Transcript New Scientific Knowledge - what is that ? - GETA

How to get a Ph.D.:
Experiences of a senior
researcher
Erkki Oja
Professor
Department of Information and Computer Science
Helsinki University of Technology
Email: [email protected]
Home page: www.cis.hut.fi/~oja
_____________________________________________________
GETA boat seminar, Helsinki, January 14, 2009
1
- My talk is about getting a doctoral degree. It is
based on personal experience rather than
textbooks (many very good texts exist; a very good
source is the course ”How to get a Ph.D.”, Oulu
University)
- What experience:
- Dr. Tech. in 1977, HUT
- Research work in 6 universities, 4 countries in
3 continents
- Advisor or supervisor for 35 doctors
- Official opponent, pre-examiner for many
2
So, instead of him ...
Listen to him in this lecture!
3
CONTENTS:
__________________________________
1. WHAT IT TAKES
(How do you succeed or fail?)
2. WHAT IT IS
(What is scientific research?)
3. WHERE IT COMES FROM
(What is creativity?)
4
1. WHAT IT TAKES
How do you succeed or fail?
5
Why get a Ph.D. (or D.Sc., TkT) at all?
- It is necessary if you are planning a career as researcher;
guarantee of the professional status (”driver’s license”); main
focus of this talk
- It helps you get profound (and proven) expertise in a (narrow)
field even if you are not a researcher.
6
Is it very hard to get a Ph.D.?
- It is very hard to make an exceptionally good Thesis, but not
very hard to make an average Thesis
- With enough motivation and willpower, most people who have
managed to get an M.Sc. can also make a Ph.D.
- Especially nowadays it is getting easier and easier due to the
graduate schools offering secure financing for many years and
good supervision (and, e.g., courses like this one).
7
What are the main requirements?
- Motivation and will. You must want to become a doctor. Role
models help a lot.
- Material resources: time and money. Good research groups
have money or can get it for you
- A thesis advisor who wants to take you as apprentice.
- A suitable problem: not too easy, not too hard.
- Certain personal skills and talents, especially:
-Ability to write fluent text in English
-Ability to make schedules and stick to them
-Ability to get at least one really good idea in your topic.
8
How can you fail to get a Ph.D.?
- Motivation declines. There are more important things in life
(industry job, raising a family, becoming a sheep farmer, ...)
- You take a part-time job in industry while ”finishing” the
Thesis
- Your self-criticism grows faster than your accomplishments
- It is so pleasant to be a grad student (or scary to be a Ph.D.)
that you do not want to change your life
- Money runs out in your lab. You have to go
- Your thesis advisor leaves, you do not want to follow, and
there is nobody to substitute for him/her
- Your problem turned out to be unsuitable.
9
Teamwork is the key to success
• It may be possible to make a Ph.D. without supervison, but
it is hard
• If nobody follows what you are doing, will the thesis be
accepted?
• Ideally, the research work is done in a group with a few
post-graduate students and one or two post-docs (or senior
researchers/professors) who work in the same lab
• Professors are very busy so there should be some other
doctor-level person who is the thesis advisor
• Ideally, the team has daily contacts.
10
11
Responsibilities
Graduate student:
• Makes a plan and schedule for the thesis and the post-grad
studies with the supervisor and then just works hard
• Passes the courses on time, nobody is following this!
• Participates actively in the team meetings (about weekly,
at least monthly)
• Gives talks in the lab seminars and international conferences
• Writes papers together with the other team members
• When the end is approaching, thinks carefully what to do next
(your life is your own responsibility, not your supervisor’s!)
12
Advisor (typically a doctor-level researcher):
• Does his/her own research work
• Writes good papers with the students and teaches them
how to write and submit to good journals/conferences
• Meets with his/her students about weekly, at least monthly
(a regular scheduled meeting time is best), gives time when
they have problems
• Goes with them to good international conferences,
introduces them to ”big names” if you know them.
13
Supervisor (professor):
• Follows what the teams are doing by going to meetings
every now and then, perhaps participating in some papers
• Is in charge of the doctoral thesis: when is the
quantity/quality of the papers enough to make up a thesis
• Handles conflicts, e.g. two students should not have all the
same papers in their theses
• Is in charge of the coursework and paperwork for graduate
studies
• Carefully reads and polishes the thesis.
14
2. WHAT IT IS
What is scientific research?
15
The three properties
of new scientific knowledge
- A good discussion has been given by Academician
Teuvo Kohonen (my own supervisor in the 70’s)
- New scientific knowledge has three essential properties:
1. Originality
2. Correctness
3. Impact
- Without all three of these, what you have found is not
new scientific knowledge.
16
1. Originality (novelty)
• Knowledge should be new
• Not just something that your supervisor did not know
but something that nobody in the world knew
• With many conferences, easily accessible papers,
and Web search engines (see other lectures of this
course), it is easier to check the originality today
than it used to be
• Once you submit your paper, competent reviewers
(referees) should be able to check this.
- In the Ph.D. Thesis, you must yourself clearly make a
difference between novel results and known results !
17
2. Correctness (truthfulness)
- Knowledge should be correct and true
- This is of course very hard! We can never prove that
anything is ”absolutely true”, but only relative to the
present state-of-the-art as defined by the international
scientific community
- Competent paper reviewers should be able to check this
but you cannot rely on them; the responsibility is with
the authors (consider some recent frauds).
18
3. Impact (influence, significance)
- Knowledge should be influential
- This is even harder because the impact only comes in the
future; so it has to be predicted
- This is where the help from senior researchers (supervisor)
is absolutely necessary
- Over time, a good measure of impact is number of citations
but they come too late for a Ph.D. Thesis
- Competent paper reviewers have a subjective opinion on
the impact but it may be totally different from yours. 19
… so remember:
The Empire Strikes Back !
There has never been a revolution without opposition.
20
To maximize the
impact, you must
work hard to
impress people
with your results
and papers in
conferences etc.,
and your Thesis
advisor (supervisor)
must help.
21
- Young researchers may have a too idealistic notion
of how to make the impact
- Research, too, is a human activity, as one clearly learns
when studying the history of science
-”The history of science should be X-rated”
- But fortunately, the race for impact usually starts
only after the Ph.D.
22
Assuming correctness, we could rank scientific results
according to the originality and impact as follows:
1. Reporting obvious facts (zero research)
2. Reproducing results given by others
3. Systematic coverage of a problem field
OK
4. New theoretical or experimental observations
OK
5. Creation of new concepts or systems
6. Starting a new research field or paradigm
23
Basic research vs. applied research
- Sometimes, research is divided in two parts:
1. Curiosity (basic) research:
understanding the world
2. Useful (applied) research:
mid-term economical use
In most engineering research, both aspects are found but
the second one is predominant
- Note that both are scientific research as opposed to other
kinds of research, not producing new scientific knowledge.
24
3. WHERE IT COMES FROM
What is creativity?
25
Where does the new scientific knowledge
come from?
- From a human brain – hopefully, yours
-You have to learn the research field thoroughly by reading
lots of books and papers – but not too many
- You have to discuss the problems within the research group,
in seminars and conferences – with good people only
- You can learn from the senior researchers what is good,
what is not good – if they are good scientists
- And then you just have to think hard !
26
What is creativity and is it
absolutely necessary?
- Remember that in the Ph.D. you are only practicing and
developing scientific creativity, you are not yet a
professional researcher
- Creativity can be learned
- It is not the same thing as doing well at school
- Needs continuous thinking of your problem (conscious
and subconcious).
27
Thomas Alva Edison:
”5 per cent inspiration,
95 per cent perspiration”
Louis Pasteur:
”Luck favours
a prepared mind”
28
THANK YOU FOR
YOUR ATTENTION !
29