Choosing a Statistical Test A2

Question 1

What are the 3 things to look for when choosing a statistical test?

Answer 1

1) Type of data/ Level of Measurement
2) Test of Difference or Correlation
3) Experimental Design

Question 2

What is Nominal Data?

Answer 2

Data in the form of categories, discrete data

Question 3

What is Ordinal Data?

Answer 3

Data that can be put in order, has no equal intervals

Question 4

What is Interval Data?

Answer 4

Data that uses public scales of measurement* and is on a numerical scale

*Accepted units of measurements

Question 5

What is Unrelated Design?

Answer 5

Independent Groups

Question 6

What are the Related Designs?

Answer 6

Matched Pairs

Repeated Measures

Question 7

What is Nominal Data’s measure of central tendency and measure of dispersion?

Answer 7

Central Tendency - Mode

Measure of Dispersion - N/A

Question 8

What is Ordinal Data’s measure of central tendency and measure of dispersion?

Answer 8

Central Tendency - Median

Measure of Dispersion - Range

Question 9

What is Interval Data’s measure of central tendency and measure of dispersion?

Answer 9

Central Tendency - Mean

Measure of Dispersion - Standard Deviation

Question 10

Why are statistical tests conducted?

Answer 10

To determine whether the difference or correlation was found in a particular in investigation is statistically significant

Question 11

What does the outcome of statistical tests have implications for?

Answer 11

Whether we accept/reject the null hypothesis

Question 12

Draw the table for Choosing a Statistical Test.

What is its acronym?

Answer 12

Carrots Should Come

Mashed With Swede

Under Roast Potato’s

Question 13

For a Sign Test - What does the calculated value need to be for it to be significant?

Answer 13

calculated value needs to be EQUAL TO or LOWER than the critical value to be significant

Question 14

What is the Sign Tests calculated value sign?

Answer 14

S

Question 15

What is the acceptance level for an investigation to be classed as significant?

Answer 15

0.05

Question 16

What are the 5 Non-Parametric Tests?

Answer 16

Chi-Square
Sign Test 
Mann Whitney 
Wilcoxon
Spearman's Rho

Question 17

What are the 3 Parametric Tests?

Answer 17

Unrelated T-test
Related T-test
Pearson’s R

Question 18

What is a Type I error?

Answer 18

When the null hypothesis is is rejected and that alternative hypothesis is accepted when it shouldn’t have been as the null is ‘true’

Question 19

What is a Type I error also known as?

Answer 19

False Positive

Question 20

What is a Type II error?

Answer 20

When the null hypothesis is accepted and the alternative hypothesis has been rejected, but the alternative hypothesis is ‘true’ and should have been accepted

Question 21

What is a Type II error also known as?

Answer 21

False Negative

Question 22

When is a Type I error more likely to occur?

Answer 22

If the significance level is too lenient (aka too high)

Question 23

When is a Type II error more likely to occur?

Answer 23

If the significance level is too stringent (too low)