卓越實證概述 Best Evidence in Brief
What is the impact of pay-for-performance?

Teacher Incentive Fund (TIF) grants were awarded in 2010 by the US Department of Education to support performance-based compensation systems for teachers and principals in high-need schools.

In order to assess the impacts of pay-for-performance on educator (teachers and principals) and pupil outcomes, an experimental study design was used in ten US school districts to randomly assign elementary and middle schools to treatment and control groups. Both groups implemented the same performance-based compensation system, but in the control schools, the pay-for-performance element was replaced by a one percent bonus paid to all teachers and principals regardless of performance. A fourth and final report from this evaluation has now been published, covering all four years of the programme (between 2011 and 2015).

Among the key findings are that:

  • Pay-for-performance had small, positive impacts on pupil achievement by the second year of implementation.
  • From that year onward, reading and maths achievement was higher by 1 to 2 percentile points in schools that offered performance bonuses than in schools that did not.

However, it was not entirely clear how this improvement was achieved. The impacts of pay-for-performance on classroom observation ratings did not appear to explain the impacts on pupil achievement, and in treatment schools as many as 40% of teachers were unaware that they could earn a performance bonus.

 

Source: Chiang, H., Cecilia, S., Mariesa, H., Kristin, H., Paul, B. & Alison, W. (2017). Evaluation of the teacher incentive fund: final report on implementation and impacts of payfor-performance across four years. Washington, DC: National Center for Education Evaluation and Regional Assistance, Institute of Education Sciences, U.S. Department of Education… Read the rest

A thirty-year look at studies on computer-assisted maths

During the past 30 years, thousands of articles have been written about technology’s effects on pupil achievement. In order to quantify technology’s effects on maths achievement, Jamaal Young at the University of Texas conducted a meta-analysis of all of the meta-analyses on the topic during the last three decades. His second-order meta-analysis was comprised of 19 meta-analyses representing 663 primary studies, more than 141,000 pupils and 1,263 effect sizes. Each meta-analysis that was included had to address the use of technology as a supplement to instruction, use pupil maths achievement as an outcome measure, report an effect size or enough data to calculate one, have been published after 1985 and be accessible to the public.

The author found that:

  • All technology enhancements positively affected pupil achievement, regardless of the technology’s purpose.
  • However, technology that helped pupils perform computational functions had the greatest effects on pupil achievement, while combinations of enhancements demonstrated the least effects on pupil achievement.
  • Study quality and the type of technology used in the classroom were the main influencers on effect sizes. The highest-quality studies had the lowest effect sizes, which he attributes to their more rigorous analysis procedures. The high-quality reviews gave an overall effect size for the use of technology of +0.16 (compared with +0.38 for low- and +0.46 for medium-quality reviews).

As limited by the availability of effect size, this secondary meta-analysis did not assess meta-cognitive, motivational, and affective outcomes. The author suggested that the accumulation and aggregation of effect sizes must improve and evolve to support policy and praxis, keeping technology integration relevant and valuable in mathematics classrooms.

 

Source: Young, J. (2017). Technology-enhanced mathematics instruction: A second-order meta-analysis of 30 years of research. Educational Research Review, 22, 19–33.… Read the rest

Preschool language skills a predictor of later reading comprehension

systematic review published by the Campbell Collaboration summarises the research on the correlation between reading-related preschool predictors, such as code-related skills  and linguistic comprehension, and later reading comprehension skills.

Sixty-four longitudinal studies met the eligibility criteria for the review. These studies spanned 1986 to 2016 and were mostly carried out in the US, Europe and Australia. Overall, the findings of the review found that:

  • Code-related skills (rhyme awareness, phoneme awareness, letter knowledge and rapid automatised naming) are most important for reading comprehension in beginning readers.
  • However, linguistic comprehension (grammar and vocabulary) gradually takes over as children become older.
  • All predictors, except for non-word repetition, were moderately to strongly correlated with later reading comprehension.
  • Non-word repetition had only a weak to moderate correlation to later reading comprehension ability.

These results suggest a need for a broad focus on language skills in preschool-age children in order to establish a strong foundation for reading comprehension.

 

Source: Hjetland, H. N., Brinchmann, E.l., Scherer, R., & Melby-Lervåg, M. (2017). Preschool predictors of later reading comprehension ability: a systematic review. Oslo, Norway: Campbell Systematic Reviews.… Read the rest

How do parents influence children’s mindset?

Children with a fixed mindset believe that they have a fixed amount of intelligence that they cannot change. As a result, when work becomes difficult they may question their ability, stop trying, and achieve less. In contrast, children with a growth mindset see their intelligence as malleable and something that can be developed through hard work, good strategies, and teaching. As a result, when work becomes difficult they are more likely to increase their efforts and achieve more.

To date, no clear link has been found between parents’ mindsets and their children’s. A series of experiments, published in Psychological Science, has found that parental response to failure is influential. Parents who believed that failure is a debilitating experience that inhibits learning and productivity had children who tended to have a fixed mindset. This occurred because these parents reacted to their children’s failures by focusing more on their children’s ability or performance than on their learning.

It may not be sufficient, therefore, to teach parents a growth mindset and expect that they will pass this on to their children. Instead, an intervention could target teaching parents that failure can be beneficial, and help them to react appropriately to their children’s setbacks in order to support their children’s future motivation and learning.

 

Source: Haimovitz, K., & Dweck, C. S. (2016). Parents’ views of failure predict children’s fixed and growth intelligence mind-sets. Psychological Science, 27(6), 859–869.… Read the rest

Predicting reading ability from DNA analysis

Researchers from King’s College London have used a genetic scoring technique to predict reading performance throughout school years from DNA alone.

Saskia Selzam and colleagues calculated genetic scores for educational achievement in 5,825 individuals from the Twins Early Development Study (TEDS) based on genetic variants identified to be important for educational achievement. They then mapped these scores against reading ability between the ages of 7 and 14.

The report, published in Scientific Studies of Reading, found there was a correlation between children’s DNA and their reading ability. Genetic scores were found to explain up to five percent of the differences between children in their reading ability. The children with the highest and lowest genetic scores had reading abilities almost two years apart. This association remained significant even after accounting for cognitive ability and family socioeconomic status.

The researchers note that although five percent may seem a relatively small amount, it is substantial compared to other results related to reading. For example, gender differences have been found to explain less than one percent of the differences between children in reading ability. The use of genetic scoring in education as an early screening tool to help identify those at particular genetic risk of reading problems, they propose, could lead to tailored intervention and prevention according to an individual child’s needs.

 

Source: Selzam, S., Dale, P. S., Wagner, R. K., DeFries, J. C., Cederlöf, M., O’Reilly, P. F., … Plomin, R. (2017). Genome-wide polygenic scores predict reading performance throughout the school years. Scientific Studies of Reading, 21(4), 334–349.… Read the rest