Home Biology Chemistry Physics Controlled Assessments Long Answer Questions 11-14 Science Big Moments in Science Linkup

Variables and scientific evidence

1. What is evidence?

-After coming up with a Scientific question, scientists have to gather evidence.

-Evidence is the information  that you collect to answer a scientific question. It may be simple Qualitative observations that you see, feel, hear or it may be Quantitative data that you collect by measuring.

To pass that scientific test, the evidence has to have certain characteristics. It has to be:

  • Reliable
  • Accurate

-Evidence that does not pass this test cannot be used to make a valid conclusion.

2. But how do we collect Scientific evidence?

Evidence can be collected by making observations and by measuring. When making these observations and measuring, we normally look for a change.

In most investigations, there should be a change otherwise there is no point in doing the investigation.

Hint: If you realise that nothing is changing, stop and check your experiment method and materials.

Example of questions to ask when making observations:

  • Does it get warmer?
  • Does it get colder?
  • Does the colour change? From which one to which one?
  • What can you hear or smell?
  • Does it get faster/slower

3. Variables and factors

We have talked about making observations and measuring. But what do we observe? What do we measure?

a. Factors

An investigation, has several 'factors' that directly or indirectly affect it. These are also called Factors.

 For example;

What affects the rate at which grass on the school field grows?

If you were to carry out this investigation, you would realise you have more than a handful of things to deal with such as amount of water, light intensity, amount of nutrients in the soil, pH of the soil, type of soil, ... the list goes on.

b. Variables

To vary is to be different. Some of the factors above can change or become different during the investigation. These are called variables.

i. Independent variable

Scientists normally hold the other variables constant and change only one at a time.

*The variable that you deliberately change is called the independent variable.

ii. Dependent variable

*The outcome that you measure is called the Dependent variable. Its change depends on the variable that you deliberately change.

You can use a measuring instrument to measure this. These are Quantitative observations (numbers or quantities).

At times we make qualitative observations (in words) about the dependent variable e.g. colour changes.

iii. Control variable

Science investigations can be like squeezing a ballon; squeeze one part, the other expands! There may be many variables affecting the investigation, to stop the ballon effect, you need to hold some of the variables constant. These are called control variables.

If you do not keep them the same , they will effect your dependent variable so you will not get a fair comparison. The values of the dependent variable would be different.

Caution: Although we know that we keep these the same to make it a fair test, it is not enough to just say '... to make it a fair test' in exams or controlled assessments without mentioning effects on the dependent variable.

5. Reliability

What does reliability mean?

Let us start from a simple question:

What is a reliable friend?

You probably think that they can be trusted. So must be scientific evidence. Let us clarify what we mean by trusted.

A trusted friend behaves consistently no matter what happens. They do not show unexpected 'strange' changes in their behaviour.

*Reliable evidence can be trusted, it can be used to make a valid conclusion.

* Reliable evidence is replicable or reproducible. It does not show wide variations or fluctuations.

In science, these variations or changes are called anomalies (anomaly- not in agreement with something else)

In most cases, an anomaly is above the 10% error margin.

It is important to note that we identify a point as an anomaly only if there is no reason to explain or justify its occurance. Some points that may seem like anomalies may occur naturally especially in Biology. This has implications on how we draw graphs

How can we make evidence reliable?

We have already met the word 'consistent'. This word tells us that the event has to happen more than once and each time it happens, it has similar features.

1. Repeat measurements

-In science to make sure that our evidence is reliable, we 'repeat the test'.

Most students know about this but it is not enough to answer most questions this way in an Exam because it leaves some questions unanswered! Why repeat, Then what?

Why repeat?

-We repeat in order to test the consistency, repeatability or replicability. If we get similar data without wide variations  after repeating, then our data is probably reliable.

-With some investigations, it is not possible to identify and lateron be able to control all variables. These unidentified confounding variables may cause the data to flactuate within certain limits. To reduce the effect of these variables, you can do repeat tests and make the data more reliable.

-If there is an anomally, then we need to retest/repeat in order to correct the anomally.

- besides, we need to ask ourselves: Why did we get the anomaly?

Is it really an anomaly or is it just an atypical sample for example, a tomato fruit that is just too small compared to others because its branch could not get enough water or has its food reduced by pests.

Is it a systematic error (caused by the measuring instrument e.g. a zero error) or a random error (eg. human error in reading the instrument or lack of precision on the instrument.)

The mean

As a consequence of doing repeat tests, we suddenly come up with a large set of data that we need to make sense of. The starting point is to calculate a mean or average.

  • A large set od data minimises the effects of anomalies
  • To calculate the mean, add up the numbers and divide by how many there are.
  • Leave out the anomaly in your calculation

2. Repeat using a different technique

This is particularly useful if you have doubts about your method. Which type of error does it correct?

3. Compare

There are other ways to test if your data is reliable. We can check with other students/scientists who have done the same investigation and compare to see if the results are similar. Even real scientists check their evidence with their peers. They call it peer review.


Dealing with evidence: Accuracy and Precision

Scientific evidence needs to be accurate. This is a hard thing to get but inaccurate evidence is of no use to scientists and the public.

What is accuracy?

Many of us immediately think of words like correct, exact and true when we are asked about accuracy. Accurate data reflects the size of the thing that we are measuring.

In science, we normally say an accurate measurement is near the 'true value'.It is a bit tricky to explain what the true value is but this does not prevent us from tyring to gather accurate evidence.

Accuracy is limited by 2 things:

1. Limitations of the measuring instrument

2. Care with which the instrument is used-  human error.

So to gather accurate data, we must check:

-that the instrument is working well, appropriately zeroed (or adjust for the zero error after)

-we correctly used the instrument and read it properly.


This is another feature related to accuracy. Precise data is to the point! This is determined by the scale on the measuring instrument. The smaller the markings on the instrument, the more precise. This also means that the more the decimal points, the more the precision.

Caution: It is a common mistake to assume that digital instruments are more precise than manual ones. This is not true. It all depends on the decimal points on each.

How precise should our measurements be?

Good question! You should not thoughtlessly aim to achieve the highest degree of precision. For example; What is the point in measuring the distance between the London and New York in millimeters (mm)? What would you possibly use this data for?

*The level of precision must be judged by the nature of your investigation; what kind of data you require.

Getting ready to gather data: Doing a preliminary test

A preliminary test is also called a trial run, a dummy run so-to-speak

Why is it necessary to do a trial run?

In most experiments and investigations a preliminary test helps to:

  • * Figure out what sort of result you can expect
  • * Identify and problems with the method- what quantities you will need
  • * Identify problems with equipment
  • *Set the range of your results and judge its appropriateness.