Statistical Tests

Question 1

When is a Student’s T-Test used?

Answer 1

To find out if the difference between two means is due to chance.

Question 2

How should you carry out a Students T-Test?

Answer 2

1. Write a null hypothesis in context of the data given
2. Calculate the test t value using the formula given
3. Determine the number of degrees of freedom in context of the question (n1 + n2) - 2 and, choose the appropriate critical value
4. Interpret the test t value by comparing it to the critical value and, accept or reject the null hypothesis

Question 3

When is a Chi-squared test used?

Answer 3

To find out if the difference between observed and expected frequencies is due to chance.

Question 4

How should you carry out a Chi-squared test?

Answer 4

1. Write a null hypothesis in context of the data given
2. Calculate the test value using the formula given
3. Determine the number of degrees of freedom ( number of categories - 1 ) and determine the appropriate critical value
4. Compare the test value to the the critical value and hence, accept or reject the null hypothesis

Question 5

When is a correlation coefficient test used?

Answer 5

To find out if the association between two variables is due to chance.

Question 6

How should you carry out a correlation coefficient test?

Answer 6

1. Write a null hypothesis in context of the data given
2. Calculate the test R value using the formula given
3. Determine the appropriate critical value using the number of pairs of measurements taken
4. Compare the test value with the critical value and hence, accept or reject the null hypothesis

Question 7

What can you conclude if the test value is greater than the critical value?

Answer 7

• less than a 5% probability that the difference/ association is due to chance
• the association/ difference is therefore not due to chance and is SIGNIFICANT
• REJECT null hypothesis

Question 8

What can you conclude if the test value is less than the critical value?

Answer 8

• there is more than a 5% probability that the association/ difference is due to chance
• the association/ difference is due to chance and therefore is NOT SIGNIFICANT
• ACCEPT null hypothesis