1
00:00:00,350 --> 00:00:05,378
There are the statistical analyses that weigh up the differences that were found in the data.

2
00:00:05,379 --> 00:00:11,179
The results of the analyses are presented in many ways, but usually a p value is reported.

3
00:00:11,193 --> 00:00:15,271
This is the probability that the differences found were due to chance.

4
00:00:15,271 --> 00:00:25,016
If the p value is less than 5 per cent (usually written as p < 0.05) then the differences are said to be statistically significant.

5
00:00:25,038 --> 00:00:27,452
In other words, they are not due to chance.

6
00:00:27,454 --> 00:00:35,001
Remember statistical significance has a precise statis‒ tical meaning of something not being due to random chance alone.

7
00:00:35,007 --> 00:00:43,759
‘Significant’ results are not necessarily important, newsworthy or exciting (although the researcher who found them probably feels that they are).

8
00:00:43,796 --> 00:00:49,554
So when writing essays about research, be careful to be clear about which sense of the word you are using.

9
00:00:49,569 --> 00:00:53,142
In science, we can never really prove that a theory is true.

10
00:00:53,158 --> 00:00:57,899
This may sound pessimistic, but it is an essential part of the scientific method.

11
00:00:57,911 --> 00:01:03,975
All we can do in science is use evidence to reject or falsify a hypothesis or a theory.

12
00:01:03,983 --> 00:01:07,982
An experiment is like an axe: all it can do is chop things down.

13
00:01:08,008 --> 00:01:19,531
Experiments never directly prove that a theory is right; all they can do is provide indirect support by rejecting all the other theories until only one plausible theory remains.

14
00:01:19,535 --> 00:01:32,863
For example, sometimes you hear people say things like ‘evolution is only a theory: science has never proved it.’ Well, that’s true, but only in the sense that science never proves that any theory is positively true.

15
00:01:32,863 --> 00:01:41,122
But what the theory of evolution has done is assemble an enormous amount of convincing data proving that other competing theories are false.

16
00:01:41,168 --> 00:01:48,371
So though it hasn’t been ‘proved,’ overwhelmingly, evolution is the best theory that we have to explain the data we have.

17
00:01:48,371 --> 00:01:51,384
And that’s as good as we get with the scientific method.