HOW TO CRAFT A RESEARCH QUESTION. Jan Weisenberger Senior Associate Vice President for Research

Transcription

1 HOW TO CRAFT A RESEARCH QUESTION Jan Weisenberger Senior Associate Vice President for Research weisenberger.21@osu.edu June 5, 2013

2 RESEARCH IDEAS ARE EVERYWHERE

3 SOME SOURCES OF RESEARCH QUESTIONS Courses, classes, labs Other experiences (internships, clinical rotations) Previous research papers Something your advisor is working on Support or challenge to a theory Just stuff (have you ever wondered...?)

4 SOME SOURCES OF RESEARCH QUESTIONS - 2 Unexpected occurrences Chilean Earthquake, Hurricane Sandy, Oklahoma Tornado Ohio State research team documents shift of major cities after Chilean earthquake

5 SOME SOURCES OF RESEARCH QUESTIONS - 3 Incidental findings: COMMON ANTI-INFLAMMATORY COAXES LIVER CANCER CELLS TO COMMIT SUICIDE COLUMBUS, Ohio The antiinflammatory drug celecoxib, known by the brand name Celebrex, triggers liver cancer cell death by reacting with a protein in a way that makes those cells commit suicide, according to a new study

6 SOME SOURCES OF RESEARCH QUESTIONS - 4 Answer to a need from a business or industry How do you design instrument clusters to minimize driver distraction?

7 SOME SOURCES OF RESEARCH QUESTIONS - 5 Maybe you have a fabulous new idea Something innovative that no one has ever done But how do you know if it is any good? EXCMR s non-ferrous treadmill 3-D Sensor Design Tech4Imaging, LLC

8 SO HOW DO YOU REFINE THESE GENERAL QUESTIONS INTO A RESEARCH TOPIC? Your research advisor research program A roadmap of research studies covering a span of years, designed to address an issue Previous literature Intro section: May identify unanswered questions Discussion section: Often outlines next studies that need to be done

9 The #1 error of young researchers: ASKING TOO BIG A QUESTION! How to make it smaller? Do I need to take a step approach to address this problem? What kind of timeline do I have? What kind of resources do I have? What kind of expertise do I have? SO HOW DO YOU REFINE THESE GENERAL QUESTIONS INTO A RESEARCH TOPIC?

10 THE STEP APPROACH STRONG INFERENCE Often, big questions are answered in a series of smaller steps: Strong inference: What are all the possible ways a study can come out? What does each outcome mean explanations? What follow-up study allows us to eliminate competing explanations? o What are all the possible ways that study can come out? o Etc.

11 Hypothesis-driven research DESIGN THE RESEARCH QUESTION SO THAT YOU GET A USEFUL ANSWER - 2 Null hypothesis: o Your manipulation will have no effect Alternative hypothesis: o Your manipulation will have an effect o But which way (directionality)? Example: Best material for spintronics applications GaMnAs or ZnCdSe?

12 DON T DUPLICATE THE PAST Do your homework! Has your study already been done by someone else? Your textbooks are way out of date Reference databases: Medline, PsychInfo, hundreds of others Recent review articles: Reference lists are great for working backwards Proceedings from recent professional meetings in your area of study the latest findings Use Google or other search engines carefully!

13 Is the increase in shootings at schools and other large group venues caused by excessive exposure to violent movies and video games? What do I need to define? SOME EXAMPLES OF VALID RESEARCH QUESTIONS o Has the incidence increased? o What constitutes excessive? o What do I mean by violent movies and video games? What other factors might impact these results? o Medical or mental health factors of the shooters o Home environment factors, history of abuse o Life events (school bullying, work layoff)

14 SOME EXAMPLES OF VALID RESEARCH QUESTIONS Why is a Mediterranean diet effective in fighting cancer? What do I need to define? What other factors might impact these results? Photo credit: Jerzy Strzelecki

15 SOME EXAMPLES OF VALID RESEARCH QUESTIONS Are harmful algal blooms on Lake Erie caused by agricultural runoff (excess phosphorus)? What do I need to define? What other factors might impact these results? Lake Erie, July 2011

16 SOME EXAMPLES OF VALID RESEARCH QUESTIONS Can a specific intervention with 3 to 5-year-olds prevent later difficulties in learning to read? What do I need to define? What other factors might impact these results?

17 OTHER QUESTIONS TO ASK Will anyone care about the results of my study? What do I want people to learn? How can they use the results? What do I do with the results of my study? Publication where? Newspapers, TV, Internet Patent?

18 FINAL ANALYSIS: THERE ARE LOTS OF VALID RESEARCH QUESTIONS BUT MAKE YOURS COUNT!

19 Contact Dr. Janet Weisenberger Senior Associate Vice President for Research The Ohio State University Office of Research