The lab is organized
around a small number
of open questions.
The question is whether reflection, historically a one-on-one, high-touch practice, can be made systemic without becoming hollow. We're testing this primarily at Dartmouth College, where the lab's tools sit inside the rhythm of an academic term.
The interface question. Most reflection tools are conceptually ambitious and graphically generic. We treat the surface, spatial map vs. ranked list, single word vs. detailed profile, paper card vs. drag-and-drop, as part of the cognition, not a wrapper around it.
The lab started from individual reflection. We're now studying whether the deliberative skills that grow from sustained personal reflection, listening, holding contradiction, suspending judgment, also surface in group contexts. Participatory design research is the early site for this.
LLMs are good at saying confident things about people. That's the wrong job description. We're studying how to use models as mirrors, surfacing the gap between what someone says about themselves and what their resume, writing, or card sorts actually show, without prescribing what to do with it.
Built tools work across populations. Field-specific tools work deeper but narrower. Threshold gave us a chance to ask: which parts of the work generalize, and which parts have to be re-built for each profession? The answer matters for whether the lab's work can scale beyond Dartmouth.
The principles below are not hypotheses; they are commitments that organize the lab's design decisions. Each one is testable; each remains under active examination. The full design vocabulary (reflection, interpretation, visualization, action) is articulated in the lab's framework paper, The Architecture of Becoming.
The methods catalog below describes the lab's intended evidence program. At the current stage, the lab operates without a consent-reviewed research-data substrate; deployments are documented through instructor field notes, design-team review, and design-rationale write-ups. The empirical studies that would require consent and review infrastructure are documented as open research lines in each working paper.
Every project starts as a prototype. We deploy it with students in the lead author's working contexts (Dartmouth, CAED), observe through facilitator and design-team review, document what is and isn't working in instructor field notes, and use what we learn to build the next version. Each tool is a design hypothesis; each academic term is a design-iteration cycle. We treat career development as a design problem, not a service-delivery problem.
Quantitative (intended once consent infrastructure is in place): engagement traces from Synapse and DartWorld; pre/post measures of clarity and decision-confidence; behavioral observation of map vs. list interaction. Qualitative (current): instructor field notes, design-team review, student-facing deliverables produced as program work. Computational text analysis of student-produced material is in scope under a future consent process.
Everything is iterated in real working contexts, not simulated in a lab. Primary site: Dartmouth College. Secondary: Kent State University's College of Architecture and Environmental Design (CAED). The lab designs alongside the students it works with; formal human-subjects research with those students is a planned next stage, not a current condition.
The lab draws on five intersecting literatures: learning sciences (reflection, metacognition, distributed cognition), identity development theory (self-authorship, possible selves), narrative psychology (life stories, narrative identity), sensemaking research (Weick, Klein), and human-computer interaction (external representations, cognitive load, ambient interfaces). We treat career development as a form of learning, not a transaction, which means it needs environments that support sustained reflection over time.
Baxter Magolda (2001) on self-authorship; Bransford, Brown & Cocking (2000) on how people learn; Schön (1983) on reflective practice; Weick (1995) on sensemaking; Mezirow (1997) on transformative learning; Kolb (1984) on experiential learning; Markus & Nurius (1986) on possible selves; Norman (1993) and Kirsh (2010) on external cognitive supports; McAdams (1993) on the stories we live by.
On the design-management and firm-side traditions, the work also draws on Shulman (2005) on signature pedagogies, Person & Snelders (2010) and Person (2016, 2023) on design integration in firms, Junginger (2009, 2015) on design absorption, and Liedtka (2015) on design thinking in management.
The Architecture of Becoming: Designing Career Development Systems as Longitudinal Learning Environments
The paper argues that most career platforms are built for the institution, not the student, prioritizing placement metrics and operational efficiency while ignoring the reflective work students actually need to do. We propose a four-part framework: reflection (structured writing and exercises), interpretation (tools that surface patterns without prescribing meaning), visualization (maps and diagrams that externalize thinking), and action (job search, mentorship, skill-building, sequenced after self-knowledge, not before).
Read the working paper · See also Publications.
Lo/Be Lab works within several intersecting research traditions. On the education side: Donald Schön's reflective practitioner (1983) and his analysis of the architecture studio as exemplar of professional education (1987); Lee Shulman's account of signature pedagogies (2005); situated learning and communities of practice (Lave & Wenger, 1991; Wenger, 1998). On the firm side: the empirical design management tradition associated with Oscar Person and the SDIM group at Aalto University, Sabine Junginger, Jeanne Liedtka, and Rachel Cooper & Mike Press. Methodologically, the lab is anchored in design-based research (Brown, 1992; Barab & Squire, 2004) and reflective practice.