.

Is there a global Enem score?
No, Inep does not calculate a global performance average; it only presents the averages separately.

The Enem exam has five scores: one for each area of knowledge assessed – Natural Sciences, Human Sciences, Languages, and Mathematics – plus the average for the essay. The calculation of the averages in each of the four areas used the Item Response Theory (IRT) methodology, which seeks to measure knowledge based on observed behavior in tests. In the case of the essay, the criteria are the same as the traditional Enem.

For the distribution of spots in the Unified Selection System (Sisu) of the Ministry of Education, institutions will use the set of Enem scores following specific aggregation and weighting criteria.

How the Enem score is calculated using IRT
Unlike a standard test, the Enem score in each area does not simply represent the proportion of questions the student got right. In each of the four areas assessed, the average obtained depends not only on the number of questions answered correctly but also on the difficulty of the questions missed and answered correctly, and the consistency of the answers. Therefore, people who get the same absolute number of items correct may obtain different performance averages.

What the score represents
On the scale built for Enem, within each of the assessed areas, the score 500 represents the average obtained by high school graduates who took the exam (excluding those who have already graduated and those taking it for practice). Therefore, the further the student's score is from 500, upwards, the higher the performance obtained in relation to the participants' average. The same reasoning applies to performance lower than 500, which indicates worse performance in relation to the average.

Scale
The limits of the scale, within each area, vary according to the difficulty level of the exam questions and the students' behavior on each question. Therefore, the minimum and maximum for each assessed area are not pre-fixed.

In the Natural Sciences and its Technologies exam, the IRT analysis showed that the lowest observed proficiency average was 263.3. This number represents the beginning of the scale for this area, that is, the lowest level of proficiency measurable by the exam questions. The highest proficiency was 903.2.

For Human Sciences and its Technologies, scores vary between 300.0 and 887.0.

For the area of Languages, Codes and their Technologies, the averages are between 224.3 and 835.6.

In the case of Mathematics and its Technologies, scores range from 345.9 to 985.1.


Enem 2009 – Lowest and highest observed IRT score

Enem 2009 – Distribution of participants across performance ranges, by area

Enem 2009 – Average observed in the Essay: 601.5

*With changes at 19:40 on 01/28/10

Inep/MEC Press Office