Part 4: Hypothesis Testing And t-Tests

Question 1

P-values determine how likely the obtained result would occur if the null hypothesis was:

Answer 1

true.

Question 2

All statistical tests are calculated by:

Answer 2

Treatment effect / error

Treatment effect is the group differences.

Question 3

5 steps of hypothesis testing:

Answer 3

1. State hypotheses.
2. Covert score to z-score.
3. Compare obtained value with critical value of +/-1.96.
4. Make statistical decision.
5. Write up in APA format.

Question 4

What should we do if the variance of the population is known when we are dealing with our sample?

Answer 4

We can incorporate calculating the standard error into our z-score.

Question 5

What should we do if the variance of the population is unknown when we are dealing with our sample?

Answer 5

Run a t-test.

Use t-table to make statistical decision.

Question 6

Rejecting the Ho when it is true, is called a:

Answer 6

Type 1 error.

Question 7

Retaining the Ho when is false, is a:

Answer 7

Type 2 error.

Question 8

To avoid making type 1 or 2 errors, we want the power of our test to be around __, which gives us an __% chance of being correct.

Answer 8

To avoid making type 1 or 2 errors, we want the power of our test to be around .8, which gives us an 80% chance of being correct.

Question 9

What 3 things can influence the power of our test?

Answer 9

1. Alpha (significance) level.
2. Effect size.
3. Sample size.

Question 10

The standard difference between sample means is called:

Answer 10

Cohens d

Pearsons r

Question 11

What are the 4 assumptions of parametric tests?

Answer 11

1. Normally distributed data.
2. Similar variances between groups.
3. Equal distance between scale points.
4. Behaviour of one participant doesn’t influence another.

Question 12

One continuous DV
One categorical IV with 2 levels in which participants are exposed to the same levels.

What type of test should you use?

Answer 12

Paired samples t test

Question 13

What are the 9 steps to a paired samples t-test?

Answer 13

1. Calculate the difference scores.
2. Calculate the mean difference score.
3. Compute the SD of each difference score.
4. Compute the estimated SE.
5. Compute the df.
6. Compute t obtained.
7. Look up t critical and make statistical decision.
8. Compute effect size (d).
9. Write up findings in APA format.

Question 14

One continuous DV.
One categorical IV with 2 levels in which different participants are exposed to only one of the two levels.

What test should we use?

Answer 14

Independent t test.

Question 15

What are the 9 steps to an independent t test?

Answer 15

1. Check parametric assumptions are met, with similar variances in each group.
2. Compute means and variances.
3. Compute pooled variance.
4. Compute estimated SE.
5. Compute df.
6. Compute t obtained.
7. Look up t critical and make statistical decision.
8. Calculate effect size (d).
9. Write up findings in APA format.