Research Methods - Factors Affecting Choice Of Statistical Test

Question 1

What 3 pieces of information should be known when picking statistical test

Answer 1

Is the researcher looking for a difference or relationship (correlation)

If a difference is being investigated, what type of experimental design has been used?

What level of measurement is being used?

Question 2

How would you work out if it is a difference or correlation

Answer 2

Difference hypothesis = difference, reference to conditions, experiment
• Correlation hypothesis = relationship, association

Question 3

What are three different types of design and explain them

Answer 3

• Independent groups design – each participant performs in one condition of an experiment
• Repeated measures design – each participant performs in all conditions of an experiment
• Matched pairs design – participants are put into pairs based on key characteristics relevant to
the study. One member of each pair performs in one condition of an experiment

Question 4

Repeated measures and matched pairs designs are referred to as what

Answer 4

Related designs

Participants in conditions related in some way

Question 5

What is independent groups design referred as

Answer 5

Unrelated design

Question 6

What are the 3 levels of measurement

Answer 6

Nominal
Ordinal
Interval

Question 7

What is nominal data

Answer 7

Data represented in the form of categories
Crudest and most uninformative
Eg it was either a sunny day or rainy day

Question 8

What is ordinal data

Answer 8

Data that is ordered in some way
No equal intervals
Lacks precision

Eg ranking, order of scaling ,
You were first, second, third person into classroom

Question 9

What is interval data

Answer 9

Data based on numerical scales that include units of equal, precisely defined size
Differences between MEASUREMENTS
Eg, heart rate, degrees Celsius , seconds

Question 10

If test of difference with unrelated designs and nominal data then which test is it

Answer 10

Chi squared

Question 11

List what test you would choose If test of difference with unrelated design and
Nominal :
Ordinal :
Interval :

Answer 11

Nominal : chi squared
Ordinal : Mann Whitney
Interval: unrelated t test

Question 12

List what test you would choose if data if test of difference with related design 
And 
Nominal :
Ordinal
Interval

Answer 12

Nominal: sign test
Ordinal: wilcoxon
Interval: related t test

Question 13

Wha test would you choose for association or correlation and
Nominal
Ordinal
Interval

Answer 13

Nominal : chi squared
Ordinal: spearman’s rho
Interval : Pearson’s r