 | |
| |
| Module 3: Interpreting Data
4. Using data to support an argument; making inferences
4.4 Does statistically significant mean practically significant?
For the test that was conducted on Vitamin C, the results obtained are presented in the table below:
| | % with no colds | % with colds |
Placebo | 18 | 82 |
Vitamin C | 26 | 74 |
Analysis of this result indicated that the difference was statistically significant. Statisticians can evaluate that a difference this large would arise by chance in 1% of studios of this size and design thus the result is statistically significant.
However, we could also ask if the difference in the table would be enough to convince people to take 1 gram of Vitamin C every day of their life? Would people change their behaviour on the basis of these results? If they wouldn't, then you need to ask whether these results are practically significant even though they are statistically significant.
Also, as mentioned earlier, the larger the sample size the smaller the association between variables can be and still be assessed as statistically significant. However, the judgement as to practical significance is related to effect size, which does not depend on sample size, whereas statistical significance does. | | |
| | | |
|
| | |
|
This page last updated 1 September 2009 | 
