To explore the effect of reduced ventilation on
student performance, observed ventilation rates (l/sperson)
were first categorized into four groups using
cut points of 2.25, 3.50 and 4.50 (25th, 50th and 75th
percentiles). Table 1 presents calculated means in math
and reading scores, suggesting an increasing trend with
increasing ventilation rate. Results from statistical
testing related to these results are reported by Shaughnessy
et al. (2005).
Using math test score as dependent and continuous
ventilation rate as independent variables, univariate
linear regression analysis resulted in a crude regression
coefficient of 1.30 (P ¼ 0.32) for the ventilation rate.
The regression coefficient was 2.27 (P ¼ 0.06), adjusted
for confounding factors (% free lunch, % limited
English, and % mobility rate). Using GAM functions