Loading presentation...

Present Remotely

Send the link below via email or IM

Copy

Present to your audience

Start remote presentation

  • Invited audience members will follow you as you navigate and present
  • People invited to a presentation do not need a Prezi account
  • This link expires 10 minutes after you close the presentation
  • A maximum of 30 users can follow your presentation
  • Learn more about this feature in our knowledge base article

Do you really want to delete this prezi?

Neither you, nor the coeditors you shared it with will be able to recover it again.

DeleteCancel

Make your likes visible on Facebook?

Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.

No, thanks

S&T 03 The Modern Scientific Method

From Peter Bock's Getting It Right: R&D Methods for Science and Engineering.
by

Ale Ibarra

on 23 January 2014

Comments (0)

Please log in to add your comment.

Report abuse

Transcript of S&T 03 The Modern Scientific Method

The Modern Scientific Method
It focuses on the formulation of hypotheses that
can be rejected
.
This method argues that
no hypothesis can ever be completely proved
, but it can be disproved or rejected (Popper and Medawar).
Often it can be adapted and modified so that it gradually converges more and more closely to the truth.
Analysis
Validation
Synthesis
Hypothesis
“A scientist is a searcher after truth, but complete certainty is beyond his reach”
Peter Brian Medawar
Describe Problem
Investigate Related Work
State Objective
Set Performance Criteria
Implement Solution
Conduct Experiments
Reduce Results
Design Experiments
Draw Conclusions
Solicit Peer Review
Prepare Documentation
Compute Performance
Block Design
Data Management
Experiment Laboratory
Primary Project Documentation
Secondary Project Documentation
The scientific method comprises four sequential phases –
Analysis
,
Hypothesis
,
Synthesis
, and
Validation
– which are applied to a task
iteratively
and recursively to achieve the objective of the task.
Iterative execution
Even the simplest tasks often
require several attempts
to achieve the objective.
Reach over to pick up a glass of lemonade
Surface covered with a thin film of condensation (slippery)
You change your tactics for holding it
Once you grasp it, you try to lift it and it is stuck to the table, you change your tactics again
As you lift it, you observe that it is very full and reduce the speed of movement to your mouth…
Each iteration incurs an
additional cost
in both time and money
Thus, each iteration following a failure to achieve the task objective should be preceded by a
decision
whether
to continue
to pursue the task objective
or to abandon
the entire task as a lost cause.

Although abandoning a task is usually an unattractive and discouraging conclusion, task failure is a very common occurrence.
The vast
majority of research
and development tasks are
unsuccessful.
It is important, therefore, to find out as
quickly
as possible whether a task is likely to be successful or not.
The objective of the Analysis Phase of the Scientific Method is to gain a thorough
understanding of the components of the problem domain
, leading to the formulation of a single specific and reasonable task
objective.
A
problem statement
is expressed as an interrogative sentence, a declarative sentence, or an imperative sentence that summarizes a
question, complaint, or requirement
.
The name of this step in the Analysis Phase is
Describe Problem, not State Problem
and it is essential that the summary problem statement be accompanied by a
full description of the problem
including sufficient
background information
.
Customers complain that the battery in the portable drill (Part #156A-90) loses its charge very quickly when it is not being used.
My dog has fleas
Design a more effective and reliable windshield wiper
Land a team of astronauts on the Moon and bring them safely back
How well can humans recognize the characters on license plates when the surfaces of the plates are severely degraded by dirt or debris?
Why are red and green used for traffic lights to indicate “stop” and “go”, respectively?
Bock Peter; Getting it right R & D Methods for Science and Engineering, Chapter 9. Academic Press; 2001.
Performance criteria are
requirements
that any proposed solution to the problem
must fulfill
. The requirements are normally characterized by
restrictions, limitations, or exclusions
.
Performance criteria are very common in the course of our everyday lives:
You may watch television for another half-hour, but only if you have already completed all your homework for tomorrow.
Stop on your way home and pick up a quart of milk, but be home no later than 7 pm.
Please buy us two tickets for the concert on Saturday evening, but don’t spend more than $45 each.
When setting performance criteria, one must take care that the criteria not be
brittle
and/or not too
excessively precise
.
Brittle performance criteria are often highly counterproductive or unfair.
Performance criteria should be “soft” rules or more
flexible,
allowing judgment calls based on the final success or failure of the overall project objective.
Ex. of Brittle Performance Criteria: If the definition of the maximum income level for receiving food stamps is $18,000 for a family of four, then the family earning $18,000 eats adequately, while the family earning $18,001 goes hungry.
Ex. of Flexible Performance Criteria: If your family is hungry, you are entitled to receive food stamps.
What are some examples of performance criteria that can be set to your research or project task?
To find out as much as possible about
what has been
done to address the problem and achieve similar objectives in the past.
A review of the literature may reveal that the task objective has already been achieved, and the results of that past project may simply be acquired and applied to the current project.
Professional journals
Conference proceedings
Books
Professional studies and investigations
Internet reports and databases
Newspaper and magazine reports
Theft
Interrogation
scholar.google.com
http://www.springerlink.com/content/v642443g7254r64r/
http://www.sfi.mtu.edu/publications.php
http://www.biblioteca.uson.mx/digital/tesis/dig_TesisDigitalWeb.asp
http://www.me.mtu.edu/~jwsuther/Publications/28_tcm_article_sutherland.pdf
Once the
problem
has been clearly described, the
performance criteria
have been clearly and completely defined, and the
related work
has been thoroughly investigated, then presumably the task domain has been appropriately and sufficiently focused down and constrained.
At this point, a
specific, clear, and detailed
task objective may be formulated.
The project objective is
to create a sociable robot that interacts with a human based on the biometric feedback from the human being.
The heart plays a vital role in affecting emotional, cognitive, psychological, and physiological changes within the human body and as such, we will focus on the heart rate variability as the main input for the robot.
The robot can respond and react to the user's emotional state through speech, movement, and facial expressions.
Through these interactions, we believe that using a sociable robot in this particular interaction will help people better manage their physical and emotional states than with a specific monitoring device.
The task objective is a statement of what the task is intended to achieve, expressed as an infinitive phrase.
The task objective is the
baseline
by which the
success or failure
of the task will be evaluated.
The accuracy of the task will be
measured
by how well a task achieves its stated objective.
We should formulate an objective that expresses a
single, highly specific, and realistic expectation,
tightly constrained by the specific governing propositions that have been incrementally imposed in the preceding steps of the Analysis Phase.
From the Greek word “analusis” meaning
“dissolving”.
This etymological root imply that the fundamental process of analysis is
to disassemble something into its parts
.
From the Greek hupotithenai meaning
“to suppose or to speculate”
.
Thus, the fundamental process of hypothesis is
to “speculate” about an underlying cause-effect relationship.
The objective of the Hypothesis Phase is to propose a solution to achieve the task objective, a set of goals and hypotheses for this solution, and the factors and performance metrics for testing the validity of the solution.
We must establish its
validity
through a formal process of Synthesis and Validation. Without such rigor, an hypothesis remains idle speculation
The solution to achieve your objective consists of:
a mechanism
(tools)
and
a procedure
(method)

Objective: Measure the length of a table
Mechanism: Use a Tape Measure
Procedure: (the procedure is straightforward)

Objective: Landing astronauts on the Moon and bringing them safely back to Earth
Many mechanisms
Procedures had to be designed from scratch.
Once a
solution
for a problem has been
specified
in complete detail, the next step is setting the
goals.
The goals specify how the
response
of the task will be determined when the
solution
is applied to it.
Every task has at least one goal to determine the response of the task unit to the application of the solution, expressed as an infinitive phrase.
Goals may be decomposed into a hierarchy of subgoals, in which the descendant goals, taken together, accomplish the parent goal.
An
hypothesis
is a declarative sentence that asserts a
desired, expected, or possible
conclusion of a goal.
Primary Objective
: To identify which highway carries faster traffic, Highway A or Highway B.
Possible Hypotheses
:
The average traffic speed on Highway A is higher than on Highway B
The average traffic speed on Highway A is lower than on Highway B
The average traffic speed on Highway A is the same as on Highway B
Parameters and condition, that will not or cannot be varied are known as fixed parameters or
fixed
conditions.
Factors, on the other hand, are parameters and conditions that could be
varied
in an experiment protocol, and the researcher selects a subset that will be varied to observe the effects of performance of the task.
When determining the factors for a task, the first step is to
make a complete list
of all the parameters and conditions.

Then, once planning reaches the Hypothesis Phase, a subset of these parameters and conditions can be moved to the list of factors.
A performance metric is a postulate that
transforms the results of the task
into
measures of performance
for
drawing conclusions
about the task
objective.
Selecting the right performance metric can change a failed task into a successful one.
The metric should only be changed if the new metric is more representative of performance than the one that indicated failure.
It is always a great temptation to tinker with metrics until you get the “right” results, but this must never be done at the expense of objectivity
Eating breakfast increases levels of productivity in students:
Age of students
Activity performed
Food ingested as breakfast: type, quantity
Time from breakfast to activity
Levels of performance
Performance environment
The word “synthesis” derives from the Greek word
suntithenai
meaning “
to combine separate elements to form a whole”

assemble and test something.
The objective of the Synthesis Phase is
to implement the task method
(solution and experiments) to
accomplish
the goals and
validate
the hypotheses of the task.
Before the goals and hypotheses for the specified solution can be tested, the solution must be implemented.
There are 2 different methods for implementing solutions:
It can be manufactured.
Or it can be acquired.
Or a combination of the two.
Computer spreadsheets (Excel) or word-processing (Microsoft Word) are common tools for implementing solutions (
acquired solution
).
Also, purchasing a computer desktop or mainframe machine
(acquired)
for implementing a solution is much easier and cost efficient than to
manufacture
a computer system from scratch.
Measure-Character Recognition Case:
Solution Implementation: The software application for acquiring the character-recognition data during the experiment trials was constructed in Microsoft Excel…
In this case, the solution implementation was acquired via Microsoft Excel…
Hiring a consultant (
acquired solution
) versus training a member of a task team (
manufactured solution
).
If it meets the needs of a task, using an existing solution is almost always more cost-effective than manufacturing it yourself.
Before beginning the design of the experiments, carefully inventory all the components of the task.
Goals & Hypotheses
Solution
Experiments
2.1 Laboratory
2.2 Block design
2.3 Data management
The purpose of this step is to design a series of experiments whose results will eventually be used to estimate how well the task objective has been achieved.
An experiment acquires data to measure the performance of the solution under controlled conditions in a laboratory.
Personal Safety first!
Security of the lab too (keep out sign)
Maintenance people, CEO’s, etc.
The laboratory for a task is the set of physical regions containing the task unit and all the task resources required during the experiments.
This definition certainly does include the conventional image of the laboratory:
A large facility filled with electronic, mechanical, chemical, medical, or computer equipment
However, this stereotype turns out to be the exception rather than the rule
Generally in the field, in several different locations.
Also, some might not need such an area, just a person and a PC.
A
cohort
is a group of subjects who have shared a particular event together during a particular time span(e.g., people born in Europe between 1918 and 1939; truck drivers who smoked between age 30 and 40); a member of a cohort is usually called a subject.
Coming back to the ‘factors’…
Each factor is assigned a set of values.
The total number of combinations of values that may be assigned to all the factors is the product of the number of values assigned to each one of them
This number is called the extent of the factor space.
Data is what the entire project is all about, its raison d’etre
Avoid the evolution of data empires
Organizations regard data as intellectual property.
It is an age-old dilemma between access and control
Options for continuous monitoring of the lab
Video Cameras
Sound and vibration
Power line activity
Temperature
Humidity
Smoke and gas
A
sample
is a task component consisting of a set of nonliving objects or concepts, all of which are assumed to be equivalent for the requirements of the task objective; a member of a sample may be called: instance, response, data point, specimen, item, or case.
Temperature
Amount of water
Light time
Levels
25oC, 30oC, 35oC
5 ml, 15 ml, 25 ml
1 hr, 5 hrs, 10 hrs.
27 possible unique combinations of the values of the factors.
An experiment
treatment
is a combination of one level or setting for every factor.
25oC, 15 ml, 5 hrs
An experiment
trial
is a complete set of treatments applied to a member of the task unit during the experiment.
An experiment
block
is a set of trials that provides a cover of the factor space that is appropriate and adequate for achieving the task objective.
Typically, a blocking factor is a source of variability that is not of primary interest to the experimenter.
An example of a blocking factor might be the sex of a patient; by blocking on sex, this source of variability is controlled for, thus leading to greater precision.
The basic concept is to create homogeneous blocks in which the nuisance factors are held constant and the factor of interest is allowed to vary.

Within blocks, it is possible to assess the effect of different levels of the factor of interest without having to worry about variations due to changes of the block factors, which are accounted for in the analysis.
The general rule is:
"Block what you can, randomize what you cannot."
Blocking
is used to
remove the effects
of a few of the most important nuisance variables.
Randomization
is then used to
reduce the contaminating effects
of the remaining nuisance variables.
Protect data from unauthorized access with public-key encryption.
Maintain logs of every copy of important data files.
Copyright and publish all experiment data promptly.
Record ALL experiment data
Establish clear organizational and documentantion conventions for data.
Monks often took an ancient Greek or Roman scroll, scraped off the old ink, and wrote their daily prayers on the erased parchment.
Recently, using sophisticated technology, some of the original works of the Greek mathematician Euclid were detected under the religious scribblings of a minor 13th century monk.
Now we have CDs!!!
Internet!!!
Risk of data theft
The plans have all been laid. Now is the time to put them to the test.
Resist the temptation to improvise
If the experiment does not run smoothly or successfully, stop it, and reenter the Scientific Method to prepare a new and improved set of plans that solve or avoid the problem.
Regard a failed experiment as a pilot.
Often the performance values from which the conclusions of the task will be drawn cannot be computed directly from the raw results of the experiments.
First, the results must be reduced, combining and/or transforming them in some way to generate values suitable for the domains of the performance metrics.
Distance and counts are direct (no transformations)
Time Conversion (Hours -> Minutes)
Distance (m) and Time (s) -> Speed (m/s)
Which are the factors?
Which are the levels of each factor?
How many possible combinations can we have?
Sophomore, 16, ITESM … is a ____________
In which way can we reduce the following results:
12 Abercrombie
10Hollister
8 Aeropostale
Which are the factors?
School, Student Class, and Age of Students
Which are the levels of each factor?
ITESM, Vanguardia, Regis
Freshmen, Sophomore, Junior, Senior
15, 16, 17, 18, 19
How many possible combinations can we have?
(3 )(4)(5) = 60 unique possible combinations
Sophomore, 16, ITESM … is a treatment.
In which way can we reduce the following results:
12 Abercrombie (12/30) x 100 = 0.40 x 100 = 40% Abercrombie
10Hollister (10/30) x 100 = 0.33 x 100 = 33% Hollister
8 Aeropostale (8/30) x 100 = 0.27 x 100 = 27% Aeropostale
From the latin
validus,
meaning
“strong”.
It refers to the
capacity of knowledge to resist challenge or attack
.
The objective is
to decide whether the objective of the task has been achieved,
based on formal conclusions about its goals and hypotheses and a rigorous peer review of the task methodology.
Ex: Identifying Faster Traffic
Performance Metrics:
Assuming Normal (Gaussian) Distribution,
The Averages of Traffic Speeds of Highway A and B
The Standard Deviations of each Traffic Speeds
Null Hypothesis
(Ho)
When the decision criterion refers to a single, unique point in the decision space (e.g. zero)
Alternative Hypothesis
(HA)
Not null (Research Hypothesis)
Histograms
and
Pareto plots
can reveal unusual trends
For example, if one assumes normal distribution, then the recorded data should reveal a symmetrical distribution where the peak value is at the mean.
However, a Histogram chart may reveal that the recorded data may be “skewed”, and thus the assumption of normal distribution will not work.
Suppose that we have obtained a sample of measurements of the fuel efficiency for the new engine (1st example).
Regardless of the value of the average fuel efficiency, we cannot conclude that it is statistically greater than 40 mpg, because we do not know how much greater it must be to support this claim.
However, we can compute the confidence of rejecting the null hypothesis that the average measurement is exactly 40 mpg.
A Pareto plot is type of plot used in quality control applications that combines a bar chart displaying percentages of categories in the data with a line graph showing cumulative percentages of the categories.
Create a histogram and a Pareto diagram using the data from high school preferences.
If the methods and knowledge of the project task remain locked in research notebooks, then all the planning and hard work go for nothing.
The primary project documentation must be complete and well-organized and must explain every detail of every task.
The secondary project documentation is based on the primary project documentation, and can be in the form of written technical reports, professional journal articles, or conference presentations.
Title
Compose a brief, but accurate description of the project, so that when the literature on the subject is perused, the researcher may make a fairly reliable initial decision whether or not the report is of interest, simply by reading the title.
Sometimes up to 20 or 30 words!
“The successful Insertion into Low Earth Orbit of a Manned Mercury Spacecraft Launched with a USAF Atlas Rocket” (18)
“A rigorous and Complete Mathematical Proof of Fermat’s Last Conjecture” (10)
Abstract
It is a summary of the application of the four phases of the Scientific Method to the project.
It is generally about ten times longer than the title, (about 200 to 300 words) equivalent to a long paragraph.
Introduction
To summarize the entire project at a higher level of precision than the abstract
To define the typographic conventions used in the report
To report the levels of effort and costs of the project.
Often managers and executives have no time to read any further.
It must tell the whole story in as much detail as possible in no more than 10% of the total length of the report.
Analysis, Hypothesis, Synthesis and Validation
Recommendations
The recommendations for follow-on efforts are likely to be linked specifically to the objective of some parent project
“Based on the encouraging conclusions of this project, it is strongly recommended that the parent development project be fully funded in the next fiscal cycle”
Appendix
To include material that is not appropriate to be included, or physically cannot be included in the body of the report itself.
Diagrams, blueprints, computer programs, data sets, etc.
Index
Written Reports
Shorter than the primary reports
Editing down the primary report to an appropriate length.
6-8 pages long.
Oral Reports
They require special design and delivery skills.
PowerPoint.
Acceptance requires an explicit and unbiased confirmation that the three criteria have been met:
Reproducibility
Completeness
Objectivity
Unfortunately the time delay from submission to publication is usually at least one year.
Peer review is the final step in the process of the Scientific Method, the Supreme Court, if you will, that issues the imprimatur for a research and development project, either recommending the provisional acceptance of its conclusions as reproducible, complete, and objective; or remanding the project back to the R & D team for reconsideration in some specified respects.
By definition, the
reduce results files
recorded in the last step of the
Synthesis Phase
are the basis for computing the performance of the task unit using the
performance metrics
postulated in the
Hypothesis Phase
.
Note: The average and standard deviation of the traffic speeds are examples of reduced results as discussed in the Synthesis Phase.
Thus, the importance of applying the correct performance metrics is essential.
Because stepwise refinement is not only allowed, but expected as part of the Scientific Method, the performance metrics that you apply may have to be revised.
Once the results have been reduced and the required measures of performance computed (i.e. in tables, charts, and/or graphs), the conclusions for the task may be drawn.
It is these conclusions that determine the extent to which the primary objective has been achieved.
There are two types of conclusions that may be drawn:
Formal Conclusions
– those conclusions that are required to achieve the primary objective and are based on rigorous methods for acquiring and reducing the results (i.e. statistical methods).
Informal Observations
– those conclusions that may be drawn based on obvious trends in the performance values (very-low risk conclusion).
There is a 95% confidence that the temperature of the room lies between 62 and 68 degrees F.
At a significance level of 5%, the investigator´s claim that students in Tec prefer the BlackBerry over the iPhone may be correct.
Suppose an elementary school teacher is trying to decide which flavor of ice cream (chocolate or vanilla) is preferred by the school children in preparation for a party.
The teacher asks a group of 20 children and 19 indicate a preference of chocolate over vanilla.
Therefore, the teacher can simply reject the null hypothesis that the children have no strong preference. The teacher does not need to apply statistical methods to compute the confidence with which she can reject the null hypothesis.
She simply can conclude “Chocolate is Better!”
Surprise Quiz!
part of the analysis phase.
complaint
question
request
The desired situation, contrasting the given problem.
It is the reason the research is performed.
1) Pose a Question
2) Develop a Hypothesis
3) Plan the procedure
Identify the manipulated variable and the responding variable. Describe every step of the changes on the manipulated variable, and record your observations of the responding variable on a prepared data table. Control the variables and establish operational definitions.
4) Interpret the data
5) Draw conclusions
Compare that interpretation with your hypothesis and decide whether it was true or false.
Remember, as one of the first steps in planning your investigation, you may need to narrow your original question. Then write a hypothesis and design an experiment to answer the question. Be sure to include all the necessary parts of an experiment, such as naming the manipulated and responding variables and identifying the variables you will control.Write any operational definitions that are needed. Include a data table you could use for recording your observations
1. How is heart rate affected by exercise?
2. How are bean seedlings affected by water that has been polluted by detergent?
3. What effect does acid rain have on marble statues?
4. Does sand in the wheels of my in-line skates affect how fast they roll?
5. Will a wet sheet become dry when hung outdoors on a freezing day?
6. Is a family’s health affected by using a dishwasher?
7. How is gas mileage affected by the type of gasoline used?
8. Does the presence of plants growing on a hillside change the amount of soil
erosion?
9. Does cold water freeze faster than hot water?
10. Does the type of shampoo I use have an effect on how long my hair stays
clean?
28 February 1915 – 2 October 1987
British-Brazilian biologist
whose work on graft rejection and the discovery of acquired immune tolerance was fundamental to the practice of tissue and organ transplants
WHY?
APA Style
references
format
ethical?
legal?
Validity
Citation
Traditionally, investigating the related work is carried out by performing literature searches of the
relevant publications
, either in print libraries or at sites on the Internet.
Sam has been spending his summer boating on the great lakes. However, he’s not too bright (after all, the brains of worms are pretty small). He’s never learned how to swim, and he never wears his life preserver. The worst has happened! His boat has capsized and he’s stuck! Fortunately, his life preserver is in the boat, but unfortunately he does not know how to reach it without falling off and drowning.
SAVE SAM!
Sam, the boat, the life preserver can be touched only with the paper clips. NO HANDS.
Do not stab Sam or the lifejacket, please.
Upload a Vine #SavingSam
Full transcript