You find several studies that report upon the effectiveness of physiotherapy for improving range of motion flexion following shoulder impingement pathology. They report upon the outcome of range of motion via:

a) goniometry (degrees);

b) distance marking on a wall (height from floor to peak of 'finger climb', in cms); and

c) a self-reported scale from 0 - 20 (higher scores denoting better freedom of flexion movement). 

Which of the following statements are correct?

You seek to evaluate, among all athletes referred to you for physiotherapy evaluation and management, the proportion of patients with confirmed (or highly likely) patellofemoral joint pain / 'runners' knee'. 

What kind of research question would best describe this purpose?

We wish to evaluate whether teaching research skills to students affects their ability to critically evaluate evidence. We implement a knowledge test (continuous / ratio data) at the beginning and end of a semester in a cohort of 100 students and compare the pre- vs post- scores.

Which statistical test should we use for analysis of our data in order to determine the effect of our teaching on student knowledge?

You examine the effectiveness of manual therapy technique on pain-free range of motion (ROM) and compare it to ultrasound (n=20 for both groups). The mean (SD) post-treatment ROM was 65 (10) degrees in the intervention group and  50 (10) in the ultrasound group.

Calculate a Cohen's d effect size (one decimal place).

You read a paper in which the study authors are evaluating their newly developed self-reported physical activity questionnaire. They wish to evaluate its 'proof of concept' and see how well self-reported scores relate to objectively measured step counts.

What kind of research question is being examined here?

You read an interesting article that describes some of the key advancements in physiotherapy education over the past 10 years. It is written by an esteemed academic in the field based on their awareness of the 'hot topics' they have witnessed change in their practice. No protocol was deemed necessary by the author as they were already aware of the major papers in the field.

True or false: This is a likely example of a systematic review.

The independent and dependent t tests are different from each other because the units of interest for the independent t test are the mean and variance of difference scores (i.e. change) whereas the units of interest for the dependent t test are the mean and variance for each group.

True or false?

You hypothesize that physiotherapy will improve pain levels in patients with chronic back pain and undertake an evaluation of treatment outcomes. You compare pain scores measured after treatment with pain scores measured pre-treatment in a cohort of 50 patients. Assuming an alpha of 0.05, use MS Excel's =T.INV function to work out the reference critical t value. 

Report your value to one decimal place (do not enter any + or - sign in front of the number).

Several different approaches can be used to calculate effect sizes, with the most common type (Cohen's d) derived from use of the pooled standard deviation (across the intervention and control groups) as the denominator in the mathematical equation.

Under which condition might you opt for an alternative calculation method such as Hedge's g or Glass's delta?

Review the forest plot below and answer the following question:

From visual inspection of the forest plot alone (no data presented), select all study/studies that demonstrated a statistically significant difference between the intervention and control group.