What Is Analysis and Discussion?

The analysis and discussion section is where you explain what your experimental results mean.

The results section answers:

What happened?

The analysis and discussion answers:

What do the results tell us, and why might this have happened?

A strong analysis does more than repeat numbers from a table. It identifies patterns, uses evidence, connects the results to scientific ideas, discusses unusual results, and considers how convincing the evidence is.

---

Results vs Analysis

This distinction is one of the most important parts of scientific writing.

Results:

The mean reaction time decreased from 84.2 s at 20°C to 31.6 s at 50°C.

This reports what happened.

Analysis:

The decrease in reaction time indicates that the reaction rate increased as temperature increased.

This interprets the results.

Discussion:

This pattern can be explained because particles have greater average kinetic energy at higher temperatures, increasing the frequency of collisions and the proportion of collisions with sufficient energy to react.

This connects the interpretation to scientific theory.

A useful progression is:

Data → Pattern → Interpretation → Scientific Explanation

---

Start with the Data

Analysis must be based on evidence.

Suppose an experiment produces:

A weak analysis might say:

Temperature affected the reaction.

This is too vague.

A stronger analysis is:

As temperature increased, reaction time decreased. Increasing temperature from 20°C to 60°C reduced the mean reaction time from 82 s to 27 s.

The claim is now supported by numerical evidence.

---

Identify the Overall Pattern

Begin by asking:

What happens to the dependent variable as the independent variable changes?

Common patterns include:

• increasing
• decreasing
• remaining approximately constant
• increasing and then leveling off
• decreasing and then leveling off
• reaching a maximum
• reaching a minimum
• showing no clear relationship

Describe the pattern actually shown by the evidence rather than the pattern you expected.

---

Positive Relationships

A positive relationship occurs when one variable tends to increase as another increases.

For example:

As force increases:

extension increases

A suitable analysis might state:

There was a strong positive relationship between applied force and spring extension over the measured range.

---

Negative Relationships

A negative relationship occurs when one variable tends to decrease as another increases.

For example:

As temperature increases:

reaction time decreases

A suitable statement is:

Reaction time showed a negative relationship with temperature over the range tested.

Be careful with terminology. A decreasing reaction time may indicate an increasing reaction rate.

---

Linear Relationships

A relationship is approximately linear when the data follow a straight-line pattern.

For example:

1 N → 2 cm

2 N → 4 cm

3 N → 6 cm

4 N → 8 cm

The extension increases by approximately the same amount for each additional newton.

A useful analysis might state:

Extension increased approximately linearly with applied force, increasing by about 2 cm for each additional newton.

---

Nonlinear Relationships

Not every relationship is linear.

Suppose:

The response increases but begins to level off.

This is a nonlinear relationship.

A good analysis should describe the curve rather than incorrectly calling it linear.

---

No Clear Relationship

Sometimes the results do not show an obvious pattern.

That is a scientifically valid result.

For example:

No consistent relationship was observed between fertilizer concentration and plant height over the range tested.

Do not invent a trend because the hypothesis predicted one.

Scientific analysis must follow the evidence.

---

Use Numerical Evidence

Strong analysis uses actual values.

Weak:

The plants grew more in brighter light.

Better:

Mean plant growth increased as light intensity increased.

Stronger:

Mean plant growth increased from 3.2 cm at the lowest light intensity to 8.7 cm at the highest light intensity.

Numbers make scientific claims more precise.

---

Compare Values

Analysis often involves comparing measurements.

Suppose:

Treatment A = 12.4 cm

Treatment B = 18.6 cm

Difference:

18.6 − 12.4 = 6.2 cm

You could write:

Plants receiving Treatment B grew an average of 6.2 cm more than plants receiving Treatment A.

This communicates the difference more clearly than simply repeating both values.

---

Percentage Change

Sometimes a percentage provides a more meaningful comparison.

Suppose a measurement increases from:

40 g to 50 g

Increase:

50 − 40 = 10 g

Percentage increase:

(10 ÷ 40) × 100 = 25%

A suitable analysis is:

The measured mass increased by 10 g, corresponding to a 25% increase from the initial value.

---

Use Graphs to Identify Patterns

Graphs make relationships easier to identify.

When examining a graph, consider:

• direction of the trend
• steepness
• shape
• maximum or minimum
• plateau
• anomalies
• clusters
• variation
• line or curve of best fit

Do not simply say:

The graph goes up.

Describe what the graph represents:

Mean extension increased approximately linearly as applied force increased.

---

Tables and Graphs Have Different Roles

Tables are useful for showing:

• exact measurements
• repeated trials
• calculated values
• precise comparisons

Graphs are useful for showing:

• patterns
• trends
• relationships
• anomalies
• changes across a range

A strong analysis may refer to both.

For example:

The graph shows a clear decreasing trend, while the table shows that mean reaction time decreased from 82.1 s at 20°C to 27.3 s at 60°C.

---

Do Not Simply Repeat the Entire Table

An analysis should not rewrite every measurement.

Instead, select evidence that supports the important patterns.

Weak:

At 20°C it was 82 s. At 30°C it was 63 s. At 40°C it was 46 s. At 50°C it was 34 s. At 60°C it was 27 s.

Better:

Reaction time decreased consistently as temperature increased, falling from 82 s at 20°C to 27 s at 60°C.

The second version communicates the important relationship efficiently.

---

From Pattern to Scientific Explanation

After identifying a pattern, explain why it may have occurred.

Suppose increasing temperature decreases reaction time.

Pattern:

Higher temperature produced shorter reaction times.

Interpretation:

The reaction rate increased.

Scientific explanation:

At higher temperatures, particles have greater average kinetic energy. This leads to more frequent collisions and increases the proportion of collisions with sufficient energy to react.

This is the difference between simply describing data and scientifically analyzing it.

---

Claim–Evidence–Reasoning

A useful structure for analysis is CER.

Claim: What pattern or relationship does the evidence show?

Evidence: Which measurements support the claim?

Reasoning: Which scientific ideas explain the relationship?

For example:

Claim: Increasing temperature increased reaction rate.

Evidence: Mean reaction time decreased from 82 s at 20°C to 27 s at 60°C.

Reasoning: Higher temperatures give particles greater average kinetic energy, increasing the frequency and effectiveness of collisions.

---

Example: Spring Investigation

Research question:

How does applied force affect the extension of a spring?

Results:

Analysis:

Spring extension increased approximately linearly with applied force. Increasing the force from 1 N to 5 N increased mean extension from 1.9 cm to 10.1 cm. Each additional newton produced approximately 2 cm of additional extension. This is consistent with Hooke's law over the measured range, where extension is proportional to applied force provided the elastic limit is not exceeded.

---

Example: Pendulum Investigation

Research question:

How does pendulum length affect period?

Suppose:

20 cm → 0.90 s

40 cm → 1.27 s

60 cm → 1.55 s

80 cm → 1.79 s

100 cm → 2.01 s

A suitable analysis is:

Pendulum period increased as pendulum length increased. Increasing the length from 20 cm to 100 cm increased the mean period from 0.90 s to 2.01 s. However, the relationship was not linear because doubling the pendulum length did not double its period.

This describes both:

direction

and:

shape

of the relationship.

---

Example: Density Investigation

Suppose students measure the mass and volume of several samples of the same material.

As volume increases:

mass also increases

If the samples are made of the same uniform material:

m = ρV

A graph of mass against volume should show an approximately linear relationship.

The gradient represents:

density

A strong analysis connects the observed graph to the relevant scientific equation.

---

Comparing Results with a Hypothesis

The discussion should consider whether the evidence supports the original hypothesis.

Suppose the hypothesis was:

Increasing temperature will decrease reaction time.

If the results show this pattern, write:

The results support the hypothesis.

Do not simply write:

The hypothesis was correct.

Scientific evidence supports or does not support a prediction; a single experiment rarely proves a general scientific statement.

---

When the Hypothesis Is Not Supported

Suppose you predicted:

Increasing salt concentration will increase plant growth.

But the data show plant growth decreases.

A scientifically appropriate response is:

The results do not support the original hypothesis. Mean plant growth decreased as salt concentration increased.

This is not a failed experiment.

Unexpected evidence can be scientifically valuable.

---

Anomalous Results

An anomalous result does not fit the overall pattern.

Suppose:

The result at:

3 N = 9.8 cm

does not fit the overall pattern.

A good analysis should acknowledge it.

---

Discussing an Anomaly

A suitable statement might be:

The measurement at 3 N was anomalous because its extension of 9.8 cm was substantially greater than the approximately linear trend shown by the other measurements.

The discussion can then consider possible explanations.

For example:

• measurement error
• incorrect mass
• ruler misreading
• movement of the apparatus
• recording error

Do not claim to know the cause unless there is evidence.

---

Do Not Automatically Remove Anomalies

An unusual measurement should not be deleted simply because it is inconvenient.

Instead:

1. identify it
2. check for recording or calculation mistakes
3. repeat the measurement if possible
4. compare it with other trials
5. explain how it affects the interpretation

If there is a justified reason to exclude a measurement from a calculation, that decision should be explained.

---

Variation Between Trials

Suppose repeated measurements are:

25.1 s, 25.3 s, 25.2 s

These show little variation.

Now compare:

22.4 s, 28.9 s, 25.1 s

These show much greater variation.

The amount of variation provides information about the consistency of the measurements.

---

Range

A simple way to describe variation is the range.

Range = maximum value − minimum value

For:

22.4 s, 28.9 s, 25.1 s

Range:

28.9 − 22.4 = 6.5 s

A larger range indicates greater spread among the measurements.

---

Error Bars

If a graph contains error bars, the analysis should consider what they represent.

They may show:

• range
• standard deviation
• standard error
• measurement uncertainty

Do not simply state:

The graph has error bars.

Explain what the variation suggests about the measurements, while being careful not to make claims the error bars cannot support.

---

Correlation and Causation

If two variables change together, they may be correlated.

However:

correlation does not automatically prove causation

Suppose students observe that plants receiving more light are taller.

If other variables such as water, temperature, or soil quality were not controlled, it may be difficult to conclude that light alone caused the difference.

Controlled experiments provide stronger evidence for causal relationships.

---

Use Careful Scientific Language

Avoid claims stronger than the evidence supports.

Instead of:

X definitely caused Y.

you may need:

The results suggest that X influenced Y.

or:

The data show a relationship between X and Y under the conditions tested.

Scientific language should reflect the strength and limitations of the evidence.

---

Discuss the Quality of the Evidence

A strong discussion considers whether the observed pattern is convincing.

Ask:

• Are there enough data points?
• Were trials repeated?
• Are measurements consistent?
• Are there major anomalies?
• Is the trend strong?
• Is the measured range large enough?
• Are important variables controlled?
• Is uncertainty large compared with the observed effect?

These questions help determine how much confidence should be placed in the interpretation.

---

Evidence Strength Example

Experiment A:

2 trials

Values:

15.2 s and 28.7 s

Experiment B:

10 trials

Values clustered between:

15.0 s and 15.6 s

The second data set provides stronger evidence of a consistent measurement.

Quantity of data alone is not everything, but consistency and repeated measurements can strengthen an analysis.

---

Accuracy, Precision, and Reliability

These terms should not be treated as identical.

Accuracy

How close a measurement is to an accepted or true value.

Precision

How finely a measurement is reported or how closely repeated measurements agree, depending on context.

Reliability

How consistently an investigation or measurement produces comparable results.

A data set can be very consistent while still being systematically inaccurate.

---

Uncertainty

Every measurement has uncertainty.

For example:

A ruler may have millimetre divisions.

A digital balance may display:

0.01 g

A stopwatch may display:

0.01 s

But the overall experimental uncertainty may also depend on:

• reaction time
• reading technique
• equipment calibration
• experimental design

Analysis should consider whether uncertainty is large enough to affect the interpretation.

---

Absolute vs Percentage Uncertainty

Suppose a measurement is:

50.0 ± 0.5 cm

Absolute uncertainty:

±0.5 cm

Percentage uncertainty:

(0.5 ÷ 50.0) × 100

= 1%

Percentage uncertainty is useful when comparing measurements of different sizes.

---

Discussion and Scientific Theory

The discussion should connect experimental evidence to relevant scientific concepts.

For example:

Physics

Observed: acceleration increased with resultant force.

Theory:

F = ma

---

Chemistry

Observed: reaction rate increased with temperature.

Theory:

collision theory

---

Biology

Observed: enzyme activity increased with temperature before decreasing sharply.

Theory:

increased molecular motion followed by enzyme denaturation

The scientific explanation should be relevant to the actual investigation.

---

Compare with Expected Theory

Sometimes experimental results can be compared with a theoretical relationship.

Suppose theory predicts:

F = kx

Your experimental graph shows an approximately straight line through the origin.

You might write:

The approximately linear force-extension relationship is consistent with Hooke's law over the tested range.

If the graph curves at larger forces:

The deviation from linear behaviour suggests that the spring may be approaching or exceeding its limit of proportionality.

---

Comparing with Accepted Values

Some investigations produce a value that can be compared with an accepted reference value.

For example, suppose an experiment measures:

g = 9.5 m/s²

and the reference value used is:

9.8 m/s²

Difference:

9.8 − 9.5 = 0.3 m/s²

Percentage difference relative to the reference value:

(0.3 ÷ 9.8) × 100 ≈ 3.1%

This gives a quantitative way to discuss agreement.

---

Do Not Force Agreement with Theory

Experimental results sometimes disagree with expectations.

Do not change data to make them fit the theory.

Instead, discuss:

• what the evidence actually shows
• how it compares with expectations
• possible limitations
• whether more evidence is needed

Unexpected results are part of science.

---

Analysis vs Evaluation

These sections overlap, but they have different main purposes.

Analysis and Discussion:

What do the results mean?

Evaluation:

How good was the investigation, and how could it be improved?

For example:

Analysis:

Reaction time decreased as temperature increased.

Evaluation:

Temperature fell during some trials, so the intended temperature was not maintained consistently.

---

Discussion Can Lead into Evaluation

Sometimes the analysis reveals a problem.

For example:

Measurements at 40°C showed much greater variation than measurements at other temperatures.

This belongs naturally in the discussion.

You might then evaluate:

The variation may have resulted from inconsistent stirring speed, which was controlled manually. A mechanical stirrer could provide a more consistent stirring rate.

This connects evidence directly to evaluation.

---

Presenting Data Clearly During Analysis

Even though the main tables and graphs usually appear in the results section, the analysis should refer to them clearly.

For example:

As shown in the force-extension graph, extension increased approximately linearly between 1 N and 5 N.

or:

Table 1 shows that mean reaction time decreased from 82 s to 27 s across the tested temperature range.

This allows the reader to connect the interpretation with the evidence.

---

Scientific Conventions Still Matter

Analysis should use:

• correct scientific terminology
• appropriate units
• sensible significant figures
• variable names
• equations where useful
• references to tables and graphs
• quantitative comparisons
• objective language

Avoid vague phrases such as:

It changed a lot.

Instead:

The mean value increased by 38%.

---

A Strong Analysis Paragraph

A useful structure is:

1. State the pattern.

2. Give numerical evidence.

3. Interpret the pattern.

4. Explain it scientifically.

5. Identify important anomalies or limitations if relevant.

For example:

Mean reaction time decreased as temperature increased, falling from 82 s at 20°C to 27 s at 60°C. This indicates that reaction rate increased with temperature. According to collision theory, particles at higher temperatures have greater average kinetic energy, producing more frequent collisions and a greater proportion of collisions with sufficient energy to react. Although the overall trend was clear, the 40°C trials showed greater variation than the other temperature groups.

This paragraph moves logically from evidence to interpretation.

---

Worked Example 1: Plant Growth

Results:

Low light → 3.1 cm

Medium light → 5.8 cm

High light → 8.4 cm

Analysis:

Mean plant growth increased from 3.1 cm under low light to 8.4 cm under high light. This suggests that greater light availability increased growth under the conditions tested. Increased light availability can support a greater rate of photosynthesis when other required factors are available, providing more chemical energy and materials for growth.

---

Worked Example 2: Cooling

Results:

0 min → 82°C

5 min → 64°C

10 min → 52°C

15 min → 45°C

20 min → 41°C

Analysis:

Temperature decreased throughout the experiment, but the rate of cooling became slower over time. The sample cooled by 18°C during the first five minutes but only 4°C between 15 and 20 minutes. This is consistent with the temperature difference between the sample and its surroundings becoming smaller as the sample cools.

Notice that the analysis discusses more than simply:

temperature decreased

It examines how the rate of change changed.

---

Worked Example 3: Density

Results:

50 cm³ → 135 g

100 cm³ → 270 g

150 cm³ → 405 g

The ratio:

mass ÷ volume

is approximately:

2.7 g/cm³

Analysis:

Mass increased directly with volume, and the mass-to-volume ratio remained approximately 2.7 g/cm³. This indicates that the samples had an approximately constant density of 2.7 g/cm³.

---

Worked Example 4: Unexpected Result

Hypothesis:

Increasing fertilizer concentration will increase plant growth.

Results:

0% → 6.2 cm

1% → 8.4 cm

2% → 9.1 cm

4% → 7.3 cm

8% → 3.8 cm

Analysis:

Plant growth initially increased with fertilizer concentration, reaching a maximum mean growth of 9.1 cm at 2%. Growth then decreased at higher concentrations, falling to 3.8 cm at 8%. Therefore, the relationship was not simply linear. The evidence suggests that moderate fertilizer concentration benefited growth under the tested conditions, while high concentrations reduced growth.

This is more informative than claiming:

More fertilizer makes plants grow more.

---

Worked Example 5: Anomalous Data

Results:

20°C → 75 s

30°C → 58 s

40°C → 92 s

50°C → 31 s

60°C → 24 s

Analysis:

Reaction time generally decreased with increasing temperature, but the 92 s result at 40°C did not follow the overall pattern. This result should be investigated because it may reflect experimental variation or an error in that trial. Repeating measurements at 40°C would help determine whether the value is reproducible.

---

Common Mistakes

Mistake 1: Repeating the results without interpreting them

20°C was 80 s, 30°C was 62 s, 40°C was 47 s...

Instead, identify the overall relationship and use selected values as evidence.

---

Mistake 2: Explaining results without evidence

The reaction was faster because the particles moved faster.

First establish what the data showed.

---

Mistake 3: Saying "my hypothesis was correct"

Use:

The results support the hypothesis because...

---

Mistake 4: Ignoring anomalous results

Discuss important results that do not fit the overall pattern.

---

Mistake 5: Inventing explanations

Possible explanations should be described as possibilities unless evidence identifies the actual cause.

---

Mistake 6: Claiming causation from weak evidence

Use language appropriate to the strength of the experimental design.

---

Mistake 7: Treating every relationship as linear

Examine the shape of the graph carefully.

---

Error Analysis

A student writes:

The graph shows that the temperature was 20°C, 30°C, 40°C, 50°C, and 60°C.

This identifies values but provides almost no analysis.

A stronger statement is:

Reaction time decreased consistently as temperature increased from 20°C to 60°C, with the largest reaction time occurring at the lowest temperature.

Even stronger:

Reaction time decreased from 82 s at 20°C to 27 s at 60°C, indicating an increase in reaction rate across the tested temperature range.

---

Another Error Analysis

A student writes:

The 40°C result was wrong because someone made a mistake.

This makes an unsupported claim.

Better:

The 40°C result was substantially different from the overall trend and may be anomalous. Possible causes include timing error or inconsistent experimental conditions, although the available evidence does not identify the exact cause.

Scientific discussion distinguishes evidence from speculation.

---

A Reliable Analysis Strategy

Use this sequence:

Step 1: Identify the overall trend.

What happened to the dependent variable?

Step 2: Support the trend with data.

Use specific values.

Step 3: Describe the relationship.

Positive, negative, linear, nonlinear, constant, or no clear relationship?

Step 4: Interpret the relationship.

What does the pattern indicate?

Step 5: Explain the science.

Connect the pattern to relevant scientific concepts.

Step 6: Examine variation and anomalies.

Are all measurements consistent?

Step 7: Compare with the hypothesis or expected theory.

Does the evidence support the prediction?

Step 8: Consider the strength of the evidence.

How confident should we be?

---

Analysis and Discussion Checklist

Before completing your analysis, check:

• Have I identified the main trend?
• Have I used numerical evidence?
• Have I referred to appropriate tables or graphs?
• Have I described the type of relationship?
• Have I interpreted what the relationship means?
• Have I explained the pattern scientifically?
• Have I discussed important anomalies?
• Have I considered variation between trials?
• Have I compared the results with the hypothesis?
• Have I compared the results with scientific theory where appropriate?
• Have I avoided claims stronger than the evidence supports?
• Have I distinguished evidence from possible explanations?
• Have I used correct units and scientific terminology?
• Have I avoided simply rewriting the results table?

---

Did You Know?

Scientists can collect exactly the same data and still disagree about its interpretation.

This is why scientific arguments require evidence and reasoning.

A strong scientific discussion makes the reasoning visible:

Here is the evidence.

Here is the pattern.

Here is what we think the pattern means.

Here is the scientific explanation.

Here are the limitations of that interpretation.

This allows other scientists to evaluate the argument rather than simply accepting the conclusion.

---

Key Terms

• Analysis: Examination and interpretation of experimental data.
• Discussion: Explanation of what results mean and how they relate to scientific ideas.
• Trend: Overall pattern in a data set.
• Positive relationship: Relationship in which one variable tends to increase as another increases.
• Negative relationship: Relationship in which one variable tends to decrease as another increases.
• Linear relationship: Relationship represented approximately by a straight line.
• Nonlinear relationship: Relationship that does not follow a straight-line pattern.
• Correlation: Association between variables.
• Causation: Relationship in which a change in one factor produces a change in another.
• Anomaly: Result that does not fit the general pattern.
• Range: Difference between the largest and smallest values.
• Uncertainty: Limitation in the precision or certainty of a measurement.
• Error bar: Graphical representation of variation or uncertainty.
• Accuracy: Closeness of a measurement to an accepted or true value.
• Reliability: Consistency of measurements or results.
• Evidence: Observations or measurements used to support a scientific claim.
• Scientific reasoning: Use of scientific concepts to explain evidence.
• Hypothesis: Testable prediction supported by reasoning.
• Claim: Statement or interpretation supported by evidence.

---

Key Relationships

A strong scientific analysis follows:

Data → Pattern → Interpretation → Scientific Explanation

A strong analytical statement combines:

Claim + Evidence + Reasoning

Results answer:

What happened?

Analysis answers:

What does it mean?

Discussion asks:

Why might it have happened, and how well does the evidence support that interpretation?

Evaluation asks:

How good was the investigation, and how could it be improved?

---

Key Takeaways

• Analysis and discussion explain the meaning of experimental results.
• Results and analysis are not the same thing.
• Results primarily present evidence.
• Analysis interprets that evidence.
• Discussion connects interpretations to scientific ideas.
• Analysis should begin with the actual data rather than the expected result.
• Overall patterns should be identified clearly.
• Relationships may be positive, negative, linear, nonlinear, constant, or unclear.
• Strong analysis uses numerical evidence.
• Tables provide exact values, while graphs make patterns easier to recognize.
• An analysis should not simply rewrite every value in a results table.
• Differences, ratios, percentages, gradients, and rates can help quantify relationships.
• Scientific explanations should connect directly to observed evidence.
• Claim–Evidence–Reasoning is a useful structure for scientific discussion.
• Hypotheses should be described as supported or not supported by the evidence rather than simply "proved."
• Unexpected results can provide useful scientific information.
• Anomalous results should be identified and investigated rather than automatically removed.
• Possible explanations for anomalies should not be presented as confirmed facts without evidence.
• Variation between repeated trials provides information about measurement consistency.
• Uncertainty should be considered when deciding how strongly the data support a relationship.
• Correlation does not automatically demonstrate causation.
• Scientific language should reflect the strength of the available evidence.
• Experimental results can be compared with theoretical relationships or accepted values where appropriate.
• Data should never be altered simply to make results agree with theory.
• Analysis focuses mainly on what the evidence means, while evaluation focuses on the quality and limitations of the investigation.
• Scientific conventions such as units, appropriate precision, equations, graph references, and correct terminology remain important during analysis.
• Strong scientific discussion makes the reasoning from evidence to interpretation clear enough for another person to evaluate.