A comparison of embedded total task instruction in teaching behavioral chains to massed one-on-one instruction for students with intellectual disabilities: accessing general education settings and core academic content.

Jameson et al. (2012) compared two ways to teach long chains of school tasks to students with intellectual disabilities.

One group got the whole chain every trial while sitting with a teacher. The other group got the same chain broken into the regular class schedule.

The authors never say which group learned faster or kept the skill longer.

The paper only tells us the study happened. No scores, no graphs, no winner.

We cannot tell if embedded trials beat massed trials or vice versa.

Neuringer et al. (1968) showed that tiny timeouts cut errors when typical adults learn lab chains. Their work is the grandparent of all later chaining studies, including this one.

Wilkins et al. (2009) used chaining to teach children with autism to retell stories. They got clear gains, unlike the missing data here.

Kellett et al. (2015) found that computer concept maps helped students with learning disabilities in the same general-ed rooms. They give positive numbers, again showing what Matt et al. left out.

You cannot pick embedded or massed chaining from this paper alone. Read the full text or wait for a meta-analysis before you change your lesson plan. Use it as a reminder to collect your own data every time you test a new format.