Now that I am retired, I will be providing research support on
Mondays for UNM medical students.
SCHEDULING:
Please help me by scheduling an appropriate block of time with my assistant, Mary Armijo; phone: 272-5062; e-mail: mararmijo@salud.unm.edu. Typical estimated time requirements for each activity are listed below. Be sure to let her know your project's title and what type of help you'll be needing (see below). I prefer to be able to review a brief written description of your proposal before we meet. Always come to your scheduled sessions carefully prepared so we do not waste time. Try to bring a formatted, empty, floppy disk, CD, or memory stick. My office hours will typically be on Mondays in Basic Medical Sciences Building B65H. You can check for open time blocks by going to my public schedule at http://calendar.yahoo.com/tandberg51 .
LITERATURE SEARCHES:
Estimated time: one to two hours. Bring copies of previous similar or related research papers that you have found. A hand search done before (index medicus, textbooks, resident conference files, your own files, etc.) will improve the quality and efficiency of any machine search immeasurably. Be sure to try and identify the relevant major subject headings.
PLANNING AND DESIGN OF EXPERIMENTS AND STUDIES:
Estimated time: two hours. Bring copies of all previous similar or related research papers that you have access to. Before we meet, consider carefully the issues of study design, subject selection, exclusions, blinding, reduction of bias, and controls. Consider funding and human subjects issues. Write down your ideas.
SAMPLE SIZE ESTIMATION:
Estimated time: one to two hours. This will require a clear (usually written) idea of your study design and copies of all previous similar papers. This is so that we can intelligently estimate how much variability to expect in your own, yet to be collected, data. You will also need to have a clear idea about how big of a difference in your main outcome variable you wish to be able to detect.
PREPARATION FOR DATA COLLECTION:
Estimated time: two hours. The way in which you collect and record your data will effect the ease with which it can later be error checked and analyzed. This is one of the most important things that I can help you with.
PREPARATION OF DATA FOR ANALYSIS:
Estimated time: two to twelve hours. All collected data should be on magnetic media. Tab delimited ASCII, database, or spreadsheet format is the easiest to translate to statistical package formats. In general, data should be entered and then hand checked for errors. In most cases you should leave any missing values blank (not zero or N/A).
STATISTICAL ANALYSIS OF DATA:
Estimated time: two or more hours. At this point, the data should be completely collected, absolutely error-free, and perfect! This will require us sitting together in front of the computer and working interactively. All written material and raw data should be easily available. Complex or exotic analyses will often require several such sessions or outside consultation.
DESIGN OF TABLES AND FIGURES:
Estimated time: two to six hours. Beautiful and compelling graphic displays or tables of your data require considerable time and thought before they are actually prepared. You should always look through ER Tufte's three books on this subject before we meet. Chambers, Cleveland, Kleiner, & Tukey is another excellent place to get ideas. Be sure to bring previously published tables or figures from similar studies that you think are especially good (or bad).
PREPARATION OF MANUSCRIPTS:
Estimated time: two or more hours. I would like to assist you in this. I can be especially useful in helping you approach writing your project up, efficient handling of references, and selection of appropriate journals to submit your work to. Also, come talk to me if you have finished collecting your data, but seem to have "writers' block". Most helpful is a previous paper, similar to how you envision your own looking when it’s done and published in the journal that you have selected.
REVIEW OF MANUSCRIPTS:
Estimated time: two to four hours. In general, I can only read a paper carefully once. (By the second or third reading, I am just too familiar with it to be effectively critical.) The best time to have me (or other interested persons) read a paper is when you believe it's just about ready to submit. All papers should be free of spelling and grammatical errors before asking anyone else to read them. What you want is evaluation and criticism of your ideas and their presentation, not correction of your run-on sentences and overuse of the passive voice.
HELP WITH REVISIONS OF ACCEPTED (OR REJECTED) MANUSCRIPTS:
Estimated time: two to twelve hours. In general, this is the most difficult problem I have to help people with. It will usually take two or more sessions. It is very helpful to have a chance to review the manuscript and the referees' comments a few days prior to our meeting. Always bring all study material including disks, raw data sheets, drafts, and copies of references.