In this paper the challenge of promoting computational thinking for all by contextualized computing education is addressed. The two phases learning session we designed was implemented in the AlgoRythmics environment which includes ten algorithmic dance choreographies (and attached interactive computer animations). In addition to previous studies that had focused only on supporting students in assimilating the strategy of the studied algorithms, this study examined whether they are able to build on this knowledge by extracting from visualizations some algorithm efficiency related concepts too. To this end, the learning environment was complemented/supplemented with targeted questioning (with and without teacher guidance). Participants (first year undergraduate students) were grouped based on the number of years they had learned programming in high school (0, 1/2 or 4 years). We performed two analyses: (1) group-0 vs. group-1/2 and group-4 (N=181; self-paced setting: algorithm visualization complemented with teacher prepared questions); (2) group-0 (one-group pretest-posttest design, N=46; supplementary Socratic questioning). Research results revealed that the AlgoRythmics environment, if complemented/supplemented with tutorial questionasking, could be an effective instrument in introducing students with no prior knowledge in computing (group-0), even with deeper Computer Science concepts such as algorithm efficiency.