The work of Mary Beth Bruder, Carl Dunst, and Carol Trivette on "learning opportunities" has documented that ordinary routines in a child's life are where learning naturally happens. And the work of Tom Weisner and Cindy Bernheimer has shown that everyday routines are the chunks of time that families construct to make their lives work, and that these ecocultural niches are the context of child and family development.
The discussion here is about substantial time. I've used 15 hours a week as an informal, unscientific criterion for deciding whether a caregiver should be included in a Routines-Based Interview (RBI). If the relationship between time spent and influence on child functioning is linear, those spending fewer than 15 hours a week can't be said to have no influence. The issue becomes one of efficiency. If the amount of influence is small, then it's not worth their time or ours to be interviewed. Families, however, can always have anyone they like at the RBI.
Who spends substantial time with the child? Almost always the custodial adults, such as parents. Custodial adults, therefore, are the parent figures in the child's life, so, from now on, I'll just call them "parents." Adults who are responsible for care of the child, perhaps while the parent is working, also spend substantial time with the child; these are child care providers or preschool teachers, for example.
What about influential people who don't spend the requisite amount of time, such as grandparents. They can give children enjoyable times, new opportunities, special moments, and so on--all of which can create memories, if the child is old enough. But early intervention should focus on those adults who are around during the everyday routines: They have enough teaching opportunities distributed through the day that can make a difference. Furthermore, they're teaching when, where, and with whom the child needs the skills. By teaching in the moment, the child is learning to respond to the naturally occurring "discriminative stimuli." When someone walks in the room, the child learns to say hi. That's a lot better than repeating hi in multiple trials in an early intervention session with a professional. Children can learn context-specific behaviors that reinforce the adults in charge of those contexts, and here I'm talking about the short-time caregivers, such as visiting grandparents or early intervention professionals. The child can learn not to touch the ornaments at Grandma's house or to take his seat in the therapy room. The issue now is one of functionality. How do these help with the child's functioning in everyday routines? If they generalize, fine, but young children, especially if they have developmental disabilities, do not transfer well. If the adult also "transfers," then you get a bit more bang for your buck. For example, if an early intervention visit focuses on the parent interacting with the child, perhaps practicing how to teach the child something during a routine, the parent can then use the same task directions and prompts during regular life. Therefore, because the adult generalized his or her strategies from the short-time session to the long-time real life, the child has a chance of learning the skill.