Research Methods and Evidence-Based Practice
What to know
Candidates should understand study designs (RCTs, quasi-experimental, cohort, case-control, single-case designs), threats to validity (internal, external, statistical conclusion), and bias sources (selection, attrition, detection, publication). Know effect sizes (Cohen's d, r, NNT), confidence intervals, statistical versus clinical significance, and how to read a meta-analysis forest plot. Evidence-based practice is the integration of best research evidence, clinical expertise, and client values and preferences-not the rote application of manuals. Understand levels of evidence and practice guidelines, and the role of moderators and mediators.
Traps
Questions often confuse statistical with clinical significance, or ask which design supports causal claims (randomization and control). Watch for items that treat a correlational finding as causal, or that confuse a moderator (changes the strength of an effect) with a mediator (explains the mechanism).
Rule of thumb
Ask three questions of any study: was it randomized and controlled, how large was the effect, and does the sample resemble my client?
Worked example
A trial shows a 4-point symptom reduction with p<.001 but a tiny effect size and no functional improvement. Report it as statistically significant but not clinically meaningful, and weigh it against client values.