Statistical testing
- Provides a way of determining whether hypotheses should be rejected or accepted.
Criteria needed to look at:
1. If researcher is looking for a difference or a correlation
2. In the case of a difference, what experimental design is being used such as independent
groups, repeated measures or matched pairs
3. level of measurement (quantitative data) such as nominal, ordinal or interval.
Sign test
- Looked for a difference not an association.
- Used RMD
- Collected nominal data (organised into categories)
How to conduct:
1. State hypothesis (one tailed/ two tailed)
2. Subtract scores on table if its negative use – if its positive use + and if equal it will be
ignored.
3. Add up plusses and minuses = N
4. Find calculated value which is S (the number of times the less frequent sign occurs)
5. Compare S (calculated value) with N (critical value), to do this you need:
- Significance level of 0.05 (p)
- The number of participants in the investigation
- Whether it’s one tailed or two tailed (directional or non-directional)
- USE TABLE PROVIDED IN QUESTION
6. Calculated value must be equal or lower (REFER TO QUESTION) than critical value for
the data to be significant.
7. Conclusion, refer to hypothesis mentioning IV and DV and support conclusion with
values of critical value, calculated value, and what probability you used (0.05).
When choosing an inferential test, you must think about 3 factors:
1. Design of study unrelated design (IMD) or related design (RMD, MPD)
2. Level of data (ordinal, nominal or interval)
3. Whether difference of correlation is being measured.
Type of data Test of difference Test of association/
- Provides a way of determining whether hypotheses should be rejected or accepted.
Criteria needed to look at:
1. If researcher is looking for a difference or a correlation
2. In the case of a difference, what experimental design is being used such as independent
groups, repeated measures or matched pairs
3. level of measurement (quantitative data) such as nominal, ordinal or interval.
Sign test
- Looked for a difference not an association.
- Used RMD
- Collected nominal data (organised into categories)
How to conduct:
1. State hypothesis (one tailed/ two tailed)
2. Subtract scores on table if its negative use – if its positive use + and if equal it will be
ignored.
3. Add up plusses and minuses = N
4. Find calculated value which is S (the number of times the less frequent sign occurs)
5. Compare S (calculated value) with N (critical value), to do this you need:
- Significance level of 0.05 (p)
- The number of participants in the investigation
- Whether it’s one tailed or two tailed (directional or non-directional)
- USE TABLE PROVIDED IN QUESTION
6. Calculated value must be equal or lower (REFER TO QUESTION) than critical value for
the data to be significant.
7. Conclusion, refer to hypothesis mentioning IV and DV and support conclusion with
values of critical value, calculated value, and what probability you used (0.05).
When choosing an inferential test, you must think about 3 factors:
1. Design of study unrelated design (IMD) or related design (RMD, MPD)
2. Level of data (ordinal, nominal or interval)
3. Whether difference of correlation is being measured.
Type of data Test of difference Test of association/
