In the international educational context, the negative impact of environmental noise in scholar’s achievement has been studied for the past forty years. This is consistent with world concern upon global environmental change where noise pollution is the main factor that most affects quality of life as highlighted by the World Health Organization (2004).

In a compilation, Evans (2006) states that children exposed to noise showed significant lags in reading skills; he also reports that several studies with kids found links between chronic noise exposure and diminished attention. Considering empiric evidence reported in specialized literature, the present comparative study aims to compare three different schools with differential noise exposure. Study design tested the physical independent variable of environmental noise and attributive independent variables of participant’s sex and age. The psychological variables (dependent) were evaluated with two second-order factors: a) the emotional impact conformed by the annoyance, the interference in the communication with their teachers and with their peers and b) the cognitive impact conformed the speech intelligibility (understanding of the spoken words) and the attention. In addition, the educative variable of reading comprehension was assessed.

The sample was not probabilistic and included 521 elementary school students: 49.3% female and 50.7% male, age 8 to 13 years old, with an average of 10 years. Two types of instruments were used, psychometric scales to assess psychological attributes and a standardized sound level meter to measure environmental noise inside the classrooms. The study was conducted in three public elementary schools that because of their geographical location are exposed to different intensities of environmental noise. The noisiest school recorded an average of 63 decibels, the one with middle noise reached an average of 59 and the one with lower noise averaged 54 decibels. In order to know the simultaneous relationship between acoustic and psychological variables the statistical method of structural equation modeling system was applied.

A Multiple Indicators and Multiple Causes (MIMIC) model was used to compare schools. For the observed data, the tested model has an acceptable adjustment, as indicated by the tests of goodness of fit: x2 = 29.34, p = 0.01, CFI = 0.99, RMSEA = 0.05. Since the model have a significance level of 0.01 an additional criteria was required, the rescaled chi-squared (x 2/df) statistic was also reported and it should be near two (29.34/14). The quotient is a small value and therefore the model is considered acceptable. Finally, the importance of this study relies on the facts that the tested second-order model is supported on theoretical assumptions and that this model fully explains how the noise in the classrooms of the evaluated schools impacts differentially and negatively the psychological and educational processes of students.