Lesson 4: Kit-Build Framework

You built a Kit-Build platform. You need to explain what Kit-Build is, why it exists, and how it differs from open-ended concept mapping — in about 60 seconds.

What is Kit-Build?

Kit-Build is a close-ended concept mapping framework developed by Hirashima et al. (2015). It solves a fundamental problem: open-ended concept maps are hard to assess automatically because every learner creates a different representation.

The core idea: The teacher creates a "goal map" (the correct concept map), decomposes it into a "kit" (individual nodes and links), and learners reconstruct the map using only those components. Because both maps use the same components, the system can automatically compare them at the proposition level.

The Kit-Build Workflow

Teacher creates
Goal Map

→

System decomposes into
Kit (nodes + links)

→

Learner sees
Kit components

→

Learner reconstructs
Map

→

System compares to
Goal Map

→

Automatic
Diagnosis

In your study, the goal map contained 7 propositions representing spatial and functional relationships in the "わたしのうち" (My House) passage from Minna no Nihongo.

Scoring Formula

A score of 53.9% means approximately 3.8 out of 7 propositions were reconstructed correctly on average.

Kit-Build vs Open-Ended Concept Mapping

Open-Ended Concept Map	Kit-Build Concept Map
Learner creates nodes from scratch	Learner uses provided nodes only
Learner creates links from scratch	Learner uses provided links only
Difficult to assess automatically	Automatic proposition-level diagnosis
High cognitive load (recall + organize)	Focused load (organize + connect)
Assesses recall AND structural understanding	Assesses structural understanding primarily

In Japanese context, there's an additional layer: Learners must also decode Kanji to place the correct nodes. This means Kit-Build score in Japanese measures both structural understanding AND script-decoding ability — which may weaken its correlation with pure comprehension tests compared to English/EFL contexts.

Why Kit-Build for Japanese?

Japanese reading comprehension is uniquely challenging because of the hybrid writing system (Hiragana + Katakana + Kanji), SOV sentence structure, and conceptually distant relationship from Indonesian. Traditional memorization-based approaches are insufficient when learners need to understand relationships between concepts in a text.

Kit-Build forces learners to focus on propositions — the smallest meaningful unit of knowledge — rather than vocabulary in isolation. This aligns with meaningful learning theory (Ausubel): knowledge is constructed by connecting new information to existing structures.

Key Questions and Answers

What does Kit-Build measure that a comprehension test doesn't?

Kit-Build measures structural understanding — whether learners can identify and organize the relationships between concepts in the text. A comprehension test measures whether learners can answer questions about those relationships. They test overlapping but different things.

Why is Kit-Build score not correlated with score change?

The Spearman correlation was not significant, but the study was not powered to detect this relationship. Kit-Build score measures reconstruction accuracy, while score change measures improvement on a comprehension test. They may tap different cognitive processes, and the small sample limits correlational analysis.

What is Yomilink's contribution compared to the original Kit-Build?

Yomilink integrates the full research workflow: consent collection, pre-test/post-test delivery, TAM questionnaire, qualitative feedback, analytics, and data export — all in one web-based platform. The original Kit-Build platform doesn't support these integrated research administration features. However, this study does not empirically compare the two platforms.