Published by Unseen Progress, an independent publisher of caregiver research. Last reviewed 2026-05-10. Part of the TBI caregiver research overview.
Short answer. Family disagreement about TBI recovery is structurally inevitable, not a sign that someone in the family is being unreasonable. Different family members see different slices of the survivor's week — the spouse sees cognitive fatigue and the worst hours, the weekend visitor sees the rally version, the teenage child sees the personality changes, the adult child checking in by phone sees an articulate phone manner — and each forms a different mental model of "how recovery is going." The TBI family research (Kreutzer and colleagues; Chan and colleagues on caregiver burden) treats this divergence as a defining feature of long-arc TBI caregiving and finds that the families who resolve it best are not the ones who argue their way to consensus but the ones who build a shared record.
Five forces converge:
Rehabilitation teams have long recognised that survivors often pull executive function together for a 30–60 minute appointment in a way they cannot sustain for a typical evening. Social scripts are preserved longer than executive function in TBI; the rally version exists. The same effect distributes across family. Family members who interact with the survivor for short, structured windows — Sunday lunch, Wednesday phone call — see something closer to the rally version. Family members who interact for unstructured hours see something closer to the underlying picture.
The 10 a.m. version of the survivor is often substantially better-functioning than the 4 p.m. or 8 p.m. version. Family members who see the survivor primarily in one window form their model around that window. The Brain Injury Association of America's guidance on cognitive fatigue treats this distribution as a defining feature; it makes "what is the survivor like?" a poorly-formed question without a time-of-day specifier.
Cognition shows up in tasks. Behaviour shows up under stress. Speech shows up in conversation. Mood shows up in unstructured time. A family member who interacts with the survivor mostly through one of these channels has a high-resolution picture in that domain and a low-resolution picture in the others.
A spouse with primary responsibility for the survivor's day-to-day reality has a different incentive to notice incremental gains than an adult child who is grieving the pre-injury parent. The TBI family research (Kreutzer and colleagues) finds these motivational differences shape attention asymmetrically across long-arc caregiving.
Each family member's mental model is a noisy weighted average of their last few interactions. With different last few interactions, different averages.
The result is the canonical TBI family experience: each person believes they are describing "the survivor's recovery"; each is actually describing the slice they see; each interprets the others as missing the obvious; chronic low-grade conflict.
The way TBI families describe this on community forums maps onto the research:
Each is correct. None is complete.
The intuitive response to "we see this differently" is to argue toward a consensus version. This typically fails for two reasons.
First, the disagreement is not actually about the survivor — it is about which slice each person sees. Arguing the conclusion does not address the underlying difference in input. Each person leaves the argument unconvinced because each is still right about their slice.
Second, in the absence of shared data, the argument resolves to whoever is most confident, most senior, or most exhausted. None of these is the same as accurate. Hetherington's longitudinal work on family functioning under stress (a different population, but the dynamic generalises) and Kreutzer's TBI-specific family research both find that confidence-driven resolution produces worse outcomes than data-driven resolution.
The TBI family literature converges on a small set of interventions for divergent mental models. The strongest is the simplest: build a shared record across domains and time windows that everyone in the family can read.
A shared record is not a disagreement-resolution tool in the conventional sense. It does not produce one mental model; it makes the differences visible. Once the spouse can see "behaviour at 4 p.m. on weekdays" alongside the weekend visitor's "behaviour at 1 p.m. on Sundays", the disagreement reframes from "who is right about him?" to "these are both true; the survivor is differently regulated across windows." That reframe is itself most of the resolution.
Three components matter:
Cognition, behaviour, mood, speech, energy. The blended "how is he today" question loses information. The separated record retains it.
The 10 a.m. and 4 p.m. ratings carry the cognitive-fatigue distribution that any single rating loses.
Not every family member rating every day; that produces fatigue and conflict over the rating itself. One primary recorder, with periodic input from secondary observers (a 1–2 line note from the weekend visitor describing what they saw) is enough to surface the divergence.
Without data, this is unresolvable. With data, both are usually accurate within their windows. The conversation moves from "who is right?" to "what would make the 4 p.m. version look more like the 10 a.m. version?" — which is a tractable question.
Without data, this is a values argument. With data — particularly cognitive scores tracked separately from behavioural scores across rolling 90-day windows — it usually becomes clear which domain is moving and which is not. The conversation becomes a strategy conversation rather than a credibility contest.
These are the same observation viewed at different time scales. Long enough record reveals the truth most TBI families eventually find: the personality changes are partially neurological signature of frontal injury (which heals partially, on a multi-year curve) and partially environmental (which responds to load management). Both family members were partially right.
1. Stop arguing the conclusion. Each family member is reporting their slice accurately. 2. Start logging across domains. Cognition, behaviour, mood, speech, energy — separately, with time-of-day where it matters. 3. Hold a monthly family read of the record. Not a debate; a read. Each family member describes what their slice has looked like; the record provides the connective tissue. 4. Use the rehab team as the tiebreaker on objective questions. Neuropsychological retesting and structured functional assessment are the externally-validated measures that no family member's slice alone can produce. (See bringing data to rehab appointments.) 5. Accept that some disagreement will not resolve. Different family members are grieving different versions of the survivor on different timelines. The shared record narrows the disagreement to the part that is actually about facts; the rest is grief and is allowed.
---
Unseen Progress publishes long-form caregiver research and builds research-backed daily trackers for the families covered. See the full TBI caregiver research overview for the complete framework.