As a detailed analyses of the PISA 2009 report the OECD has published feb 2011: Quality Time for Students. Learning in and out of School, Paris, OECD Publishing

What are the ideal conditions to ensure that students use their learning time efficiently? What can schools do to maximise the learning that occurs during the limited amount of time students spend in class? In what kinds of lessons does learning time reap the most benefits? And how can this be determined?

The report draws on data from the 2006 cycle of the Programme of International Student Assessment (PISA) to describe differences across and within countries in how much time students spend studying different subjects, how much time they spend in different types of learning activities, how they allocate their learning time and how they perform academically.

It can be seen as a very interesting enhancement of the PISA data. It focuses on three aspects of learning time:

Learning time in regular lessons and performance
Learning time in out of school time lessons and performance with a school teacher and/or non school teacher (where the last tends to reinforce the impact of socio-economic background on performance).
Learning time in individual study and performance

Results to take notice of:

1. 'The basic premise of the report is that were Carrolls model of 1963 addressed the issue of enough and more learning time as a condition for performance. In the mean time, as Carroll in 1989 and other researchers have statued, learning time is necessary for learning, but not sufficient for acquiring knowledge. (p. 20), and depends on the way in which it is organised, the proportion of time dedicated to full engagement of students and other conditions'. In other words, the learner plays a fundamental role in determining how well learning time is used" says the OECD (p.20)
2. Across countries, findings show that students tend to perform better if a high percentage of their total learning time, including regular school lessons, out-of-school-time lessons and individual study, is dedicated to regular school lessons (p. 13) between 2 and 4 hours per week!
3. Where the report mention a minimum of 2 but maximum 4 hours the number they use for the data is then the medium of 3, p.28.
4. "This positive relationship between learning time in regular school lessons and performance is even more pronounced when the time students spend in regular school lessons is considered as a share of total learning time. Countries with low relative learning time in these lessons also share some educational system characteristics that are related to low overall performance: less human and material resources, less school autonomy, and fewer standardised external examinations of student performance" (p.14).
5. Within country relationship between performance and learning time in regular school lessons increases towards 2-4 hours per week, increases slightly and most for science towards max. 6 hours and decreases onwards 6 hours for mathematics and reading. "the curvilineair relationship can be also observed after adjusting for the socio-economic background of students and schools, and for students learning time in mathematics in out of school time lessons and individual study", but is best seen between 0 and 3 hours pro week (3, as the middle of 2-4).
6. "The general pattern observed in many countries indicates that students who spend long hours in mathematics in regular school lessons perform much lower than students who spend a moderate amount of hours. For example in Spain, Norway, Luxembourg, Germany, the Netherlands and Croatia students who spend 6 or more hours per week in regular school lessons in mathematics perform at least 44 points lower than students who spend from 4 to less than 6 hours per week in lessons" (p. 65). This is only different for Asian countries where there is still a linair relation above 6 hours between performance and hours of regular school lessons (like Korea, Chinese Taipei and Hong Kong China), p. 65.
7. "Reading performance increases steadily as learning time in regular school lessons increases up to six hours per week, but beyond six hours per week, the relationship becomes negative. For example, across OECD countries, students who spend two to less than four hours per week in lessons tend to perform 80 score points higher than students who do not spend any time learning the language of instruction in regular school lessons; students who spend four to less than six hours per week tend to perform 87 score points higher; and students who spend six or more hours per week in lessons tend to perform 60 score points higher. This relationship can also be observed when adjusting for the socio-economic background of students and schools, and for students’ learning time in the language of instruction in out-of-school-time lessons and individual study (Table 4.2c)." (p. 66).
8. OECD suggests that why more time in science doesn't lead to an decreasement of performance like in reading and mathematics is due to the fact that students do this for different reasons, science by choice of interest, mathematics and reading because they are obliged to do so for remedial purpose. In more detailed analyses of this hypotheses they find evidence for this reason. p. 66-67.
9. Learning time in out of school time lessons and performance with a school teacher can work quite well, but if these lessons are done by non school teachers it tends to reinforce the impact of socio-economic background on performance (p.74).
10. Students who spend more then 2 hours pro week but not more then 4 hours pro week tend to perform 41 points higher, whereas students who spend more then 4 hours a week tend to perform less higher. But spending individual study time is not necessarily related with better performance p. 75-76
11. "There are countries like Switzerland, Austria, Luxembourg, Czech Republic and Iceland where students who spend 6 hours or more per week in mathematics tend to perform 28 to 72 points lower then student who do not spend any time in individual study!...And in Switzerland, Czech Republic & Israel the same goes on for 24 to 40 points lower performance for students who spend 4 to 6 hours pro week in compare with students who do not spend any time in individual study!"(p.76).
12. The relations between a decreasement of performance when studying more then 4 yours pro week individual is moderate for science, higher for mathematics and highest for reading until 6 hours or more pro week. The performance increasement starts with 2 hours pro week (30 points higher), until 2-4 hours (41 resp. 36 and 33). On behalve of this report students shouldn't spend more then 2 hours pro week on individual study of reading!!, p. 76.
13. When it comes to comparison of relation between higher scores of students who don't spend any time in reading instruction in individual study towards students who perform 6 hours or more per week with 24 to 105 points lower the reports makes the following reservation because of the lack of data about characteristics of students and teachers to relate with: "there is no evidence to conclude that the efficiency of learning time diminishes after a certain amount of time dedicated to individual study". But the opposite can also be considered!!

Huge lessons can be learned from this sub analyses of the PISA results of 2009, while already different nation states have proposed further measures to invest in more time for reading and mathematics.