Peer review is wonderful in theory. Scientists reviewing other scientists’ work to evaluate whether the science was applied thoroughly, implemented well, and interpreted effectively can be a wonderful way to allow the best science through. But the Reproducibility Project clearly demonstrated that something is broken – when over a quarter of the published studies reviewed could not be replicated.
It is not a surprise to most. Humans are involved. We make mistakes…
Let’s see how simple we can make this…beware, I may go too far. If you really want to know about this topic, click the “Human Subjects Protections Update” button to the right, and you will have plenty to read…
Most agree that people have numerous motivations to participate in research. In a recent article, Florian Keusch thoroughly details the various reasons why people participate in web surveys. It is clear that there is no single reason — societal characteristics, individual characteristics, and survey characteristics all play a role. But like most strong literature reviews do, it left me with this…
Bear with me — but I have an idea. I may have stumbled upon the solution to a troubling trend in our science. Response rates are dropping. We used to say that it was critical to get high response rates in our studies. We spoke of our fight against our nemesis of “Non-Response Error”!
But then as it became clear that we were losing that fight…
Why do we NOT routinely capture data about respondents’ local current temperatures and cloud cover, recent or upcoming extreme weather, regional pollen counts, and other related data while we collect survey data?
Rarely a day passes without me saying (or hearing someone else say), “I wonder how this may impact data quality,” or, “I wonder why we keep doing it this way,” or my favorite, “I wonder if anyone else has wondered this too!“
I have seen the power of the scientific method. I have seen that ideas can grow and expand, and how they usually twist and turn. I know that science is not linear.
So it is time for me to spin off my little wonderments to the world…
If my office location being in Ann Arbor, Michigan doesn’t tip you off to my collegiate sports allegiance, then this will. In 1983, University of Michigan football coach Bo Schembechler addressed his team with a motivational talk…
When Apple’s ResearchKit was announced, almost immediately some squawked about the ethical problems that may be raised by its use. One argument suggested that because teens use iPhones, and because teens “aren’t supposed to take part in medical studies without parental consent” this would be bad. The argument proceeds…
Use this as a checklist for your own studies – or as a tool to evaluate survey research collaborators. Researchers who do not consider each of these elements will make disastrous mistakes. A solid researcher is not one who is perfect – it is one who learns from past mistakes. Here are the final six (of twelve)…