Statistical Testing Flashcards
Why do we use statistical testing?
To find out if there is a significant difference and if the difference has occurred by chance.
What is significance?
If the results can be accepted because they aren’t due to chance.
What is the probability level?
- Mostly accepted ar 0.05 level
- P <- 0.05
- 95% sure results aren’t due to chance
What does a statistical test produce?
A number we compare on a critical values table
What are the 3 things we need to know for a critical value table?
- Number of participants
- One or two tailed hypotheses (direction or non- directional)
- Significance level desired (usually 0.05 or 5%)
If results are significant what do we accept?
The experimental hypothesis because there was a less than 5% probability that the results occurred but chance which means the difference was found because of the manipulation to the IV.
If no significance is found what do we accept?
The null hypothesis
Why would we use a stricter significance level e.g. 0.01 / 1%?
- When research involves human cost e.g. drug trials.
- When an investigation is a one-off and cant be repeated.
What are the 3 conditions for using sign-test?
- Difference rather than association
- Repeated measures design
- Nominal data
Method of sign test:
Subtract the 2nd condition from the 1st.
1. For each ppt not whether the result is negative or positive using + or -.
2. Count the number of + and -.
3. See which is the lower of the two.
4. The lower = your S value.
5. Compare S value with the critical values table.
6. If you value is equal to or less than the CV value we can accept the result as significant.