Sanz (2003). It is sensible to think that the influence of corrective feedback treatment will rely on the selection of the aimed structure (Mackey, 2007). Pienemann (1998), for instance, offers that the order in which grammatical structures are acquired as implicit knowledge include the processing implementations. Many of the aimed structures studied up to now included intricate processing implementations that some of the students may not have been developmentally ready for. The structures also varied considerably in how easy they were to acquire explicit knowledge. For instance, Spanish noun- adjective agreement (Leeman, 2003) constitutes a rather easy rule to understand whereas dative alternation is much more difficult. In line with what was said, Mackey (2007) reports that, there is a clear need to regard carefully the choice of target structure, being aware of both the developmental stage of the students and the conceptual complexity of the structure selected. In general, explicit instruction is more influential with simple rules. Nevertheless, this may just demonstrate the fact that the testing instruments only presented measures of explicit knowledge (Mackey, 2007). For intricate rules the picture is mixed. In three studies In two studies (Scott, 1989; de Graaff, 1997) the explicit instruction was more influential for acquiring complex structures. 2.3. Empirical Background A well-known study on the relationship between corrective feedback and learner uptake is by Lyster and Ranta (1997), who studied second language learners in immersion classrooms in Canada. They examined six corrective feedback types in terms of their frequency and distribution, as well as their effects on learner uptake. Lyster and Ranta (1997) discovered that the teachers had a strong tendency (55% of all occurrences) to use recasts as the strategy for corrective feedback (Lyster&Ranta 1997, p.53), even though it was the least likely strategy to elicit student-generated repair (only 31% of all occurrences) (Lyster&Ranta 1997, p. 54). They concluded that of the six feedback types, elicitation, repetition, clarification requests and metalinguistic feedback were the more successful in evoking student generated feedback (Lyster&Ranta 1997, p.56). In Lyster and Ranta’s data, the teachers provided corrective feedback on 62% of the students’ erroneous utterances on average, and the researchers did conclude that more fr>