Blog Summary for Tues July 12th Journal Club Presentation
Strong Inference
John R. Platt
Science, New Series, Vol. 146, No. 3642. (Oct. 16, 1964), pp. 347-353.
Strong inference (SI) identifies, explains and supports a systematic use of inductive thinking. The paper describes SI as a four step method that if adopted will enhance scientific progress in both speed and accuracy of findings. SI steps include 1) devising alternative hypotheses 2) devising crucial experiment(s) to test among the alternatives for the purpose of eliminating the alternatives which are inconsistent with the data 3) running a “clean” experiment that supports the conclusions 4) Recycling the procedure by repeating steps/observations with the intended purpose of revising, rejecting or accepting the accepted hypothesis. The implementation of multiple hypotheses, as noted in the paper, is intended to protect against bias to try to confirm our first “pet” idea. And, it forces the researcher to think outside of the box and about “all” possibilities.
In other words, imagine a world where we can look up the answers to everything! Hum! Yes, like having access to Google or Wiki or Dr. Mueller’s blog! So pretend that in this everything-at-your-finger-tips world, a young lad has car trouble. He looks up the trouble he is having by typing it in the famous search engine (no pun intended). He then determines what is wrong with his car and happily provides his personal diagnosis to the mechanic who has been hired to repair the car. In the world of Google, it is possible that the car mechanic would do whatever the customer desired. However, in a world of SI, the mechanic would follow a method for determining the car trouble. She would
1) listen to the car owner, use her background knowledge in automechanics to determine possible problems
2) run a series of tests based on the car symptoms, the owner’s complaint, and his automehcnaic knowledge
3) eliminate from the list of possible problems based on the tests
4) refine the tests as needed until a firm diagnosis was reached
5) repair the car ($$$)
The strong-inference protocol: not just for grant proposals
Sara M. Hiebert
ADV PHYSIOL EDUC 31:93-96, 2007
Department of Biology, Swarthmore College, Swarthmore, Pennsylvania http://advan.physiology.org/content/31/1/93.full
In this paper, we focus only on the Appendix—Specific content for the chicken embryo metabolism experiment. The author of this paper contends that students in the Animal Physiology course are able “to apply the general form of the strong-inference (SI) protocol to specific experiments, with the proviso that they benefit…in the early stages from seeing a specific example on which to model their own protocols” (p94). The students are armed with an invaluable tool—their personally designed SI protocols— to respond to the overarching question and support thinking throughout the experiment. Prior to the student- designed experiment, students are required to complete the introduction and background, methods, and outcomes (alternative hypotheses) sections of their lab reports. In this way, students know ahead of time and throughout the lab why and how the experiment is to be done. And the outcomes section provides a list of differential diagnoses for the upcoming experimental results. The paper asserts that the SI approach to conducting experiments
1) supports the laboratory’s direction/purpose by keeping students focused on their protocol
2) benefits student thinking by developing big idea understanding
3) frontloads the actual writing of the lab report thereby taking the pressure off of completing an arduous task at the close of the experiment
4) maintains the focus on the larger questions of why something is occurring and what are the implications of our experiment, as opposed to being lost in the method of step by step cookbook instructions
No comments:
Post a Comment