180 Understanding Rational Decision Making
Easy-to-Compare Slot Values
Audiences will integrate information more accurately and effi ciently when comparable slot values
for each alternative are expressed in comparable ways, for example when the values for several
international currencies are all expressed in U.S. dollars.^387 Comparable scaling (e.g., Good, Very
Good, Excellent) of slot values is another way to improve the accuracy of the audience’s informa-
tion integration process. Noncomparable scaling (e.g., Good, B+, *****) across alternatives tends
to decrease the audience’s use of more normative compensatory choice rules.^388
The Right Graph for the Comparison
The best graph for any particular task is one that minimizes the number of steps needed for making
the intended comparison.^389 For example, bar charts are better than pie charts when viewers need
to compare quantities because bar charts display quantities as lengths on a common scale, whereas
pie charts display quantities as areas of a circle or slices and require viewers to mentally rotate the
slices before making a comparison.^390 By the same token, pie charts are better than bar charts when
viewers need to make accurate part/whole judgments because pie charts give viewers a better sense
of the size of the whole at a glance.^391
For a graph to be effective, it must also activate the appropriate schema for interpreting the data
in it.^392 Viewers may misinterpret the meaning of a graph if its graphic elements do not map onto
their schema as expected. For example, viewers expect the dependent variable in a line graph to be
plotted on the y axis, such that a steeper slope implies a faster rate of change. When a line graph
violates this expectation, graph viewers tend to misinterpret the meaning of the slopes.^393
The Elimination of Irrelevant and Inconsistent Information
Irrelevant information can have a dilutive effect on the audience’s information-integration pro-
cess.^394 Irrelevant information mixed with relevant information on complex charts and graphs
impairs viewers’ ability to interpret them correctly.^395 Visuals and text that are topically related to a
teacher’s lesson but irrelevant to the learning goal decrease student learning.^396 Such irrelevant but
“seductive” details interfere with learning by priming inappropriate schemata which students then
use to organize the relevant information in the lesson.^397
Another reason irrelevant information has a dilutive effect on audience decisions is that audi-
ences often use a weighted averaging rule to make their predictions and other judgments. Thus,
when audience members receive irrelevant information mixed with information that is relevant,
their predictions become more regressive and less based on the relevant information presented to
them.^398 For example, when college students were asked to predict another student’s grade point
average (GPA), they predicted the student would earn a high GPA when they were given only rel-
evant information about the student’s good work habits.^399 But they predicted a lower GPA when
the relevant information about the student’s work habits was presented along with irrelevant infor-
mation about the student’s age, hair color, etc. Although expertise can mitigate dilution effects,^400
it does not always do so. For example, irrelevant information provided during performance reviews
distorts supervisors’ assessments of employee performance even when supervisors recognize the
information is irrelevant to the assessment and should not matter.^401
Irrelevant information also dilutes the persuasive impact of strong arguments in much the
same way that weak arguments dilute the persuasive impact of strong arguments^402 and mildly
favorable information dilutes the persuasive impact of highly favorable information.^403 Irrelevant
information added to highly relevant and factual policy arguments, for example, reduces the