New research on dual-language immersion programmes

A new research brief by Jennifer L Steele and colleagues, published by the RAND Corporation, presents new research on dual-language immersion (DLI) programmes. These programmes provide both native English speakers and children learning English as an additional language (EAL) with general academic teaching in two languages from kindergarten (Year 1) onwards.

In partnership with the American Councils on International Education and the Portland Public Schools in Oregon (PPS), the authors conducted a random-assignment study of DLI education. The goal was to estimate the causal effects of the district’s DLI programmes on pupil performance over time in reading, mathematics and science, and on EAL pupils’ reclassification as English proficient.

PBS allocates immersion slots using a random-assignment lottery process for those who apply to the programmes. The study focused on 1,625 DLI lottery applicants in the kindergarten cohorts from 2004–2005 to 2010–2011. Pupil achievement was tracked until 2013–2014.

Key findings of the study were as follows:

  • PPS pupils randomly assigned to dual-language immersion programmes outperformed their peers on state reading tests by 13% of a standard deviation in grade 5 (Year 6) and by 22% of a standard deviation in grade 8 (Year 9).
  • Immersion-assigned pupils did not show statistically significant benefits or deficits in terms of mathematics or science performance.
  • There were no clear differences in the effects of dual-language immersion according to pupils’ native language.
  • EAL pupils assigned to dual-language immersion were more likely than their peers to be classified as English proficient by grade 6 (Year 7). This effect was mostly attributed to EAL pupils whose native language was the same as one of the two languages taught.

Source: Dual-language immersion programs raise student achievement in English (2017), RAND Corporation Research Brief, RB-9903

Poor literacy skills hold poorer pupils back in science

A report, published by the Education Endowment Foundation and the Royal Society, has reviewed existing studies to identify interventions and teaching approaches that have a positive impact on pupil learning in science, particularly for pupils from disadvantaged backgrounds.

The researchers from the University of Oxford analysed data in the National Pupil Database in England to measure the extent of the gap in the performance between economically disadvantaged pupils and pupils from higher socioeconomic status (SES) backgrounds on national science tests. This analysis confirmed that disadvantaged pupils (pupils who have been entitled to free school meals at least once in the last six years) had much lower scores and made poorer progress in science, at every stage of their school career, than pupils from higher SES backgrounds. The gap, they suggest, first becomes apparent at Key Stage 1 (ages 5–7) and only gets wider throughout primary and secondary school. The gap for science is as wide as it is in English and maths, and grows particularly strongly between the ages of 5–7 and 11–16.

The study also found that the strongest factor affecting pupils’ science scores was how well they understood written texts. According to the report, poor literacy skills affect how well a pupil is able to understand scientific vocabulary and to prepare scientific reports. This suggests that strategies to boost disadvantaged pupils’ reading comprehension could have a positive impact on their achievement in science too. The authors write: “In correlational studies of science learning, the strongest and most consistent predictor of pupils’ scientific attainment has undoubtedly been how literate they are”. They add that there is a “strong relationship” between pupils’ socioeconomic status and their literacy.

A study, which we covered in a previous edition of Best Evidence in Brief, found a similar relationship between literacy and science achievement gaps for pupils in US elementary and middle schools.

Source: Review of SES and science learning in formal educational settings: A report prepared for the EEF and the Royal Society (September 2017), Education Endowment Foundation and Royal Society

Talking in class boosts progress in maths, science and English

An intervention that trained teachers to improve and monitor the quality of classroom talk had a positive impact on primary pupils’ test scores in English, maths and science, a report published by the Education Endowment Foundation (EEF) reveals.

Seventy-six primary schools with higher-than-average proportions of disadvantaged pupils took part in a randomised control trial of the Dialogic Teaching intervention, which is designed to improve the quality of classroom talk as a means of increasing pupils’ engagement, learning and achievement. Year 5 teachers in 38 schools (2,493 pupils), and a teacher mentor from each school, received resources and training from the delivery team and then implemented the intervention over the course of the autumn and spring terms in the 2015/16 school year. A control group of 38 schools (2,466 pupils) continued with business as usual. Following the intervention, pupils were tested in English, maths and science.

The results showed that pupils in the intervention schools did better in the main outcome measures of English (effect size = +0.16), science (+0.12), and maths (+0.09) when compared with pupils in the control schools who didn’t receive the intervention. For pupils who received free school meals, the intervention had a higher impact on maths (+0.16), but around the same for English (+0.12) and science (+0.11). Teachers reported positive effects on pupil engagement and confidence, and on the whole the intervention was highly regarded by participating schools. However, some teachers felt that it would take longer than two terms to fully embed a Dialogic Teaching approach in their classrooms.

The Dialogic Teaching intervention was developed by the Cambridge Primary Review Trust and the University of York. This University of York news story has more.

Source: Dialogic teaching: evaluation report and executive summary (July 2017), Education Endowment Foundation

Science professional development and pupil achievement: A cluster-randomised trial

Joseph Taylor of Abt Associates and colleagues conducted a rigorous study of the Science Teachers Learning Through Lesson Analysis (STeLLA) professional development (PD) programme.

STeLLA is designed to increase elementary (primary) teachers’ science knowledge. Instead of the standard practice of teaching pupils to memorise science concepts and then perform activities that prove these concepts, STeLLA teachers lead pupils to discover science concepts through experience and experimentation. One of STeLLA’s main tenets is to have pupils think through science problems aloud so that teachers can respond to pupils’ ideas and guide them to scientific conclusions and specific learning goals. Its other distinguishing feature is that during the course of a year, groups of 5–10 teachers led by a PD coach watch and critique videos of experienced science teachers’ lessons, later moving on to their own and their colleagues’ lessons, to analyse them regarding science content, teaching and learning. In addition, STeLLA teachers are taught by university-level science teachers the summer prior to implementation to provide them with greater science content knowledge, a process called “content deepening”.

In the current study, researchers used a cluster-randomised design to compare STeLLA to The Content Deepening Program, a PD programme that deepens teachers’ science knowledge through university faculty-led science teaching, like STeLLA does, but without STeLLA’s analysis-of-practice component. Seventy-seven schools, with 144 teachers and 2,823 fourth and fifth grade pupils (Years 5 and 6) in Colorado, were randomly assigned either to STeLLA (n=42 schools) or to The Content Deepening Program (n=35 schools) in two cohorts, the first in 2011–12 and the second in 2012–13. Teachers in both conditions experienced 88 hours of PD and had the same learning goals for their pupils. Pupils were pre- and post-tested on a science measure based on established assessments. Although the control group demonstrated a slight achievement advantage at baseline, results showed that pupils in STeLLA classes scored higher (effect size = +0.55) at post-test than pupils in classes whose teachers had been through The Content Deepening Program.

Source: The effect of an analysis-of-practice, videocase-based, teacher professional development program on elementary students’ science achievement (2017),  Journal of Research on Educational Effectiveness, Volume 10:  Issue 2

Three-year achievement gap between poor pupils and their better-off peers

Research published by the Sutton Trust shows that for schools in the UK, the achievement gap in maths, science and reading between the top-performing pupils from low and high socio-economic backgrounds is around two years and eight months.

Global Gaps by Dr John Jerrim of the UCL Institute of Education analyses the 2015 test scores from the Organisation for Economic Co-operation and Development (OECD) PISA tests to assess how well the top 10% of pupils in the UK’s schools are doing. In England, the highest-achieving pupils score above the median score for OECD countries in maths, science and reading. However, in Scotland, Northern Ireland and Wales, high-achieving pupils perform, on average, below the OECD median scores.

For girls in England, the achievement gap in science and reading is even greater. High-achieving girls from low socio-economic backgrounds are around three years behind their more advantaged, high-achieving peers. This is around eight months greater than the equivalent gap for boys for science, and nine months greater for reading. There is no significant gender difference in maths, with an achievement gap of around two years and nine months for both girls and boys.

Source: Global Gaps: Comparing socio-economic gaps in the performance of highly able UK pupils internationally (February 2017), The Sutton Trust

Predicting success in STEM

The Regional Educational Laboratory Southwest has conducted a literature review to determine what predictors from primary school of postsecondary STEM success have been identified in peer-reviewed studies, with a focus on predictors for Hispanic students.

The review defined postsecondary STEM success as enrollment in, persistence in, and completion of postsecondary STEM majors or degrees. Twenty-three relevant studies were identified, but only four examined factors predictive of success specifically for Hispanic students.
Key findings from the review included:

  • The number of high school maths and science courses taken and the level of those courses predict postsecondary STEM success for all student subgroups, but racial/ethnic minority students were less likely than White students to take the highest level maths and science courses.
  • Interest or confidence in STEM showed statistically significant predictive relationships with students’ postsecondary STEM success, and the relationships were evident as early as middle school. Racial/ethnic minority and White students had similar interest and confidence in STEM.
  • Statistically significant high school predictors of postsecondary STEM success included schools’ academic rigour, percentage of students enrolled in college preparatory programmes, students’ satisfaction with their teachers, and levels of parent participation.

Source: A Review of the Literature to Identify Leading Indicators Related to Hispanic STEM Postsecondary Educational Outcomes (2016), Regional Educational Laboratory Southwest.