Study: Culture of learning begins at home
Science Daily reports on a collaborative study from the IEA and the International Study Center at Boston College's school of education using a huge international data set from TIMSS and PIRLS tests (some formatting added):
The study, titled "TIMSS and PIRLS 2011: Relationships Among Reading, Mathematics and Science Achievement at the Fourth Grade -- Implications for Early Learning," is the first report looking at the issue of cultural excellence -- what parents, schools, and students are doing to improve success in reading, math, and science. Researchers used data from 180,000 students, 170,000 parents, 14,000 teachers, and 6,000 principals who participated across 34 countries.
"The culture of educational excellence starts in the home," says BC's Dr. Ina Mullis, co-executive director of TIMSS and PIRLS and the study's co-author. "It follows with a school that has a focus on educational success by all the parties concerned -- the teachers, the administration, the parents, the students themselves. It continues into the classroom with a teacher that is holding student engagement. We know then we will have students in the end that have a higher achievement, a higher motivation, and actually I think have a higher probability of becoming life-long learners."
Chapters 1 and 2 in the full report, authored by University of Gothenburg researchers GradeJan-Eric Gustafsson, Kajsa Yang Hansen, and Monica Rosén, review the data and analyses from the assessments. Chapter 3 identifies characters of effective schools, while Chapter 4 focuses on the effects of parents.
The following diagram (page 209 in the printed report; page 217 in the PDF) shows all the statistically significant relationships found between parents and learning outcomes:
In summarizing the results, the authors write:
These direct effects represent effects of Parental Education that the path model cannot account for via mediating variables. It is obvious from the model, however, that Books is an important mediating variable, with a strong relationship (0.47) between Parental Education and Books, and a direct effect of Books on the achievement variables of 0.22, similar to that of Parental Education.
Parental Education also had an indirect effect via the sequence Books, Activity, and Ability to achievement. All links in this chain were fairly strong and this indirect effect agrees with the theoretical expectations and with findings in previous empirical research.
There was, however, a pattern of indirect effects of NumLitAct on the three achievement variables that went via Ability. There also was a negative direct effect of NumLitAct on Mathematics achievement. These results mean that homes which reported a stronger emphasis on literacy activities than on numeracy activities also reported a higher level of Ability, which in turn had a positive direct effect on achievement in all three domains. This is an interesting result, and one possible interpretation is that emphasis on literacy activities has a positive effect on development of both literacy and numeracy skills. A partially different interpretation is that numeracy skills at the beginning of primary school tend to involve both reading and writing, because expression of numeracy skills often requires use of literacy skills.
It may seem strange that there was a negative direct effect of NumLitAct on mathematics achievement. However, there was a positive indirect effect of NumLitAct on mathematics achievement, which was mediated via Ability. This positive indirect effect of NumLitAct on mathematics achievement thus partially balances out the negative direct effect of NumLitAct. Because there was no negative direct effect of NumLitAct on science or reading achievement, the net effect is that the emphasis on literacy activity will cause a profile of achievement with a relative strength in reading and science compared to mathematics.
There was significant variance of the effects of parents between countries (starting on page 212, with a detailed data table on page 213), which is not surprising given the cultural differences between the countries represented in the data:
For Hungary, Iran, Romania, Poland, and Botswana Parental Education had total effects which exceeded .40 in all three domains. The lowest impact of Parental Education was observed for Azerbaijan and Hong Kong SAR, whereffects were lower than .16 in all three achievement domains. Thus, there were considerable differences in the amount of relationship between Parental Education and achievement across countries, even though it also may be noted that for many countries effects were between .30 and .40. From the list of countries with high and low impact it is not possible to determine any simple and clear grouping of countries which may explain the differences. Among participants with high impact, some were East European countries. However, the Russian Federation was among the countries with lowest impact, thus the pattern is far from clear. Among East Asian countries there were both examples of countries with the highest impact (Singapore) and the lowest impact (Hong Kong SAR). Similarly among developing countries, there were examples of high impact of Parental Education (Botswana) and low impact (e.g., Morocco). These examples indicate that the amount of effect of Parental Education on educational achievement cannot be accounted for in simple terms.