NC Standard Course of Study · Kindergarten ELA

The Kindergarten Playbook

What software can genuinely do across all thirty-nine live kindergarten standards — and why almost none of it looks like a quiz.

01Start from the constraint

A kindergartener cannot read the instructions and cannot type the answer. Everything they produce comes out as speech, drawing, or pointing. That single fact disqualifies most of the software that gets sold into primary classrooms, which assumes a reader operating a screen alone.

So the question is not “how do we assess these standards.” It is: given a child who communicates by talking and drawing, and an adult who is already in the room, what can software do that is actually worth the interruption?

Three of the forty-two codes are dormant. RL.K.8 is marked not applicable to literature, RI.K.8 begins in grade 1, and L.K.3 begins in grade 2. Thirty-nine are live. Any coverage tool that shows a teacher forty-two empty boxes is lying to them three times.

02Six capabilities

Organised by what the software does, not by which standard it touches — because the same capability usually serves a dozen standards across three strands.

1. Capture

RL · RI · SL · W — about 24 standards

The largest unlock, and the least glamorous. Nearly everything a kindergartener knows is demonstrated out loud and then gone. A child retells The Three Billy Goats Gruff at the carpet, nails the sequence, and there is no record except the teacher's memory and a sticky note.

Fifteen seconds of audio, tagged to a child and a standard, timestamped. That is the whole feature. No scoring, no transcription necessary — the teacher already made the judgement in the moment; the software just stops it evaporating.

Tap child, tap standard, hold to record. Under five seconds of teacher time.
Photograph a drawing and it becomes evidence for W.K.3 and SL.K.5 at once.
The same clip can satisfy several standards. One retell is RL.K.2, RL.K.3 and SL.K.4.

Hardest part: making it faster than a sticky note. If it is not, it will not be used.

2. Discrete practice

RF · parts of L — about 6 standards

Finite, right-or-wrong items with spaced retrieval. This is the only family where a machine can legitimately keep score, because the answer set is closed: there are 52 letterforms, 21 consonant sounds, 10 vowel sounds, a known list of high-frequency words.

It is also the only family where adaptivity means anything. Everywhere else in kindergarten, “adaptive” is marketing.

This is the tool that already exists. It covers two of the thirty-nine.

3. Composition without writing

W.K.1–4 · SL.K.5

Kindergarten writing is explicitly draw, dictate, write — in that order. A canvas plus a microphone plus live transcription is a genuine writing instrument for a child with no spelling: they draw, they say what it is, and they watch their speech turn into print.

That last moment is not a side effect. Watching your own words become letters is a print concepts lesson (RF.K.1.b) delivered by the act of composing.

Child keeps the authorship. The software is a scribe, not a co-author.
Show the words appearing as they speak, left to right, with spaces.
Let them hear it read back. Revision becomes possible at an age where it usually is not.

W.K.4 literally requires digital tools — it is the one standard that cannot be met without software.

4. Matching

RF.K.5 · teacher workflow

Given the fourteen sounds this child has actually been taught, which books in the room can they read without guessing? That is a computation, and nobody does it by hand because it is tedious and exact — which is the definition of work to give a machine.

Decodable text matched to the phonics scope and sequence a child has covered, not to a reading level.
Grouping: cluster the class by shared gap, so small groups form around a lesson rather than a label.
Book-to-standard: which text on the shelf gives you a clean shot at RI.K.9 this week.

5. Facilitation

SL.K.1–4 · RL.K.10 · RI.K.10 · W.K.5

Software as the object on the table that structures a conversation, rather than something a child operates. Turn-taking tokens, a visible talk timer, partner pairing that does not repeat the same two children, sentence stems displayed as pictures.

The entire Speaking and Listening strand is about collaborative talk. No software can judge whether a conversation went well. Software can make the conversation happen more often and more evenly, which is the part teachers actually struggle to do while also teaching.

6. Teacher judgement support

All 39, indirectly — and where the real time is

Thirty-nine standards times twenty children is 780 judgements per reporting period. The exhausting part is not making them. It is knowing which ones you have evidence for and which child you have somehow not watched closely since September.

Coverage grid. Thirty-nine by twenty. Colour by evidence age. The holes are the product.
The quiet-child alarm. Who has the least evidence this fortnight? That is usually the child who most needs looking at.
Report assembly. Pull the captured evidence for one child into a draft a human then edits.
Conferring notes that attach to a standard as you type them, so filing is free.

Highest leverage of the six, and it never touches a child's hands.

03Which capability serves which strand

Strand	Live	Capture	Practice	Compose	Match	Facilitate	Teacher
RL Literature	9
RI Informational	9
RF Foundational	5
W Writing	6
SL Speaking & Listening	5
L Language	5

Strong fit Partial Not the right tool

Read the columns rather than the rows. Capture and teacher support are the only two capabilities that serve every strand. Discrete practice — the thing most people picture when they hear “educational software” — serves exactly one.

04What software must not do

Score a retell. Keyword matching against a story summary will reward a child who says the right nouns in the wrong order and punish one who understood it perfectly in their own words.
Judge a conversation. SL.K.1 is about taking turns and building on what someone said. There is no signal for that in a microphone.
Produce a reading level from a probe. Ten letter sounds is not a Lexile, and the moment a number looks official it ends up in a file that follows the child.
Replace the adult listening. Most of these standards say “with prompting and support” in the text. The prompting is the standard.
Put a five-year-old alone with a screen to hit a metric. If the design depends on sustained solo screen time, it is wrong for this age regardless of what it measures.

05Where to start

If the goal is the most good for the least build, the order is not the order people expect.

Capture. One screen, a class list, the 39 standards, hold-to-record. It is a week of work and it serves two dozen standards on day one.
The coverage grid on top of it — which is nearly free once capture exists, and is the thing that makes a teacher open the app on a Sunday.
Draw-and-dictate for writing. Satisfying for children, and the only route to W.K.4.
Decodable matching, if a phonics scope and sequence is already in the system from the practice tool.
Facilitation odds and ends — timers, turn tokens, pairing. Small, cheap, used daily.
More discrete practice last. It is the most fun to build and the narrowest in reach.

The test for any of it: does this save a teacher time, or does it spend a child's? Kindergarten has roughly nine hundred instructional hours in a year. Anything that takes some of them has to give back more than it took.

Standard codes, counts and dormancy are taken from Kindergarten: ELA Standards, Clarifications and Glossary, NC Department of Public Instruction, adopted 2017, implemented 2018–19. Companion to Mapping the Alphabet Code, which covers the foundational strand in depth.