In the Netherlands, we have a problem. A quarter of the pupils who leave school are functionally illiterate and children from deprived areas are disproportionately represented. On the one hand, the problem is that children don’t learn to read well. They don’t learn to read technically well: to recognize letters, to learn the letter sounds, to combine letters with their sounds into syllables and words, and to automate all this in the end. But there is a second problem, namely: instruction in ‘reading comprehension’ mainly focuses on so-called analytical skills and not on text comprehension.
The standard approach to this instruction is the teaching of generic skills and strategies, such as ‘finding the main idea’ of a text or ‘recognising inferences’. Students then practice these skills on random texts with the aim of mastering these skills so that, theoretically, they’re able to understand complex texts. The assumption is that students do not need to acquire much content knowledge in order to achieve reading comprehension.
A fox is not caught twice in the same snare.…
This approach is so entrenched that it has persisted, despite the fact that it doesn’t yield any gains in reading test scores and research shows that it’s no good (Note: An incidental fact is that children experience reading not as fun but as something that requires using tricks to get good grades). In the face of stubbornly low scores, the recipe has often been to double it; and spend even more time learning these tricks.
A recent study (Grissmer et al., 2023) shows that building students’ general knowledge (i.e., their knowledge of the world/knowledge-rich curriculum) can lead to dramatic long-term improvements in reading comprehension. This study of more than 2,000 students showed that children who received a knowledge-rich curriculum in grades 1-4 not only performed significantly better than their peers who received ‘normal education’ on standardised reading tests but also that students from low-income families made such dramatic gains that their performance equalled that of children from higher-income families.
Natural experiment
Natural experiment
In the late 1940s, a British Medical Research Council doctor, Jeremy Morris, became convinced that the increasing incidence of heart attacks and coronary disease was related to levels of activity, not just age or chronic stress, as was almost the case at the time. universally thought. With Britain still recovering from the war, research funding was tight, so Morris had to come up with a cheap way to conduct an effective large-scale study. One day while commuting to work, it occurred to him that every double-decker bus in London was a perfect laboratory for his purposes, as each had a driver who spent his entire working life sitting down and a conductor who was constantly on his feet. In addition to moving sideways, conductors also climbed an average of 600 steps per shift. Morris could hardly have come up with two more ideal groups to compare. He followed thirty-five thousand drivers and conductors for two years and found that, after adjusting for all other variables, the drivers, however healthy, were twice as likely to have a heart attack as the conductors. It was the first time anyone had demonstrated a direct and measurable link between exercise and health (p. 209; Bryson, B. (2019). The body: A guide for occupants . Penguin.)
This brings us to the natural experiment that I want to discuss here. In this case, the experimental and the control groups are formed by a coincidental combination of circumstances, for example regulations or population changes (see examples above). For some reason, the state of Colorado in the US has an unusually high number of elementary schools that use a knowledge-rich curriculum. There were nine in the school district that was surveyed. Therefore, lots had to be drawn for which children in the school district could attend those nine schools. This shortfall enabled the researchers to compare a “treatment group” (the so-called lottery-enrolled children) with a “control group” (the so-called lottery-excluded children).
The 688 enrolled children entered received a knowledge-rich curriculum. The other (excluded) pupils attended ‘ordinary’ primary schools where ‘ordinary’ education was provided; in other words, no knowledge-rich curriculum was used there.
The researchers were patient. They waited four years for the children to reach 5th grade — the first year of state standardized testing — before measuring the results. They continued to look at test scores for those drawn and drawn students, through 8th grade.
And…?
They found that the enrolled students made significant progress on reading tests in each grade compared to the excluded students. To show just how great, the measured progress of the enrolled students was large enough that when translated to American students as a whole, the US would move up to a position in the top five countries on the PIRLS ranking. At that time (2016 PIRLS), the US was ranked 15th out of 58 participating countries.
If this was the only result, it would be great, but that’s not all. A breakdown of the results by income level led to additional insights. Eight of the nine schools were in middle to high-income areas. Although these children probably acquired a considerable amount of (academic) knowledge at home, they still benefited from the knowledge-rich curriculum. The effect size – which measures the difference between their performance and their respective control groups – was medium [1](0.445). But the effect size for students from low-income families was spectacular (1,299!). And the students who received a knowledge-rich curriculum also made a big gain in their math scores on the standardized test that students in group 7 received. In fact, their gains were so great that, according to the researchers, they erased the differences between students from low- and high-income families!
Imagine what this would mean if the Netherlands and all other countries that drill analytical skills made the transition to a knowledge-rich curriculum!
Grissmer, D., Buddin, R., Berends, M., Willingham, D., DeCoster, J., Duran, C., Hulleman, C., Murrah, W., & Evans, T. (2023). A kindergarten lottery evaluation of core knowledge charter schools: should building general knowledge have a central role in educational and social science research and policy? . (EdWorking Paper: 23-755). Retrieved from Annenberg Institute at Brown University: https://doi.org/10.26300/nsbq-hb21
Inspectorate of Education (2020). The State of Education . State publishing house, https://www.onderwijsinspectie.nl/documenten/rapporten/2020/04/22/staat-van-het-onderwijs-2020
For a fun blog about knowledge-rich curricula, check out Tom Sherrington’s blog “What is a knowledge-rich curriculum? Principle and Practice.” https://teacherhead.com/2018/06/06/what-is-a-knowledge-rich-curriculum-principle-and-practice/
[1] A Cohen’s d value of 0 indicates that there is no difference between the groups, while a Cohen’s d value greater than 0 indicates that there is a difference, and the larger the value, the greater the difference. In general, a Cohen’s d-value of about 0.2 is considered small, 0.5 medium, and 0.8 large, although this interpretation may vary depending on the field of study and research area.
So, if you read in a text that the effect size is, for example, “medium” or “spectacular”, this refers to how large the difference or effect is between the groups or conditions examined in the study. A larger effect size indicates a significant difference, while a smaller effect size may indicate a less striking difference between the groups.
3-Star learning experiences 
2 thoughts on “Duh!!! Knowledge can increase text comprehension and close knowledge and learning gaps”