# A closer focus: Inference

#### By Sarah Gibb, Research Manager

Thursday 31 October 2019

It will come as no surprise that many of the questions in the key stage 2 reading national curriculum tests* that pupils found challenging involved inference, and the majority of this article will focus on skills that pupils have not yet mastered. There is not just one level of inference, however, and the following exploration of the types of inference that pupils typically can and cannot make will hopefully be a useful aid to teaching. This piece will also consider some interrelated skills which pupils need in order to demonstrate their inferential understanding.

Pupils can…

make simple inferences which are grounded in a basic general knowledge/understanding of the world.
For example, in the 2018 test, pupils were asked [how they could tell from the text] that giant pandas could be dangerous animals. Almost all pupils (96 percent) were able to infer that a panda’s razor-like claws and powerful jaws posed a potential threat. It is reasonable to suppose that most pupils would understand the concept that both claws and jaws can be used by animals to attack, although in this particular instance pupils are further guided to the correct ideas by the use of another piece of evidence in the question itself (Pandas can grow up to 1.5 metres and weigh up to 150 kilograms.) which directly precedes the target information in the text. The text also clearly supports the inference that there is some threat through the use of the expression They might look cute but… which immediately precedes the mention of the giant panda’s razor-like claws and powerful jaws.

Pupils find it harder to…

In both 2017 and 2018, there were two 3-mark questions based on higher level fiction texts. Regardless of their order in the test paper, in both years pupils were more likely to gain marks for questions focusing on character, rather than on something a little more abstract such as the mysterious nature of an animal or a game, and were also more likely to attempt to answer these character-based questions. Overall, the 3-mark question which was most successfully answered by pupils was from 2018, asking pupils to explain the impressions they had formed of a character at a particular point in the extract. Whilst it was still difficult for pupils to attain all three marks, they were more likely to score one or two marks for this question than for the other 3-mark questions, perhaps because the character featured only in a short section of the extract as a whole. Many pupils were also able to successfully identify evidence to support characteristics already identified in a question, as discussed below…

find evidence to support inferences
Responding to the 2018 higher level text, ‘Albion’s Dream’, pupils were asked to Give one piece of evidence that show[ed the main character’s] determination to find a game. From the three possible options that were available to select, 67 percent of pupils were able to identify a relevant piece of evidence. Although all of the potential evidence could be found in a relatively condensed section of the text, the question itself had no locator directing pupils to this section or to the relevant page, meaning that successful scanning of the text was also required for pupils to find this particular passage, which was rich in suitable evidence. The detailed diagnostic commentary from NFER’s autumn reading test also highlights that pupils are sometimes able to distinguish between whether appropriate evidence should take the form of a quotation from the text or their own paraphrase of the relevant point. For one particular question asking pupils to complete a table with evidence showing two different things about a character, lower achieving pupils were less likely to use a direct text lift than middle or higher achieving pupils however, perhaps indicating a lower level of confidence in this area.

Further commentary on questions assessing pupils’ ability to find evidence can be found in the previous article in this series – A Closer Focus: Locating Information and Understanding Ideas.

Pupils find it hardest to…

make multiple inferences, connecting linked ideas together
When pupils were asked to Give two impressions of the poet’s grannie based on the poem ‘Grannie’ (the mid-level text in 2018), only some pupils were able to give two distinct points to receive full credit. There was a considerable difference in the relative difficulty of attaining the first mark and of attaining the second mark, indicating that many pupils were able to make only one inference about the character from the evidence in the designated verse. It is very probable that some pupils struggled with distinguishing different character traits from one another, such as the subtleties of the difference between being understanding and being tolerant, two of the acceptable points listed in the mark scheme. This ability to both identify and then to verbalise the exact nuances of what a text conveys is a particularly challenging task, and one that most pupils appear unlikely to have mastered by the end of Year 6. This task may have been particularly difficult for pupils because of the limited text available to comment upon. However, it is clear from other examples, such as one of the 3-mark questions in the 2018 paper for which the third mark was considerably harder to achieve than either of the first two, that most pupils are unlikely to be able to express more than one inference effectively. In this case, it is interesting to consider this particular mark scheme more closely as there is one quotation – give me that immediately, Edward – which serves as acceptable evidence for five out of the seven specified acceptable points. Many of the other pieces of acceptable evidence are also valid for more than one acceptable point and there are indeed a number of ways to interpret these lines. Encouraging pupils to look for multiple interpretations, of a particular action or a piece of dialogue for example, is a useful activity which may help them to understand that there can be more to explore than just their initial idea or first response.

Based on the same text, pupils were also asked to Give two reasons why [the main character] does not want to part with [the game he had found]. Only 46 percent of pupils were able to answer this successfully, although there was evidence in the text to support four plausible reasons for the character’s reaction. In 2018, a similar question on the mid-level text asked pupils Why does [his grannie] hesitate? Only 34 percent of pupils attained a mark for this question, although this question was particularly challenging, requiring pupils to make a considerable mental leap between the available evidence and the grandmother’s likely thoughts and feelings. Empathy is clearly a crucial element in both of these questions; and tasks that allow pupils an opportunity to develop their ability to empathise with characters in a text can often be some of the most enjoyable and creative. Perhaps this should serve as a reminder that such activities should not be considered an indulgence when they remain a valuable means of improving pupils’ ability to respond thoughtfully to texts.

make / explain more conceptual inferences
Also in 2018, pupils were asked to Explain what the poet finds weird about his grannie in the last verse. This was a 2-mark question requiring pupils to first make reference to the grannie being small/frail and for the second mark to link this to the poet’s perception of his grannie and how this has changed. Unsurprisingly, there was a substantial difference in the difficulty of attaining the second of these marks, a far more conceptual idea relying on a pupil’s ability to consider the time frame of the poem and how the passage of time affects people and their perceived traits. The mark scheme itself includes exemplar answers which demonstrate how some pupils were able to successfully crystallise this notion. Again, some level of empathy is required here, as at 11 years old, most pupils are unlikely to have experienced such a shift in their own perception of an adult.

The final question in 2018, again based on the text ‘Albion’s Dream’, required pupils to explain How [they could] tell that there was something strange about the game. To be credited with the full 3 marks awarded for this question, pupils were required to give two acceptable points which showed this, as well as to provide evidence from the text to support one of these. A similar question in 2017, based on the higher level text ‘An Encounter at Sea’, asked pupils how the whale is made to seem mysterious. From the raw data alone it is not possible to ascertain whether the considerable difference between pupils obtaining the second and the third mark in these questions was more commonly as a result of a failure to provide two acceptable points or to provide supporting evidence for one of these. A substantial number of students also chose not to attempt these questions, a more significant fact perhaps for the second example which was not the final question in the test paper but which 20 percent of pupils omitted answering. However, an example from NFER’s own autumn assessment, based on a non-fiction text about an orangutan rescue centre, is another interesting case. The most common error made here, across all attainment groups, was providing textual evidence on its own, rather than in support of an acceptable point, with pupils selecting examples of persuasive language or repeating some of the more worrying facts included in the article. There was also evidence of this in another 3-mark question from the same test, this time based on a higher level fiction text. This suggests that for some pupils the difficulty lies in verbalising what the writer is doing or expressing an abstract point about a text/character: they can identify the evidence and see the text at work but are unable to synthesise this into a statement which summarises their overall understanding of the salient point in question.

search for complex/inferential ideas across a whole text
As has already been discussed in the previous article in this series – A Closer Focus: Locating Information and Understanding Ideas, scores for 3-mark questions in the national curriculum tests and NFER’s own Year 6 tests generally reveal that relatively few pupils manage to access the third mark. It has also already been mentioned in this article that most pupils still need to learn to make multiple inferences from a text, whether that be considering different possible readings of one detail or section, or finding a variety of distinct points across a text. Practising this skill is essential to prepare pupils for the literary criticism that they will be expected to perform as they progress through secondary school.

Moving forwards
Although this is merely a snapshot of the different levels of confidence pupils appear to have in making and/or supporting a range of different types of inferences, some clear common threads are apparent. Inferences are easier to make if we ourselves have some experience or knowledge that can be used to inform our understanding of texts. This is a vital aspect for reading as a whole, as any relevant background information or cultural capital that a pupil can bring to bear on a text inevitably makes it more accessible. Empathy too is a crucial skill, especially when exploring character-based texts, as is the ability to view texts from different angles in order to try to perceive alternative meanings and interpretations. The ability to express their ideas, synthesising what they have gleaned from a text into an effective summary, is also a particularly demanding aspect of reading that pupils need support with. The final article in this series will address pupils’ ability to understand writers’ uses of language and their effects but will also consider the importance of pupils’ vocabulary and written expression as well. For a more in-depth exploration of inference, see NFER researcher Anne Kispal’s report on Effective teaching of inference skills for reading.

Return to the main Implications for Teaching page to access the handy summaries for year 6 and 7 teachers.