A [hypothetical] new drug was created and evaluated in a large international, rigorously designed randomised controlled trial (RCT) to determine its ability to lower blood pressure in people with risk factors for cardiovascular disease. The study showed very promising results. When administered in routine clinical practice, however, blood pressure improvements were not observed. Further examination led to a hypothesis that the new drug did not perform well in the more generalist population due to their many differing chronic disease comorbidities (and likely associated drug interactions) compared to the participants in the RCT (people with comorbidities were not eligible for recruitment).

True or false: This observation is a good example of an intervention proving to have good effectiveness but poor efficacy.

The following studies investigated the effect of treatments on patients following total knee replacement surgery using the following outcomes:

Study A: Examined the effect of early mobilisation on active range of knee extension (via goniometry)

Study B: Examined the effect of continuous passive mobilisation on knee pain (via 1-5 Likert scale questionnaire)

Study C: Examined the effect of different forms of analgesia on knee pain (via 0-10 visual analogue scale)

Study D: Examined the effect of a multidisciplinary 'care bundle' on quality of life (Short-Form-36 questionnaire)

Study E: Examined the effect of more intensive physiotherapy on knee swelling (measured circumference, cms)

Study F: Examined the effect of TENS on subjective measures of knee pain (custom questionnaire, score range 0-100).

It would be appropriate to calculate an effect size to summarise ('pool') results from which of the above 6 studies?

We evaluate exam data from two independent samples of students (group A and B) at the end of a semester. The mean score for group A was 82% while the mean score for group B was 85%. Findings from an independent samples t-test (two-tailed, alpha 0.05) reveal the following:

Critical t: 2.04

T-statistic: 1.93

True of false: Based upon these findings, we can conclude the mean difference in exam scores was highly unlikely (<5%) due to chance and the difference is 'statistically significant'.

You read a published report of a randomised controlled trial of cryotherapy compared to no treatment (control) and try to gauge how large the the treatment effect was. You calculate a Cohen's d standardised mean difference (effect size) and the result is 0.4.

Based on this result, we would conventionally describe the magnitude of this treatment effect as being:

You conduct an unpaired (independent samples) t-test to evaluate treatment responses between two groups of patients who received treatment A (n=20) or treatment B (n=19).

How many degrees of freedom should be used for this test? Indicate a single number only (no decimals).

Which one of the following sets is paired correctly?