In the usual life, people have theories about the world around them and themselves. They have a “meaning system”, a whole of beliefs about fields as varied as work, school, sport or relationships through which they understand the world (Molden & Dweck, 2006). Today, several studies have manipulated these implicit theories in order to show the causal role that they play in expectations about the future, motivation and performance, particularly how self-theories (implicit theories about self) can be influential in the academic domain. Dweck and her colleagues (Dweck, 1999; Dweck & Leggett, 1988; Elliott & Dweck, 1988) have proposed that people have different “theories” about intelligence: Those who focus on stability and believe that intelligence is fixed (entity theorists) and those who focus on malleability and believe that intelligence is flexible (incremental theorists). Many results suggest that people's achievement goals often follow their implicit theories of intelligence (Dweck, 1999). Generally, "entity theorists" worry about their level of intelligence and pursue performance goals whereas "incremental theorists" try to attain learning goals. In comparison to entity theorist, incremental theorists "have been found (a) to focus more on learning goals versus performance goals [...] (b) to believe in the utility of effort versus the futility of effort given difficulty or low ability [...] (c) to make low-effort, mastery-oriented versus low-ability, helpless attributions for failure ; and (d) to display mastery-oriented strategies (effort escalation or strategy change) versus helpless strategies (effort withdrawal or strategy perseveration) in the face of setbacks" (Blackwell, Trzesniewski, & Dweck, 2007, p. 247). Thus, these two ways of thinking about intelligence can have important consequences for students in an achievement context. In this line, Muller and Dweck (1998) have supposed that praises for intelligence given by teachers or parents may influence children's beliefs about intelligence (i.e. their theories of intelligence) and may affect their motivation and performance. In order to test these hypotheses, they realized several studies with American fifth graders. Results showed that stable praises for intelligence (to be smart) had more negative consequences for students' achievement motivation than praises for effort (to work hard). Do children from other countries, confronted with different educational systems, react to praises similarly?

In line with this idea, we examined the effects of praise through two studies conducted with French fifth graders (aged 10 to 11). First, pupils were asked to realize a first easy set of exercises (Raven’s matrices) and received a positive feedback (intelligence praise versus effort praise) or no praise (control). Then, their achievement goals were assessed. Second, pupils were asked to realize a very difficult set of exercises. Subsequently, they received a failure feedback, and pupils’ motivation and attributions were measured. Third, they were asked to realize a new easy set of exercises. In accordance with Muller & Dweck (1998) we expect that children who received effort praise should have more learning goals, higher performance and higher motivation than pupils who received intelligence praise. In addition, we predict that children with effort praise should attribute the failure more to a lack of effort than the other groups.

Generally, we didn't replicate the pattern of results obtained by Mueller & Dweck (1998) on French children. We found no significant differences, whatever the parameters, between the groups of children (intelligence praise versus effort praise versus no feedback). However in our first study we differentiated two groups of children from the causal explanations for failure: Those who attributed their failure to lack of effort and those who attributed their failure to lack of ability. Results revealed an interaction between the attributionnal preference and praise. Children had better performances when they produced effort explanations rather than ability explanations, but this effect only appeared for children who received praise, especially children praised for ability (i.e. smart). Concerning our second study, it appeared that effort praise conducted children to less stable causal attributions than praise for intelligence or no praise. Why such results: Experimental artefact or intercultural effect? We will discuss it.