Is Multiple-Choice Spelling Practice Good for Kids? What the Research Says (and How We Use It)

We build spelling practice for children, and one of the modes in SpellCrush, Spot the Spelling, asks a child to hear a word and pick the right spelling from three or four. Before we built it we read the research, because the obvious objection is a good one: every question shows a child at least two wrong spellings. This article sets out what we found, as plainly as we can, including the parts that cut against us.

Two classic studies with adults found that seeing a misspelling can leave a trace.

In one, 220 undergraduates either produced or just saw misspellings between two spelling tests, and that experience was found to be harmful to how they spelled afterwards. The author concluded that teaching or testing spelling with wrong spellings is a questionable method (Brown, 1988). In another, adults who read correctly spelled words were more likely to spell those words correctly later, and adults who read misspelled words were more likely to misspell them (Jacoby & Hollingshead, 1990). The title of that second paper is a warning to anyone who marks a lot of student essays.

That is the heart of the concern, and we took it seriously. It is why the rest of this article, and the quiz itself, treat wrong spellings as something to keep on screen for as short a time as we can.

Those two studies were with adults. A later study used a similar method with children whose average age was 10. The pattern was different. The children showed a clear benefit after seeing a correct spelling, and no measurable effect, good or bad, after seeing a misspelling. The children also did better when they had processed the word more deeply at the moment they saw it. The researchers suggested that children use explicit, deliberate processing of what they have just seen, where adults' spelling is nudged by an automatic, implicit effect (Dixon & Kaminska, 2007).

What it does give us is a useful lever: the thing that seemed to matter was how much the child worked with the word. So we make sure that a child who picks a wrong spelling still has to look at the right one and then type it. We follow the cautious reading of the evidence, and design as if the adult risk applied to children too.

Being tested helps memory, and this holds for multiple-choice tests too. In one set of experiments, taking a multiple-choice test gave a large benefit on a later test of the same facts. But the same work found a cost: the more wrong options people had read, the smaller that benefit became, and the more often they later gave one of those wrong options as an answer (Roediger & Marsh, 2005).

A follow-up tested a simple remedy. Students took a multiple-choice test with immediate feedback, delayed feedback or none. Both kinds of feedback increased correct answers and reduced repeats of the wrong options on a later test (Butler & Roediger, 2008). In other words, when the right answer follows every question, most of the downside shrinks.

A third group of researchers asked why multiple choice is often dismissed as "just recognising". They found that it depends on the wrong options: when they are believable, multiple-choice questions can prompt people to retrieve what they know rather than simply recognise the answer (Little et al., 2012). With obviously silly options, a child could spot the answer without knowing it, so we avoid those.

Real spelling means producing a word from memory, not spotting it. A set of experiments with 1st and 2nd graders compared "rainbow writing", where children copy a word in several colours, with retrieval practice, where they try to write the word from memory. The children who practised retrieval spelled better, both the next day and five weeks later, and said they liked it at least as much (Jones et al., 2016).

Choosing a spelling from a list is easier than producing one. That is exactly why it is useful for some children: a five-year-old who is not ready to type a word can still show that they know how it looks, and a child who freezes at a blank box can start gently. But it is a stepping stone. So in SpellCrush a Spot the Spelling answer never counts towards a word being mastered and never schedules a word to come back for review. Mastery comes from typed answers in the regular Practice mode.

Here is each finding next to the design choice it led to. The right-hand column describes the quiz as it works today. The left-hand column says no more than the studies did.

1. Adults who saw misspellings later spelled worse; seeing the correct spelling helped. (Brown, 1988) (Jacoby & Hollingshead, 1990)

   So: Wrong spellings fade out within about half a second of the tap, and are struck through. The last and biggest thing on screen is the correct spelling, in green. After a wrong pick it is read aloud again. The results list correct spellings only, never the wrong ones.

2. Children averaging 10 saw a benefit from correct spellings and no measurable harm from misspellings; how deeply they worked with the word mattered. (Dixon & Kaminska, 2007)

   So: A wrong pick isn't just a miss. The child sees the right spelling while a voice spells it out letter by letter, it is hidden after a few seconds, and they type it from memory (look, cover, write). That step earns nothing; it is there so the child works with the correct form.

3. Reading more wrong options weakened the benefit of multiple-choice testing and led people to repeat those wrong options later. (Roediger & Marsh, 2005)

   So: In the app, children aged 7 and under get three choices (two wrong), not four. On the public quiz the number of choices is a setting; it starts at four, and the page suggests three for the youngest.

4. Feedback, immediate or delayed, raised later correct answers and cut later repeats of wrong options. (Butler & Roediger, 2008)

   So: Feedback after every question, always. There is no mode that saves the answers for the end.

5. Multiple-choice questions with believable wrong options made people retrieve what they knew, not just recognise the answer. (Little et al., 2012)

   So: Wrong options are believable mistakes (a dropped letter, a swapped pair, the wrong vowel), not random strings. As our own rule, never a real word, so a child can't pick a word that exists, such as a homophone, and be marked wrong.

6. 1st and 2nd graders spelled better after producing words from memory than after copying them, a day later and five weeks later. (Jones et al., 2016)

   So: Spot the Spelling never counts towards mastery and never schedules a word for review. In the app it earns a little XP (about half what a typed word earns), no stars, and counts as practice for the day. The public game's stars only reward finishing a round; they never touch a word's progress. Typing from memory stays the main event.

The quiz comes in two forms, and they share the same question: the word is read aloud, then its definition, and a wrong pick leads to the look, cover, write step. They differ in how they are played. Spot the Spelling is a practice mode inside SpellCrush. It looks like the rest of practice, with the word's picture and definition on screen, none of the game's bells and fanfares, and no timer. It goes through every word in the lists you have assigned, unmastered words first, and with no list it keeps going with standard words. It pauses for a breather after every ten words, and a list run picks up where it stopped. It is free on the standard word library, and running it on a list you have assigned follows the same Premium rule as our other modes. The public Spelling Quiz is the game version, also in our spelling games. It needs no account for five quizzes a day and gives ten words at a time from grade and year sets, with stars for each round. Premium members can also quiz on a saved word list, or type up to 20 words of their own.

• None of the seven studies tested a multiple-choice spelling quiz with young children. We are combining findings from neighbouring questions, and that is a judgement, not a proof.
• The one study with children found no harm from misspellings, but it was one study and one age group.
• We haven't measured whether our quiz improves spelling. We have designed it around the evidence above, and we think that is the sensible way to start.

If your child ever seems to be copying a wrong spelling they saw, or the quiz feels like it is doing more harm than good, stop and go back to typing practice. You know your child best.