• <More on Intel.com

Assessing Arguments

Assessing arguments

Make sure that students are secure with the idea that there is no right answer. They must feel safe to make a claim and support it without the threat of a lower grade if they pick the “wrong” claim. Ensure that students understand that the quality of their case is based on the quality of their argumentation, not right or wrong answers like a multiple-choice test.

Creating and defending effective arguments is not something learned in one lesson. Discuss throughout the project what constitutes a strong—as well as a weak—argument. Consider problems that could occur in evidence while constructing an argument:

Problem Example
The conclusions do not follow logically from the evidence given. The candy bar was here on the table last night. This morning, it’s gone. Johnny must have taken it.
The argument is based on analogy that doesn’t work. Baseball bats can be as deadly as guns, but we don’t ban bats.
Fact and opinion are intermingled, opinions are presented as facts, or it is not clear which is which. People who regularly wear seatbelts are more responsible and have fewer accidents.
Celebrity is used as authority.  Film star endorses new diet
Vague sources are used in place of specific references.  "leading doctors say ... ," "science has shown that ... ," "compared to some other states ... ," "the scientific community recommends that ..."
Care is not taken to guard against deliberate or subconscious distortion, in self-reported opinions or information. “In a survey of our college students, 87% are rated as ‘above average.’” No notice is given that it was a self-reported survey with no outside confirmation.
No mention is made, in evidence said to come from an experiment, of control groups very much like the experimental group. Scientific studies proved that the new drug was effective for treating depression.
Graphs are used that distort the appearance of results. Chopping off part of the scale, using unusual scale units, or using no scale at all
Categories are over-generalized—implying that all members of a group have nearly identical characteristics. All "teenagers," "consumers," "immigrants"
Average results are reported, but not the amount of variation around the average. The average income of college graduates from private universities exceeds the average income of college graduates from public universities
A percentage or fraction is given, but not the total sample size. "9 out of 10 dentists recommend..."
Absolute and proportional quantities are mixed.  “We had 3,400 more robberies in our city last year; whereas, other cities had an increase of less than 1 percent.”    
Results are reported with misleading preciseness. Representing 13 out of 19 students as 68.42 percent. Using the percentage distorts the fact that the actual sampling is a very small number.
Explanations or conclusions are represented as the only ones worth consideration, with no mention of other possibilities. The experimental data prove that aluminum is the best conductor (when only 3 conductors were tested).

Adapted from
Science for All Americans Online
Chapter 12: HABITS OF MIND