Why mission length is a separate question from isolation itself
Isolation and mission length are not the same variable. A crew can be isolated for a short mission or a long one, and the question of whether shared mental models keep predicting creativity as a mission stretches on requires evidence that tracks both the mental-model measure and the outcome over time. One study of team performance in space crews examined the effects of isolation and confinement and considered creativity tasks alongside complex intellective problems, and it discussed expectations about how crews would perform on a 30-day mission versus longer time frames [5]. But that study did not explicitly treat shared mental models as the predictor [5], so it cannot tell us whether the shared-mental-model–creativity link strengthens, weakens, or holds steady with mission length.
A separate line of work asked how shared mental models can be maintained over long-duration exploration missions and considered how isolation affects team cognition [2]. That work is about preserving team cognition across a long mission, not about whether shared mental models predict creative problem solving, and it did not investigate creativity, problem solving, or the impact of mission length [2]. So the maintenance question and the prediction question have been studied in different places, and the bridge between them—does the predictive relationship survive a long mission?—is not yet directly tested in these studies.
What Earth-based construction and human-robot teams suggest
Some of the most concrete evidence about shared understanding and problem solving comes from teams that are not in space at all. A review of team dynamics in complex construction scenarios on Earth, used to inform training for lunar construction crews, describes crews living and working in isolation and points to shared mental models and transactive memory—a team's shared awareness of who knows what—as part of how knowledge is coordinated [4]. This is a review of Earth-based teams, so it informs what to train and what to watch for, but it does not measure whether shared mental models predict creativity in an isolated space crew.
Work on human-robot collaboration in safety-critical situations is also relevant because it examines shared mental models as predictors of creativity and problem solving in crews under prolonged isolation [1]. That framing extends the shared-mental-model idea beyond human-human coordination to teams that include autonomous assistants, which is closer to the mixed crew that lunar or long-duration missions may involve [1][4]. The abstract-level evidence supports the prediction claim in that setting but does not report how the relationship behaves as mission length increases [1].
The practical takeaway, and the gap that remains
For someone designing or training a crew, the usable message is that building a shared mental model looks like a reasonable target if you care about creative problem solving under isolation [1][3]. The evidence is strongest for the existence of the link in isolated analog crews and for the importance of shared mental models in real mission problem solving; it is weakest for any claim about how many days or months into a mission that link persists.
The mission-length question remains open in a specific way: the studies that address long-duration missions focus on maintaining team cognition rather than on whether shared mental models still predict creativity [2], and the study that examines isolation and creativity over mission time frames does not use shared mental models as its predictor [5]. Until a study tracks both together across a long mission, the honest answer is that shared mental models predict creative problem solving in isolation, but whether mission length changes that prediction is not yet established by these studies.
Sources used in this answer
Toward enabling crew resource management objectives in human-robot collaboration in safety-critical situations
Shared mental models predict creativity and problem solving in analog crews experiencing prolonged isolation and confinement, in a human-robot collaboration context.
Maintaining shared mental models over long-duration exploration missions
The paper considers how to maintain shared mental models over long-duration exploration missions and how isolation affects team cognition, but does not investigate creativity, problem solving, or mission-length effects.
Teamwork in the Vacuum of Space: The Team Processes Guiding Apollo Missions' Collaborative Problem Solving
Apollo mission teamwork analyses point to shared mental models as important for collaborative problem solving, especially during high workload or isolation.
Team dynamics and collaborative problem-solving for lunar construction: Lessons from complex construction scenarios on Earth
A review of Earth-based complex construction teams working in isolation identifies shared mental models and transactive memory as part of how knowledge is coordinated, to inform lunar construction crew training.
Team performance in space crews: Houston, we have a teamwork problem
A study of space crew team performance examined isolation and confinement effects on creativity and complex problem solving and discussed expectations for a 30-day mission, but did not explicitly test shared mental models as a predictor.