Whether I receive a difference in volunteers' performance if I put chunk (pre learned part of sequence) at the beginning of the sequence, in the middle of the sequence or at its end?
I quite agree with Joachim Pimiskern’s comments so far. What’s unclear to me, however, are the task and setting in which your subjects are going to learn and then reproduce their sequence knowledge (i.e. explicit vs. implicit learning; random or regular order of sequence elements other than those in the pre-learned “chunk”; semantic meaning attachable to substrings of sequence elements (as in words; see Joachim’s last comment), or not; etc.).
Considering for instance implicit learning (as e.g. repeatedly performing a serial reaction time task on the key board following a randomly series of key-specific visual stimuli presented on a computer screen), then over time any subsequence of elements bearing some regularity eventually will yield shorter reaction times (and less errors) than all the rest – at the end of (sufficient!) acquisition, as well as at retention. And this will be more or less independent of that particular subsequence’s positon in a particular practice/retention run of, e.g., 30 seconds or so. I’d suppose you get pretty much the same results if you incorporate any pre-learned (and thereby “regular”) subsequence. Only you wouldn’t need that much practice to reach the well known “implicit learning effect” as usual.
Dear Klaus, thank You for long and detailed comment.
In my experiment my subjects had to learn their chunks of hand movements explicitly. Other elements of movement sequence are random.
But I collect different views and different thoughts about "heterogeneous" (i.e. consisting of chunks and other elements) sequence. Because there are huge literature about recency and primacy effects in "homogeneous" sequences, but not in "heterogeneous" sequences.