Topic 1 - The Scientific Method

Question 1

5 Stages of the Scientific Method

Answer 1

1 - Make observations
2 - Form a testable question
3 - Propose a hypothesis (testable, falsifiable - you can prove it wrong)
4 - Test hypothesis (experimental manipulation)
5 - Assess: reject or accept all or part of the hypothesis and return to 1

This is idealized, it doesn’t always work this way. Sometimes the order is different or there was luck involved etc.

Question 2

What is a testable question?

What words should these qs avoid?

Answer 2

One that allows you to set up an experiment to learn the answer. It must be measurable.

Do not use words like best/better/worse

Question 3

How to turn a question into a hypothesis

Answer 3

Use the format:

If… then… because

If I put my fingers in warm water then the water will dissolve the oil so warm water will wrinkle my fingers faster than cold water

Question 4

The pirate example

Answer 4

Observe that the temperature has gone up globally as the number of pirates has decreased - a correlation

Create a hypothesis: if we increase the number of pirates then the global temperature will drop

Experiment - manipulate the number of pirates, observe temp drop. You would have to control for all other variables
.
Assess - did the number of pirates influence global temperatures

Question 5

Correlation vs. causality

Answer 5

Correlation does not equal causality
Ice creams and drowning. Correlated. Actually both caused by summer.

Sometimes correlations DO point out causal relationships. We just cant assume it.

Always be weary of the media presenting things as causal when they are actually only correlated.

Question 6

Sample size

Answer 6

Make sure the sample size is appropriate to show what you are interested in.

E.g. A coin tossed once shows 100% of the time it lands heads or tails. With repeated measurement, it gets close to 5050

Question 7

Sampling bias

Answer 7

If your sample is biased, you can make wrong conclusions.

WW2 fighter planes examined for bullet holes. Do you put armor where the holes are or elsewhere?

Elsewhere - the observations are all done on planes that made it back. In reality, the armor needs to go where there are none as they are not surviving.

Question 8

Internal validity

Answer 8

Study is well designed and free from biases or confounds. DOes the study really measure what it wants to?

Eg sample men who take tylenol for headaches and women who take advil for headaches

Worthless - we dont know if differences are due to medication or sex.

HAS internal validity as long as:
The same selection criteria applied to everyone
Double blind

Question 9

External validity

Answer 9

Does it generalise?

If the sample is too specific, it might not apply to everyone

Question 10

Watch the Y axis

Answer 10

Different y axes can be used to make a point seem bigger or smaller than it really is. Always check for y axes.

Question 11

Watch what data is being included

Answer 11

Are they reporting things that are reasonable to make the point they wish to make or data that specifically makes their point?

Question 12

What is being compared?

Answer 12

Comparisons must be reasonable or they distort the impression an article gives.

Question 13

Who is paying for it?

Answer 13

Good - government or unbiased sources

Bad - industry or biased source

Question 14

What makes a good science media source?

5 things

Answer 14

• Clear links/reference to the original research so you can track it down
• Description of the data, how it was collected. Not just the interpretation
• Interviews with the author(s) of the study
• Interviews with other scientists not involved with the study