Investigating Psychology – lecture 4 – Probability and an introduction to
hypothesis testing
Descriptive statistics – describe the data we have from our sample
Inferential statistics – allow us to make inferences and generalise our results to
the rest of the population – tell us about the probability that the certain pattern
of data we can see in the sample has come from a certain underlying population
Null hypothesis significance testing (NHST) –
- Hypothesis – a statement / prediction about the data
- Null hypothesis – there is no difference / relationship
- NHST tells us if a certain pattern of data we can see in our sample has come
from an underlying population distribution where the null hypothesis is
true (where there is no difference)
- P = the probability of getting data patterns that we see in our sample if in
reality we measured everyone in the population and there is no difference /
relationship
- P can range from 0 to 1
- If the null hypothesis was true and there was no difference or relationship,
p < 0.05 ( there would be less than 5% chance of getting the pattern in the
data in the sample if the null was true) – if p is more than 0.05 then the null
hypothesis must be accepted
Type 1 and 2 errors –
- Type 1 error – no effect in population but we can see that there is an effect
from the sample data – null is true but there is something affecting the data
- Type 2 error – there is an effect in the population so the null isn’t true but
we decide that there isn’t an effect from our sample data
hypothesis testing
Descriptive statistics – describe the data we have from our sample
Inferential statistics – allow us to make inferences and generalise our results to
the rest of the population – tell us about the probability that the certain pattern
of data we can see in the sample has come from a certain underlying population
Null hypothesis significance testing (NHST) –
- Hypothesis – a statement / prediction about the data
- Null hypothesis – there is no difference / relationship
- NHST tells us if a certain pattern of data we can see in our sample has come
from an underlying population distribution where the null hypothesis is
true (where there is no difference)
- P = the probability of getting data patterns that we see in our sample if in
reality we measured everyone in the population and there is no difference /
relationship
- P can range from 0 to 1
- If the null hypothesis was true and there was no difference or relationship,
p < 0.05 ( there would be less than 5% chance of getting the pattern in the
data in the sample if the null was true) – if p is more than 0.05 then the null
hypothesis must be accepted
Type 1 and 2 errors –
- Type 1 error – no effect in population but we can see that there is an effect
from the sample data – null is true but there is something affecting the data
- Type 2 error – there is an effect in the population so the null isn’t true but
we decide that there isn’t an effect from our sample data
