Skip to main content

General Strategies to Support Your Child’s Speech Development

Modelling and Recasting

Modelling means to provide a ‘model’ or example to a child on how to communicate. A model can involve demonstrating an example to your child of how a specific speech sound, word or sentence should be produced. 

Recasting involves hearing your child produce a phrase containing an error, then repeating it back to them using the correct speech sounds, words or grammar. This is best done naturally, around what the child is interested in at the time. The child does not have to repeat your model. It is important for the child to simply hear the correct model. 

Providing a ‘good’ model all the time requires you to use:

  • natural speech rate (avoid speaking too fast or too slow) 
  • clear speech
  • short sentences
  • appropriate vocabulary
  • more repetition and restating than you would use with an adult 

Provide ‘good’ listening conditions by:

  • reducing background noise
  • getting face to face
  • participating in shared interaction
  • demonstrating shared attention and interest 

Ineffective Modelling

There are both helpful and unhelpful ways to model the correct production of speech sounds to your child. Ineffective modelling occurs when a parent tries to correct a child’s speech sound difficulties but does not provide a model for the child to learn from. Examples of these are

Pointless Modelling

Child: “That’s a tat”

Adult: “Not a tat, a cat”

In this example the parent has cancelled themselves out as the child has heard both a correct and an incorrect model of how to say the word ‘cat’.

Exaggerated Modelling

Child: “I see a bue bird?”

Adult: “You mean buh-lue. I see a buh-lue bird.”

This sort of exaggerated modelling is not effective as it distorts the sound so that the child doesn’t hear the word correctly. When you model it’s important to say the word clearly without over exaggerating the sounds. You can give the target word a little extra emphasis.

Modelling with No Model

Child: “He hurt his binger.”

Adult: “Hurt his binger? What are you supposed to say?”

In this example no speech model has been provided. The child hears the incorrect sound once, but no example of the correct production has been given for them to learn from.

Effective Modelling

Effective modelling for speech sound errors requires a parent to repeat the correct production of the word back to their child during natural conversation. The effectiveness of the model depends on how many times you can repeat the correct word back to your child. The more times the better!

Weak modelling

Child: “That’s a tat”

Adult: “It is. It is a cat.”

It comes naturally to many parents to model this way. While this is okay for a typical speech sound learner it is just not ‘powerful’ enough for children with speech sound difficulties. They need MORE repetitions.

Strong modelling

Child: “I like the bid bird”

Adult: “I like the big bird too. It’s really big. A big bird. What do big birds like to do?”

This model is most effective as it provides lots of opportunities for the child to hear the correct production. Using modelling and recasting when your child says a word incorrectly during general conversation provides a correction:

  • without any criticism, judgment, or testing
  • without disrupting the ‘flow’ of conversation
  • without getting in the way of listening

Frequency: Remember to Model or Recast Again and Again!

After modelling or recasting a speech sound for your child, revisit it a few more times that day. Below is an example of how to do this. Keep in mind - your child will learn best from regular repetition. Frequency of modelling:

  • for the same sound or sound pattern
  • aim for 12 to 18 correct models per minute
  • 3 to 4 minutes a day

An example: If your child has trouble with the ‘k’ sound at the start of words, you might start by modelling “cup” 15 times during breakfast after your child has said “tup”. Later, in the day your child might be playing with a toy car so you can use this opportunity to model “car” repeatedly, as it also starts with the ‘k’ sound. Try to model target sounds as frequently as possible during everyday interactions with your child. You do not need to recast constantly throughout the day, but ensure you are modelling the correct productions more often than you usually would.


Modelling and recasting are effective strategies for encouraging correct production of speech sounds in children; however, speech and language development is a gradual process. Expect your child’s progress to take time and practice.

Adapted from: Delivering Feedback - Modelling and Recasting. Caroline Bowen, 2011

More strategies

Auditory bombardment

Find lots of objects in the environment that begin with the target sound and highlight them to your child.

Focused Stimulation

Say a target word 10-20 times in a very short amount of time (i.e. one minute).


Provide occasional feedback throughout the day regarding your child’s production of target sounds.

Fixed Up Ones

Occasionally make a speech error yourself and then “fix it up”. e.g. “That was lots of bun…oops, silly me! I meant FUN”. This will teach your child about ‘fixing up errors’ which is an important skill to support change in their speech.

Last updated 15 Dec 2023