Improving how students write conclusions in Science


I teach a Year 8 mixed ability Science group, ranging from NC level 4 to level 8. Previously, I have noticed that when writing up science experiments, conclusion writing is one of the areas where there is lots of room for improvement from students across the ability range. Conclusion writing is where students get to show off their scientific understanding – making sense of the data, comparing their results to their prediction, and really working out what’s going on.

As I began teaching a new unit on ‘rocks and weathering’ which includes a number of practical experiments, I wanted to understand more about conclusions writing among students from across the ability range, in the hope that it would help inform my ability to teach all students to improve their ability to write high level conclusions in future.


Following a coaching discussion with a colleague, I identified that conclusion writing is an area I would like to develop with my Year 8 Science class. I chose this topic for my inquiry because conclusion writing requires students to look at evidence, and then try to explain what they see in terms of their Scientific understanding. They have to think analytically and critically, and think carefully about the language that they use to explain their thinking.

Year 8 students can sometimes find such abstract thinking difficult, However, with a broad range of ability in my year 8 class, I wanted to explore how different students approach conclusion writing, with a view to sharing good practice and improving my teaching.

Based on previous experience, with Year 8 I find it useful to structure tasks as a series of levelled questions. Although this is quite a simplistic model, it helps students to understand that highly developed conclusions are really driven by the questions you are trying to answer.  It also helps them to structure their writing.

Research question:

What strategies are most effective in improving the writing of conclusions in a year 8 mixed ability Science class?

Research method:

I decided to focus on two experiments, carried out over two successive weeks as part of the ‘Rocks’ unit. One experiment explored the relationship between temperature and crystal size, and the other related to the relationship between particle size and the formation of sedimentary rocks. After the experiment, I gave students 10 minutes in which to write a conclusion, based on a series of levelled questions as follows:

Crystals experiment: Conclusions level ladder:

  • Level 4                Which slides (hot or cold) made the larger crystals?
  • Level 5                Draw a diagram of the two crystal sizes.
  • Level 6                Why do you think there is a difference in the crystal sizes?
  • Level 7/8            In which parts of a volcano would you expect to find larger and smaller crystals?

Sediment experiment: Conclusions level ladder:

  • Level 4                Why did the larger sediment sink to the bottom first?
  • Level 5                Where will the most pressure in the bottle be? Why is this?
  • Level 6                How do the layers within the bottle vary?
  • Level 7/8            How does this model show the formation of sedimentary rocks?

I chose a random sample of students’ work to analyse, from a range of abilities within the class. The students’ responses were typed up into a rubric to allow easy comparison (see attached – students’ names have been changed). I then levelled each conclusion against criteria that were shared with the students.


Please see below for a summary of the study students’ conclusions and levels. Please note, levels awarded here are based on whether the student met the levelled criteria set for this specific task, and may not reflect wider interpretations of NC levels.



This discussion takes the form of a series of reflective questions, which I worked through with my coach. Here are the questions and responses:

Why did the conclusions level ladders take the form of a series of levelled questions?

I find they respond best to that. I’ve tried instructions and tick lists – e.g. “I have done this to get to this level” but they are so used to questioning… I regularly question them during lessons, so to see those kinds of questions on the board, they respond well to it. I focus heavily on questioning, standing in front of a class, asking questions, going deeper – why do you think this – they are conditioned to work in this way by this time in the year, and so I know they respond to it well.

What did you notice about the students’ conclusions in these two tasks?

I found that more than anything, it supports the lower and middle ability students. With the higher ability students, for example Matthew and Jess, they have targets of an 8c and they are getting level 7s. It helps the lower and middle students in seeing where they can make those improvements. If they don’t know what to do, all they need to do is look at the level 7 box to know how they can improve. So I noticed that it was particularly helpful for the lower and middle ability students.

What surprised me for the sediments task is that the students struggled with linking models together. I think that’s maybe something we need to work on as a class. Because we use models quite a lot and it’s a skill that I now realise that they are lacking, so I was surprised that they didn’t attempt that part of the conclusion.

What was expected?

I was expecting students to aim for their target level – the minimum expected was that they should improve on their target grade. Mary and David are both level 6 students and they aimed for level 7.

What was unexpected?

What’s not really predictable is some of the middle ability students achieving higher than expected grades – David and Mary achieving level 7s – allowing them to raise themselves. Matthew and Jess are very intelligent, but the level ladder levels the playing field because students who know how to structure an answer were able to achieve highly, and narrow the gap.

The way a question was phrased makes an impact – the way the higher level question on modelling was phrased maybe put some of them off answering it.

What was good about the students’ conclusions in the rocks unit?

What was good was that students were encouraged students to look at the why, beyond just straightforward observations, and justify their decisions. Why is there a difference in crystal sizes for example? So naturally we’re progressing through Bloom’s taxonomy – they’re describing, then explaining, then modelling – to deepen their thinking.

What do the students need to do more of?

They need to push themselves harder. To try to achieve higher. They need a timer and they need to be given a sense of challenge – how high can you get up the level ladder in 8 minutes.

What do they need to do less of?

I think they could take a lesson from Matthew in terms of what they could do less of. Because he’s very concise and he’s hitting level 7. So they need to cut a lot of the waffling out. For example “Why do you think there’s a difference in crystal sizes?” he answered “The larger crystals cooled over a longer period of time” so he’s hit that, and exam wise he’s hitting the marks. To help the other students become more concise I could introduce a word limit – it would be a shame to limit what they write, but less of the needless words. Say 15 words per question – think about what key words you need to include, and then write it.

Does this vary according to ability?

Actually I don’t think it varies according to ability, I think it varies according to the students; look at Mary and Mo, similar ability, hugely different conclusion writing. So I don’t think it varies according to ability, but to another strange factor. Like a student’s willingness to write paragraphs of work. Matthew has written one of the least and he’s the highest ability there. And Mary has written the most and she’s middle ability.

What’s the best way to structure conclusion level ladders in a group with a wide range of abilities?

Starting really basic – so look at what I started with – “Which slides are hot and cold” – low Bloom’s level which only involves identifying which is hotter or colder. You could do it in less than 10 words. So everyone can access that. If you started with “which parts of a volcano are the different crystals found in and why” then you would start losing students. Following that, ask them to draw a diagram and they’re halfway up the level ladder. Then start stretching them. Why are the crystals different, and then how does this link to volcanoes. You start accessible and work up and up, so at the end you’re catering to the most intelligent students, and stretching the rest. And everyone should be able to do the first half, and so it caters for all abilities.

What do the students need in order to take their conclusion writing to the next level?

I don’t know about the students but I think there are things I could do to take their consclusion writing to the next level. For example, time limits – giving them 8 minutes and asking “how high can you get on this?” Peer assessment too, or magpieing so they can see how other students have approached the work, and seeing how the marks are awarded to different examples of conclusions. I like the idea of introducing a word limit for each question, if students are writing too much and struggling with time keeping. Giving the lower ability students key words, or sentence starters. Separating levels 7 and 8. Because I could add an extra part onto it to really stretch them, and that might help them see how they can improve.


Have you found this process of inquiry useful?

Definitely. Through talking it through, I begin to realise that if I did this activity again I would change aspects of how I structured the task. I suppose that’s the benefit of reflective practice!

Featured image: Pixabay

Leave a Reply