Transcript Scientific Papers

Scientific Papers
Announcements

Office hour will be changed from next week
 Monday




and Wednesday 11:30AM – 1:30PM
Lab on Friday – Individual Assignment
Write lab report using the data we collected from
our own personal computers and the analyses
you did last week.
Lab Reports are similar to what scientists
publish in research journals
Follow format of scientific papers
Key Parts
Title
 Abstract
 Introduction
 Methods
 Results
 Discussion/Conclusion
 Tables and Figures
 Literature Cited

Title
Specific and brief description of contents
 Or raise a major question addressed in the
paper.
 What makes a good title.

 Precision
 Catchy,
in a scientific way. well…if possible.
Example of Titles




Natural Image Statistics and Neural
Representation.
What does the retina know about natural scene?
Emergence of Topography and Complex Cell
Properties from Natural Images using Extension
of Independent component analysis and
Bayesian Probability Models
Sex with Support Vector Machine. (NIPS 1999)
Abstract







Overviews of facts, results, conclusions.
Function – between the Title and the article itself.
Very important – read more widely than article.
Placed at beginning of article, just below the Title.
“I” never used.
Passive voice commonly used (active voice in the
articles).
Should be self-contained. Don’t include footnotes,
references (e.g. Lennon and Ono (1998) concluded
that… )
Introduction

The purpose of Introduction:
 Defines
problem, scope, and
purpose/objective of the study.
 Provides theoretical or historical background.
 Proposes Hypothesis and predictions
 Reveals plan of development of article.
Literature review


Part of the introduction.
Nothing is totally new, the literature review
serves
 review
scientific literature.
 explain choice of materials or methodology.
 show rationale for investigation.


Paraphrase ---Don’t steal other peoples’ word.
Quotation --- Keep it minimum.
Hypothesis


Plural is hypotheses
A tentative explanation that accounts for a set of
facts and can be tested by further investigation;
 Dell
computer is better than HP’s. (not specifically
defined)
 people who commit acts of violence are under orders
from the Devil (untestable)


Tested and supported, not proved!
Come up with hypotheses for study of computer
hardware and performance
Hypotheses of Computer
Performance
The startup speed of a computer is
determined by processor speed.
 The startup speed of a computer is
determined by the amount of RAM.
 The stability of a computer is affected by
its operating system.

Null Hypothesis

The null hypothesis is often the reverse of what
the experimenter actually believes; it is put
forward to allow the data to contradict it.

H0: "There will be no difference in readability
between 15" and 17" monitors."

H0: "There will be no relationship between
processor speed and startup time, indicated by a
line with a slope of 0.“
Don’t put Null Hypothesis in the paper.

Hypothesis Testing

Hypothesis testing is using statistical
values to determine whether the Null
Hypothesis can be rejected.
Common Statistical Tests for Hypotheses:
t-Test, ANOVA, Regression, X2 (Chi –
Square)
 Beyond scope of this class

Methods
Explain what steps were taken in
collecting data and why
 Use past tense
 Usually described using passive voice

 “The

image was digitalized…”
Using active voice occasionally is fine.
 “we
conducted out experiment… ”
Results




Present general trends without comment, bias or
interpretation
Present all relevant results, even those that do
not support the hypotheses
If statistics are used, report statistical value and
probability in parentheses
Refer to tables and figures
 “Startup
speed decreased as processors speed
increased (Figure 1).
 “Most computer used a version of Windows as shown
in Table 1
Discussion/Conclusion
Discuss the results and whether they
support the hypotheses
 Discuss relevance to work by others
 Avoid redundancy with results
 End with a summary of the significance of
your work – a conclusion paragraph

Tables and Figures





Each table and figure has a name, e.g. Table 1,
Table 2, … and Figure 1, Figure 2, …
Tables are data tables
Figures are graph, maps, photos, drawings, etc.
Can be pasted into MS Word document from
Excel
Each should have a Caption, e.g. a description
of what the table or figure represents
Captions
Captions are descriptions of the table or
figure, include details necessary to
understand the item
 Tables: Caption above
 Figures: Caption below

Example
Figure 1. Amount of RAM in MB vs. average startup time for 45
computers with Windows XP operating system. Trendline shows
a negative relationship.
Literature Cited



Called Bibliography or Reference section.
Provide details of any citations in paper
Format varies from field to field and journal to
journal
…
observed by Hubel and Wiesel (1962)
 …got the similar result (Smith and Johnson, 1990)
 …many research work [1][2][3]…

Only list resources cited in paper, not general
references
A few words
Don’t consider these conventions as
unbreakable.
 The function of a scientific paper is
communication.
 Follow format in Guidelines for Writing
Scientific Papers used in LBS 158H (will
posted on website later.)

Title Page
Title of project
 Your name
 Course number
 Date
 Honesty statement with signature

Putting it all together
Title Page
 Abstract Page
 Body: Introduction – Methods – Results –
Discussion

 Tables
and figures can be embedded or at
end of body

Literature Cited starts on new page
Friday
Work on formal scientific paper on
computer hardware and performance
 I will provide details on methods, other
analyses expected (Posted Friday
Morning)
 Due next Friday (September 26)
