Statistical Testing and Probability

Question 1

What is statistical testing

Answer 1

It is a method used to see if a hypothesis should be accepted or rejected

Question 2

Why do we use statistical testing (2 reasons)

Answer 2

.they work out the probability of whether a particular set of data could have imply occurred by chance
.find a difference between 2 soles but we want to know if the difference is big enough for it be be significant

Question 3

What information do you need to know to determine what statistical test to use

Answer 3

3 D’s:
.level of Data (nominal,ordinal or interval)
.type of Design (related or unrelated)
.test if Difference or correlation

Question 4

What experimental designs are ‘related’ designs

Answer 4

Repeated measures or matched pairs

Question 5

What experimental designs are ‘unrelated’ designs

Answer 5

Independent group or single participant

Question 6

What is the mnemonic for determining which statistical test to use

Answer 6

Simon Cowell Wants More Singers Receiving Unanimous Praise

Question 7

What does the mnemonic correspond to for the statistical tests

Answer 7

Sign tests, Chi squared, Wilcoxon, Mann whitney, Spearman’s rho, Related T, Unrelated T and Pearsons

Question 8

When do we use the sign test

Answer 8

When it is a test of difference and related design and when the data is nominal

Question 9

What is the sign test

Answer 9

It is a calculation used to determine whether a difference is statistically significant

Question 10

What is the first step for the sign test

Answer 10

Put the pairs of related data into the table (going to be already done for you most times)

Question 11

What is the second step of the sign test

Answer 11

Subtract the values in the pair and label it with their + if it is a positive number or a - if it is a negative number

Question 12

What is step 3 of the sign test

Answer 12

Count how many - and + there are and the one which has a lower number is the S value

Question 13

What is step 4 of the sign test

Answer 13

Compare the calculated value (S) with the critical value

Question 14

What criteria’s do you need to follow to find the critical value from the critical value table

Answer 14

.One or two tailed hypothesis?
.How many total participants? (Take away those who have no difference)
.What level of probability are you using?

Question 15

What is a one tailed hypothesis

Answer 15

A directional hypothesis

Question 16

What is a two tailed hypothesis

Answer 16

non directional hypothesis

Question 17

What level of probability is always used (unless said otherwise in question)

Answer 17

0.05 (5%)

Question 18

What do you use when comparing the S and CV and determining whether the hypothesis is significant

Answer 18

The statement below the critical value table

Question 19

What other 2 probability values may be used by psychologists

Answer 19

0.1 or 0.01

Question 20

Why is 0.05 the probability level that psychologists work at

Answer 20

Because they need to be 95% sure of their research and so that errors are reduced but also so that the data isn’t too strict that it leads to more type 2 error

Question 21

What type of error is more likely to be cause because of 0.1 probability level

Answer 21

Type 1

Question 22

What type of error is more likely to be cause because of 0.01 probability level

Answer 22

Type 2

Question 23

What is a type 1 error

Answer 23

When the investigator rejects a hypothesis or data but it is actually right (false positive)

Question 24

What is a type 2 error

Answer 24

When the investigator thinks a hypothesis or data is true but it is actually false (false negative)

Question 25

What is a null hypothesis

Answer 25

A hypothesis that states that there is no relationship between the 2 variables

Question 26

Do you have to know how to calculate the other stats test other than sign tests

Answer 26

No

Question 27

What do you have to know about the other stats tests

Answer 27

Know when to use them and compare critical and calculated value and state if its significant

Question 28

When do you use wilcoxon

Answer 28

Test of differences, related designs and ordinal data

Question 29

When do we use Mann Whitney

Answer 29

Test of difference, unrelated design and ordinal data

Question 30

When do we use spearman’s rho

Answer 30

Correlation and ordinal data

Question 31

When do we use chi squared

Answer 31

Test of difference, unrelated design and nominal data

Question 32

When to use Pearsons r

Answer 32

Correlation and interval data

Question 33

When to use T tests (related and unrelated)

Answer 33

Test of difference, corresponding design and interval data

Question 34

When is 0.01 probability level used

Answer 34

When the researcher needs to be more confident and eliminate as much chance of uncertainty as possible. Also mainly when there is a human cost such as drug testing, you want to be more careful and be 99.99% sure