Question

Problems and Research

Healthcare professionals need to answer three essential questions about the groups studied:

1. Did something different occur between the groups (usually due to some intervention or treatment)?

2. If something different occurred between the groups, was it due to chance?

3. If something nonrandom occurred between the groups, how large was the effect (was the difference clinically meaningful)?

From these questions, could you answer:

1) How do these questions pertain to the null hypothesis?

2) Should these professionals be interested in observing differences or similarities between study groups to achieve statistical significance? Explain your answer.

3) If the mere chance was behind the observed study results, do we accept or reject the null hypothesis? Explain your answer.

4) What is the "power" of a statistical test? Please, describe what it refers to.

Answer 1

1. The first two questions pertain to the null hypothesis in such a manner that if something different has really occurred between the groups, the null hypothesis is rejected and the alternate hypothesis willl be accepted. If nothing different has occurred, healthcare professionals can't reject the null hypothesis.

2. Yes, the professionals should be interested in observing the differences using different test statistics such as the t-statistic or F-ratio obtained from different procedures. Statistical significance is all about the likelihood of observing the results- whether it is not due to any chance, it shouldn't be compared to clinical significance which the professionals interpet after the study results.

3. First, we should know that no hypothesis test is 100% certain and there is always a chance that we are making a wrong conclusion. When we say that we are conducting the test with 95% confidence interval, we are saying that we are willing to accept a 5% chance that we might be wrong when we reject the null hypothesis. Hence, if there is a mere chance behind the results, we fail to reject the null hypothesis. We always try to avoid conclusions by just chance.

4. The power of a statistical test is the probability of correctly rejecting the null hypothesis when it is really false. It is dependent on various factors such as the significance level of the test, the size of the effect being measured etc. It is the probability of not committing a Type II error which might be very costly in the medical sector. Hence, it is a quality parameter.