W HAT IS SCIENCE?



1. A BODY OF FACTUAL KNOWLEDGE about the natural world.
2. The PROCESS by which this body of knowledge is gained and modified: The search for truth about the natural world.

This process embodies a PHILOSOPHICAL SYSTEM of logical thinking and actions to determine natural realities and their underlying mechanisms.

INCLUDES

REQUIRES

THE SCIENTIFIC PROCESS


INITIAL OBSERVATIONS
& prior knowledge
|
|
Inductive reasoning
|
V

HYPOTHESIS
Statement of general relationships & causal mechanisms

|
Deductive reasoning
|
V

PREDICTION
Specific testable expectation
"If...is true, then...should happen when...is done."

|
|
V

EXPERIMENT
appropriate controls, elimination of bias,
statistically valid sample sizes

|
|
V
RESULTS
 Negative: falsifies hypothesis (proves it wrong)
Positive: supports hypothesis (but does not necessarily prove it true)


EXPERIMENTAL CONTROLS


--In any experiment to investigate the effect of a particular treatment or other variable on a group of subjects (e.g. in the case of biology, on a population of cells or organisms), there must be a control group in addition to, and identical to, the experimental group.

--Both groups must be subjected to all of the same experimental conditions except for the specific variable to be tested.

--The purpose of the control group is to account for any unexpected "side effects" of the experimental conditions not caused by the test variable, that could be mistaken for, or mask, the effects of the test variable.

For example, to test the effects of a new drug on a group of mice:

Experimental group
The drug is injected hypodermically in a saline solution.

Control group (same age, raised under identical conditions)
The same volume of saline is injected without the drug in the solution. This controls for any effects of the injection process or the saline vehicle.

BIAS

What is it and how do we deal with it?

Bias is

Bias may be

Bias is difficult to eliminate (and some is justified), but
for conducting proper science (and fair citizenship),
bias (no matter whether it is correct or incorrect) must be

Manifestations of bias in science

GOAL FOR ELIMINATING OR CONTROLLING BIAS :

Equal opportunity for every observation to be seen for what it really is, not what it is expected to be.

EXPERIMENTAL ERROR & PROBABILITY


--There is some degree of randomness in the movement and /or behavior of matter at all levels of organization, and therefore some degree of unpredictability in this behavior. Therefore,

1. 100% results can rarely be expected or obtained, and random variations can be large enough to cause erroneous conclusions or obscure real "treatment" effects.

2. One must be able to know or determine what portion of the results can be expected and explained on the basis of random "error". This is a form of "experimental error", but it is not in itself a mistake committed by the investigator.

--Accuracy and precision of measurement technique must also be known (i.e. degree of "sloppiness" of data gathering). This form of "experimental error" can include investigator mistakes, and also must be accounted for.

--Knowledge of and ability to use statistical techniques is essential for doing good science (or any other use of quantitative information), and for judging the quality of otherís work. It is necessary for many aspects including

1. Experimental design; sample size, etc.
2. Revealing data variance
3. Properly determining level of significance of apparent differences between data groups

THE TRUTH TABLE

If a hypothesis is ... Results of testing its predictions
TRUE always be positive

[Thus one negative result will prove a hypothesis to be false.]
FALSE be negative for some or all predictions, but may be positive for one or more predicitions. [I.e. a hypothesis can be partially true, under some but not all circumstances.)

[Thus one (or some) positive result(s) do not necessarily fully prove a hypothesis to be true.]


Because of the above relationships, the following governs the progress of a scientific endeavor:

--Hypotheses that are of any use or importance must be testable, i.e. must generate predictions that can be shown by doable experiments or observable phenomena to falsify the hypothesis (if it is in fact false).

--Limited testing with positive results will support, but not prove, a hypothesis.

--Extensive testing by many investigators with all positive results can elevate a hypothesis or set of hypotheses to the status of general theory.

An hypothesis (or theory) found to be false under certain circumstances can be modified and restated (and of course retested).

CORRELATION vs. CAUSATION

Correlation - Two or more phenomena that regularly occur at the same times and that vary quantitatively in a consistent relation with each other (in either positive or negative directions).


Phenomena that are correlated with each other may or may not be causally related to each other; they both could be caused by another unrecognized variable. One may appear to cause the other, but correlation alone never proves causation.


Correlation can be determined descriptively by recording a series of observations.

Causation, which would be the basis of an hypothesis explaining the correlation, must be determined by experimentation, i.e. by manipulating the appropriate variable(s) in the system and observing whether the predicted outcome occurs.

SCIENTIFIC SERENDIPITY

Many important scientific discoveries arise partly as a result of "accidents", or unexpected variations in the outcomes of experiments.

Example: Minkowski's discovery of the role of the pancreas (and its insulin) in diabetes.

Failure to notice (and to be curious, rather than just annoyed, about) the extra urine and the flies would have prevented (or postponed) this pancreatic function from being discovered.


"Accident" is an inadequate description of such discoveries:

--Experiments are usually being conducted in at least the same general area of inquiry.

--The observations are interpreted by someone with the appropriate background, knowledge and perspective to notice the implications and to pursue them to useful conclusions (or to alert others who can). This is where curiosity and imagination become so important.

"Discovery consists of seeing what everyone has seen and thinking what no one has thought" (Albert Szent-Gyorgyi)

BIOETHICS


DECISIONS ABOUT ACTIONS WITH BIOLOGICAL CONSEQUENCES

Examples of issues involving bioethics:


Decisions such as these require more than knowledge of the biology involved. Other value systems in addition to science play a role, including one or more of the following:


However, bioethics cannot properly be addressed without knowledge and understanding of the appropriate biological science.

THE MAJOR VALUE SYSTEMS (as compared to science)

SCIENCE - A system of logic in which the truth about the natural world is sought based on factual, unbiased, repeatable observations and experiments, through the use of testable hypotheses. An hypothesis is presumed true until proven false. Evidence is sought to falsify it; all facts must be considered.

LAW - The process or pursuit of justice based on legal rules of evidence. An accusation is legally presumed to be false until proven true. Evidence is sought to prove it; only legally admissible facts can be considered.

PHILOSOPHY - A search for truth through an internally consistent system of logical reasoning (which may or may not include factual observation). Such a system can be based on any premise or agreed upon set of rules or descriptors of a situation or point of view.

SOCIOLOGY - The analysis and description of sets of behavior patterns that have evolved in populations as "social norms", i.e. the usual, accepted ways of doing things, with and against which all behaviors are compared and judged.

POLITICS - The pursuit of understanding and/or influencing the policies of institutions and/or governments. Upon erecting an internally consistent premise, an entire prescription for institutional and social behavior can be devised (e.g. federal democracy, fascism, communism, etc.). The premises are based on sets of assumed sociological and socioeconomic "truths".

THEOLOGY - The intepretation of religous truths expressed in bodies of written and/or spoken histories and/or mythologies, generally accepted by proponents of a religion as revelations of the relationship of mankind to the universe and the meaning of life. Especially applicable to things that are scientifically unknowable.

G ENERAL UTILITY OF SCIENTIFIC APTITUDE

Familiarity with science and adopting a more scientific frame of mind can benefit non-scientists in many ways (both personally and as citizens), including the following:

Everyone should make more use of the two questions


Return to the BIO 103 Mainpage...