*3.6. Airtightness*

The average value of the infiltration rate at 50 Pa (n50) was 6.97 h<sup>−</sup>1, with a standard deviation of 2.06 h<sup>−</sup>1. Models with the lowest n50 values were those in the C3 climate zone, where the lowest average temperature values are recorded in the winter. The values of n50 ranged from 10 h−<sup>1</sup> (maximum) to 2.6 h−<sup>1</sup> (minimum), both recorded in the B4 climate zone (Table 10).


**Table 10.** Average values and standard deviation of n50.

### **4. Discussion**

This section is focused on the analysis of the relationships between the symptoms described by occupants and the rest of the parameters under study (physical, building operation, and votes).

### *4.1. Relationship between Physical Parameters and Classroom Operation*

It could be assumed that manual opening of windows in naturally ventilated buildings should depend on outdoor conditions, as this is the main element of control. However, it was observed that, despite the fact that in midseason windows remain open longer than in winter, no clear linear trend can be observed. In winter, the need for ventilation or indoor air changes is considered more important than the need to control the entry of outdoor cold air. Analysis by categories of the opening of windows (Figure 12) showed this occurs mostly in mid temperatures, although it was also observed in cooler conditions when necessary. Furthermore, no progressive growth was observed with the increased temperature, as could be expected. In midseason, it is more common to open windows, although there was no clear correlation with temperature, some of which was similar to winter, where more windows are opened in comparison.

**Figure 12.** Cross-tabulation for windows' opening and outdoor air temperature. (**a**) Cross-tabulation for windows opening and outdoor air temperature in winter: windows closed (gray) and open (blue). (**b**) Cross-tabulation for windows opening and outdoor air temperature in mid-season: windows closed (gray) and open (blue).

It could be deduced that users are psychologically or culturally conditioned to some extent as to how and when they open windows. Although it would be preferable for the classroom windows to remain open, the act of opening was seen as a reaction to poorer indoor air quality, which was more noticeable for the same thermal conditions in spring. It, therefore, appears that there is an adaptation process.

As it can be seen in Figure 13, although there was statistical significance between the CO2 concentration and the outdoor-indoor air temperature differential (*p*-value < 0.05), the correlation was somehow weak and more clear in winter time (R<sup>2</sup> = 0.249) than in midseason (R*<sup>2</sup>* = 0.145), with a better fit to a y-reciprocal relation. However, the predictive mathematical model lacked enough accuracy to be of utility to forecast actual situations. Besides the wide dispersion on values, there was a trend in the worsening of the indoor environment as DT increased, as can be usually expected, due to the lack of a controlled ventilation system.

**Figure 13.** Fitted regression model plot for CO2 indoor concentration related to indoor-outdoor air temperature differential, closed windows (red), and windows open (blue).

Is noteworthy to highlight that moderate DT winter and midseason trends were very similar, which matched with the foreseen windows' operation patterns, when most of the apertures occurred around cold-mild external temperatures.
