- By conducting an inferential test we are checking how far we can accept the null hypothesis.
- To asses the probability that the results are due to chance, an inferential statistical test is used.
-Inferential statistics tell us whether the difference between two sets of scores is significant or due to chance.
-In psychology, we accept the null hypothesis as the best explaination of out of results unless there is 5% probability or less of the results being due to chance.
Three Types of Data
- Nominal Data :- Categories
Eg: Hot day or Cold day
- Ordinal Data :- Ranking
Eg: How Hot or Cold (Ranked 1-10)
- Interval or Ratio :- Measuring
Eg: Numerical (the exact numerical value)
> Independent measures - An experimental design whereby each group involved contains different people. for every conditionof the experiment, a different group of participants will be used.
> Repeated measures design - Where the same participants undergo all conditions of the study.
> Matched pairs design - An independent measures design where the experimental groups are matched on important characteristics, eg: Background, Age, etc.
> Correlation design - Where one participant provides data for two measures which are then tested to see if they show a relationship. There are two variables, (but not an IV or a DV) both are measured and both are of interest.
In Inferential Tests
You should be looking either
- For a Difference
- For a Correlation
Spearman's Rank Correlation
- If a study looks for a correlation, it should be Spearman's Rank.
- It should contain Ordinal data
- The participant design should be a correlation design.
Chi- Squared Correlation Coefficient
- If a study looks for a difference, it can be a Chi-Squared test.
- It should contain Nominal data.
- The participant design should be Independent groups design.
Mann Whitney U Test
- If you arw lokking for a difference, it can be Mann Whitney U test.
- It should contain Ordinal data.
- The participant design should be Independent groups.
Some Important Points
- Identify the critical value; check 0.05 or 5%
Calculated Value > Critical Value = Positive correlation
- Accept alternative/experimental hypothesis and reject null hypothesis
Critical Value > Calculated Value = Negative Correlation
- Accept Null hypothesis and reject alternative/experimental hypothesis