Perceived Ability and School Choices

We design and implement a field experiment that provides students from disadvantaged backgrounds with individualized feedback on their academic performance during the transition from middle to high school.

2014-02-15

2014-03-31

school choices (school preferences), enrollment, and graduation on time

participation in assignment mechanism, placement in school alternative, score in the admission exam, size of school rankings

Among a restricted universe of schools in disadvantaged neighborhoods, we randomize the treatment within 12 strata, as defined by the geographic location of the school and its academic performance in the national performance evaluation in 2012. In the treatment group (n=44 schools) we administered the mock exam and provide face-to-face feedback on performance. A second arm, the placebo group (n=46 schools), took the mock exam but did not get information about test results. Finally, we also included a pure control group (n=28 schools).

We impose two restrictions to select the experimental sample from the universe of potential COMIPEMS applicants. First, we focus on schools with a considerable mass of applicants in 2012 (more than 30). Second, we consider schools located in neighborhoods with high or very high poverty levels (according to the National Population Council in 2010). Students in these areas are less likely to have access to previous informative signals about their own academic potential in general, and about their performance in the COMIPEMS exam in particular. Indeed, data from the 2012 edition of the assignment system shows that, on average, 44 percentage of the applicants in schools located in more affluent neighborhoods took preparatory courses before submitting their school rankings but this figure drops to 12 percent among schools in high poverty areas. Among the applicants in our sample, 16 percent report previous exposure to a mock test of the admission exam with performance feedback.

Schools that comply with the criteria imposed are grouped into four geographic regions and terciles of school average performance amongst ninth graders in a national standardized test aimed at measuring academic achievement (ENLACE, 2012). Treatment assignment is randomized within strata at the school level. As a result, 44 schools are assigned to a treatment group in which we administer the mock exam and provide face-to-face feedback on performance, 46 schools are assigned to a "placebo"group in which we only administer the mock exam, without providing information about the test results, and 28 schools constitute a control group. Within each school in the experimental sample, we randomly pick one ninth grade classroom to participate in the experiment.

Randomization done in office by a computer

School

No

118

3,644

44 treatment schools, 46 placebo schools, 28 control schools