Books designed for a purpose
Every text states its code demand (C1–C4, against an explicit sequence of Spanish) and its language profile (vocabulary, syntax, discourse, knowledge, image). Two descriptions, never a single band that labels the child.
Books children can actually read. Assessment that teaches. A multigrade classroom that works.
A national proposal for early literacy in Spanish for Peru's rural multigrade schools.
The problem
One classroom, very different reading histories
In a rural multigrade school, children with very different reading journeys share a single classroom. The materials available were not designed for that reality.
- There are no leveled Spanish texts for that reality. Either progressions built for English are imported, or texts are «simplified» without criteria.
- Assessment labels instead of guiding. It often assigns a «level» rather than telling the teacher what to teach next.
- Equity arrives late. Readers who come with another language or variety are treated as a special case, not as the starting point of the design.
The approach
Three pieces that work together
None works alone: the books state their demand, screening turns into teaching, and the kit organizes the classroom around that priority.
Screening that becomes teaching
A short task route yields a teaching priority (code, fluency, language-comprehension, extension) and a state of the evidence. No scores, no levels: deciding nothing is a legitimate outcome.
Multigrade classroom kit
Temporary groups by objective, certified independent tasks, re-entry for those who missed class, functional routines and teacher support. Priority organizes mobile groups, not fixed identities.
What makes it different
Five design decisions
These are decisions, not promises: they define how books are written, how learning is measured, and how teaching is organized.
- No child has a level. Priority is mobile and revisable: it organizes temporary teaching groups and changes with the response to teaching.
- Variety is not error. Andean Spanish is accepted as a legitimate variety; only what changes the word is corrected. Without these rules, an instrument would read a child's pronunciation of a different variety as deficit.
- Culture is not decoration. Situated content goes through community consultation and authorization before publication; values are embodied in the story, never as a moral.
- Build once, verify light. Scale, corpus, books, instruments and kit are national assets that are not rebuilt per district; each site runs a light, standardized verification.
- Honesty about status. Decisions, hypotheses and open issues are registered and kept separate: no hypothesis is presented as a result.
The project in numbers
What exists today
20 chapters of active manuscript, across seven parts
7 technical dossiers (PT1–PT7) with evidence and validation strategy
22 print-ready books produced (world W1 complete, 14/14; W2 at 8/12)
100 planned titles across six worlds
The collection
Books with original watercolors
Twenty-two titles produced, each described by its code demand (C1–C4) and its language profile. The series follows Nina the vizcacha across worlds.
First titles
World 1 · The hillside
World 2 · The fair
Status and transparency
Architecture complete, validation pending
Honesty about status is part of the method. A complete documentation architecture does not turn a hypothesis into a validated result.
What is complete
- Documentation architecture PT1–PT7 with evidence, design and validation strategy.
- A 20-chapter manuscript explaining the complete design.
- 22 print-ready books produced and a world guide with teaching sheets.
- A v1 screening instrument with a teacher sheet and a map of 100 titles.
Not yet validated in the field.
The printed editions are first working editions. Before any data on minors is collected, the project must close:
- Expert and community review of the national assets and situated texts.
- Testing of the titles with real readers.
- A closed and authorized ethics and data plan.
- Territorial agreements (UGEL/DRE) and EIB coordination.
The road to the first entry
Six steps to the first entry
The goal is not an impact pilot. It is a feasibility, process, exposure, cost and adverse signal study: how much local adjustment the national core needs, so it can be standardized.
-
National assets
Reference sequence for code, Andean children's corpus, variety rules, governance and the ethics and data plan. They unblock everything.
-
Authors and illustrators
Briefs and early handoff: writing, illustrating and cultural authorization are the critical path.
-
Instruments
Screening, teacher sheet and leveled stimuli, in parallel with the books.
-
Tested titles
6–10 contrast titles with independent double leveling and testing with real readers.
-
Classroom kit
Assembled last, from tested books and instruments: integrated and verified in a multigrade classroom.
-
First entry
The hard test of the light protocol in a demanding territory. If it works there, it travels.
Join the project
We are looking for partners to close the national assets and carry out the first entry. Write to us to receive the project brief and the territorial phase plan.
Semillas de lectura is a project of LIDES, a social and educational innovation lab.
