Summary of results from multiple linear regression models.

a)Independent variables included exposure status(0 = school A, 1 = school B), age(years), years of education(years), familiarity with computer games (0 = little to none, 1 = much), sex(0 = male, 1 = female), length of residency(years), body mass index(BMI), second-hand smoking (0 = no,1 = yes), sleep status(0 = bad, 1 = not bad, 2 = good), household fuel (0 = coal, 1 = liquified petroleum gas, 2 = electricity, 3 = others), high-tension lines surrounding residence (0 = no,1 = yes) and frequent use of household appliances(0 = no,1 = yes).b)4 different multiple linear regression models were done considering NAI scores of VRT, DSB, VSRT and PAT as dependent variables. Backward selection for independent variable was used in the model with the entry significance level set to 0.05 and a removal level of 0.10.#Independent variable reserved in the model was with a p-value *Independent variable reserved in the model was with a p-value Beta = standardized coefficients; VRT = visual retention test; DSB = digit symbol; VSRT = visual simple reaction time; PAT = pursuit aiming test.