Abstract
Executive summary
Introduction and background
Primary science is important for pupils as individuals and for society: amongst many other reasons, it can help children to understand and reason about themselves and the world, enable them to live healthy lives, and make informed choices (Harlen, 2018). It is important for society to have a scientifically literate population to respond to major social and environmental threats to human wellbeing such as climate change, food supply, pandemics and energy production (Royal Society, 2010). Concerns have been expressed by Ofsted (2023) amongst others that the status of science in primary schools has been lower since the removal of national tests in 2009, with pupils in some schools going an entire half term without learning science. England’s position in the grade 4 (age 10-11) TIMSS (Trends in International Mathematics and Science Study) comparative tests, whilst above the average of participating countries, is outside the top ten (Mullis et al., 2020). It is therefore important to understand how to best organise, teach and assess primary science. Whilst several reviews have focused on identifying the effectiveness of specific approaches such as inquiry (Slavin
et al., 2014), the purpose of this study was to review the effectiveness of different approaches to primary science teaching on a range of pupil outcomes.
Approach and methods
Informed by a review of practice to identify outcomes important to teachers, we systematically searched electronic bibliographic databases and other educational sources for studies reporting on the effects of different approaches to primary science curriculum, teaching and assessment, published and including data collected between January 2007 and September 2021. A total of 15,476 studies were screened. We included studies if they had a sample of children aged between 5
and 11 and reported a science-specific teaching approach or a general teaching approach used in a primary science context. Studies were included if they had a counterfactual and where they reported pupil-level outcomes. The final body of literature for review included 90 studies. The majority of studies used quasi-experiments (66 studies), with a smaller number of randomised
control trials (23 studies) and one randomised field trial. The majority of studies (77) involved children aged 8 or above, with only 13 including children under 8 in the sample. Studies were conducted in a range of countries, with over half carried out in the USA (36) and Taiwan (16), and the remainder elsewhere. Only 6 studies were conducted in the UK. Studies were assigned to one of nine clusters determined by the review team in consultation with the EEF and Guidance Panel: assessment and feedback; context-based and cross-curricular approaches; co-operative and
collaborative approaches; critical thinking and argumentation; explicit instruction and related approaches; ICT-supported and online teaching and learning; language, literacy and text-based approaches; learning outside the classroom; and practical work, inquiry and investigation. In each cluster, we identified the constituent approaches, nature of the comparison group, outcomes measured and the summary of effects, along with identifying requirements for implementation, gaps in the evidence base and applicability of the evidence base to the English context. We found a diverse range of approaches, desired outcomes and measures across many
different primary science topics and as a result, it was not meaningful to conduct a meta-analysis or to rank approaches. Few approaches were independently evaluated or used existing research instruments.
Findings and implications
We found a relatively large number of studies, multiple high-quality studies, with large sample sizes. Positive effects were reported across a range of outcomes including attainment and scientific language. Approaches included dialogic teaching, science and language integrated instruction and science for language learners. We found evidence to suggest that integration of science and literacy over an extended period can improve cognitive and affective outcomes for pupils in science and
literacy, particularly for pupils who are less proficient in the language.
A connection between language, thinking and science was common in a number of approaches across clusters, indicating that increased talk enables pupils to identify prior knowledge and evidence to support their thinking, integrate new ideas, practise articulating these and receive feedback from peers and teachers. Relatedly, a number of studies involving cooperation and collaboration reported positive effects on attainment and attitudes. These included collaborative writing, discussion, online collaboration, and collaboration during differentiation. Few studies reported on the effectiveness of different approaches to assessment, but those reporting on formative assessment approaches which create opportunities for meaningful classroom
interactions focused on the learning goal, and time for thinking and talking (embedded formative assessment, guided peer feedback, bidirectional peer assessment) reported positive effects. Several studies across clusters pointed to positive effects on attainment of giving children opportunity to learn outside the classroom, whether linked to school grounds and local habitats or
shorter-term field visits to science centres, planetaria, or nature parks.
Few studies reported differential outcomes for different groups of pupils. For those eligible for free or reduced school meals, approaches which provide opportunities to make thinking explicit through words, whether spoken or written, seem to show promise, for example, dialogic teaching and the use of writing for argumentation. For pupils with low prior attainment compared with peers, approaches with additional benefits included those involving thinking time and classroom talk. Approaches which reported specific benefits for pupils who speak English as an additional language
include scaffolded instructional discourse and instruction which integrates scientific content with literacy instruction. Finally, approaches reporting gains for children with special educational needs included guided inquiry involving activation of prior knowledge, explicit instruction of vocabulary, read-alouds and investigation activities. A frequent moderating factor across approaches was the provision of CPD for teachers, and whole school support for the approach. Gaps in research evidence exist in relation to approaches that are used by teachers and those that have been investigated using research designs with a counterfactual. The review points to the need for a research agenda aligned to the aims of primary science education, and which is dedicated to accumulating an evidence base on approaches to the curriculum, teaching and assessment.
Limitations
The review included studies published and containing data collected between 2007 and 2021 and published in English, meaning that high-quality studies of approaches in areas of research that have been less popular in recent decades, and those published in languages other than English are not included. The predominance of small-scale studies reporting statistically significant positive results of approaches used in interventions suggests some publication bias. Finally, the review did not
include studies that focused only on the early years foundation stage, so the review will have limited applicability to teachers of the Reception year in primary schools in England.
Introduction and background
Primary science is important for pupils as individuals and for society: amongst many other reasons, it can help children to understand and reason about themselves and the world, enable them to live healthy lives, and make informed choices (Harlen, 2018). It is important for society to have a scientifically literate population to respond to major social and environmental threats to human wellbeing such as climate change, food supply, pandemics and energy production (Royal Society, 2010). Concerns have been expressed by Ofsted (2023) amongst others that the status of science in primary schools has been lower since the removal of national tests in 2009, with pupils in some schools going an entire half term without learning science. England’s position in the grade 4 (age 10-11) TIMSS (Trends in International Mathematics and Science Study) comparative tests, whilst above the average of participating countries, is outside the top ten (Mullis et al., 2020). It is therefore important to understand how to best organise, teach and assess primary science. Whilst several reviews have focused on identifying the effectiveness of specific approaches such as inquiry (Slavin
et al., 2014), the purpose of this study was to review the effectiveness of different approaches to primary science teaching on a range of pupil outcomes.
Approach and methods
Informed by a review of practice to identify outcomes important to teachers, we systematically searched electronic bibliographic databases and other educational sources for studies reporting on the effects of different approaches to primary science curriculum, teaching and assessment, published and including data collected between January 2007 and September 2021. A total of 15,476 studies were screened. We included studies if they had a sample of children aged between 5
and 11 and reported a science-specific teaching approach or a general teaching approach used in a primary science context. Studies were included if they had a counterfactual and where they reported pupil-level outcomes. The final body of literature for review included 90 studies. The majority of studies used quasi-experiments (66 studies), with a smaller number of randomised
control trials (23 studies) and one randomised field trial. The majority of studies (77) involved children aged 8 or above, with only 13 including children under 8 in the sample. Studies were conducted in a range of countries, with over half carried out in the USA (36) and Taiwan (16), and the remainder elsewhere. Only 6 studies were conducted in the UK. Studies were assigned to one of nine clusters determined by the review team in consultation with the EEF and Guidance Panel: assessment and feedback; context-based and cross-curricular approaches; co-operative and
collaborative approaches; critical thinking and argumentation; explicit instruction and related approaches; ICT-supported and online teaching and learning; language, literacy and text-based approaches; learning outside the classroom; and practical work, inquiry and investigation. In each cluster, we identified the constituent approaches, nature of the comparison group, outcomes measured and the summary of effects, along with identifying requirements for implementation, gaps in the evidence base and applicability of the evidence base to the English context. We found a diverse range of approaches, desired outcomes and measures across many
different primary science topics and as a result, it was not meaningful to conduct a meta-analysis or to rank approaches. Few approaches were independently evaluated or used existing research instruments.
Findings and implications
We found a relatively large number of studies, multiple high-quality studies, with large sample sizes. Positive effects were reported across a range of outcomes including attainment and scientific language. Approaches included dialogic teaching, science and language integrated instruction and science for language learners. We found evidence to suggest that integration of science and literacy over an extended period can improve cognitive and affective outcomes for pupils in science and
literacy, particularly for pupils who are less proficient in the language.
A connection between language, thinking and science was common in a number of approaches across clusters, indicating that increased talk enables pupils to identify prior knowledge and evidence to support their thinking, integrate new ideas, practise articulating these and receive feedback from peers and teachers. Relatedly, a number of studies involving cooperation and collaboration reported positive effects on attainment and attitudes. These included collaborative writing, discussion, online collaboration, and collaboration during differentiation. Few studies reported on the effectiveness of different approaches to assessment, but those reporting on formative assessment approaches which create opportunities for meaningful classroom
interactions focused on the learning goal, and time for thinking and talking (embedded formative assessment, guided peer feedback, bidirectional peer assessment) reported positive effects. Several studies across clusters pointed to positive effects on attainment of giving children opportunity to learn outside the classroom, whether linked to school grounds and local habitats or
shorter-term field visits to science centres, planetaria, or nature parks.
Few studies reported differential outcomes for different groups of pupils. For those eligible for free or reduced school meals, approaches which provide opportunities to make thinking explicit through words, whether spoken or written, seem to show promise, for example, dialogic teaching and the use of writing for argumentation. For pupils with low prior attainment compared with peers, approaches with additional benefits included those involving thinking time and classroom talk. Approaches which reported specific benefits for pupils who speak English as an additional language
include scaffolded instructional discourse and instruction which integrates scientific content with literacy instruction. Finally, approaches reporting gains for children with special educational needs included guided inquiry involving activation of prior knowledge, explicit instruction of vocabulary, read-alouds and investigation activities. A frequent moderating factor across approaches was the provision of CPD for teachers, and whole school support for the approach. Gaps in research evidence exist in relation to approaches that are used by teachers and those that have been investigated using research designs with a counterfactual. The review points to the need for a research agenda aligned to the aims of primary science education, and which is dedicated to accumulating an evidence base on approaches to the curriculum, teaching and assessment.
Limitations
The review included studies published and containing data collected between 2007 and 2021 and published in English, meaning that high-quality studies of approaches in areas of research that have been less popular in recent decades, and those published in languages other than English are not included. The predominance of small-scale studies reporting statistically significant positive results of approaches used in interventions suggests some publication bias. Finally, the review did not
include studies that focused only on the early years foundation stage, so the review will have limited applicability to teachers of the Reception year in primary schools in England.
| Original language | English |
|---|---|
| Publisher | Education Endowment Foundation |
| Publication status | Published - 24 Nov 2023 |
Keywords
- Primary science
- primary education
- elementary education