Research methods - Inferential testing

Question 1

How do you see if our result is significant?

Answer 1

• Work out the probability that what we have found is due to chance.
• This is expressed as a value, p.
• The smaller the value of p is, the more significant your results are.

Question 2

What are the steps to find out if your data is significant?

Answer 2

• Collect your data
• Carry out a statistical test
• This gives you the calculated value
• Compare calculated value with the critical value in a statistics table
• Read the table to see if your data is significant, and if so, what level of significance (p)

Question 3

In psychology, what percentage do we usually aim to be sure that our results are due to a real effect?

Answer 3

• 95% sure
• 5% probability that the null is true
• This is expressed as p=<0.05. This is known as the significance level.
• So p=<0.05 means there is an equal to or less than 5% probability that our results are due to chance.

Question 4

What does it mean by ‘the results were significant at p<0.05’?

Answer 4

• This means that the result is significant at 0.05 level

- This means there is a less than 5% probability that the difference/relationship found was due to chance

Question 5

What is a type 1 error?

Answer 5

This is a false positive.

It is accepting the alternate when we should have accepted the null.

Question 6

What is a type 2 error?

Answer 6

This is a false negative.

It is accepting the null when we should have rejected it.

Question 7

Why are statistical tests used in psychological research?

Answer 7

Statistical tests are used to determine the likelihood that the effect they have found is due to chance by allowing a significance level to be determined.

Question 8

What do statistical tests for correlations give us?

Answer 8

A correlation coefficient

calculated value

Question 9

What is the sign test?

Answer 9

• A test of difference
• Use it for nominal data
• Used to analyse repeated measures or matched pairs design (related)

Question 10

How do you carry out the sign test?

Answer 10

• For each participant, note whether the sign is +, - or = depending on whether the score in each condition goes down, up or stays the same.
• Count N, which is the number of participants, excluding those with equal (=) scores.
• Count the number of the least frequent sign (- or +).
• This is s.
• Use the sign test table
• See whether the results are significant by seeing the number of N, whether the hypothesis is 1 or 2 tailed and at what level of significance it is at.

Question 11

What test is used if:

• It is a test of difference
• It is unrelated
• It uses nominal data?

Answer 11

Chi-squared

Question 12

What test is used if:

• It is a test of difference
• It is related
• It uses nominal data?

Answer 12

Sign test

Question 13

What test is used if:

• It is a test of difference
• It is unrelated
• It uses ordinal data?

Answer 13

Mann-Whitney

Question 14

What test is used if:

• It is a test of difference
• It is related
• It uses ordinal data?

Answer 14

Wilcoxon

Question 15

What test is used if:

• It is a test of difference
• It is unrelated
• It uses interval data?

Answer 15

Unrelated t-test

Question 16

What test is used if:

• It is a test of difference
• It is related
• It uses interval data?

Answer 16

Related t-test

Question 17

What test is used if:

• It is a test of correlation
• It uses nominal data?

Answer 17

Chi-squared

Question 18

What test is used if:

• It is a test of correlation
• It uses ordinal data?

Answer 18

Spearman’s rho

Question 19

What test is used if:

• It is a test of correlation
• It uses interval data?

Answer 19

Pearson’s r