Scientific reasoning and conclusion drawing

IB Chemistry HL· 45 min read

1. Types of Error and Impact on Conclusions★★☆☆☆⏱ 10 min

All experimental chemistry data contains error, which directly affects the certainty of any conclusion you draw. It is critical to distinguish between the two main categories of error to correctly evaluate your results.

📘 Definition

Random Error

Unpredictable variation between measurements that leads to a spread of values around the true value. Reduced by repeating measurements and averaging results, but cannot be eliminated entirely.

Example:

Variation in reading a burette meniscus between different trials.

📘 Definition

Systematic Error

Consistent, repeatable error that shifts all measurements in the same direction from the true value. Cannot be reduced by repeating measurements, only eliminated by improving experimental design.

Example:

An uncalibrated pH meter that always reads 0.2 pH units too high.

📐 Worked Example

A student measures the enthalpy of neutralization of HCl and NaOH five times, getting results: -55.2, -56.1, -54.8, -55.9, -55.5 kJ mol⁻¹. The accepted literature value is -57.3 kJ mol⁻¹. Identify the dominant error type and explain its impact on the conclusion.

1
Calculate the average of the student's results:
2
$\bar{x} = \frac{-55.2 -56.1 -54.8 -55.9 -55.5}{5} = -55.5 \text{ kJ mol}^{-1}$
3
The spread of results is only 1.3 kJ mol⁻¹ across 5 trials, so random variation is small. All results are consistently lower in magnitude than the literature value.
4
This consistent directional shift means the dominant error is systematic. For enthalpy experiments, this is usually caused by unaccounted heat loss to the surroundings.
5
Impact on conclusion: The result is not caused by random chance, it reflects a systematic flaw in methodology that shifts all measurements.

Exam tip:

Always link the type of error directly to your conclusion, do not just list errors without context.

2. Drawing Valid Conclusions from Hypotheses★★☆☆☆⏱ 12 min

A valid conclusion must directly answer your original research question and be fully supported by your collected quantitative and qualitative data. You must explicitly compare your results to your hypothesis, and use uncertainty analysis to confirm if any observed differences are significant.

📐 Worked Example

A student tests the hypothesis that increasing temperature increases the equilibrium constant $K_{c}$ for $N_{2} O_{4} (g) ⇌ 2 N O_{2} (g)$ . They measure $K_{c} = 1.08 \pm 0.05$ at 25°C and $K_{c} = 1.72 \pm 0.07$ at 40°C. Is the hypothesis supported by the data?

1
Compare the average values: $K_{c}$ at 40°C is larger than at 25°C (1.72 > 1.08)
2
Calculate the difference between values and the total combined uncertainty:
3
$\text{Difference} = 1.72 - 1.08 = 0.64 \\ \text{Combined uncertainty} = 0.05 + 0.07 = 0.12$
4
Since the difference (0.64) is larger than the total uncertainty (0.12), the increase is significant and not caused by random error.
5
Conclusion: The hypothesis is supported by the experimental data.

3. Reliability and Validity of Results★★★☆☆⏱ 15 min

Reliability refers to how repeatable your results are, while validity refers to whether your experiment actually measures what you set out to measure. Both are required for a meaningful conclusion.

📘 Definition

Reliability vs Validity

Reliability: Consistency of results when the experiment is repeated. Validity: Extent to which the conclusion is free from bias and actually answers the research question.

📐 Worked Example

A student investigates how surface area of calcium carbonate affects reaction rate with HCl. They test 1g of large, medium, and small chips, measuring mass loss over 5 minutes. They find smaller chips give faster rate, but the room temperature increased by 4°C between the first and last trial. Comment on the validity of the conclusion.

1
The experiment tests the effect of surface area, but an uncontrolled variable (temperature) changed between trials.
2
Increased temperature also increases reaction rate, so the observed rate increase could be partially or fully caused by temperature change, not just surface area.
3
This confounding variable reduces the validity of the conclusion that surface area caused the change in rate.
4
Repeating trials at a controlled constant temperature would improve the validity of the conclusion.

4. Limitations and Justified Improvements★★★☆☆⏱ 12 min

A key part of conclusion drawing is identifying significant limitations of your investigation and suggesting specific, realistic improvements. Generic improvements like 'do more repeats' or 'be more careful' do not earn full marks in IB assessment.

Limitations must be significant: they must actually affect your results, not just be minor.
Improvements must directly address the limitation you identified.
Improvements must be feasible for a standard school chemistry laboratory.

📐 Worked Example

A student identifies the limitation: 'Heat loss to the air caused our measured enthalpy of combustion to be 15% lower than the literature value'. What is an appropriate improvement?

1
Poor generic improvement: 'Be more accurate when measuring heat' (does not address the limitation)
2
Poor unfeasible improvement: 'Use a professional bomb calorimeter that costs $10,000' (not possible in a school lab)
3
Good specific improvement: 'Add an insulating polystyrene layer around the calorimeter and fit a lid to reduce heat loss, then repeat trials to confirm results'
4
This improvement directly addresses the limitation of heat loss and can be implemented in a school laboratory.

5. Common Pitfalls

Wrong move:

Stating 'human error' as a source of error or limitation

Why:

This is too generic and does not explain what error occurred or how it affected your conclusion

Correct move:

Name the specific error, e.g. 'random error in reading the burette meniscus caused variation in titre values'

Wrong move:

Claiming your hypothesis is 'proven' by your results

Why:

A single experiment can never prove a scientific hypothesis; new evidence could always contradict it

Correct move:

State that the data 'supports' or 'does not support' your hypothesis

Wrong move:

Ignoring uncertainty when assessing differences between results

Why:

A difference between average values that is smaller than combined uncertainty could be entirely due to random error

Correct move:

Always compare the size of the difference to the total uncertainty to check for significance

Wrong move:

Suggesting improvements that do not link to your stated limitations

Why:

Examiners expect you to connect weaknesses directly to solutions to show evaluation skills

Correct move:

For every limitation you list, write one specific improvement that directly solves that problem

Wrong move:

Changing your conclusion to match the literature value when your results differ

Why:

IB awards marks for your analysis of your own data, not for matching expected values

Correct move:

Explain the specific errors or limitations that caused the discrepancy between your results and the literature

6. Quick Reference Cheatsheet

Concept	Key Definition	Exam Tip
Random error	Unpredictable spread around true value	Reduced by repeats, does not affect accuracy
Systematic error	Consistent directional shift in all measurements	Affects accuracy, fixed by improving method
Reliability	Repeatability of experimental results	Low spread = high reliability
Validity	Conclusion measures what it claims	Confounding variables reduce validity
Draw conclusion	Answer research question, test hypothesis	Always use uncertainty to check significance
Evaluate	List specific limitations, link to results	Improvements must be specific and feasible

7. Frequently Asked

How many limitations do I need for my IA conclusion?

For full marks, you need 2-3 specific, significant limitations that directly impacted your results, not generic vague weaknesses like 'human error'.

What if my results don't match the literature value?

Do not change your conclusion to match the expected value. Instead, explain the specific errors or limitations that caused the discrepancy, which shows good evaluation skills.

When this came up on past exams

AI-estimated based on syllabus patterns — cross-check with official past papers for accuracy. Use only as revision-focus signals.

2025 · 3
Evaluate validity of experimental conclusion
2024 · IA
Required for conclusion/evaluation criterion
2023 · 3
Identify error sources affecting conclusion

Going deeper

official guideIB Chemistry 2025 Practical Skills HandbookFull IA assessment criteria details

What's Next

Mastering scientific reasoning and conclusion drawing is the core of the internal assessment's Conclusion and Evaluation criterion, which makes up 25% of your total IA mark. These skills are also tested extensively in Paper 3 Section A, where you are required to analyze unseen experimental data and evaluate conclusions. Developing strong reasoning skills will not only boost your practical assessment marks but also help you structure clear, evidence-based answers for all exam questions. Build on these skills by reviewing uncertainty calculation and IA criteria.

→Internal assessment requirements