Upload
ruth-robertson
View
216
Download
1
Tags:
Embed Size (px)
Citation preview
Library Searching
Steven A. Jones
Louisiana Tech University
Created for Louisiana Tech’s NSF-funded Research Experiences in Micro/Nano
Engineering Program
Exercise
Research Idea:
You are interested in targeting drugs so that they will release tissue-destroying agents only at the site of a certain type of cancerous tumor.
Write down 20 questions that Relate to your research project You really want to know the answer to
State of the Art
If I have made any progress at all, it is because I have stood on the shoulders of giants.
Sir Isaac Newton
It may not be obvious where to find a giant to stand on in a literal sense, but in a figurative sense, they live at the library.
Reasons for Literature Searches
You wish your work to be original.You can design research on questions that have been
raised by other people’s work. Don’t be afraid that someone has already answered your best
question. Borrow new (probably more relevant) questions from others.
Continual reference to the vast ocean of information available will stimulate new ideas.
Learn new techniques.Even for the most novel experiment, 99% of the experimental
methods you will need will have already been developed by other people.
The Student’s Greatest Fear
What if I find out that someone has already done my research – and … It’s the only thing I know how to do. It’s the only good question I can think of. I’ve been working on the problem for 4 years! Their experiments are better than mine. Their results are different from mine.
It’s the only thing I know how to do
It’s never the only thing you know how to do.
You need to use your imagination to come up with other ideas.
Looking through the literature will give you ideas for other (better) problems.
You can probably still do the same thing, only you will do it in a more focused manner.
Look for the questions that the author left unanswered.
It’s the only good question I can think of
You will think of other things as you go through the literature.
Your questions will be more focused as a result of knowing more about what has been done.
I’ve been working on the problem for 4 years!
The odds of you working for 4 years on the same problem that someone else is doing are miniscule.
Even if their overall question seems the same as yours, there will be differences that are important and that justify publication of your results.
Their experiments are better than mine
Maybe they are. The important thing is that the experiments are different.
If you do your experiments carefully and design them so that they are meaningful, you will make a contribution.
Merely having better equipment does not make for a better experiment.
Their results are different from mine
Maybe you are right and they are wrong.
More likely, you have asked a different question.
You need to determine the difference between what you did and what they did.
Hierarchy of Literature
Up To Date Reliable
Books
Books (collections)
Review Articles
Journal Articles
Conferences
Internet ??Internet ?
5-10
3-7
2-5
1-2
½ - 1 Year
HighlyClassic Books 10+
(Questionable)
Strategy for Library Search
(Very Early) Initial Web Search (not for citation).(Early) Read review articles on the topic.
Understand basic concepts. Note who is important in the field.
(Early) Read articles & proposals given to you by your mentor.
(Later) More emphasis on refereed research articles(Mature) Look for journal articles when:
New publications come out (monthly scan of journals) A new question or idea arises. A new search term comes to mind. A new result has been obtained in your experiments. A new method is needed.
(Steady State) More emphasis will be placed on recent journal articles and conference proceedings.
Exercises
Look at your list of 20 questions. Do you wish to change any of them at this point?
Give your 20 questions to someone else to read over.
What is a Journal Article?
Peer-reviewedYou know who wrote the articleYou know the name of the journalYou know when it was publishedYou can find the journal article through a
reputable libraryContains an in-depth description of the work
(intro, methods, results, discussion, conclusion), typically ~ 10 pages.
What is a Proceedings?
Describes “papers” presented at a specific meeting.
May be short or long BMES society meeting 250 words. ASME meetings, 2 pages.
Sometimes includes only abstractsMinimal or no peer reviewNo revision as a result of peer reviewAuthor generates the type-set paper – no editing
by a professional.
Finding Journal Articles
On line databases http://www.LaTech.edu/tech/library/ia.htm
Medline Pub Med Engineering Village
B. web: scholar.google.com; www.scirus.comC. Standard web search will sometimes give
journal articlesD. Reference pages of journals you already haveE. Word of mouth
Searching and Filtering
To search, type in appropriate key words as necessary. Filter through the titles and abstracts to identify ones that
are clearly relevant to your interests (or equivalently, to eliminate those that are clearly irrelevant).
Download the citation and abstract to each article in one of two ways: Have the system email this information to you (particularly useful
if you are using the library’s computer system). Have the system save the information to disk (not so useful if
you are using the library’s computer system, but highly useful if you are connected through your own computer).
Choice of Search Words
Be aware of your search engine’s syntax.a.Does mean nitric
and oxide, nitric or oxide, or “nitric oxide”
b.Will this search also find “nitric oxides?”
c.Will also find “ambiguous?”
d.Some search engines will use a special syntax, e.g., for “protein,” “prototype,” “protection,” etc.
SEARCH: nitric oxide
SEARCH: Doppler Ambiguity
SEARCH: prot$
Primary Search Problems
Overspecifying
vs.
Underspecifying
SEARCH: “platelet adhesion to fibrinogen”
SEARCH: shells
SEARCH: “platelet adhesion” and “fibrinogen”
A Common Complaint
I cannot find any information on the use of Braille with cell phones. When I type “Braille and Cell Phone,” nothing comes up.
1. This result could be a good sign.
2. You still need to know about
a. Braille
b. Cell Phones
c. Human Interfaces/Cognition
d. Ergonomics
Getting the Full Article
The library has a hardcopy subscription. Look up the call number, find the journal, and make a photocopy.
The library has an electronic subscription (as indicated by the catalogue). Go to the web site and download the article. If you have trouble, ask a library staff member. May need to be physically in the library.
The library has no hardcopy or electronic subscription. Request the article from interlibrary loan. Must go through a faculty member. Make sure that the article truly is unavailable!
Demonstrate:
Going to a database.
Searching
Emailing articles back
Checking the library for subscriptions
Going to an electronic database
Scholar.google.com Pdf readily available Pdf not available at default site, but available at
another.
Levels of Reading Articles
Relevance: Is the article relevant to my work?
Meaning: What does the article say?
Background Understanding: What can I learn about the topic in general?
Validity: Are the results and conclusions believable?
Application: How can I use the methods to design new experiments? How can I use the results to formulate new hypotheses?
Skimming for RelevanceExamine the Title for Keywords & Ideas
Read the Abstract
Potentially Useful?
Will the Article:
1. Stimulate new ideas?
2. Clarify state of the art?
3. Contain useful methods?
4. Clarify a question or phenomenon?
Download or Copy
Skimming for Meaning1. Read the Abstract
2. Read the last paragraph of the Introduction. What did the authors set out to do?
3. Read the conclusions.
4. Look over the results (graphs and tables).
5. Look at parts of the Results section related to graphs and tables.
6. Look through the discussion section for interesting points.
7. Look through the methods section to clarify general questions you may have on the experiments.
8. Read the complete article, if it may be relevant.
Reading to Enhance Background Understanding
Focus is on the Introduction & DiscussionProcedure is:
a. Scan the Abstract.b. Scan the Conclusions.c. Scan the Introduction.d. Skim the results.e. Scan the Discussion sections.f. Refer back to results and methods as needed
to understand the discussion. (Scan – a thorough reading of that section).
Evaluating for Validity
1. Skim the article for meaning.
2. Read the complete article.
3. Re-read the article from a critical viewpoint.
4. Read articles from the reference section.
• The features you should look for in each section are described in the following slides.
• One or two problems might not negate the validity of the work.
Evaluating Introduction for Validity1. Is the provided background information accurate to your
knowledge?
2. Are appropriate references provided for major concepts?
3. Do the authors’ statements agree with your understanding of those references with which you are familiar?
4. Have the authors cited all of the references that are required to make their point?
5. Is the background appropriate to the overall purpose of the work undertaken?
6. Is the purpose of the work clear and unambiguous?
7. Are you convinced that there was a need for this work?
8. Is the general approach of the authors clear?
Evaluating Methods for Validity1. Are standard methods used, where appropriate?2. Are the methods that were adopted from others properly
referenced?3. To your knowledge, do the authors describe adopted
methods correctly?4. Where appropriate, are calibration methods described?5. Are positive controls described, ensuring proper
implementation of the methods?6. Are negative controls provided for comparison?7. Are statistical analysis methods properly described?8. Are correct units used for measurements (e.g. molarity of a
reagent)?9. Are all methods plausible, to the extent that you can
evaluate them?
Evaluating Results for Validity1. Are needed calibrations adequately presented?2. Are all plots properly labeled, including axes, units of
measurement, legends for different curves?3. Do all results generally agree with what is physically
possible (e.g., mass & energy are conserved).4. Is the range of data presented within the boundaries of the
experimental protocol (e.g. length is not given in m when it was measured with a ruler).
5. Are conversions correct (e.g. if measurements are made with a force transducer, but presented in terms of pressure, what area was used?).
6. Are curve fitting methods and data models appropriate?7. Do p-values appear to be plausible, given the scatter in the
data and the number of measurements made?
Exercise
Exchange the 20 questions. Ask each student to categorize them as: Motivational (e.g. related to the Introduction) Technical (e.g. related to methods) Scientific (e.g. related to results) Synthesis (e.g. related to discussion) Other (e.g. who works in such-and-such
area?)
Strategies for Keeping Track of Journal Articles
Maintain a database of all articles.Do not make printouts of pdf files unless you have to!Give each article a unique and permanent reference number.Mark entries for each article that you have copied with the
date you copied it and the folder in which you filed it.Mark the physical article with the date on which you copied it
and dates on which you took notes on its contents.Maintain any notes you make from the article in either a
database or in your log book.Include the article’s unique and permanent reference number
in any notes that you generate.
When Copying Articles
Be sure to copy the entire article1. References2. Commentary (if provided)
Be sure that the journal reference appears on the copy
Will usually be on the first page If not, write it on the first page
Be careful that the edge of the text is not cut off.May want to copy from books from last page to first
(depending on the copying machine).
Database and Article
In Database On Article
Complete Reference Complete Reference
Identifying Number Identifying Number
Date copied/downloaded Date copied/downloaded
Notes – State sections in which information was found
Markup & notes
Folder in which the article is kept
Name of the database file
Identifying Gaps
Begin with a general idea or questionI would like to invent a method to target cancer-
destroying beads to tumors
B. Ask yourself what questions this idea raises? Is tumor vasculature different from normal
vasculature? Are there antibodies I can use? How do I seal in the toxin until the tumor is reached? How do I trigger the toxin to be released? Will the toxin pass through the vascular barrier? What happens to the dead tissue?
Example: Platelets/Hemodynamics
Q: Where might fluid mechanics be important?A: Blood flowB: Cerebral spinal fluidC: RespirationD: Micturation
Q: What applications (in blood flow) have been examined?1. Atherosclerosis: (diagnosis/pathogenesis)2. Artificial devices: (thrombus formation, cell damage)
Q: What causes thrombus formation? Platelets Coagulation factors High shear stress Low shear stress Injury
Connection?
Example: Platelets/Hemodynamics
Q: How do platelets relate to shear stress? Direct activation by shear stress Endothelial denudation Von Willebrand factor Recirculation Chemical transport Enhanced diffusion
Interesting
Example: Platelets/Hemodynamics
Q: How does von Willebrand factor work? Adsorption of protein onto collagen Opening up of the protein Possible activating mechanism
Q: How readily does vWf absorb?
Q: What stress is required to open the protein?
Q: How does activation occur?