Elimination of Investigator Bias
We won’t throw out the baby with the bathwater. Just because we discard statistical analysis doesn’t mean we ignore the many insights on bias developed over the centuries by the discipline of statistics, for example how to avoid selection bias, investigator bias, etc. We preserve intact those practices that mechanically reduce bias in experimental results. And we have special firewalls in place to fend off the Trial Lawyers. Although it is somewhat flawed by its uncritical acceptance of the value of comparator trials, the article by Lindner (subscription required) should be one of the first sources consulted for its extensive list of potential investigator biases of which we should be aware, and for which we should develop direct methods of control (i.e., in contrast to the indirect methods of hiding behind a veil of numbers).
Again, if analysis of video tapes is a powerful means of avoiding investigator bias, where would you look to recruit individuals in order to improve your odds of catching subtle visual clues or signals?
- Login or register to post comments
- visit linked page
- Surprise Me!