3. Conducting the Investigation

Conducting the Investigation

Once your research question, hypothesis, variables, materials, and experimental procedure have been planned, it is time to carry out the investigation.

This is where careful scientific work becomes especially important.

A good experimental plan is only useful if it is followed carefully. During the investigation, you need to:

• follow the planned procedure consistently
• measure carefully
• record observations immediately
• repeat measurements when appropriate
• keep control variables as constant as possible
• respond sensibly to unexpected problems
• maintain safety
• collect enough data to answer your research question

The goal is not simply to finish the experiment. The goal is to collect reliable evidence that can be used to reach a meaningful conclusion.

Follow the Procedure Carefully

Your experimental procedure is the step-by-step plan for carrying out your investigation.

For example, a procedure for investigating how ramp height affects the motion of a toy car might include:

1. Set the ramp to a height of 10 cm.
2. Place the car at the marked release point.
3. Release the car without pushing it.
4. Measure the time required to travel 1.0 m.
5. Record the result.
6. Repeat the measurement three times.
7. Calculate the average time.
8. Increase the ramp height to 20 cm.
9. Repeat the procedure.

Following the same method each time helps make the investigation a fair test.

If the car is pushed during one trial but simply released during another, an additional variable has been introduced.

---

Consistency Matters

Experimental conditions should be kept as consistent as possible.

Suppose you are investigating:

How does ramp height affect the time taken for a toy car to travel 1 m?

The independent variable is:

ramp height

The dependent variable is:

travel time

Important control variables could include:

• same toy car
• same ramp
• same surface
• same travel distance
• same release point
• same timing method
• same person releasing the car

The goal is to make ramp height the main factor changing between trials.

Measure Carefully

Measurements are a major source of scientific evidence.

Depending on the investigation, you might measure:

• time
• distance
• mass
• volume
• temperature
• force
• pH
• plant height
• number of organisms
• gas volume
• electrical current
• voltage

Use equipment appropriate for the measurement.

For example:

• ruler or metre stick → length
• stopwatch → time
• thermometer → temperature
• balance → mass
• measuring cylinder → liquid volume
• gas syringe → gas volume
• force meter → force

Whenever possible, measurements should include the correct unit.

Instead of recording:

Temperature = 25

record:

Temperature = 25 °C

Units give measurements meaning.

Read Instruments Correctly

Using measuring equipment properly improves the quality of your data.

When reading a scale:

• identify the unit
• determine the value of each division
• view the scale from the correct position
• record an appropriate number of digits

For liquid volume in many measuring cylinders, read the scale at eye level and use the appropriate part of the meniscus.

Viewing a scale from above or below can produce a parallax error.

The same principle applies when reading rulers, analogue meters, and other graduated instruments.

---

Record Data Immediately

Do not rely on memory.

Measurements and observations should be recorded as soon as they are made.

A data table prepared before the investigation can make this much easier.

For example:

A good data table should clearly show:

• what was changed
• what was measured
• units
• repeated trials
• calculated values when appropriate

Never invent missing measurements simply to complete a table.

---

Record Observations Too

Not all useful scientific evidence consists of numbers.

Scientists also record qualitative observations.

These might include:

• colour changes
• bubbling
• formation of a precipitate
• changes in appearance
• unusual sounds
• changes in smell when safely observable without deliberately inhaling chemicals
• movement
• unexpected behaviour
• equipment problems

For example, during an acid-carbonate reaction you might record:

Rapid bubbling occurred immediately after the acid was added. The solid gradually became smaller.

These observations can help explain the numerical results.

---

Quantitative and Qualitative Data

Scientific investigations often produce two types of data.

Quantitative data involve numbers and measurements.

Examples:

• 24.5 °C
• 18.2 cm
• 35.6 s
• 42 mL

Qualitative data describe characteristics or observations.

Examples:

• solution changed from colourless to blue
• bubbles formed rapidly
• plant leaves became yellow
• precipitate appeared

A strong investigation may use both types of evidence.

---

Use Appropriate Precision

Measurements should reflect the precision of the instrument being used.

For example, if a ruler is marked in millimetres, reporting a measurement as:

12.374829 cm

would suggest a level of precision the ruler cannot provide.

Similarly, if a thermometer reads only to the nearest degree, recording:

24.736 °C

would not be justified.

Record measurements with appropriate precision, and try to use the same level of precision throughout a data set when the same instrument is being used.

---

Repeat Your Measurements

One measurement may be affected by random variation.

Repeating measurements allows you to see whether the result is consistent.

Suppose three measurements of travel time are:

2.41 s

2.38 s

2.43 s

These results are fairly close together.

You could calculate an average:

Average = (2.41 + 2.38 + 2.43) ÷ 3

Average = 2.41 s

Repeating trials and calculating an average can reduce the influence of random variation.

Look for Anomalous Results

Sometimes one result is very different from the others.

Suppose you record:

Trial 1 = 2.42 s

Trial 2 = 2.39 s

Trial 3 = 5.81 s

The third result appears unusual.

This is called an anomalous result or outlier.

Do not simply delete it because you do not like it.

Instead, ask:

• Was there a timing mistake?
• Was the equipment disturbed?
• Did the car hit something?
• Was the procedure different?
• Is the result repeatable?

The best response may be to perform another trial.

An unusual result can sometimes reveal a real scientific effect, so it should be investigated rather than automatically discarded.

---

Collect Enough Data

A science fair investigation needs enough evidence to reveal a meaningful pattern.

Suppose you want to investigate how temperature affects dissolving.

Testing only:

20 °C and 80 °C

gives very limited information.

It may be better to investigate several temperatures, such as:

20 °C, 30 °C, 40 °C, 50 °C, 60 °C and 70 °C

and perform several trials at each temperature.

This gives you more information about the relationship between the variables.

However, more data are not automatically better. Measurements should be chosen deliberately and collected carefully.

---

Range and Intervals

When planning measurements, consider both the range and intervals.

The range is the span of values tested.

For example:

10–50 cm ramp height

The interval is the difference between successive values.

For example:

10, 20, 30, 40, 50 cm

uses intervals of:

10 cm

A useful range and sensible intervals make patterns easier to identify.

---

Do Not Change Several Variables at Once

Suppose you are investigating:

How does temperature affect dissolving time?

At 20 °C you use:

• 100 mL water
• 5 g sugar
• no stirring

At 60 °C you use:

• 200 mL water
• 10 g sugar
• constant stirring

If the sugar dissolves faster at 60 °C, you cannot tell whether this was caused by:

• temperature
• water volume
• sugar mass
• stirring

Several variables changed at once.

A fair investigation should change the independent variable while controlling other important factors.

---

Unexpected Problems Are Normal

Real experiments do not always work exactly as planned.

You may discover that:

• a measurement is too difficult to make
• equipment does not work
• the reaction occurs too quickly
• the reaction occurs too slowly
• plants fail to germinate
• measurements vary too much
• your chosen range is inappropriate
• an important control variable was overlooked

This does not automatically mean the experiment has failed.

Scientific investigations often require problem solving and adaptation.

---

Adapting the Procedure

Suppose your investigation asks:

How does ramp height affect the time taken for a toy car to travel 20 cm?

You discover that the car travels 20 cm so quickly that accurate stopwatch timing is nearly impossible.

You could modify the procedure by increasing the travel distance to:

1.0 m

This may make the timing easier.

However, once you change the procedure, you should use the revised method consistently for the relevant comparable trials.

You should also record the change and explain why it was made.

Good science is transparent.

Do not quietly change the procedure and pretend that nothing changed.

---

Pilot Tests

A small pilot test can help identify problems before the main investigation begins.

For example, before investigating five different ramp heights with five trials each, you could perform a few preliminary tests.

You might discover:

• the lowest ramp is too shallow for the car to move reliably
• the highest ramp causes the car to leave the track
• the travel distance is too short
• stopwatch timing is difficult

You can then improve the procedure before collecting the main data.

A short pilot investigation can save considerable time later.

---

Record Changes to Your Method

If the procedure changes, record:

What changed?

Why did it change?

How might the change affect the results?

For example:

Original method: Measure the time for the car to travel 20 cm.

Problem: The travel time was too short to measure reliably with a handheld stopwatch.

Change: Increased the travel distance to 1.0 m.

Reason: Longer travel times reduced the relative effect of human reaction time.

This information can later become part of your evaluation.

---

Safety Throughout the Investigation

Safety is not something you consider only before starting.

You should continue monitoring risks throughout the investigation.

Before and during practical work:

• follow teacher instructions
• wear appropriate protective equipment
• keep the workspace organized
• handle glassware carefully
• keep liquids away from electrical equipment
• use heat sources appropriately
• clean spills promptly using the correct procedure
• tie back long hair when appropriate
• avoid loose clothing near flames or equipment
• never taste laboratory chemicals
• never deliberately inhale chemicals
• wash hands after appropriate practical activities
• report accidents, spills, breakages, or injuries immediately

Stop if Conditions Become Unsafe

If something unexpected creates a safety risk, stop the investigation.

For example:

• equipment becomes damaged
• electrical equipment gets wet
• glassware cracks
• unexpected fumes appear
• a chemical spills
• equipment becomes dangerously hot
• someone is injured

Do not continue simply because you want to finish collecting data.

Inform the teacher or supervisor and reassess the situation.

Safety takes priority over completing the experiment.

---

Keep the Workspace Organized

Good organization improves both safety and data quality.

Keep:

• unused equipment away from the working area
• chemicals clearly identified
• containers labelled
• electrical cables organized
• notebooks away from spills
• measuring equipment accessible
• waste in appropriate containers

An organized workspace reduces the chance of accidentally:

• mixing samples
• recording the wrong measurement
• knocking over equipment
• contaminating materials
• creating hazards

---

Label Your Samples

If your investigation uses several samples, label them clearly.

Suppose you are investigating plant growth at different salt concentrations.

You might label containers:

0% salt

0.5% salt

1.0% salt

1.5% salt

2.0% salt

Without clear labels, samples can easily become mixed up.

Never rely on remembering which container is which.

Record Data Honestly

Scientific integrity means recording what actually happened.

Do not change a measurement because it does not fit the expected pattern.

Do not invent a result because a trial was missed.

Do not remove unusual values simply to make a graph look better.

Suppose you predicted that increasing temperature would always decrease dissolving time, but one measurement does not fit the trend.

Record it.

Then investigate possible explanations.

Science depends on accurate and honest evidence.

---

Keep Raw Data

Raw data are the original measurements and observations collected during the investigation.

For example:

These original measurements should be kept even if you later calculate:

• averages
• percentages
• rates
• differences
• other processed values

Raw data allow you and others to see what was actually measured.

---

Record Enough Detail

Your notes should contain enough information that you can understand what happened later.

Useful information may include:

• date and time
• measurement values
• units
• equipment used
• environmental conditions
• observations
• unusual events
• changes to the method
• failed trials
• safety issues

This record is sometimes called a laboratory notebook or research log.

Reliability and Repeated Trials

Reliability describes the consistency of measurements or results.

If repeated measurements are similar, confidence in the result generally increases.

Consider two sets of results.

Set A

4.2 s
4.3 s
4.1 s

These results are quite consistent.

Set B

2.1 s
7.8 s
4.4 s

These results show much greater variation.

Set B may require investigation.

Possible causes include:

• inconsistent technique
• uncontrolled variables
• measurement error
• equipment problems
• natural variation

Repeated trials help reveal these problems.

---

Accuracy, Precision and Reliability

These terms are related but have different meanings.

Accuracy describes how close a measurement is to the true or accepted value.

Precision describes how closely repeated measurements agree with each other or the resolution with which measurements are reported, depending on context.

Reliability describes the consistency or dependability of the results across repeated measurements or trials.

Imagine arrows being fired at a target.

A tight cluster away from the centre can be precise but inaccurate.

A scattered group around the centre may have a reasonable average but poor precision.

Good experimental technique aims for measurements that are as accurate, precise, and reliable as reasonably possible.

---

Know When You Have Enough Data

There is no single number of trials that is correct for every experiment.

The amount of data required depends on:

• the research question
• natural variation
• measurement uncertainty
• available time
• experimental design
• number of independent-variable values
• type of analysis

For a simple school investigation, three repeated measurements at each condition are often a useful starting point, but some investigations require more.

Ask:

Do I have enough evidence to identify a pattern and answer my research question reasonably?

That is more important than simply reaching a predetermined number of measurements.

---

Example Investigation

Consider:

How does salt concentration affect bean seed germination?

A student prepares solutions containing:

• 0% salt
• 0.5% salt
• 1.0% salt
• 1.5% salt
• 2.0% salt

Each treatment uses several seeds.

The student controls:

• type of seed
• number of seeds
• volume of solution
• temperature
• light conditions
• container type
• investigation duration

Every day, the student records:

• number of seeds germinated
• visible changes
• unusual observations

At the end, the student calculates the percentage germination for each concentration.

This produces much stronger evidence than using only one seed at each concentration.

---

From Investigation to Conclusion

The investigation produces the evidence that will eventually support your conclusion.

The overall process is:

Research question

↓

Hypothesis

↓

Variables

↓

Procedure

↓

Careful measurements and observations

↓

Repeated trials

↓

Recorded raw data

↓

Data analysis

↓

Patterns and relationships

↓

Conclusion

If the experimental data are poor, the conclusion will also be weak.

This is why careful data collection is one of the most important stages of the entire scientific investigation.

---

Common Mistakes

Relying on memory

Record measurements immediately.

Even a small delay can lead to confusion or forgotten details.

Changing the procedure between trials

If the method changes unnecessarily, the trials may no longer be directly comparable.

Recording numbers without units

35 is incomplete.

35 cm communicates an actual measurement.

Performing only one trial

One measurement provides little information about variation or reliability.

Deleting results that do not fit the hypothesis

Unexpected results must be investigated, not hidden.

Changing several variables simultaneously

This makes it difficult to determine which variable caused the observed effect.

Quietly changing the procedure

If a method needs to be modified, record and explain the change.

Using too narrow a range

Testing only two very similar values may make a relationship difficult to detect.

Using unrealistic precision

Do not report more decimal places than your equipment can reasonably measure.

Ignoring qualitative observations

Numbers are important, but observations can provide valuable additional evidence.

Continuing when something becomes unsafe

Stop the procedure and seek appropriate assistance.

Inventing missing data

Missing data should remain missing or the trial should be repeated when appropriate.

Never manufacture results.

---

Key Terms

Investigation — A systematic scientific process used to answer a question by collecting and analyzing evidence.

Experimental procedure — The step-by-step method followed during an investigation.

Independent variable — The factor deliberately changed.

Dependent variable — The factor measured or observed.

Control variable — A factor kept as constant as reasonably possible.

Fair test — An investigation in which important variables other than the independent variable are controlled.

Measurement — A numerical value obtained using an appropriate measuring method or instrument.

Observation — Information noticed and recorded during an investigation.

Quantitative data — Numerical measurements or counts.

Qualitative data — Descriptive observations.

Raw data — Original measurements and observations before processing or calculation.

Data table — An organized arrangement used to record and display data.

Trial — One complete performance of an experimental condition or procedure.

Repeated trial — Repetition of a measurement or experimental condition.

Average — A representative value calculated from several measurements, commonly the arithmetic mean.

Anomalous result — A result noticeably different from the general pattern or other repeated measurements.

Outlier — A data value that differs substantially from most other values in a data set.

Reliability — The consistency or dependability of repeated results.

Accuracy — How close a measurement is to the true or accepted value.

Precision — How closely repeated measurements agree, or the degree of detail with which a measurement can be reported.

Range — The span of independent-variable values investigated.

Interval — The difference between successive values used in an investigation.

Pilot test — A small preliminary investigation used to identify problems before the main experiment.

Parallax error — A reading error caused by viewing a measurement scale from an incorrect angle.

Laboratory notebook — A record of experimental procedures, measurements, observations, problems, and changes.

Experimental evidence — Measurements and observations used to answer a scientific question.

Scientific integrity — Honest and responsible collection, recording, analysis, and reporting of scientific information.

---

Key Takeaways

• Follow your experimental procedure carefully and consistently.
• Change the independent variable deliberately.
• Measure the dependent variable carefully.
• Keep important control variables as constant as reasonably possible.
• Use appropriate measuring equipment.
• Always record appropriate units.
• Record measurements immediately.
• Record both quantitative and useful qualitative observations.
• Prepare data tables before practical work when possible.
• Repeat measurements to investigate variation and improve confidence in the results.
• Do not automatically remove anomalous results.
• Investigate unusual results and repeat measurements when appropriate.
• Use a suitable range and interval of independent-variable values.
• Collect enough data to reveal meaningful patterns.
• Use a pilot test when it can help identify problems with the planned procedure.
• Unexpected problems are a normal part of experimental science.
• Procedures can be adapted when necessary, but changes should be recorded and justified.
• Keep your experimental conditions consistent after making an appropriate change.
• Maintain safety throughout the entire investigation.
• Stop if unexpected conditions make the investigation unsafe.
• Keep samples clearly labelled and the workspace organized.
• Keep your original raw data.
• Never invent, alter, or hide results to make them agree with your hypothesis.
• Good conclusions depend on good evidence.
• The purpose of conducting an investigation is not to prove your prediction correct.

The purpose is to collect accurate, honest, and sufficient evidence to answer your research question.

---

Check Your Understanding

1. Why is it important to follow the same experimental procedure during repeated trials?

2. What is the difference between quantitative and qualitative data?

3. Why should measurements be recorded immediately?

4. Explain why units must be included with measurements.

5. Give three examples of control variables in a toy-car ramp investigation.

6. Why are repeated trials useful?

7. What is an anomalous result?

8. A student records three times:

3.21 s, 3.18 s, 8.74 s

What should the student do about the third result?

9. Explain why the student should not simply delete the 8.74 s measurement.

10. What is a pilot test, and why might it be useful?

11. A reaction occurs too quickly to measure accurately. Give one appropriate way the procedure might be adapted.

12. Why should changes to an experimental procedure be recorded?

13. Explain the difference between raw data and processed data.

14. Why is changing several variables at the same time a problem?

15. A student is investigating temperature but changes both temperature and the amount of water between trials. Explain why this makes the results difficult to interpret.

16. Why should an experiment be stopped if unexpected conditions make it unsafe?

17. Explain why more data are not automatically better data.

18. What is scientific integrity?

19. Why is it scientifically acceptable for your results to disagree with your hypothesis?

20. Challenge: During an investigation, you discover that your original method produces highly inconsistent measurements. Explain how you would respond while maintaining the quality, transparency, and safety of the investigation.