Randomized controlled trials (RCTs) are the gold standard in evidence-based research. At first glance, they seem simple enough: Give some patients the real drug, give some the placebo, and see what happens.
But the devil is in the details, and each small detail in designing an RCT is critical to the researcher’s end goal of finding meaningful results from each test. Choose the wrong study design, and your results lose meaning. Choose the wrong test subjects, you lose meaning. Choose the wrong variables to measure, the wrong hypothesis to test, or the wrong statistical analysis, and you lose meaning. In fact, flawed scientific research is much more abundant than the relatively rare paper that is sound on every one of these criterion.
This article is designed to make you a research-report-reading whiz, arming you with the ability to recognize the different components of an RCT, understand how they contribute to the overall study design, make sense of the results, and know how those results fit in to your world based on the way they were produced.
To do this, below is an example of an LPRC research study with key components labeled and defined. After a quick intro to the study, we’ll tackle the components of the study design, followed by components of all-important Results sections.
The Study: CVS InVue T-1000 Anti-Sweep Peg Hook R&D
Scope of Study: The LPRC worked with CVS to test the effectiveness of a new anti-sweep peg hook called with T-1000 developed by InVue. Effect on shrinkage was measured through cycle counts, and effect on sales was measured through POS data in two test stores and two control stores, a small-scale RCT design. See test store setup in image 1 and control store setup in image 2. Two product categories, Men’s Wet Shave and Printer Ink, were selected. Both categories had been protected by locking peg hooks. We therefore hypothesized a sales increase moving from locked to unlocked, while hypothesizing no change to a small increase in shrink.
Time to test your knowledge! To access the LPRC Research Report outlining the results of this test, go here. (LPRC login required).
Please contact mike [at] lpresearch.org with any questions or comments. Don’t forget to register for the LPRC Impact conference to see results from the other sixty 2018 research projects! http://lpresearch.org/impact/