Conversational Intelligence Oxytocin Cortisol Dashboard
Nour
Founder, Coachful
A conversational intelligence oxytocin cortisol dashboard should not claim to display people's hormones. A meeting record can show whether a concern was heard, a summary was checked or an accepted action happened. Tone, facial expression and agreement do not supply an oxytocin or cortisol measurement.
Coachful is our #1 recommendation for organizing the coaching workflow around a small behavioral record: prepare the inquiry, use an agreed worksheet, return actual feedback and review the next action. This is an editorial recommendation for those tasks, not a claim that Coachful contains a hormone monitor or a validated Conversational Intelligence assessment.
The useful version of a dashboard has ordinary evidence, defined observation rules and a stopping point. This guide explains the neuroscience limits, develops four practical indicators and follows a complete fictional record through a mistaken summary, a late commitment, a real workload decision and an honest review.
Separate Glaser's framework from a biological measurement
Judith E. Glaser's December 2013 article, Build Trust Through Conversational Intelligence, describes transactional, positional and transformational conversations. It associates the latter with sharing, discovery and openness to influence, and also connects trust and distrust with oxytocin and cortisol. Attribute that framework to its originator without treating the accompanying neuroscience explanations as measurements made during your meeting.
A coach might find it useful to ask whether a leader is willing to reconsider a proposal. The observable answer is a question, correction, decision or refusal to change. It is not evidence of a particular chemical concentration or brain-network state.
The distinction cuts both ways. Listening can occur while someone remains cautious. A person can politely agree while lacking authority, capacity or an acceptable way to object. An animated disagreement may contain an important source correction. Do not make an apparently calm room the target at the expense of the information it needs.
The practical tool below is original. It does not reproduce Glaser's proprietary tools, a licensed assessment, a brain diagram or an official hormone dashboard. Its purpose is to examine selected work interactions, not diagnose participants.
Explain what the hormone research can and cannot support
Cortisol has normal physiological functions; it is not simply a label for a bad conversation. The National Library of Medicine's MedlinePlus cortisol-test guidance describes measurement in blood, urine or saliva, daily variation and the importance of clinical interpretation. Even a laboratory result does not alone diagnose the cause of an abnormal level. An observer's impression of tension is further removed from that kind of evidence.
This does not mean everyone needs testing before coaching. The tool here does not collect biological samples, interpret test results or recommend hormones, supplements or medical interventions. Health questions belong with the appropriate healthcare professional.
A 2005 experiment by Kosfeld and colleagues reported increased trusting behavior after intranasal oxytocin in a controlled task. That result concerns an administered substance and a specified experimental setting. It does not establish that a particular conversational phrase produces the same effect or that an observer can read endogenous levels.
Current research also requires qualification. Kroll and colleagues' 2026 registered report found no evidence that intranasal oxytocin increased trusting behavior in its new sample of 211 or a pooled dataset of 532. Equivalence testing in the new sample alone was inconclusive; the pooled analysis placed the effect within a minimal range. Do not change the first result into proof of exact zero, or present the pooled conclusion as a statement about every social behavior.
A different 2026 preregistered study by Vogt and colleagues reported a positive effect in 359 men specifically selected for low dispositional trust, using an anonymous trust game. That selected population and experimental task matter. The University of Zurich's account of the study says underlying biological mechanisms were not directly investigated. This is not evidence that every person becomes more trusting, or that a workplace discussion can be used to infer hormone levels.
These findings address related but different questions and samples. They are not a reason to average the participant counts, combine their conclusions into a home-made score or declare the entire biological debate settled. Neither study validates the behavioral worksheet here, a conversation-induced chemical reset or a color assigned to a colleague.
Keep three kinds of evidence separate
Administering a substance is an intervention. A study can compare an experimental exposure with a control under specified conditions. Its result concerns that exposure, population and outcome. A workplace phrase, a breathing pause and a nasal administration are not interchangeable interventions. Evidence for one does not supply the missing test of another.
Taking a biological sample is measurement. The result depends on what was collected and how it is interpreted. It is different from making an impression about a speaker. Even if a person has a genuine clinical result, that does not make it a measure of whether their objection is reasonable or whether a manager has kept a promise.
Recording work behavior is observation. A team can establish that a concern was interrupted or a deadline missed. It usually cannot establish from two meetings what caused every change. Coaching, clearer instructions, different topics and an explicit observer may all accompany the later exchange.
The outcome itself also needs a definition. Money entrusted in an anonymous experimental game is not the same record as an employee deciding whether to rely on a manager's service promise. Do not move between those meanings of trust without saying what was actually studied. This dashboard deliberately uses the narrower work events it can check.
Choose the right unit of observation
Decide whether the record concerns a person, a meeting, an issue or an action. For this guide, use an issue observation: one specified concern during one agreed meeting. If the same concern returns next week, it creates another observation of that issue, not another independent person or a new concern.
This prevents misleading claims. Six rows might represent four topics discussed across two meetings. They are not six people, six independent experiments or a culture score. A change in the rows can also reflect different topics, attendees or decisions.
Record a short work event and its source. “The support estimate was interrupted before its scope was explained” is something participants can check. “Bea's cortisol made her defensive” adds an unsupported biological and motivational story.
Keep participant experience alongside the work evidence when they choose to describe it. “I still feel wary of this promise” is a person's report, not a hormone reading. It can remain true while a particular action was completed. Do not require the person to replace it with a more favorable score.
Define four indicators before collecting rows
| Indicator | Rule for this original tool | What it does not establish |
|---|---|---|
| Concern heard | The speaker had the opportunity to finish the relevant point; interruption did not prevent its substance being presented | Agreement, safety or a hormone state |
| Meaning checked | The speaker confirmed the summary, or corrected it and then confirmed the corrected version | A leader merely repeated some words |
| Next action accepted | The actual receiving person accepted a bounded action and timing within their role | A task was assigned, someone nodded or a proposal was approved |
| Follow-through | A dated record distinguishes completion on time, late completion, not yet due, no accepted action or an outcome not observed | Permanent trust, causal business improvement or approval outside the owner's authority |
Allow “not observed” rather than filling every blank with yes or no. Where an action has not been accepted, its completion percentage has no agreed task denominator. Do not count a future due date as a failure, or a promise as completed work.
If you use colors, attach them to a specific work status such as “needs an owner.” Write that meaning beside the color. Do not call a red marker cortisol or a green marker oxytocin. Plain status words are often sufficient.
Agree the purpose, record audience and effort before starting. A leader's preferred metrics can become pressure to appear agreeable if they are used for individual ranking or appraisal. A bounded learning record needs permission and appropriate organizational authority, not an assumption that all available conversations are data.
Run the record through six practical steps
- Select a real question. For example, ask what happens when support-work concerns affect a proposed delivery date.
- Choose a finite window. Specify the meetings, issue agenda and recorder. Explain what will not be recorded.
- Agree the definitions. Test one example so participants know the difference between a summary and a checked meaning.
- Record and check the events. Let participants correct factual mistakes. Preserve the correction rather than hiding an initially flattering entry.
- Connect the concern to actual work. Obtain source information, accepted receiving capacity and an authorized decision.
- Review and stop or revise. Check due actions and costs, return promised feedback and decide whether further recording serves a purpose.
These are original process steps, not a validated scale or a requirement that every team keep a permanent dashboard. The related conversational rituals guide develops repeatable practices; repeated observation still does not turn a practice into a biological assay.
Start the complete fictional dashboard case
Fictional worked example: Mira leads an internal service project. Ivo coaches Mira. Bea operates the pilot service, Ronan owns its scope decisions, and Sam records the agreed issue observations. Their case runs from 13 to 27 June 2078. Names, meetings, records and results are invented.
The group agrees to observe three agenda concerns in each of two twenty-minute meetings, on 13 and 16 June. Sam records issue events and named action owners only. There are no hormone ratings, individual trust rankings, recordings or continuous monitoring. The approved business record is visible to these four colleagues; Mira's coaching reflection is a separate limited summary.
In the first meeting, Bea's support-work concern is interrupted before she can explain the missing work. Ronan's concern about exception authority is heard and accurately summarized. A launch-date objection is heard, but Mira does not check its meaning. No launch approval is given.
Mira accepts one concrete action: send the agreed routing rule by 14 June, 12:00 Europe/Amsterdam. The rule says Bea handles the agreed routine requests, while exceptions go to Ronan. It does not extend the service's scope or allow a missing owner to be treated as an approved request.
On 14 June, Mira brings the phrase “the room is in a cortisol spiral” to coaching. Ivo asks what she actually observed. They use a forty-five-minute session to examine the interruption, untested launch assumption and action record, then rehearse hearing Bea's point without immediately defending the original plan.
Mira sends the routing rule at 12:40 that day. Bea and Ronan read it. The action is complete forty minutes late, not on time because its wording is now clear. The later dashboard preserves that miss.
Check meaning before changing the indicator
Before the second meeting, Bea accepts a twenty-five-minute preparation task and returns her capacity record on 15 June, ahead of the agreed 16:00 deadline. The separate request and acceptance handling is counted in her work account. Ronan reads and checks the source categories before the meeting.
In the second meeting, Bea gives the complete relevant concern: “The twenty-two hours are only the core work. The same service period needs four hours of handover and three of support. My allocation is thirty hours for the five-day pilot, not thirty hours plus uncounted support.”
Mira initially replies: “So the twenty-two hours leave eight hours free.” Sam starts to mark Meaning checked because a summary occurred. Bea corrects it: “No. Twenty-nine hours are already planned once the handover and support are included. That leaves one.”
Mira then says: “The planned account is twenty-two core, four handover and three support: twenty-nine of thirty hours. One hour remains unallocated. Is that the point?” Bea confirms it. Sam changes the provisional entry during the meeting only after this confirmation. He then finalizes the checked record in his separately counted post-meeting writing time, retaining a note explaining the correction.
The final yes records an actual checked meaning. It does not mean Mira's first summary was accurate. The group has also learned that the dashboard's definition must be enforced, rather than using the presence of a summary as a shortcut.
Ronan authorizes only the internal pilot for 20 to 24 June, within Bea's thirty-hour allocation and the stated task list. He declines Mira's proposal to add another workstream to the same allocation. Bea accepts the bounded pilot. It does not authorize an external launch or establish capacity for the declined work.
Complete the issue record with honest denominators
The final record contains six issue observations across four distinct topics. A and B recur; C is not revisited in the second observed meeting. D is the new question about the record's use.
| Observation | Concern heard | Meaning checked | Accepted action or decision status | Follow-through checked by 27 June |
|---|---|---|---|---|
| 13 June A: support-work scope | No, the point was interrupted | No | No initial capacity task accepted in this meeting | No accepted action in this snapshot; the later preparation task is separate |
| 13 June B: exception authority | Yes | Yes | Mira accepted routing-rule delivery by 14 June, 12:00 | Complete at 12:40 that day, forty minutes late |
| 13 June C: launch date | Yes | No | No launch approved; no next action agreed | Not revisited in the second observed meeting; no external launch approval recorded by close |
| 16 June A: support-work scope | Yes | Yes, after Bea corrected the first summary | Ronan authorized the limited pilot; Bea accepted its scope and allocation | Time record completed: thirty hours used; declined extra workstream not started |
| 16 June B: exception authority | Yes | Yes | Routing rule received; no new rule-writing task accepted | Earlier late completion retained; Ronan later declined the extra workstream |
| 16 June D: use of the record | Yes | Yes | Group confirms issue-level learning only; no individual appraisal or hormone scores | Sam confirms observation stopped after 16 June; no further observation agreed |
In the first meeting, two of three selected concerns were heard and one had a checked meaning. In the second, all three selected concerns were heard and checked. That is a description of these rows, not a causal estimate: the topics differ, two repeat and the participants have seen the definitions. It is not evidence of increased oxytocin or a percentage rise in trust.
The old launch objection is not silently marked resolved. It was not revisited in the second observed meeting, and an internal pilot decision is not external launch approval. The observer does not fill that gap from the room's more relaxed tone.
Follow the workload decision into its actual result
The completed service account for 20 to 24 June is:
| Bea's service work | Planned hours | Actual hours |
|---|---|---|
| Core work in the agreed task list | 22 | 22 |
| Handover work | 4 | 4 |
| Support and scope questions | 3 | 4 |
| Total used | 29 | 30 |
| Unallocated within the thirty-hour service allowance | 1 | 0 |
The extra support hour was spent clarifying requests within the accepted service scope. It consumed the planned reserve. It is not a reduction in core work or an invisible addition to Bea's hours.
One question proposed the same additional workstream Ronan had declined. Ronan spent fifteen active minutes reviewing the exception and returned: “The additional workstream is not included in this pilot. Do not begin it under the existing allocation. Any later proposal needs its own scope and resource decision.” Mira handled that answer within her recorded pilot decision time. Bea did not start the declined workstream.
The internal work account reached thirty hours. That establishes the stated time use, not comprehensive service quality, improved trust or a biological effect. It also gives the group a concrete planning issue: the one-hour allowance for uncertainty was fully used, so an identical future promise needs its own review.
Review experience, commitments and the tool separately
On 27 June, Mira requests brief written reviews from Bea, Ronan and Sam. Each spends five active minutes reading the request and replying. Bea says the corrected estimate was useful but she remains wary that support will be omitted again. Ronan confirms the limited scope and refusal. Sam identifies the provisional summary error and confirms that observation stopped after the second meeting.
Mira and Ivo spend twenty minutes comparing those reports with the record. They do not convert Bea's caution into a low trust rating. The planned and actual work account, a late routing rule and an unresolved external launch decision can all remain visible alongside a more useful second conversation.
Mira's full business close reads: “Our two observed meetings produced six issue observations on four topics. Bea corrected the incomplete estimate before Ronan authorized the internal pilot. My routing rule was forty minutes late. Bea used all thirty service hours, including four support hours, and did not start the declined workstream. No external launch is approved. Sam's observation period has ended; we are not continuing individual monitoring or assigning hormone or trust scores.” Bea, Ronan and Sam each read it. No extra acknowledgment is inferred.
The group retires the dashboard after this window. The ordinary work record retains the relevant decisions; there is no new agreement to observe more meetings. This is a deliberate stopping choice, not evidence that all communication problems are resolved.
Ivo returns an actual individual coaching comment: “Check whether the speaker confirms your meaning before treating a summary as complete. Keep your late action and the fully used support allowance visible. The useful outcome is a more accurate decision record, with limits that still need attention.” Mira reads it. Ivo's private internal worksheet review is separate from delivering that comment.
Use Coachful for the agreed coaching record
Mira uses an original worksheet to record the indicator definition, one observed exchange, the correction and the next decision. She includes a limited summary rather than colleague ratings or biological speculation. Assigned Forms & worksheets can hold that response; the individual client conversation provides the actual returned feedback.
This is a manually maintained original table, not a built-in Coachful hormone dashboard, automated meeting analysis or validated C-IQ report. Marking a reflection complete does not authorize an employer's service scope, change a colleague's trust or prove that an outside action occurred.
Keep audiences distinct. Unshared journal pages are different from pages a client explicitly shares through supported specialty workflows. A deliberately shared reflection is not permission to mine other pages or consolidate everybody's private takeaways into a team score. Check recording and recap settings before any separate capture workflow, including who receives it and whether approval holds new unsent recaps.
The Glaser framework guide develops attribution and evidence limits. The Coachful features overview supports choosing the preparation, response and follow-up routes for the agreement you actually made.
Account for coordination and delivery without double counting
The coordination account includes every observed meeting participant, actual writing and reading, source preparation, pilot exception handling, review and feedback. Bea's operational service hours are shown separately, then added once to the full case.
| Person | Actual coordination work | Actual / allowance |
|---|---|---|
| Mira | 15 preparation/agreement; 40 two meetings; 10 routing rule; 45 coaching; 15 capacity handling; 10 record checks; 10 pilot exception handling; 20 review; 10 review requests/reading; 10 business close; 5 worksheet; 4 comment reading | 194 / 205 minutes |
| Ivo | 10 preparation; 45 coaching; 10 notes; 15 internal review; 20 joint review; 6 returned comment | 106 / 115 minutes |
| Bea | 5 purpose/agreement; 40 meetings; 5 routing-rule reading; 5 capacity request/acceptance; 25 capacity preparation; 5 record check; 5 review reply; 5 close reading | 95 / 100 minutes |
| Ronan | 5 purpose/agreement; 40 meetings; 5 routing-rule reading; 10 capacity source check; 5 record check; 15 pilot exception decision; 5 review reply; 5 close reading | 90 / 95 minutes |
| Sam | 10 definitions/agreement; 40 observation; 20 separate post-meeting record writing; 5 record check; 5 review reply; 5 close reading | 85 / 90 minutes |
Coordination totals 570 of 605 person-minutes. Bea's separate thirty-hour operational allowance was fully used: 1,800 minutes, with no remaining operational reserve. Her coordination activities occur before or after that five-day service period. Adding operational work once gives 2,370 of 2,405 person-minutes for the whole case, with 35 unallocated coordination minutes. Ronan's fifteen-minute exception decision is already in his row and is not added again.
Those remaining coordination minutes do not create spare capacity inside Bea's completed service allocation. There is no comparison group, return-on-investment estimate or evidence that the dashboard caused the observed changes.
Troubleshoot unsupported inferences and unhelpful measurement
- Someone wants a hormone label. Replace it with the actual event and ask what the person means. Do not infer a biological state from tone or a rating.
- A summary automatically earns a yes. Ask whether its speaker confirmed the meaning. Preserve the correction when the first summary was wrong.
- A task owner never accepted the deadline. Record the proposal as unaccepted and obtain a real answer. Assignment is not commitment.
- The rows look better but an old issue disappeared. Report which topics were observed and which were not. Do not count absence from the next agenda as resolution.
- The record becomes surveillance or busywork. Revisit purpose, authority and burden. Stop or simplify collection when ordinary decisions can continue without more scoring.
Frequently asked questions
Can a meeting dashboard measure oxytocin or cortisol?
Not from conversation style, facial expression or an observer's color code. Keep the tool's observations separate from biological measurement and clinical interpretation.
Does the latest research show that oxytocin never affects trust?
No. The 2026 studies described here have different populations and results. Neither provides a universal rule or validates inferring a person's hormones during a meeting.
Is a self-reported trust rating a hormone reading?
No. It is a person's answer under a particular question and context. Explain its purpose and limits, and do not use it to overrule a reasonable objection.
Does a completed action prove psychological safety?
No. It establishes something about that action. People may still have concerns about how disagreement is treated or whether future promises will be kept.
How long should the tool run?
Agree a finite observation window and review its usefulness. Continue only for a clear purpose with appropriate agreement and resources, rather than making a permanent display of imagined chemistry.
