Introducing Inferential Tests

HideShow resource information
What is the main inference of significance?
We're interested statement about the population from which the sample are drawn rather than the samples themselves.
1 of 24
How do we work out whether a difference is significant?
We use inferential tests which permit you to work out, at a given probability, whether a pattern in data could have arisen by chance or because there's a real difference/correlation in the populations.
2 of 24
Define chance.
Something with no cause.
3 of 24
Why do we 'risk' a probability?
Because you can't be 100% certain that an observed effect wasn't due to chance, but you can state how certain you are.
4 of 24
In general, what percentage of probability do psychologists use and why?
They use a probability of 95%, which expresses the degree of uncertainty, meaning that there's a 5% chance of the results occurring if the null hypothesis is true, or a 5% probability that the results would occur even if there was no real difference.
5 of 24
How is probability recorded?
The probability of 5% is recorded as p=0.05.
6 of 24
What would a psychologist do if they wanted to be more certain about their results?
They would use a more stringent probability/significance level.
7 of 24
Define null hypothesis.
A statement of no difference or correlation, or that 'nothing is going on'.
8 of 24
Why do psychologists use a 5% level of significance?
Because it's a compromise between type 1 and 2 errors.
9 of 24
What do type 1 and 2 errors concern?
Whether we make a mistake about accepting/rejecting the null hypothesis.
10 of 24
Consider the null hypothesis 'there's no difference in the map reading abilities of men and women'. A study is conducted and a small difference found. What are the 2 correct conclusions and 2 possible errors?
There's no difference but because we use a lenient level of significance we reject the null hypothesis, which is a Type 1 error; we reject a null hypothesis that's true. There's a difference but because we're using a significance level that's too...
11 of 24
Previous card continued
stringent, we accept the null hypothesis when it's false, which is a Type 2 error.
12 of 24
Why do we use inferential tests?
They help us draw inferences about populations based on the samples.
13 of 24
What do you do during a inferential test? Give examples of the outcome.
Each test involves taking data and doing calculations to produce the test statistic: in Spearman's Rank Order Correlation Coefficient it's called rho whereas for Mann-Whitney it's U.
14 of 24
How do we decide if the observed value is significant?
The test statistic is compared to another number found in a table of critical values, which are different for each test. The critical value is the number that a test statistic much reach in order for the null hypothesis to be rejected.
15 of 24
What 4 things do you need to know to find the appropriate critical value?
Degrees of freedom - found by looking at the number of participants; 1 or 2 tailed tests - if the hypothesis was directional you would use a 1 tailed, if it was non-directional you would use a 2 tailed; significance level; and whether the observed...
16 of 24
Previous card continued.
value needs to be greater or less than the critical value for significance to be shown.
17 of 24
What 2 factors effect the choice of inferential tests?
The research design and the level of measurement.
18 of 24
What test do you use if the data is nominal?
Chi-square test.
19 of 24
What test do you use if it is a correlation?
Spearman's Rho test.
20 of 24
What test do you use if independent groups were used?
Mann-Whitney U test.
21 of 24
If the data isn't nominal, it's not a correlation and independent groups weren't used, what test would you use?
Wilcoxon T test.
22 of 24
What 3 things do you have to mention when justifying your choice of test?
Identify the level of measurement with reference to the data, state whether a test of correlation or difference is needed and justify, and state whether it's independent groups or repeated measures and justify.
23 of 24
Name the 3 levels of measurement and outline each one.
Nominal - The data are in separate categories, ordinal - data are ordered, and the 'difference' between each item isn't the same, and interval - data are measured using units of equal intervals.
24 of 24

Other cards in this set

Card 2

Front

How do we work out whether a difference is significant?

Back

We use inferential tests which permit you to work out, at a given probability, whether a pattern in data could have arisen by chance or because there's a real difference/correlation in the populations.

Card 3

Front

Define chance.

Back

Preview of the front of card 3

Card 4

Front

Why do we 'risk' a probability?

Back

Preview of the front of card 4

Card 5

Front

In general, what percentage of probability do psychologists use and why?

Back

Preview of the front of card 5
View more cards

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Research methods and techniques resources »