Impact Journal Logo

A metacognitive approach to improving academic performance and study skills in sixth form students

9 min read

Many students find the transition to A-level study challenging. In our most recent cohort of chemistry students, those with low prior attainment made less progress than their peers. Observing the underperforming students offers two reasons for this: 1) poor study habits in terms of the type of activities undertaken as independent work, and 2) overestimating their understanding by judging it based on a specific context rather than the ability to transfer it to a different situation. Metacognition has been identified as having the potential to close the gap between students with low prior attainment and their peers, as well as being recommended by the Education Endowment Foundation (EEF) (Quigley et al., 2018) to improve study habits. The EEF report also identified two aspects of planning, activate relevant prior knowledge about the task and select appropriate strategies. These are similar to the process required for transfer of knowledge (Georghiades, 2000), which suggests that metacognition may also provide a solution to this element of the problem.

In the light of this, I set myself the following research question: How does switching lunchtime sessions from exam questions to metacognitive discussions over a two-month period affect the academic performance, metacognition and study skills of students in Year 12 chemistry?

Metacognition strategies

The term ‘metacognition’ is commonly divided into two components: metacognitive knowledge and metacognitive regulation (Cambridge Assessment, 2017). Metacognitive knowledge may be broken down further into knowledge of self, knowledge of tasks and knowledge of strategies (Cambridge Assessment, 2017) or, alternatively, declarative knowledge, procedural knowledge and conditional knowledge (Winne and Azavedo, 2014). The latter categorisation emphasises an important point: that it is insufficient for students to know about tasks and strategies unless they can pick the most appropriate strategy for the task.

A review of the literature, focusing on science-specific studies, suggests, among others, the following ways of improving students’ metacognition, study skills and academic performance.

  • Explicit instruction in study skills (Cook et al., 2013).
  • Concept maps, where students make explicit connections between different ideas, help students to reflect on their understanding (Zohar and Barzilai, 2013) and have a particular benefit for students with lower prior attainment (Haugwitz et al., 2010).
  • Explanations can improve the accuracy of students’ monitoring. Chiu and Linn (2012) found that students who explained concepts in their own words rated their understanding of online visualisations lower and more accurately.


The intervention took place in 30-minute lunchtime sessions. Students studying A-level chemistry were divided into intervention and control groups; the groups had the same number of sessions to ensure that contact time did not affect the outcome. The sessions were based around particular content, with the comparison group doing exam questions and the intervention group learning a relevant strategy (see Table 1). The format of the intervention sessions was: 1) explicit instruction and modelling of the strategy; 2) students applying the strategy; and 3) reflection on metacognitive prompts, e.g. what strategies did you use? What independent work would best allow you to improve?

Session Content Strategy
1 Mole calculations Students were taught a problem-solving process (problem, parts, prior knowledge, proceed, post-mortem) based on the work of Polya (1957), and discussed different mathematical problem-solving techniques relevant to the topic (Mountstevens, 2019).
2 Ionic compounds Students were presented with some of the evidence on retrieval practice (Roediger and Karpicke, 2006) and its challenges. They applied these to memorising the colours of precipitates.
3 Organic chemistry A concept map was modelled, highlighting the importance of the nodes and connections (Taber, 2002), and students then completed their own.
4 Enthalpy experiments Students were taught about the techniques of elaborative interrogation (why is this true?) and self-explanation (how does this link to what I already know?) (Dunlosky, 2013). They worked in pairs, asking each other these questions whilst they completed an enthalpy calculation.
Table 1: Session content and metacognitive skill

Random assignment was not possible in this case, due to students’ lunchtime availability. Instead, groups were matched by class, attainment (measured by their performance in the pre-test) and, where possible, gender. Results have been included in the study for students who attended three or more sessions.

Academic performance was assessed by 35-mark pre- and post-tests. Averages of the raw marks were calculated for the control and intervention group, as well as the change in average mark between pre- and post-test. A positive value shows an increase in test mark. The use of a post-test shortly after the intervention session may not accurately reflect a long-term improvement in academic performance, but was necessary due to time restrictions.

Metacognition is challenging to assess because it is not directly observable. A number of methods have been developed: self-report questionnaires, such as the Metacognition Awareness Inventory (MAI) (Schraw and Dennison, 1994), confidence judgements (Winne and Azevedo, 2014) and concept mapping (Haugwitz et al., 2010). This study used a version of the MAI with a reduced number of questions. It was analysed by calculating the average score for each student for the pre- and post-test questionnaires and then the change in metacognitive awareness; a positive value shows an increase in metacognitive awareness. Students also completed retrospective confidence judgements (RCJ) after each multiple-choice question, rating their confidence between 1 (very unsure) and 5 (very confident). The accuracy of confidence judgements was determined by calculating the difference between the average RCJ for correct answers and those for incorrect answers. A larger number indicates that the candidate was more discriminating. A positive value for the change in accuracy indicates an improvement during the study.

The effectiveness of students’ independent work was assessed using the independent work logs. The types of activity undertaken were coded as effective or ineffective and the average percentage of effective activities before and during the intervention was calculated. Activities taught during the metacognitive sessions (practice problems, retrieval practice and concept maps) were deemed effective, as were exam questions, which contain many of the features of these effective strategies.

Key findings

Academic performance

The change in average mark between pre- and post-test is shown in Figure 1. The average mark for the control group decreased during the intervention period and the average mark for the intervention group increased, although a decrease in the mark does not necessarily mean a deterioration in performance, due to the variation in difficulty of the tests. Students with lower prior attainment appeared to benefit more from the sessions, although the differences are very small and not statistically significant.Figure 1 is titled "Change in average mark" and shows a bar graph with the bars labelled "control" and "intervention". The horizontal axis is labelled "all", "low prior attainment" and "high prior attainment". The vertical axis is labelled "average mark post-test minus average mark pre-test. It ranges from minus 2 to plus 3.5. For the group "all", the "control" bar is negative and at around minus 1.2. The "intervention" bar is positive and at around 1.4. For the group "low prior attainment", the "control" bar is negative and at around minus 1.4. The "intervention" bar is positive and at 3. For the group "high prior attainment", the "control" bar is negative and at minus 1. The "intervention" bar is negative and at around minus 0.6.


Figure 2 shows the results of both the metacognitive awareness inventory (MAI) and the confidence judgements (RCJ). Both measures show an improvement in metacognition over the course of the study. The improvement in RCJ is particularly encouraging, as it shows that students adapt quickly to the increased rigour of A-level. However, the two measures do not show the same trend, with the MAI showing a greater improvement for the control group and the RCJ showing a marginally greater improvement for the intervention group. This reflects the complexities of measuring metacognition and the different facets of the concept assessed by the two different measures. It is therefore not possible to draw any conclusions on the impact of these sessions on the metacognitive ability of the students.

Figure 2 is titled "Change in accuracy of confidence judgments (RCJ) and metacognitive awareness (MAI)" and shows a bar graph with the bars labelled "control" and "intervention". The horizontal axis is labelled "RCJ" and "MAI". The vertical axis is labelled "MAI - average after minus before. RCJ - accuracy of confidence judgement after minus before. It ranges from 0 to 0.45. The RCJ control group stands at around 0.36, the RCJ intervention group a 0.4. The MAI control group stands at 0.1 and the intervention group at around 0.03.

Study skills

Figure 3 shows the change in the percentage of effective independent work during the intervention. Most groups demonstrated an improvement in the effectiveness of their independent work, which suggests that such explicit teaching of useful strategies does improve study skills. The most commonly used effective strategy was answering questions, but there was also evidence of reviewing past topics, daily self-testing and concept maps. Overall, the intervention sessions had a greater impact on the effectiveness of students’ independent work.

Figure 3 is titled "Change in the percentage of effective independent work" and shows a bar graph with the bars labelled "control" and "intervention". The horizontal axis is labelled "all", "low prior attainment" and "high prior attainment". The vertical axis is labelled "percentage effective independent work during minus percentage of effective independent work before". It ranges from minus 10 to plus 50. For the group "all", the "control" bar is at around positive 22, the "intervention" bar at around positive 38. For the group "low prior attainment", the "control" bar is negative at minus 10. The "intervention" bar is positive and at around 43. For the group "high prior attainment", the "control" bar is positive at around 32. The "intervention" bar is positive at around 38.


There is some indication that switching lunchtime sessions from exam questions to metacognitive discussions had a small positive effect on academic performance and a slightly larger effect on study skills. A possible mechanism for this change was that the sessions improved the effectiveness of independent work, which led to better academic performance. Different effects were observed on the two measures of metacognition, so it is not possible to draw conclusions on this aspect.

A correlation was observed between retrospective confidence judgements (RCJ) and academic performance, which is consistent with the work of Nietfeld et al. (2006). This provides an alternative mechanism for the improvement in academic performance, i.e. students using the confidence judgements to better understand their strengths and weaknesses and prioritise their independent work.

As a result of this study, all lunchtime support sessions were changed to the metacognitive format and similar strategies incorporated into lessons. We will continue to monitor the performance of the cohort until students take their final exams in 2020. In terms of their application to other subjects, the strategies used will vary, but I would recommend that strategies be taught explicitly and that time is allocated for reflection and planning for independent work.


Cambridge Assessment (2017) Metacognition. Available at: (accessed 27 November 2019).

Chiu JL and Linn MC (2012) The role of self-monitoring in learning chemistry with dynamic visualizations. In: Zohar A and Dori Y (eds) Metacognition in Science Education. Contemporary Trends and Issues in Science Education, Vol 0. Dordrecht: Springer,  pp. 133–163.

Cook E, Kennedy E and McGuire SY (2013) Effect of teaching metacognitive learning strategies on performance in general chemistry courses. Journal of Chemical Education 90(8): 961–967.

Georghiades P (2000) Beyond conceptual change learning in science education: Focusing on transfer, durability and metacognition. Educational Research 42(2): 119–139.

Haugwitz M, Nesbit JC and Sandman A (2010) Cognitive ability and the instructional efficacy of collaborative concept mapping. Learning and Individual Differences 20(5): 536–543.

Nietfeld J, Cao L and Osborne JW (2006) The effect of distributed monitoring exercises and feedback on performance, monitoring accuracy, and self-efficacy. Metacognition and Learning 1: 159–179.

Quigley A, Muijs D and Stringer E (2018) Metacognition and self-regulated learning: Guidance report. Education Endowment Foundation. Available at: (accessed 27 November 2019).

Schraw G and Dennison RS (1994) Assessing metacognitive awareness. Contemporary Educational Psychology 19(4): 460–475.

Winne P and Azevedo R (2014) Metacognition. In: Sawyer R (ed) The Cambridge Handbook of the Learning Sciences. Cambridge: Cambridge University Press, pp. 63–87.

Zohar A and Barzilai S (2013) A review of research on metacognition in science education: Current and future directions. Studies in Science Education 49(2): 121–169.

      0 0 votes
      Please Rate this content
      Notify of
      Inline Feedbacks
      View all comments

      From this issue

      Impact Articles on the same themes

      Author(s): Bill Lucas