Transcript Levelofsignificance

Levels of significance
(probability, Type 1 and
Type 2 errors)
Inferential statistics
• Used to find out what the probability is that a result has occurred by
chance.
• Probability tells us how likely it is for something to happen.
• A probability of 1 (or it can be written as 100%) means that an event
will definitely take place.
• There could also be a probability of 0.01 (1%) which means there is a
1 in 100 chance an event will take place.
• Inferential statistical analysis focuses on the null hypothesis.
• At the end of an inferential statistical analysis there will be a probability value
for the null hypothesis (how likely it is that there will be NO difference).
• This can be from 1 (100%) to 0.001 (0.1%).
• A 0.05 (or 5%) level of probability is usually referred to as having statistical
significance.
• Statistical significance is the term used for a probability at which the null
hypothesis can be rejected and the experimental hypothesis accepted.
P<0.05
• This is a significance level.
• P<0.05 is the level psychologists usually set to see if the null hypothesis is
accepted.
• It means:
P
The probability of
the results being
due to chance
<
0.05
is less
than
5%
• If the probability of the results being due to chance is less than 5%, the
probability of the results being due to the independent variable is more than
95%.
• The experimental hypothesis is accepted.
Type 1 and Type 2 errors
• Sometimes psychologists make mistakes and set their significance
levels too high or too low.
• This might lead them to accept a hypothesis that isn't correct.
Type 1 errors
• Accepting the operationalised hypothesis despite the possibility of
this being due to chance.
• Causes:
• Poor research design.
• Using a level of significance that is too lenient.
Type 2 errors
• Accepting the null hypothesis despite the possibility of the
operationalised hypothesis being correct.
• Causes:
• Poor research design.
• Using too stringent significance levels.