Analogies
Df. of Reasoning by Analogy in Argument:

In the premises, we draw a conclusion from A and B to C

In the conclusion, we draw a conclusion from X and Y to C
Status of Argument from Analogy

Usually a weak argument

"When we ask So? youíre on the way to deciding if the analogy is
good"
Example of Firefighters and Soldiers

Analogous Conclusions

We donít blame firefighters for fires

We shouldnít blame soldiers for wars

Similarities

Both involved in dangerous work

Principle of not blaming ones who help end disaster if they did not start

Differences

Still fires without firefighters

Not wars without soldiers
Keys Points for Evaluating Analogies

Assess Similarities

Assess Differences
Numbers
Issue: We use numbers to be exact, but it is easy to be mislead
when reasoning with them
Percentages Can Be Misleading

Mean: Result of adding numbers and dividing by number of items

Median: Number midway between high and low in collection of numbers

Mode: Number that occurs most often in a collection of numbers
Principles Based on an Understanding of Percentages

Average Sometimes Is Used to Refer to Mean, Median, or Mode; so, Be Clear
on the Type of Average Being Used

Unless tmean is close to median and distribution is close to bellshaped,
mean does not give important information
Generalizing
Df. of Generalizing: We conclude a claim about a group, the population,
from a claim about some part of it, the sample
Df. of Generalization: Sometimes we refer to an entire argument,
not just the conclusion, as a generalization
Samples

Representative Sample: One in which no one subgroup of the whole
population is represented more than its proportion in the population

Biased Sample: One that is not representative

Haphazard Sample: One which is not based on intentional bias, but
one which is not clearly representative
Means to Attain Accurate Sample

Random Sample: One which results if at every choice there is an equal
chance for any one of the remaining members of the population to be picked

Law of Large Numbers: The larger the sample the more likely the results
correspond to the actual probability
Gamblerís Fallacy: A bad argument that a run of events of
certain kind makes a run of contrary events more likely in order to even
up the probabilities
Premises for a Good Generalization

Sample is representative

Sample is big enough

Sample is well studied
Risk

Does not change how strong an argument you have

Changes how strong an argument you want before you accept the conclusion
Cause and Effect
Claim: if the cause is present, the effect almost certainly follows
Necessary Criteria for Cause and Effect

The cause precedes the effect

The effect almost certainly will follow under normal circumstances

The cause makes a difference

Be suspicious of ìafter this, so because of thisî (post hoc; ergo, propter
hoc)

Be suspicious that a statistical correlation is confused for a causal relation
(non causa pro causa)

The cause is close in space and time to the effect