, it will not appear in the report, but will be available to the parties in the file of evidence that represents Bucket #2.


The third bucket is really the discard pile. Evidence that is neither relevant nor directly related will be stored here and is not shared with any party. For example, if texts between the parties relate to the complaint, but also discuss homework, the texts about the homework are likely Bucket #3 evidence (meaning they’re really not evidence at all, just unrelated facts hovering around evidence).

Cooking shows are a popular pastime during COVID-19 quarantine, so we offer this metaphor from the restaurant cooking world to help comprehend the purpose of the assessment of credible and relevant evidence. The investigator is like a sous-chef, preparing the raw material, amassing the right ingredients, choosing the freshest, ripest, best ingredients available, and composing them according to the recipe. The decision-maker is like the chef who takes those ingredients, assembles the dish from them in the right order, cooks it at the right temperature, and plates it artfully. Without the chef, there is no dish, but without the sous-chef, the dish would not adhere to the recipe and would not be as satisfying. The ratio of division of duties on a kitchen preparation line (40/60? 49/51?) is comparable to the ratio of emphasis between the investigator and decision-maker in the overall resolution process. The sous-chef shouldn’t try to cook the dish for the chef, and the chef shouldn’t have to source and assemble the ingredients according to the recipe, when that is the sous-chef’s job. They each respect each others’ roles on the line and work separately but in tandem to produce a high-quality result.


In an investigation, synthesizing is the step within report writing where the investigators prepare two lists. The first list is a summary of all undisputed evidence upon which the parties/witnesses are in agreement. The second list summarizes all the evidence that remains contested between the parties/witnesses. The hearing will focus primarily on the second list because the purpose of a hearing is to resolve disputed facts.

During the report review phase of investigation outlined in the 2020 regulations, the parties have the opportunity to review and comment on the investigation report before the hearing. It’s safe to assume that evidence that was undisputed during the investigation or report review period will remain so during the hearing. Of course, that’s not always the case, and if un- disputed evidence is contested during the hearing, the decision-makers will draw out and explore the nature of the new disagreement. Similarly, evidence apparently in dispute may turn out to be uncontested by the time of the hearing. Regardless, anything investigators can do to narrow the scope of the hearing’s inquiry to that which is in dispute will help to ensure the efficient flow of the hearing process to a fair determination.

The last step we recommend is a synthesis of all questions. This will be a great help to the decision-maker in determining relevance. This likely takes the form of an appendix to the report, in which the investigator summarizes questioning in a three-column table. The first column shows all questions asked by the investigator of each party/witness. The second column shows all questions suggested by the parties/advisors to be asked of other parties or witnesses. The third and final column allows the investigator to explain whether they posed these suggested questions or not, if not, why not, and whether they reframed them and then posed them. If so, they should explain why. The same questions are likely to come up at the hearing, and this appendix will help the decision-maker prepared to rule on their relevance.


Many recipients’ procedures include a step that allows a matter to be resolved informally (without a hearing) if the parties accept the result post-investigation. They can’t accept a result if there isn’t a finding. For the reasons stated above, we don’t think that investigators should engage in policy analysis, primarily because it could unduly influence an eventual hearing.

However, the parties can often see where things are likely heading after they read the investigation report, without anything more than the G.A.S. content that is provided to the parties for their review before the report is finalized.

When the report is finalized, an appropriate administrator will notify the parties that the hearing will be scheduled unless the Responding party wishes to accept responsibility for some/all of the alleged violations. If so, and the Complainant agrees, the process can shift into informal resolution. If the finding and sanctions are agreed upon by all parties, the matter can be resolved at that point. If not, it will be resolved through a hearing.


Under the new regulations, there is a continued robust role for the investigator as a witness at the hearing. Investigator testimony should be mostly factual, and they should shy away from offering opinions, just as decision-makers should avoid soliciting their opinions. If opinions are offered, or become apparent, they should not be binding on the decision-makers, who must render an independent determination.

By the end of an investigation, investigators are in a unique position to inform the overall determination, without unduly influencing it. We hope you’ll find the right balance for your school or campus, and ATIXA will be here to offer certification training specific to the functions of investigators, coordinators, and hearing decision-makers in the process.

¹ This is not an assessment of workload, but apportionment of responsibilities within the overall  resolution process.

² Setting aside the nuances of Rule 701 of the Federal Rules of Evidence and what’s admissible for now, the larger point is about due process not being offended by the investigator serving a wider role than just fact-gathering.

© ATIXA, 2020. All rights reserved. The G.A.S. model is proprietary to ATIXA and is not to be used without express permission.