Strengthening by Eliminating Alternatives

One of the most powerful ways to strengthen an argument is to eliminate alternative explanations for the evidence cited, thereby making the author's conclusion more probable. If an argument claims that X causes Y, showing that other possible causes of Y are absent or unlikely makes it more probable that X is indeed the cause. This method of strengthening is particularly relevant to the LNAT because many passages draw causal conclusions from correlational evidence, leaving them vulnerable to alternative explanations.

---

The Logic of Eliminating Alternatives

When an argument presents evidence and draws a conclusion, there are usually several possible explanations for the evidence. The argument favours one explanation (the conclusion), but others remain open.

Argument	Author's explanation	Alternative explanations
"After the new policing strategy was introduced, crime fell. The strategy works."	The policing strategy caused the crime reduction	Crime was already declining due to other factors; economic conditions improved; demographic changes; statistical regression to the mean
"Students in schools with smaller classes get better grades. Class size matters."	Smaller classes cause better grades	Schools with smaller classes are better funded, have better teachers, or attract more motivated students
"Countries with higher minimum wages have lower poverty rates. Raising the minimum wage reduces poverty."	Higher minimum wages reduce poverty	These countries also have stronger social safety nets, higher productivity, or different economic structures

Eliminating alternatives strengthens the argument because it narrows the field of possible explanations, making the author's conclusion more likely to be correct.

---

How This Works in Practice

Step-by-step approach

1. Identify the argument's conclusion — what causal or explanatory claim is being made?
2. List plausible alternative explanations — what else could account for the evidence?
3. A strengthener eliminates one or more alternatives — new information rules out competing explanations

Worked Example

"A new reading programme was introduced in primary schools across a London borough. After two years, reading scores improved by 15%. The programme is clearly effective."

Conclusion: The reading programme caused the improvement.

Alternative explanations:

• Other educational changes were introduced at the same time
• The cohort of students tested was unusually strong
• Teachers were more motivated because they were being observed (Hawthorne effect)
• Assessment methods changed, producing higher scores without genuine improvement

Strengthener: "No other curricular or pedagogical changes were introduced during the two-year period, and the assessment methods remained identical to those used in previous years."

This eliminates two alternatives (other changes and assessment changes), making the programme a more likely explanation for the improvement.

---

Types of Alternatives to Eliminate

1. Confounding variables

A confounding variable is a factor that influences both the observed cause and the observed effect, creating a false appearance of a causal relationship.

Argument: "People who eat organic food are healthier. Organic food is better for health."

Confounding variable: People who buy organic food may also exercise more, earn more (affording better healthcare), and avoid other unhealthy behaviours.

Strengthener that eliminates the confounder: "A study comparing matched groups with similar incomes, exercise levels, and lifestyle habits found that those eating organic food still had measurably better health indicators."

2. Reverse causation

The causal direction may be opposite to what the argument claims.

Argument: "Happy employees are more productive. Companies should invest in employee wellbeing to boost productivity."

Reverse causation: Perhaps productive employees become happy because they accomplish more, rather than happiness causing productivity.

Strengthener that eliminates reverse causation: "A longitudinal study tracked employees before and after a wellbeing intervention, finding that the intervention improved wellbeing first, and productivity increases followed three to six months later."

3. Coincidence or regression to the mean

An observed change may simply reflect natural fluctuation.

Argument: "After the mayor introduced a curfew for under-18s, youth crime dropped by 20%. The curfew works."

Coincidence/regression: Youth crime may have been unusually high in the period before the curfew and would have dropped anyway (regression to the mean).

Strengthener that eliminates coincidence: "Neighbouring boroughs without a curfew saw no change in youth crime rates over the same period, and the reduction in the curfew borough has been sustained over three consecutive years."

4. Selection bias

The group studied may not be representative, creating misleading results.