Forward chaining is a procedure in which the first sub-skill or step within a total skill is targeted for independent performance, while the steps that follow are prompted through.
When the first step in the chain is performed with independence and accuracy across learning trials, the second step is targeted for independent performance, and so on.
