October 25, 2025 by Doug McCurry from BooBook Education
Learning from testing: testing for learning
While working on GAT preparation materials recently, I have been increasingly aware of the way a good test has educational potential, and that a good test is a way of examining thinking processes for teachers and students. There is reason for thinking that good test items can be used to teach students things.
I am not sure how much these claims apply to numeracy tests, but I am confident that they apply significantly to literacy tests. A good reading test in particular offers opportunities for discussing and explicitly teaching higher order reading skills in a focussed and specific fashion. While we are testing learning we can also be learning from testing.
This is no new insight of course and it is implicit in the Assessment for Learning movement. While I have been interested in the work of that movement and have looked at curriculum and pedagogy developed in the spirit of Assessment for Learning, I have not come across any examples of testing material being used systematically for teaching and learning. It seems to me that preparation for the GAT can be used for explicit teaching of reading and interpretation skills.
With these issues in mind that I developed the following classification of reading skills that have a particular focus on higher order reading.

The higher order reading skills listed here might seem to be no more than a version of Bloom’s taxonomy. This is not surprising as that taxonomy was developed by classifying test items.
It would seem to me useful to teach students about the concepts of analysis, inference, interpretation, synthesis and critical reading, and that the best way of teaching them these concepts would be by doing and examining good reading items that test these skills.
I could offer a good deal of justification for what might seem outlandish claims to some, but instead I will offer a commentary on three GAT items from the test of 2024 that show what teachers can teach and student can learn from examining such testing material.
The GAT is fundamentally a reading test, but it also tests a broad range of thinking skills. I will focus on reading and interpretation skills, but other kinds of thinking could be examined in different units and items.
When working on the test I wrote a label for each of the 75 reading and humanities items in the GAT of 2024 and then asked ChatGPT to synthesise and summarise the reading skills tested in the items. The AI bot produced the following seven kinds of skills and the diagram of them.
1. Literal Comprehension (Locating and Identifying Information)
These items test the ability to locate and understand explicitly stated information.
2. Interpretive Comprehension (Understanding Meaning and Language Use)
These assess how well a reader can interpret language in context, including suggested and figurative meanings.
3. Inferential Comprehension (Reading Between the Lines)
These involve deducing meaning not directly stated, such as authorial intent, implications, and inferred viewpoints.
4. Evaluative Comprehension (Judging Content and Tone)
These assess a reader’s ability to evaluate, compare, and judge the content, intent, or quality of the material.
5. Analytical Comprehension (Understanding Purpose and Structure)
These involve understanding the author's purpose, argument structure, or text organization.
6. Visual Literacy (Interpreting Non-Textual Elements)
These items assess the ability to read and interpret presentation, images and diagrams.
7. Synthesis and Summarisation
These test the ability to extract main ideas and integrate multiple sources or statements.

The conclusion of CGPT about my classification of the GAT items is as follows.
The reading skills assessed cover a comprehensive range from basic literal understanding to higher-order skills like inference, synthesis, evaluation, and visual interpretation. The test is designed to evaluate how well a reader can decode, interpret, infer, evaluate, and synthesise information across both textual and visual forms.
The two Sections of the GAT would seem to differ with Section A focussed on more basic reading skills and everyday issues and B focussed on more higher order skills in less familiar contexts, but this difference is more a matter of appearance than substance. There is a good deal of inference and interpretation in Section A.
The most basic way of teaching students from the GAT test is by explaining and discussing the answers. Most students will get between 25% and 40% of questions wrong, and they have things to learn by examining the items they got wrong. They should work out the correct answers and consider why they got the items wrong. The better the test item the more real and informative the mistakes student make, and the more they can learn by understanding their mistakes.
In our GAT scoring service we produce the following report showing which items a student got wrong for each part of the test:

There is a good deal for this student to conveniently learn from the 66 incorrect answers for Section A of the GAT.
I will examine a unit from Section A of 2024 to show something of what might be learned from reviewing test items. The neat little unit below has a stimulus text of 113 words followed by three questions. I offer a brief explanation of the answer after each item.
Awe and wonder
The following text is taken from a book of personal reflection by Julia Baird.
The sun's rise and the sun's retreat bookend our days with awe. We often
take awe for granted, and yet it is something both modern scientists and
ancient philosophers have told us to hunt. Awe makes us stop and stare. Being
awestruck dwarfs us, humbles us, makes us aware we are part of a universe
unfathomably larger than ourselves; it even, social scientists say, makes us
kinder and more aware of the needs of the community around us.
Wonder is a similar sensation, and the two feelings are often entwined. Wonder
makes us stop and ask questions about the world, while marvelling over
something we have not seen before, whether spectacular or mundane.
98 A reader comments: 'Awe and wonder can both lead to a greater scientific interest in nature.'
Which of the following most accurately describes the position of the text in relation to this opinion?
A. The text supports this opinion in relation to awe only.
B. The text supports this opinion in relation to wonder only.
C. The text supports this opinion more explicitly in relation to wonder than in relation to awe. *
D. The text does not specifically deal with this opinion in relation to either awe or wonder.
The answer is C because the writer distinguishes awe and wonder, and wonder is seen by the writer as more scientific than awe even though the two feelings are entwined and related.
99 Which of the following would the writer of the text likely associate with observing, for the first time, a cell under a microscope?
A. mostly awe
B. mostly wonder *
C. both awe and wonder
D. neither awe nor wonder
The answer is B because the scenario involves the use of a scientific implement and is likely to lead to scientific questions.
100 A reader makes two statements about the effects of experiencing awe.
Statement 1: Awe makes us conscious of our physical limitations.
Statement 2: Awe gives us a sense of knowing about the world.
Which of the following is correct?
A. Only Statement 1 is consistent with the text. *
B. Only Statement 2 is consistent with the text.
C. Both Statements 1 and 2 are consistent with the text.
D. Neither Statement 1 nor 2 is consistent with the text
The answer is A because awe involves feelings that dwarf and humble us even though the universe itself is unfathomable.
These three items based on a brief text would be well worth discussing with students.
In a close reading exercise with a class the first question we might ask students is what the text and the items are about.
The answer is that the writer makes a distinction between awe and wonder.
It would be good if students could see that this is what the text is doing.
We could then ask what distinction the writer makes between awe and wonder.
If a student can answer these constructed response questions, then they can answer the MCQ items. The sensations of awe and wonder are ‘intertwined’ but the writer implies that wonder involves questions about the world while awe is marvelling at the world.
What can we draw students’ attention to in this text and the items?
The underpinning teaching point is that subtle distinction between similar terms that can still be distinguished from each other. An idea worth teaching and demonstrating.
I asked ChatGPT the following question about the unit and its response is below.
The following is a reading test unit with three questions about the introductory text.
The correct answers are asterisked.
Explain and justify the correct answers.
Explain what skills reading the text and answering the questions involve.
The CGPT response
Here is a breakdown of the correct answers, their justifications, and the reading skills required for each question:
Question 98
Correct Answer: C. The text supports this opinion more explicitly in relation to wonder than in relation to awe.
Justification:
- The text says awe “makes us stop and stare,” and has psychological and social effects (humility, kindness).
- It also says wonder “makes us stop and ask questions about the world,” which is more explicitly linked to curiosity and inquiry—core to scientific interest.
- Therefore, while awe might lead indirectly to scientific interest (through humility or awareness), wonder is directly tied to questioning and exploration, which are foundational to science.
Skills involved:
- Interpretive comprehension: understanding nuance in phrasing (e.g. the suggestion that to “ask questions” is key to science).
- Evaluative comprehension: comparing support for different ideas within the text.
- Inference: connecting abstract feelings (awe/wonder) to intellectual behaviour (scientific interest).
Question 99
Correct Answer: B. mostly wonder
Justification:
- Observing a cell for the first time is likely to provoke curiosity and questioning—hallmarks of wonder as defined in the text.
- While awe might also occur, the text more clearly links wonder to new observations and asking questions about the world—“something we have not seen before.”
- Therefore, the most fitting description is mostly wonder, not awe or both.
Skills involved:
- Literal and interpretive comprehension: matching a described experience (seeing a cell) to definitions given in the text.
- Inferential reasoning: applying the text’s definitions to a hypothetical but plausible scenario.
Question 100
Correct Answer: A. Only Statement 1 is consistent with the text.
Justification:
- Statement 1: “Awe makes us conscious of our physical limitations” is consistent with the phrase “dwarfs us, humbles us”—clearly emphasizing our smallness.
- Statement 2: “Awe gives us a sense of knowing about the world” is not supported. In fact, awe emphasizes the vastness and unfathomability of the universe, which implies less certainty, not more.
- Thus, only Statement 1 is textually consistent.
Skills involved:
- Close reading: pinpointing what the text says about the effects of awe.
- Evaluative comprehension: determining which statements accurately reflect the text’s message.
- Discerning consistency: judging whether a reader’s claim aligns with textual evidence.
Overall Reading Skills Assessed in This Unit
Skill | How it's used |
Literal Comprehension | Recognising explicit definitions and descriptions (e.g. what awe/wonder do). |
Interpretive Comprehension | Interpreting the relationship between feelings and behaviours (e.g. wonder and inquiry). |
Inferential Comprehension | Applying abstract ideas to hypothetical or implied situations. |
Evaluative Comprehension | Judging the strength and accuracy of reader statements compared to the text. |
Consistency and Distinction | Distinguishing between similar ideas (awe vs. wonder) and checking textual fidelity. |
With little effort on my part and with the help of CGPT this text and the three questions give me a good deal to explore about higher order reading with students in a specific and concrete way. This kind of analysis would amount to the explicit teaching of higher order reading.
Students could also do this kind of exploration of GAT items with ChatGPT for themselves.
There are quite a few students who can learn quite a lot from these three GAT items.


