L10- Factors affecting choice of statistical tests including levels of measurement and design

Question 1

What are descriptive statistics?

Answer 1

Descriptive statistics- give summaries of data collected from research- AND indicates what statistical analysis might reveal about results

Question 2

What are levels of measurement?

Answer 2

Levels of Measurement- used to try to categorise data into one of 3 types (nominal, ordinal or interval)- allows to correctly select most appropriate statistical test to analyse our results

Question 3

What are the levels of measurement?

Answer 3

1) Nominal
2) Ordinal
3) Interval

Question 4

What is nominal data?

Answer 4

• Data consists of numbers of participants that fall into different categories
• Person can be placed in 1 category only and not the other

Question 5

What is ordinal data?

Answer 5

• Data placed in rank order from lowest to highest

- Ordinal scale can consist of measurements of unequal intervals e.g. 1.20, 1.25, 1.27

Question 6

What is interval data?

Answer 6

• Data has fixed and even intervals

- Units of data are fixed (have same distance) throughout range e.g. cm etc

Question 7

How many statistical tests are there?

Answer 7

7

Question 8

What are the statistical tests and state which ones are parametric and which ones are non-parametric ?

Answer 8

Non-parametric:

1) Chi squared
2) Spearman’s Rho
3) Mann Whitney
4) Wilcoxon

Parametric:

5) Pearson’s r
6) Related t-test
7) Unrelated t-test

Question 9

What is the difference between parametric and non-parametric tests?

Answer 9

Parametric tests are ⬆️ robust and powerful than non-parametric tests
- rely on actual data collected rather than just examining rank order of data
Parametric test ALSO ⬆️ likely to detect if data significant or not

Question 10

What are the 3 factors which mean a parametric test can be conducted?

Answer 10

1) Interval level of measurement
2) Normal distribution rather than skewed distribution
3) Similar variance/spread of data- examined by looking at dispersion of data and standard deviations for both conditions and seeing if they are similar

Question 11

Describe the table headings which help to determine which test to conduct

Answer 11

Far left- type of test
2nd to left- levels of measurement
3rd to left- test of difference- independent measures
4th to left- test of difference- repeated measures or matched pairs
5th to left- test of association or correlation-

Row 1- non- parametric
Row 2- non-parametric
Row 3- parametric

Question 12

What rhyme is used to remember which statistical test to use and what does it mean?

Answer 12

Carrots Should Come (Chi squared, Sign test, Chi squared)
Mashed with Swede (Mann Whitney U test, Wilcoxon, Spearman’s rho)
Under Roast Potatoes (Unrelated t-test, Related t-test, Pearson’s r)

Question 13

What statistical should be used with an independent measure test using nominal data?

Answer 13

Chi squared

Question 14

What statistical should be used with an independent measure test using ordinal data?

Answer 14

Mann Whitney U test

Question 15

What statistical should be used with an independent measure test using interval data?

Answer 15

Unrelated t-test

Question 16

What statistical should be used with an matched pairs test using nominal data?

Answer 16

Sign test

Question 17

What statistical should be used with an matched pairs test using ordinal data?

Answer 17

Wilcoxon

Question 18

What statistical should be used with an matched pairs test using interval data?

Answer 18

Related t-test

Question 19

What statistical should be used with an repeated measures test using nominal data?

Answer 19

Sign test

Question 20

What statistical should be used with a repeated measures test using ordinal data?

Answer 20

Wilcoxon

Question 21

What statistical should be used with a repeated measures test using interval data?

Answer 21

Related t-test

Question 22

What statistical should be used with a test of association and what level of measurement does it use and what type of test is it?

Answer 22

Chi squared- nominal and non-parametric

Question 23

What statistical should be used with a test of correlation using ordinal data?

Answer 23

Spearman’s Rho

Question 24

What statistical should be used with a test of correlation using interval data?

Answer 24

Pearson’s R

Question 25

Which level of measurements come under parametric test and which come under non-parametric tests?

Answer 25

Non- parametric = Nominal
Non- parametric = Ordinal
Parametric = Interval

Question 26

What is unique about the last column of the table to determine the statistical test to use?

Answer 26

There is only 1 test for association (top row … nominal data) and the other 2 rows are both for a test of correlation (ordinal and interval data)