|
|
|
Critique of Induction |
|
|
|
Inductive
arguments can be divided into four kinds, depending on two factors. The
factors are...
(1) the sample - which can be either...
(a) simple (uncontrolled); or
(b) bifurcated (controlled)
(2) the nature of
the generalization - which can be either
(a) straight (to the population as a whole); or
(b) oblique (to other unobserved cases within the population).
These two factors create a grid with four
quadrants.
|
|
|
|
|
|
|
|
| |
Simple (uncontrolled) sample |
Bifurcated (controlled) sample |
|
Straight
generalization |
Uncontrolled studies that
generalize to an entire population.EXAMPLE:
Opinion polls
|
Controlled studies that generalize
to an entire population.EXAMPLE:
Comparison studies
(e.g. comparing
women to men)
|
|
Oblique
generalization |
Arguments from observed
cases to unobserved cases.EXAMPLE:
Analogical arguments;
trend projections
|
Controlled studies applied
to new individual cases.EXAMPLE:
Medical studies
|
|
|
|
The criteria by
which Inductive arguments should be evaluated fall into two types,
corresponding to the two premisses of an Inductive argument: the facts
observed and the characteristics of the sample. |
|
I. Critique of the Observation There
are two criteria by which the observation should be critiqued.
|
|
|
|
A. Objectivity - Is the observation performed in a
way that filters out the expectations and prejudices of those
participating in the observation, specifically the experimental subjects
(assuming they are persons) and the experimenter who makes and reports
the observations?
- An increase in the objectivity of the experimental subjects will
strengthen the argument.
- An increase
in the objectivity of the experimenter will strengthen the argument.
|
|
|
|
|
B. Decisiveness - Is the property observed a clear
and unmistakable example of the property in question, or is it a weak,
borderline example? Is the conclusion drawn from the observation
commensurate with the decisiveness of the property actually observed?
- An increase in the decisiveness of the observation (i.e. the premiss)
will strengthen the argument.
- A decrease
in the decisiveness of the expectation (i.e. the conclusion) will
strengthen the argument.
|
|
|
II. Critique of the Sample There
are four criteria that can be used to judge the adequacy of the sample.
However, only the first two apply to every sort of Inductive argument. The
third criterion applies only to oblique generalizations; the fourth applies
only to arguments based on a bifurcated sample.
|
|
|
|
A. Sample Size - The larger the sample, the
stronger the induction.
- An increase in the size of the sample will strengthen the argument.
|
|
|
|
|
B. Diversity of Sample - The more points of
difference among the members of the sample (or the fewer points of
difference among the members of the population), the stronger the
induction. There are two ways to achieve sample diversity:
randomize - make sure each member of the population has an equal
chance to be selected in the sample.
stratify - make sure that relevant segments of the population are
represented in the sample (in proportion to their occurrence within the
population). -
An increase in the diversity of the sample will strengthen the argument.
- A decrease
in the diversity of the population will strengthen the argument.
|
|
|
|
|
C. Analogy between the observed cases to the unobserved
cases (Oblique generalizations only) -The more points of
similarity, the stronger the induction.
- An increase
in analogy (similarity of observed cases to unobserved cases) will
strengthen the argument.
|
|
|
|
|
D. Comparability of the experimental group to the control
group (Controlled samples only) -
The more points of similarity, the stronger the induction.
- An increase in comparability (similarity of experimental group to
control group) will strengthen the argument.
|
|
|
|
|
