How to Read | Class No.0018 | The Causal Read — Trace Causes, Mechanisms, Conditions and Consequences Without Confusing Sequence With Explanation

Class No.0002: The Meaning Read · Class No.0003: The Evidence Read · Class No.0017: The Comparative Read

Cause-and-effect is a text structure—and a reasoning problem

IES reading-comprehension resources identify cause and effect as one of the common informational text structures worth teaching explicitly. Signal words may include because, therefore, as a result, consequently, leads to, but clue words alone are not enough.

The deeper reading job is to reconstruct the relationship the writer claims.

Sequence is not cause

Consider:

The lights flickered. A minute later, the computer shut down.

The sequence makes a causal relationship plausible, but it does not prove it.

There may have been:

• a power fluctuation;
• a failing battery;
• overheating;
• an unrelated software crash.

Time order is necessary for many causal claims, but it is not sufficient.

Cause can be direct or indirect

A direct cause sits close to the outcome.

An indirect cause works through another factor.

Example:

heavy rain → river rises → road floods → school closes

If the question asks why school closed, “heavy rain” is part of the causal chain, but “road flooding” may be the direct mechanism.

Good readers match the level of cause to the question.

Trigger is not always underlying cause

A small event can trigger a larger failure because conditions were already unstable.

A spark may start a fire, but dry vegetation, heat and wind made the system vulnerable.

A single missed payment may trigger a business collapse, but long-term cash-flow weakness may be the deeper cause.

Texts often mix triggers and structural causes.

Necessary and sufficient conditions matter

A necessary condition must be present for the outcome to occur in the stated model.

A sufficient condition is enough to produce the outcome under the relevant assumptions.

Many real-world causes are neither individually necessary nor sufficient. They change probability.

Readers should avoid turning “contributes to” into “determines.”

Risk factors are not deterministic causes

If a factor increases risk, it does not guarantee outcome.

A text may say:

Sleep deprivation is associated with greater accident risk.

Do not rewrite this as:

Sleep deprivation always causes accidents.

Probability language belongs inside the causal claim.

Mechanism explains how

A causal claim becomes stronger conceptually when the mechanism is visible.

Ask:

• What changes first?
• What process carries the effect?
• What intermediate steps are required?
• Where could the chain break?

Mechanism turns “X affects Y” into an explanation.

Mechanism is not always known

Sometimes strong evidence establishes an effect before the full mechanism is understood.

Do not reject a causal claim solely because every intermediate step is not known.

Instead distinguish:

• evidence of effect;
• evidence of mechanism;
• remaining uncertainty.

Correlation invites causal questions, not causal certainty

When two variables move together, several explanations remain:

• X causes Y;
• Y causes X;
• Z causes both;
• selection creates the pattern;
• the relationship is unstable or coincidental.

The Causal Read asks what evidence discriminates among these.

Confounders create alternative causal stories

Suppose students who read more books score higher in English.

Possible confounders include:

• home literacy environment;
• prior achievement;
• motivation;
• school resources.

The association may still matter, but causal interpretation requires more care.

Counterfactual reasoning

Causal thinking often asks:

What would likely have happened if the cause were absent?

We usually cannot observe the same person, city or country both with and without the cause at the same time.

Research designs try to approximate this missing counterfactual through comparison.

Comparison groups matter

If a programme’s participants improve, ask:

• Would they have improved anyway?
• What happened to a comparable group without the programme?
• Were the groups different before treatment?

Causal reading enters methodology.

Natural experiments and experiments

Advanced readers should recognise that different designs support causal claims differently.

Random assignment can strengthen causal inference because it helps balance alternative explanations. Natural experiments and careful observational designs can also provide evidence, but assumptions matter.

The reader need not become a statistician to ask what design produced the claim.

Causal chains can have multiple steps

Example:

higher interest rates → borrowing becomes more expensive → some spending and investment fall → demand pressure eases → inflation may slow

Each arrow is a claim.

A reader can test the chain one link at a time.

One broken link can weaken the whole explanation

If the text assumes expensive borrowing reduces spending but the affected group has little debt, the mechanism may be weaker in that context.

Causal reading attends to conditions.

Conditions are part of causation

Many causes operate only under particular conditions.

A seed needs water, but water alone is not enough.

A policy incentive may work only where people can respond to it.

A teaching strategy may work differently depending on prior knowledge.

Ask:

Under what conditions does this causal relationship hold?

Moderators change the strength of an effect

A relationship may be stronger for one group than another, or in one environment than another.

Advanced readers should notice language such as:

• especially among;
• only when;
• stronger for;
• depends on;
• under conditions of.

These phrases limit generalisation.

Mediators carry effects

A mediator is an intermediate process.

Example:

feedback → better error detection → changed practice → improved performance

If the text argues that feedback improves performance, the mediator may explain how.

Cause can be reciprocal

Sometimes X affects Y and Y affects X.

Example:

reading knowledge ↔ reading comprehension

More knowledge improves comprehension; reading can build more knowledge.

Linear cause-and-effect may be inadequate.

Feedback loops create circular causation

A positive feedback loop amplifies change.

A negative feedback loop counteracts change.

Example:

success → confidence → more practice → more success

or:

error detected → corrective action → error reduced

Class No.0020’s Systems Read will develop this further.

Delayed effects are easy to miss

A cause may operate long before the visible outcome.

Policy changes, environmental processes and learning often include delays.

If the reader expects immediate effect, they may wrongly conclude the cause had no impact.

Cumulative causes

One small exposure may do little. Repeated exposure over time may matter.

Texts about health, climate, learning and finance often involve cumulative processes.

Ask whether the causal unit is an event or an accumulation.

Threshold effects

A system may absorb change until a threshold is crossed.

Then the outcome changes abruptly.

Readers should notice terms such as:

• critical threshold;
• tipping point;
• minimum level;
• only after.

Multiple causes can converge

An outcome may require several contributing factors.

Do not insist on “the cause” when the text supports a causal set.

Ask:

• which factors are background;
• which are triggers;
• which amplify;
• which constrain.

One cause can create multiple consequences

A policy may reduce cost, increase demand and create new congestion.

Causal reading should trace branching outcomes, not only the intended one.

Unintended consequences

A strong causal reader asks:

What else changes when we intervene here?

This is essential in policy, ecology, economics and systems thinking.

Causal language has different strengths

Compare:

• causes;
• leads to;
• contributes to;
• increases the likelihood of;
• is associated with;
• may influence.

Do not flatten them.

Cause-and-effect clue words are helpful but imperfect

IES resources teach cause-and-effect structure using terms such as because, therefore, cause, effect, so, while cautioning more broadly that text structure cannot be identified by clue words alone.

The reader must test whether the relationship really is causal.

Explanatory text can contain competing causes

A history text may propose economic, political and social causes.

A science article may compare several mechanisms.

A policy report may distinguish proximate and root causes.

The reader should not reduce a multi-causal explanation to the most memorable factor.

Root cause is task-dependent

Ask why a student missed a question.

Possible chain:

misread “except” → searched wrong evidence → chose plausible distractor → lost mark

The immediate cause of the wrong answer is distractor selection. The first weak link is instruction reading.

Which cause matters depends on the intervention goal.

Causal reading in narrative

Stories depend on causality.

Ask:

• What choice changed the situation?
• What event forced the next event?
• Which consequence was unintended?
• Could the plot still happen without this event?

A plot is more than sequence; it is structured consequence.

Causal reading in character analysis

Do not assume one motive.

A character may act because of:

• fear;
• loyalty;
• misunderstanding;
• social pressure;
• immediate opportunity.

Rank motives by textual evidence.

Causal reading in science

Ask:

• what variable changes;
• what mechanism connects it to outcome;
• what controls exist;
• what conditions matter;
• what alternative explanations remain.

Causal reading in history

Separate:

• long-term conditions;
• medium-term developments;
• immediate triggers;
• actor decisions;
• contingency.

Chronology supports but does not replace explanation.

Causal reading in economics

Look for incentives, constraints, behavioural response and feedback.

Economic texts often use models where other conditions are held constant. Ask what happens when they are not.

Causal reading in education

Improvement after tuition does not automatically prove tuition caused the improvement.

Ask about:

• prior trend;
• school teaching;
• practice volume;
• motivation;
• assessment difficulty;
• other interventions.

For instructional decisions, causal humility matters.

Kindergarten: because

Use simple stories.

Ask:

Why did the character get wet?

Require a cause from the story.

Primary 1: before and because

Contrast:

before: event order
because: causal relationship

Students learn that sequence and cause differ.

Primary 2: cause chains

Use three boxes:

cause → event → result

Then change the first box and predict what might change later.

Primary 3: multiple causes

Read an informational text where one outcome has two causes.

Ask whether both are required or either can contribute.

Primary 4: cause versus coincidence

Use simple examples:

I wore red socks and it rained.

Ask whether the socks caused the rain and what evidence would be needed.

Primary 5: direct and indirect cause

Trace a longer chain and ask which link the question is targeting.

Primary 6: causal claims in nonfiction

Highlight words such as causes, linked to, associated with, may lead to.

Rank their causal strength.

Secondary 1: text structure

Identify cause-and-effect passages and generate a gist that preserves the relationship.

This aligns with IES text-structure guidance.

Secondary 2: alternative explanations

For each causal claim, generate one plausible alternative.

Then ask what evidence would distinguish them.

Secondary 3: mechanism diagrams

Turn prose into arrows.

Label:

• cause;
• mediator;
• condition;
• outcome.

Secondary 4: causal evaluation

Read a claim and ask:

• time order;
• association;
• mechanism;
• confounders;
• comparison group;
• scope.

JC: causality becomes model comparison

Two writers may agree on an outcome but propose different causal mechanisms.

Compare which mechanism better explains timing, variation and exceptions.

JC: causal claims in public debate

Policy arguments often leap from correlation to intervention.

Ask:

If the proposed cause is real, what should change when we intervene on it?

This connects explanation to prediction.

The Causal Read protocol

1. Identify the outcome.
2. List proposed causes.
3. Separate sequence from cause.
4. Trace the mechanism.
5. Identify conditions and moderators.
6. Look for alternative explanations.
7. Check evidence design.
8. Map direct, indirect and feedback effects.
9. Match claim strength to evidence.

The Causal Chain

Write:

cause → mechanism → intermediate state → outcome → secondary consequence

Then place question marks where evidence is weak.

The Alternative Cause Test

Ask:

What else could produce this outcome?

If no alternative can even be imagined, the reader may be accepting the first story too quickly.

The Intervention Test

Ask:

If X causes Y, what should happen if X is changed?

This links causal claim to observable implication.

Common failure: post hoc causation

B happened after A, therefore A caused B.

Repair: alternative explanations and mechanism.

Common failure: single-cause story

A complex event is reduced to one memorable factor.

Repair: causal set and level analysis.

Common failure: cause becomes certainty

A risk factor is rewritten as deterministic.

Repair: preserve probability language.

Common failure: mechanism without evidence

A plausible story is treated as proof.

Repair: distinguish explanatory plausibility from empirical support.

Common failure: evidence without mechanism

A strong association is described as if the pathway were obvious.

Repair: state what is known and unknown separately.

Three learners, three causal failures

Maren confuses sequence with cause. Her repair is counterfactual questioning.

Iona sees several causes but cannot distinguish trigger from background condition. Her repair is causal-layer mapping.

Leonie understands the mechanism but ignores alternative explanations. Her repair is confounder generation.

A 45-minute Causal Read tutorial

1. 5 minutes: identify outcome and claimed cause.
2. 10 minutes: build mechanism chain.
3. 10 minutes: identify conditions and alternatives.
4. 5 minutes: compare sequence versus causation examples.
5. 10 minutes: test a real-world causal claim.
6. 5 minutes: write a calibrated conclusion.

The Causal Read audit

Ask:

• What outcome am I explaining?
• What cause is claimed?
• Did the cause precede the outcome?
• What mechanism connects them?
• What conditions are required?
• Is the cause direct, indirect, necessary, sufficient or probabilistic?
• What alternative explanations exist?
• What evidence design supports the causal claim?
• Are there feedback loops or delays?
• Does the wording match the evidence strength?

Research anchors

• Institute of Education Sciences — Text Structure: Cause and Effect
• Institute of Education Sciences — PRISMS Text Structures
• What Works Clearinghouse — Cause/Effect Text Structure Intervention
• What Works Clearinghouse — Intelligent Tutoring for Structure Strategy
• OECD — Systemic Thinking for Policy Making

Continue through SETC

• How to Read | Class No.0002 | The Meaning Read
• How to Read | Class No.0003 | The Evidence Read
• How to Read | Class No.0017 | The Comparative Read
• English Argument Evaluation Performance

Final lesson

Causal reading resists the comfort of the first explanation.

Trace the chain. Separate trigger from background. Separate sequence from mechanism. Ask what conditions make the relationship work and what else could produce the same result.

Then match the strength of your causal language to the strength of the evidence.

“Because” is a small word.

It carries a very large burden.

Return to the English Learning Library · Primary English · Secondary English · Advanced English / General Paper

Continue with the How to Read collection.