In this paper we estimate the effect that informative policies about the benefits of a good performance of the Saber Pro test can have on the results of the test. It used a sample of 537 students of the Faculty of Economics and the School of Administration of the Universidad del Rosario, who presented the exam in the years 2017 and 2018. The results show that both being chosen to receive the treatment and reading it do not have any impact at a global level. However, when analyzing by components it was found that the effect of being chosen to receive treatment 1 generates a negative impact on the test of -0.152 standard deviations and the effect of reading the message since it was chosen to receive this treatment generates an effect negative of -0.181 standard deviations.
