180 likes | 202 Views
Learn how to read and assess research materials effectively. Discover different publication types like books, journals, conferences, and manuscripts. Understand the characteristics and importance of each type in the research world.
E N D
COSC2148/2149/2150 Research MethodsReading and Assessing Research Literature James Harland james.harland@rmit.edu.au
Generating PDF from Word Despite the lab exercise, I believe the easiest way is: • Download and install Open Office • Open .doc file with Open Office • Save as PDF • Close .doc file
Overview • Reading is a fundamental research activity • Assessing what is read can be difficult • Assessment is required of active researchers • 1 paper published = 3 papers reviewed, on average • Need to know where and when published • May need to investigate publication details • Need to know publishing culture • Be rational, skeptical and humble
Publication Types • Book • Journal • Conference • Workshop • Technical Report • Web page • Thesis • Manuscript • ...
Book Published because someone believes many copies will be sold. • Author/s detailed development of ideas • Gathering of scattered research results into one volume • Collection of papers by various authors from a meeting • `Handbook on X' series by invited experts • Commemorative volume dedicated to an individual • PhD thesis published as a book • Not necessarily just for the academic market
Book • Often become standard references. • Publication can be slow. • Reviews are often written by other experts. • Good sources of basic concepts and background material. • Rarely have up-to-the-minute results. • Refereeing process varies from rigourous to virtually none.
Journal Published by major companies exclusively for academia. • Viewed generally as the most important kind of publication • Often seen as an archive of research • Published at regular intervals (monthly, quarterly) • Available in most (good) libraries • No bounds on number of papers • No real-time refereeing constraints • Increasingly publishing special issues on topic X • Main customers are university libraries
Journal • Rigourous refereeing by (at least) two experts • Only accepted for publication when editor is satisfied • Can take many iterations to succeed • Generally no bounds on pages or time to (re-)submit • Anything that is appropriate will eventually be published • Can sometimes have good but unimportant papers • Any good paper should eventually appear in a journal
Conference Proceedings published for distribution to attendees, and sometimes more widely. • Probably the most common form of publication • Not always refereed (but generally are in CS) • Proceedings sometimes published well after conference • Tend to have up to date material • Quality more variable than journals • Worldwide, regional, national, local, ... • Main papers, poster abstracts, short papers, system demonstrations, invited papers, panel sessions, ...
Conference • Generally 3-5 referees, not all from the programme committee • Often use numeric scores to simplify the process • Real-time refereeing constraints exist • ``One-shot'' chance at publication • Programme committee selects best-scoring papers • No chance to re-evaluate papers later • Reports generally not as detailed as for journals • Common for acceptance rate to be under 50%
Workshop • More informal than conferences • Deadlines closer to actual meeting • More specialised audiences • Refereeing generally more lightweight • Often self-selecting • Usually has ``cutting edge'' material • Quality again more variable • Proceedings often not generally available
Technical Report • Earliest of all publications • Not refereed in any sense • Used to disseminate information quickly • Often same paper submitted to a conference or journal • Trend away from these per se to links on personal Web pages A typical evolution is TR to Workshop to Conference to Journal. Any of these steps may be missing.
Manuscript • Sometimes you may get a personal copy of some material not widely available. • This is generally from a trusted source only. • Examples including incomplete material from colleagues or working notes from a supervisor. • Generally does not include justification, background or discussion.
Theses • Paper (or book) submitted by a student towards a degree • PhD theses can be hundreds of pages • MSc theses around one hundred pages • Honours/minor thesis around fifty pages • Read in detail by at least two examiners • Designed to show that a student is ``qualified'' • Often read and criticised in greater depth than other publications • Possibly more arcane material than journal or conference publications • Don't have to be important, just good
Reading Papers • Read the abstract first • If still interested, read the introduction and conclusion • If still interested, read the rest of the paper • If still interested, look for follow-up papers • It is not necessary to understand everything in detail to get some benefit
Evaluation • What? • When? • Where? • How did I hear about it? • What's new? • What is/is not obvious? • What is/is not justified? • What confidence do I have in my judgement?
Evaluation Approaches you could use: • Discuss the paper with colleagues • Ask your supervisor about it • Work through techniques from paper • Reproduce results from the paper • Ask the author/s about the paper • Look up references mentioned in the paper