Choose and perform one local change

The team first proposes a longer verbal explanation. MDPE.14 exposes this as a poor candidate: it moves burden to participants but does not change the simultaneous Music–Dance relation that failed. The team compares three responses: keep the showcase configuration, constrain musical changes during partner-cue intervals, or keep different configurations for showcase and social-practice use. It selects the split-use response and a bounded trial of the social-practice configuration. The decision preserves the showcase branch instead of calling it wrong for every use.

MDPE.15 now carries the selected response into actual Work. The live coder changes the social-practice configuration so that large accent changes do not coincide with the partner-cue interval, and the sound operator re-aims the monitor toward the social floor. In a second round, six of eight pairs preserve partner-responsive timing; two still depend on teacher cues. The configuration and sound relation changed, and the later observation is usable for the next decision. Because two changes occurred together and the pairs had already experienced the material, the team records causal contribution as not tested. It does not report that the intervention improved the practice as a general fact.