Metodologi Penelitian Deendarlianto Jurusan Teknik Mesin dan Industri Fakultas Teknik, UGM
Penelitian (Research)? • Research is all about asking and answering a question or solving a problem……. • Identify an issue, question, or problem. • Find out what's already known about it. – Discuss with experts, read / conduct literature reviews on the topic. • Plan, cost, and do your study accordingly. • Write it up and submit it for assessment. – Better still, do a good job on it and submit it for publication. • Your work will benefit more people if you publish it. • Rule No. 1 in academia is publish or perish.
Dimensions of Research topic: computing, physical, biological, etc…. novelty: create new, review published data, info technology: develop new or use existing methods scope: study a single case or a sample mode: observe or intervene methodology: qualitative or quantitative ideology: objective or subjective politics: neutral vs partisan utility: pure vs applied reassembling the dimensions
Research topics…. Finding a good question/problem to address can be hard. It helps to have a good colleagues, and/or knowledge or practical experience of and affinity for a topic. We must read journal articles to find out what's already known. Scientists also often point out topics for future research.
All research projects are so-called original investigations. Obtain new data or information about a topic. Some research projects are reviews of the literature. Use other researchers' published data or info about a phenomenon. We should 'earn our spurs' doing original research before taking on a stand-alone review. But a write-up of an original investigation always has to include a short review of literature.
Example of literature review
Technology: develop new or use existing method(s)? • A legitimate topic for study is methodological. • For example, development or novel investigation of… – a measuring device – a protocol for a physical performance test – a method of analysis. • We should include or focus on a reliability and/or validity study of the measure provided by the method. – Validity = the relationship between observed and true values. – Reliability = reproducibility of observed values.
Pure or applied researches? Pure : to understand the cause or mechanism of a phenomenon. Applied: impact directly on health, wealth, or culture or on development of a method. Even so, try to include mechanisms in an applied project.It will help you publish in a high-impact journal, because their editors and reviewers can be snooty about pure research. Pure is sometimes lab-based, lacking naturalness. Applied is sometimes field-based, lacking control.
Additional remarks • A given research project can be characterized by topic, novelty, technology, scope, mode, methods, ideology, politics and utility. • This dimensional view may help you sort out a good approach to a specific project, but… – I may have missed or mangled some dimensions. – There may be better ways to understand research. • Your work needs to be credible to some people and preferably also published if it’s to have any impact.
Basic steps of a research project
Find a topic What, When Formulate questions What, Why Select design & measurement How Interpret evidence How & Why Publish it !!!
Types of research methodologies QuaLitative Measures –Descriptive –Numbers not the primary focus –Interpretive, ethnographic, naturalistic QuaNtitative Measures –N for numbers –Statistical –Quantifiable
Common Pitfalls in Researches
• Problems with generalizability –False conclusions –Transformations
Getting Started Finding a topic needn’t be traumatic Work projects Research studies • Library GO Bond Proposal Project • Library workshop trends • User repair strategies • Data collection involves agreement & consent • Forge partnerships • At some point you will need to leave the comfort zone of reading and literature gathering and…… (start it….)
LITERATURE REVIEW
-
A literature review can be just a simple summary of the sources, but it usually has an organizational pattern and combines both summary and synthesis. - It might give a new interpretation of old material or combine new with old interpretations. - The purpose of a review is to analyze critically a segment of a published body of knowledge through summary, classification, and comparison of prior research studies, reviews of literature, and theoretical articles. - The format of a review of literature may vary from discipline to discipline and from assignment to assignment.
Critically Analyzing the Literature Abstract the author’s description of the study related to your area of interest? conclusions relevant?
Statement of the Problem restate the topic problem stated clearly? purpose? significance?
Hypothesis or Research Question(s) Clearly stated? Relevant? Related to your area of interest?
Assumptions Explicit assumptions? Implicit assumptions Similar to yours?
Critically Analyzing the Literature Delimitations How was the study narrowed? Which considerations are relevant to your study?
Definitions Key concepts & terms?
Critically Analyzing the Literature Method research design? population & sample? measurement? procedures?
Findings make sense? what do they say about your area of interest?
Discussion Presented clearly? Meaningful interpretations? Implications discussed? Suggestions or recommendations? Limits to practical application?
Conclusion What did you learn? Re-state your new knowledge
Organizing the Literature Search: the Tree Diagram subtopic
subtopic subtopic
subtopic subtopic subtopic subtopic
Topic
Tree Diagram Example Parent involvement
Home-school communication
barriers
Homework assistance Academic achievement
volunteerism
Home visits
Involvement in decision making
Teachers’ beliefs
A ‘good’ literature review…. is a synthesis of available research is a critical evaluation has appropriate breadth and depth has clarity and conciseness uses rigorous and consistent methods
A ‘poor’ literature review is…. ….an annotated bibliography …. confined to description …. narrow and shallow …. confusing and longwinded …. constructed in an arbitrary way
Preparations in writing the literature review !!!
Clarify If your assignment is not very specific, seek clarification from your colleagues Roughly how many sources should you include? What types of sources (books, journal articles, websites)? Summarize, synthesize, or critique your sources by discussing a common theme or issue. Evaluate your sources. Provide subheadings and other background information, such as definitions and/or a history.
Find models Look for other literature reviews in your area of interest or in the discipline and read them to get a sense of the types of themes you might want to look for in your own research or ways to organize your final review. You can simply put the word 'review' in your search engine (for an example in Elsevier Science).
Narrow your topic The narrower your topic, the easier it will be to limit the number of sources you need to read in order to get a good survey of the material.
Consider your sources In the sciences, for instance, treatments for medical problems are constantly changing according to the latest studies. Information even two years old could be obsolete. You can also use this method to consider what is 'hot' and what is not.
Searching the Web
Research Question
Results
Experiment
How to review? The aim is to extract key points by comparing & contrasting ACROSS studies, instead of reading one paper after another. Key points for a review may concern areas of similarities and/or differences in: Research aim(s) or hypotheses Research design and sampling Instruments and procedures used How data were analysed Results or findings Interpretations
1st output Feature map
Classifies and categorises your thought in tabular form
Concept map
Links between concepts and processes, or shows relationship between ideas and practice
Tree construction
Shows how topic branches out into subthemes and related questions or represents stages in the development of a topic.
Writing the Literature Review Always begin with an introduction to the review & end with a summary Make the connection for the reader between the subtopics & the topic Use direct quotations infrequently
Always cite your sources Present your knowledge on the topics & subtopics Summarize each subtopic Include a transition paragraph from one subtopic to the next
Organizing the Literature Review Example paragraph 1
The purpose of this study is to determine the effect of poor coordination on the effective delivery of consultant services.
Introduction
•Reiterate the research problem •Introduce the subtopics
Example paragraph 2
•Effective organizational structures •Service coordination practices •Consultants’ professional support needs
Effective Organizational Structures In order for the coordination of services to be effective, a structure must exist within which service units are organized and can operate. Yaddah, yaddah, yaddah… Long range strategic planning characterizes organizations with effective internal structures (Brown, 1997). Key to such an effort is the identification of…
First Subtopic Discussion should include: • How is the subtopic connected to the problem? • Your knowledge, based on the literature you have studied • Specify subtopics (if any) • Summary & transition
•Mission & Purpose •Goals & objectives (planning) •Task units So, given all that, yaddah, yaddah, it is important to note the role of…
Writing tips……. Sentences
Express one idea in a sentence. Ensure that all your sentences have a subject, verb and object.
Paragraphs
Group sentences that express and develop one aspect of your topic. Use a new paragraph for another aspect or another topic.
Use sentences and paragraphs with appropriate use of commas, Consistent Grammar colours and semi-colours. Incorrect use of punctuation can affect the meaning.
Transition Words
Use words that link paragraphs and which show contrast and development to your argument e.g. ‘hence’, ‘therefore’, ‘but’, ‘thus’, ‘as a result’, ‘in contrast’.
Example……. (1) Video
Pre-processing
Classification
Genre Motion
Shot
Colour Feature Extraction
Edge Feature Extraction Event
Object Length
Pan Zoom Translation
Edge Detection Edge Clustering
Duration
Colour Histogram
Object
Edge Count
Camera
Rule-based classifiers
Recognition result
Schematic diagram for video classification Source: L.N. Abdullah et al. 2005.
Example……. (2)
See PWR Hot Leg Source: Deendarlianto et al., 2012.
Research Lin et al. 2007
Technique A priori algorithm
Features Used Audiovisual
Association rule mining
Domain Weather Sports Commercial
Pre-filtering architecture
Davis & Tyagi 2006
Probabilistic reliableinference framework Hidden Markov Model (HMM) output likelihoods and action priors Maximum likelihood (ML) and maximum a posteriori (MAP)
Motion
Walking, running, standing, bendingforward, crouchingdown, and sitting
Disadvantage / Future Direction Advantage Reduce the amount of Due to the different misclassification errors. properties of the data sets representing the Able to identify a high semantic concepts percentage of positive such as weather, instances in each concept commercial, and sports, they proposed to use different strategies to merge the rules. The system only makes classifications when it believes the input is ‘good enough’ for discrimination between the possible actions