Using panel data from the last five PISA cycles and data on various institutional variables, the role of ICT resources, parents' occupation, school characteristics and institutions in student attainment was explored. The data were collected from 2.5 million students in 94 countries/economies representing more than 140 million students worldwide. Fixed-effects (FE) models with robust standard errors were constructed and analyzed. The main findings included (i) students with mothers that attained a higher occupational status are more successful in all subjects but the same predisposition does not hold true for fathers; (ii) an effective government does not necessarily mean an effective education system; (iii) "relax" environment in politically and economically freer countries may hinder academic success; (iv) internet connections at schools may not be used properly and/or for educational purposes; (v) students in overcrowded classes perform worse than those in lightly populated ones; and (vi) use of ICT resources at schools contributes to success in mathematics. The urgent need for uniform and better micro-level data on education in terms of quality, quantity and frequency is highlighted.