Learning Lab

Why it works

Every part of Spell Jump exists because a learning technique has evidence for it. For each one: what the research found, and where it lives in the game. Where the evidence is thin or comes from a different age group, that is noted.

  1. Retrieval practice

    Spell from memory, from audio, never from a visible word.

    The evidence
    Rated a highest-utility learning technique in a major review (Dunlosky et al. 2013). A 2018 classroom study found retrieval practice beat rainbow writing for spelling, including 5 weeks later.
    In the game
    Every jump starts with a word spoken aloud and spelled from memory. No word is ever on screen while your child spells it.

  2. Spaced practice

    Practice spread over days, not crammed into one.

    The evidence
    Also rated highest-utility by Dunlosky et al. 2013; spaced retrieval helps children's word learning.
    In the game
    Words move through boxes with growing gaps: same session, next day, 2 days, 4 days. A word counts as Mastered only after 3 first-try successes on at least 2 different days. Past lists come back 2 and 4 weeks later.

  3. Test, study, test

    Pretest, study only what you missed, then retest.

    The evidence
    About twice as many studies favor test-study-test over study-then-test, from Horn's era onward; it is the weekly cycle Gentry recommends.
    In the game
    A new list starts with a quick Qualifier (the pretest). Known words skip ahead; missed words become Focus Words and get most of the practice. A Friday Test closes the week, and the parent report shows pretest vs test.

  4. Immediate self-correction

    Correct right after each word, with the child doing the fixing.

    The evidence
    Horn's corrected-test research; later single-case studies found correcting after each word beat correcting after the whole list.
    In the game
    A miss is checked at once. Your child sees which letter positions differ, then fixes it themselves.

  5. Cover-Copy-Compare

    A short routine for fixing a missed word.

    The evidence
    An established, evidence-based self-management procedure (Joseph et al. 2012 review).
    In the game
    After a miss: Compare (differences marked), Copy (type it from the model), Cover (model hidden, type it once more from memory).

  6. No exposure to misspellings

    Wrong spellings are kept out of sight.

    The evidence
    Reading misspellings lowers later spelling accuracy for both good and poor spellers, for at least a week (Dixon 1997; Brown 1988).
    In the game
    A wrong attempt is shown for at most 3 seconds with the differences marked, then cleared. The game never offers wrong spellings as choices; "pick the right one" is deliberately absent.

  7. Simultaneous Oral Spelling

    Say the letter names while writing.

    The evidence
    A multisensory routine from Gillingham and Stillman, with reported benefits for young and struggling spellers.
    In the game
    Each typed letter is spoken by name. After a miss, the correct word is spelled aloud letter by letter with the tiles lighting in sync. Letter rings in the air say their letter.

  8. Look-Say-Cover-Write-Check

    For the very first meeting with a new word.

    The evidence
    A randomized trial with Year 3 children found it as effective as a phonological program on the taught words, but weaker at transferring to new words.
    In the game
    First encounter only: Look (the word shown with its pattern chunk highlighted and syllables marked), Say, Cover, Write, Check.

  9. Patterns and word parts

    Teaching the spelling pattern, not just the word.

    The evidence
    Explicit pattern instruction transfers to new words better than rote routines.
    In the game
    An optional "pattern of the week" (for example ough or -tion) and per-word syllables from the parent, highlighted during the Look and Fix steps.

  10. Interleaving

    Mixing kinds of words instead of drilling one kind at a time.

    The evidence
    Third graders made fewer errors after interleaved than after blocked practice (Klimovich & Richter 2025).
    In the game
    The scheduler mixes patterns and mixes in 20% past-list words, and never gives the same word twice in a row.

  11. Instant feedback, few new words

    Know right away, and don't take on too much at once.

    The evidence
    Feedback beats no feedback, and practitioner guidance is to introduce only a few new words per session.
    In the game
    Every word is checked the moment your child taps LAUNCH. At most 5 new words a day, and a daily mission of about 8 words (8 to 10 minutes of play).

  12. Intrinsic integration

    The learning action is the game action.

    The evidence
    Children learned more, and chose to play 7 times longer, with games where the learning was built into the play itself (Habgood & Ainsworth 2011).
    In the game
    Spelling sets the launch power and nitro, letters are the rings, and the Friday score sets the power of the Mega Run. There is no "do homework, then play".

Sources

Mostly consulted through secondary summaries.

  • Dunlosky et al. 2013, Improving Students' Learning With Effective Learning Techniques (Association for Psychological Science)
  • The Learning Scientists, 2018 classroom spelling study (retrieval practice vs rainbow writing)
  • Horn 1947, the corrected test (via a 1978 Iowa guide and later single-case studies)
  • Joseph et al. 2012, review of Cover-Copy-Compare
  • Dixon 1997 (thesis) and Brown 1988 (Journal of Experimental Psychology) on exposure to misspellings
  • Gillingham & Stillman, Simultaneous Oral Spelling (Bradford and Swindon help sheets)
  • Waikato randomized trial on Look-Say-Cover-Write-Check vs a phonological strategy
  • Klimovich & Richter 2025 on interleaving
  • Habgood & Ainsworth 2011 on intrinsic integration
  • Gentry on the test-study-test weekly cycle

Play Spell Jump · Back to Learning Lab