Transcript Slide 1

Week 1. Scientific method and
research methodology
PhD Seminar
Authors:
Niusvel Acosta
Hussein López
Advisors:
Dr. Felipe Orihuela-Espina
Dr. Manuel Montes y Gómez
Contents
1. Scientific method
i.
ii.
iii.
iv.
v.
Introduction
A brief history
Steps
Characteristics
Hypothesis
2. Research methodology
i.
ii.
iii.
iv.
v.
vi.
21/07/2015
What is?
Different approaches
Thesis
Types of scientific research
Research process
Scientific paper
INAOE
2
Part I
THE SCIENTIFIC METHOD
(CLASSIC)
21/07/2015
INAOE
3
Introduction (1)
 In general, science aims at providing explanations
of phenomena observed in nature and society.
 This explanation is offered in terms of relations
between an observed phenomena (effect) as a
consequence of its possible origin (cause).
 These relations are established by means of
experimentation following the scientific method.
 i.e. experiments are central to science
21/07/2015
INAOE
4
Introduction (2)
 Purpose of Scientific Method:
To
build
a
representation
of
the
world/nature/society that is accurate, reliable,
consistent and not arbitrary
 Science is a rational activity operating according
to some special method or methods.
 Alan Chalmers, in his book What is this Thing Called Science (1990)
21/07/2015
INAOE
5
Introduction (3)
 There exist an objective reality which is the
same for everyone [Cotton and Sekula].
 Reality exists as an absolute goal: facts are facts
regardless of feelings, desires, hopes or fears of
people.
 Ayn Rand, “mother” of objectivism
 “If your own private reality includes a law of
gravity that is different from Newton's, any
predictions you make with it are not going to
match reality.”
 [http://www.physics.smu.edu/pseudo/SciMeth/]
21/07/2015
INAOE
6
Introduction (4)
 There exist [unchanging] laws by which the
universe works [Cotton and Sekula].
 These laws can be discovered (not invented)
through experimentation
 These
laws
may
however
not
deterministic; they might be stochastic.
be
 Unchanging does not mean static! This
becomes especially clear in social research
as society evolves with time.
21/07/2015
INAOE
7
Introduction (5)
Fact:
 A fact is an observation that has been confirmed
repeatedly and that for all practical purposes it is
accepted as true.
 Definition by the National Academy of Sciences
 "Everyone is entitled to his own opinion, but not his
own facts.”
 Daniel Patrick Moynihan (1927-2003), Senador de los EEUU
 "Facts do not cease to exist because they are
ignored.“
 Aldous Huxley
21/07/2015
INAOE
8
A brief history
The Greek naturalistic
movement
Inductive reasoning
< 1600 DC >
The nature
of science
0
< 322 BC
Deductive reasoning
21/07/2015
The scientific
revolution
INAOE
> 1900 DC
Deductive falsification
9
A brief (not exhaustive) history
of the scientific method
Aristoteles
384-322 BC: Aristotelic Empiricism
Roger Bacon
1214-1294. Roger Bacon describes the
cycle of observation, hypothesis,
experimentation and verification
~700-1000. Arabs Alhacén, Al-Biruni and Avicena
among others, develop different forms of
experimentation and quantification to
discriminate among competing theories
Ibn al-Haytham , Abū Rayhān al-Bīrūnī, Ibn Sina Avicenna
21/07/2015
INAOE
1561–1626. Francis Bacon
incorporates induction as a
rational method to reach causality
Francis Bacon
10
A brief (not exhaustive) history
of the scientific method
David Hume, John Stuart Mill, Charles Peirce
Galileo Galilei
1564-1642. Galileo uses
mathematical demonstration as a
form to obtain valid scientific results
21/07/2015
S XIX. Several contributions from Hume (inductive
reasoning), Mill (knowledge based on experience), Peirce
(scheme for hypothesis testing and randomization)
1791-1867. Faraday demands intellectual
honesty and criticism from peer (peer review)
together with scrupulous documentation of
experiments so that they can be reproduced.
S XX, Popper (necessity for falsability),
First computer simulation, Kuhn
(paradigm and Incommensurability)
Michael Faraday
Karl Popper, Thomas Kuhn
INAOE
11
Steps (1)
The scientific method in a nutshell:
1. Observation of a phenomenon
2. Formulation of a hypothesis or plausible
explanation of the phenomenon that might
explain the observations
3. To carry out an experiment altering the
conditions
and
measuring/observing
the
phenomenon under the changing environment
4. Confirmation (or refutation) of the hypothesis
based on evidence (observations) collected
during the experimentation
21/07/2015
INAOE
12
Steps (2)
Figure from: [www.studyblue.com]
21/07/2015
INAOE
13
Characteristics (1)
 The scientific method is an effort:
 Collective of all scientists
 Individual research are unavoidably influence by
personal and cultural bias.
 Standardized to minimize bias
 Consensus among the scientific community is a
central
demand
[SwanbornPG1996]
21/07/2015
in
INAOE
empirical
research.
14
Characteristics (2)
 The scientific method is an effort:
 Dynamic (time)
 It is subject to continuous revision.
 The hypothesis that has been hold across many
observations might still be refuted at any time by new
evidence (facts).
 "When the facts change, I change my mind. What do you
do, sir?"
 Attributed to John Maynard Keynes
 Science is self-corrective.
21/07/2015
INAOE
15
Characteristics (3)
 The scientific method is an effort:
 Critic and creative
 Critical thinking is capable of deciding whether an
statement is always true, partially true or false.
 Creative thinking permits developing unique and useful
ideas worth of consideration.
 http://www.umich.edu/~elements/probsolv/strategy/crit-ncreat.htm
 Without these two thinkings we would be unable to
formulate a hypothesis nor to confirm/refute them.
21/07/2015
INAOE
16
Characteristics (4)
 The scientific method is an effort:
 Skeptical
 It requires experimental evidence before doing or
accepting an statement
 …but it admits assumptions (lacking evidence) based
on previous knowledge about the phenomenon
 Skeptical does not mean denier (available evidence is
never enough)
 Trust is of utmost importance among scientist.
 In principle, you trust that others researchers are being

21/07/2015
honest
…yet you have to test/review/challenge their evidence.
INAOE
17
Hypothesis (1)
 An hypothesis is a limited statement
regarding cause and effect in specific
situations.
 Criteria of hypothesis:
 They should be statements expressing the
relation between two or more measurable
variables.
 They should carry clear implications for
testing the stated relations.
21/07/2015
INAOE
18
Hypothesis (2)
 Problems of omitting hypothesis:
 Ambiguity is major cause of referee/reader
misunderstanding.
 Vagueness
is
major
cause
of
poor
methodology:
 inconclusive evidence;
 Unfocussed research direction.
21/07/2015
INAOE
19
Hypothesis (3)
Exploration of Techniques:
 Invention of new technique,
 Investigation of technique,
 e.g. discovery of properties of, or relationships
between, techniques.
 Extension or improvement of old technique,
 New application of a technique,
 to artificial or natural systems.
 Combine several techniques into a system.
21/07/2015
INAOE
20
Hypothesis (3)
¡YOU CAN’T DEMONSTRATE THAT
A HYPOTHESIS IS
CORRECT/TRUE!
 To have positive evidence supporting a hypothesis is not
equivalent to demonstrate a hypothesis, let alone to confirm a
fact.
 …with positive evidence you only increase your confidence in
the hypothesis.
 The more experiments are made that result in evidence
supporting your hypothesis, the bigger certainty you have on
your hypothesis.
21/07/2015
INAOE
21
Hypothesis (4)
Hypothesis
Model
Scientific
thoery
Law
Fact
21/07/2015
INAOE
22
… considerations
 “In science, you encounter the disturbing
fact that, if your ‘point of view’ does not
agree with reality as determined by
experiment through the scientific method,
then your point of view is simply wrong.”
 [Cotton y Sekula, http://www.physics.smu.edu/pseudo/SciMeth/]
21/07/2015
INAOE
23
… considerations
 Having said that:
 It is acceptable that our point of view changes as
sciences gathers more reliable evidence.
 …but that change must be guided by evidence.
 If a established theory (one that has passed
many experiments) is in disagreement with new
evidence, the theory has to be reviewed maybe
even discarded as a description of reality
 It may still be valid within certain restrictions
 Classical example: Newton’s gravity laws.
21/07/2015
INAOE
24
Part II
RESEARCH METHODOLOGY
21/07/2015
INAOE
25
What is known as research methodology?
 Ordered procedure which is followed to
establish the meaning of the facts and
phenomenon to which the scientific interest is
directed for finding, demonstrating, refuting
and providing a knowledge.
 It is a set of techniques and procedures
whose primary purpose is to implement
processes of picking, classifying and
validating the data and experiences from
reality, and from which can be build scientific
knowledge.
21/07/2015
INAOE
26
Approaches of research methodology
 Qualitative, quantitative and hybrid:
 are based on the phenomenon observations
and evaluations
 conclusions are achieved as results of this
basis
 demonstrate the reality degree of the achieved
conclusions
 check the conclusions reached and are able to
generate new fundaments, based on the
founded trends
21/07/2015
INAOE
27
Approaches of research methodology
 Qualitative:
 does not takes the hypothesis proof and
numerical measurements neither.
 Is based on surveys, interviews, descriptions,
researcher's view points, reconstructions of
facts.
 Quantitative
 Hybrid
21/07/2015
INAOE
28
Approaches of research methodology
 Qualitative
 Quantitative:
 Is based on the numerical measurements
 Uses the observation to picking data and analyzes
them to answer their research questions.
 is rather used in processes which can be
measurable or quantifiable.
 Hybrid
21/07/2015
INAOE
29
Approaches of research methodology
 Qualitative
 Quantitative
 Hybrid:
 both concepts are integrated
 both processes are combined to reach a results
in a superior form.
21/07/2015
INAOE
30
What is a thesis?
 A thesis is:
 Original (creative) and significant
 A record of research (so that it is
reproducible)
 A critical survey
 Reviews existing work
 Discusses results in terms of current evidence
 The conclusion of a set of coherent
experiments for testing certain hypothesis
about a certain phenomenon of interest.
21/07/2015
INAOE
31
What is a thesis?
 A good thesis includes:
 Who? – A phenomenon being studied
 What? – A clear message, hypothesis and/or
claim
 …and evidence for this
 Why? – A motivation and justification
 How? - Methodology
 When (time) and where (spatial)?
Constraints, limitations and discussion
21/07/2015
INAOE
–
32
Common pitfalls while doing a thesis
 Being too ambitious
 Stopping reading after an initial survey of
related work
 Think that there is no related work
 You simply have not search thorough enough
 Underestimate or overestimate the relevance
or significance of the research
21/07/2015
INAOE
33
Some tips
 Keep (organized) record of everything;
 Notes of your ideas
 Notes on read papers
 Set up and results of your experiments, but also

distractions, artifacts, etc.
Documentation of programs and code
 Focus on principles and foundations, not on
implementation and performance
 Never, ever delete a file/data
 Keep different versions
 This helps to remember, but also allows you to go back if
something goes wrong
21/07/2015
INAOE
34
Some tips
 Read and write!
 For a good understanding, reading a paper once is
often not enough.
 It is fine to fail
 It is research, so you may guess what’s going to
happen, but cannot be sure.
 Admit ignorance
 …make questions, even if you think: (i) the expert is

21/07/2015
going to laugh at you, or (ii) what you are asking is
obvious for everyone else.
“Quien pregunta es tonto 5 mins., quien no
pregunta es tonto toda la vida” Spanish saying
INAOE
35
Methods and types of scientific research
 Methods and techniques:




Exploratory
Descriptive
Correlational
Explanatory
 Types:
 Academic research (basic)
 Translational research
 Applied research
21/07/2015
INAOE
36
Research process
Problem
Development of
theoretical
framework
Hypotheses
formulation
Experimental
design and data
collection
Sample
selection
Research design
selection
In case of be necessary
Data analysis
and conclusions
21/07/2015
INAOE
37
Scientific paper
Concluding
with
persuasive
conclusions
and citations
Interesting
discussion
with clear
figures
Appropriate
abstract and
keywords
Engaging
paper
Clear and
reproducible
presentation
of the
proposal
21/07/2015
Starting with
a brief and
attractive
tittle
Presentation
of the related
work
INAOE
Introduction
suitable for
target
audience
Lucid
motivation
and research
description
38
References
 Bundy, A. et al., The Researchers Bible. Teaching Paper No. 4, Department of
Artificial Intelligence, University of Edinburgh, 1986
 Keppel, G. Design and analysis: A researcher's handbook. Prentice-Hall, Inc,
1991
 Bundy, A. The Need for Hypotheses in Informatics, University of Edinburgh
 http://www.inf.ed.ac.uk/teaching/courses/irm/notes/hypotheses.html
 The scientific method:
 Cotton J and Sekula S. The Scientific Method http://www.physics.smu.edu/pseudo/SciMeth/
 http://teacher.nsrl.rochester.edu/phy_labs/AppendixE/AppendixE.html
 Williams, J. A brief history of the scientific method, School Science Review
(2007)
 Ziemski, S. The typology of scientific research, Journal for General
Philosophy of Science 6 (2):276-291 (1975)
 Ziemski, S. Two Types of Scientific Research, Journal for General Philosophy
of Science 10 (2):338-342 (1979)
 Chapman, D. (Editor). How to do Research at the MIT AI Lab, Massachusetts
Institute of Technology Artificial Intelligence Laboratory (1988)
21/07/2015
INAOE
39
THANKS, QUESTIONS?
21/07/2015
INAOE
40