Please enable JavaScript.
Coggle requires JavaScript to display documents.
Statistical Interference (Hypothesis Testing (It generates a p-value, It…
Statistical Interference
Statistical Interference: These are the methods used to determine the likelihood that observed results were due to chance.
Hypothesis Testing
-
- It tests the null hypothesis
- It makes a statement about whether or not you have sufficient evidence to reject the null hypothesis
- It establishes a null hypothesis
Null Hypothesis: This where there is no difference in outcomes in terms of exposure to the study factor(s)
P-Values
- P < 0.05 is chosen as the cut-off for significance
- 5% is common (5% likelihood of results being due to chance)
- P-value of < 0.05 means the results are ‘statistically significant’
- The researcher decides IN ADVANCE the level of ‘significance’ (the p-value cut-off)
High P-Values
- It is possible there is no difference in outcomes in terms of exposure to the study factor
- There is insufficient evidence to reject the null hypothesis
-
Low P-Values
- It is likely that there is a true difference (of the outcome) between the groups in terms of exposure to the study factor
- The null hypothesis (of no difference) is rejected
-
-
Confidence Intervals
- The level of confidence most common is 95%
This is a range around a point estimate (e.g. OR, RR, mean, etc.) within which the true value is likely to lie with a specified degree of probability, assuming there is no systematic error (bias or confounding).
- It gives us an estimation of the precision of our results
- If the “null” value for the measure of association lies within the confidence interval, then researchers can conclude that the difference observed is not ‘statistically significant’