The decline in Australia’s performance in the Programme of International Student Assessment (PISA) since 2000 has been of considerable concern to policy makers. While a number of different possible causes have been identified, a study which attempted to evaluate several of these identified a significant factor which had not previously been raised—a decline in the number of books in Australian students’ homes.
Students sitting the PISA tests are asked about how many books there are in the family home, with the results showing a strong correlation between higher test scores across all three domains of reading, mathematics and science, and the number of books (see Figure 1 below).
Figure 1: PISA reading grade performance compared to the number of books in the family home, 2000
Source: MDR Evans, J Kelley and K Sikora, ‘Scholarly Culture and Academic Performance in 42 Nations’, Social Forces, May 2014, p. 13.
So what does this mean for Australia’s PISA results?
Figure 2 confirms the relationship between having books in the home and higher scores in the PISA reading assessment for Australia. However it also shows that, until 2012, there had been very little or no decline in the reading scores of those who came from homes with more than 25 books in them, while there was a substantial decline in the performance of those with less than 10 books in the house. This suggests that much of the 13 point decline in Australia’s PISA reading score from 2003 to 2012 was due to the 31 point decline in the performance of those with less than 10 books in the house. The results for all groups were lower in 2015 than 2012.
Figure 2: PISA reading scores by the number of books in the student’s house, Australia, 2003 to 2015
Source: Parliamentary Library analysis using the PISA Data Explorer
At the same time the proportion of students with less than 25 books in the home has increased from 14 per cent in 2003 to 24 per cent in 2015, while the proportion with over 200 books in the home has declined by a similar amount. This means that if book ownership had been maintained at 2003 levels and the 2015 scores for each category remained the same, the decline in the average PISA reading score would potentially have been halved.
Interestingly, when analysed by SES, the decline in PISA reading scores is fairly consistent across all groups, suggesting that book ownership and SES are independently associated with achievement, as was found in the study by Evans et al.
Figure 3: PISA reading literacy results over time by socioeconomic background
Source: S Thomson, L De Bortoli and C Underwood, PISA 2015:Reporting Australia’s results, Australian Council for Education Research, 2017, p. 134.
There are a number of possible explanations for why having more books at home is associated with better scholastic outcomes. It may indicate parents who value books and model reading behaviours that influence the student, and may have been more likely to read to them when younger—a very significant factor in literacy development. Better readers and more academically inclined students may acquire more books themselves.
On the other hand, increasing the number of books in the home appears to have a greater impact than having parents who provide support and encouragement for educational achievement, and there is a relatively low correlation between these factors.
And electronic devises don’t appear to have the same effect, with the reading scores of those who say they use an ereader actually lower than those who don’t, while having multiple tablet computers in the house did not seem to have any impact.
There is evidence that giving books to students who don’t have them at home does directly increase reading, which in turn has a positive impact on reading test scores. And it appears that giving students books rather than lending them has a greater effect. Most of these studies have been done in the United States, although several charities in Australia supply books to disadvantaged students, and in particular, Indigenous children in remote communities, based on this work. A government-funded pilot project in 2001 was positively evaluated in 2006, but funding does not appear to have been extended beyond the initial investment. Overall, giving books to children from disadvantaged backgrounds appears to be a relatively low-cost intervention, with a potentially high return, but has not been widely adopted in Australia.
In summary, while reduced book ownership among Australian families does not entirely explain the decline in PISA reading scores, it would appear to be a factor. In addition, the achievement of those from low-book households has declined much more significantly that those of students with larger numbers of books at home, suggesting that the school system is not adequately compensating for the change in book ownership habits.