A Cognitive Interview, Branching (Beau Johnson)

A Cognitive Interview, Branching

A constructivist think-aloud protocol, with four students' real paths traced through it.

Protocol design and interviews by Beau Johnson. Subjects: Audrey (age 10) and Alexia (age 8) at Jefferson Achievement School, Mason (freshman) at Prairie Ridge High School, and Esperanza (ninth grade) at St. Auden's Academy. Recorded April 2014.

Study framing

Research question

How does the act of preparing project-based, interest-driven work reflect the literary comprehension and academic engagement of the emerging adolescent?

Theoretical grounding

Montessori frames adolescence as a sensitive period, a time of expectation and creative work when manual and intellectual effort should complete each other. The design leans on Dewey and Vygotsky for the claim that a purposeful task drives real engagement, with Piaget informing the developmental read.

The problem

Schools claim a whole-child, outside-the-box education, yet they build it on a foundation of standardized testing, and no alternative has been mainstreamed successfully. This study looks for what a test cannot show.

Subjects

The broader study drew on rising fifth through tenth graders in Texas, California, New York City, Connecticut, and South Carolina. It spanned public and private schools as well as charters, with reading levels and prior exposure to project-based work that varied widely. Audrey and Alexia are sisters from that pool. Mason and Esperanza are two more cases, both Texas high schoolers, one at a structured public school after a project-based private one, the other at a parochial school.

What we expected, and what we found

The interview should reveal what each student brings to an unfamiliar text, from her perception to her willingness to talk it through. As expected, students preferred building something to being told what to do. One put it plainly: I can get lost in a project. I think of them like art, and I like seeing my work up on a wall. The data then complicated the easy version of that story. Engagement did not always follow, and it was not always fueled by interest.

Limitations and next step

A sample this small cannot support an across-the-board claim, and real cognition would need biological and neuropsychological expertise the study did not have. The open question is pointed: whether the search for universal similarities is just another form of standardization. The next version would run longer and sit inside a shared curriculum, comparing cases over time, asking less about what students know and more about how they use it.

About the stimulus text

The story the students read is A Desperate Case for Magick by Stephanie Burgis, a short story set just after her Regency-fantasy middle-grade novel Kat, Incorrigible (published in the United Kingdom as A Most Improper Magick). Using a real but unfamiliar text kept the reading authentic while holding the material constant across subjects.

The students' think-aloud references decode cleanly against it. The heroine they call Cat is Kat Stephenson, a fiercely independent girl who discovers her own magic. Mr. Carly is Frederick Carlyle, a wealthy young man who arrives disguised as a Classics student. The witches are Kat's older sister Angeline and her true-love spell, and Angeline is the character Audrey names as her favorite.

The readings diverge in a telling way. Both sisters took the story to be about a girl trying to escape her family, although its actual engine is Angeline's spell. Mason, outside his preferred genre, did not settle on a reading and built images to stay afloat instead. Esperanza gave the most complete reading of the four, reconstructing the world from clues and prizing the very in-media-res opening that Audrey rejected. Four readers meeting one text four ways is the point.

The protocol is a decision tree. Four real interviews run through it here: two sisters at one school, and two Texas high schoolers at two others, all reading the same story. They agree on the early moves, then split on how they read the text and how they would rebuild it. Those divergences are flagged. What this demonstrates: interview design, branching follow-up logic, and reading four very different minds against one instrument.
Audrey, age 10 Alexia, age 8 Mason, freshman Esperanza, ninth grade Response branch

Subjects are pseudonymous, and their schools are anonymized. Quotes are lightly cleaned from recorded interviews, with wording preserved; auto-transcription had swapped some speaker labels, corrected here. The interviewer is anonymized throughout.