Understanding Psychology (10th Ed)

replicated research
Research that 
is repeated, sometimes using other 
procedures, settings, and groups of 
participants, to increase confi dence 
in prior fi ndings.
P
er
centage helping 
2 3 6 
Size of group 
90 
80 
70 
60 
50 
40 
30 
20 
10 
0 
The smaller the number 
of bystanders, the 
greater the degree of 
helping 
FIGURE 4
The Latané and Darley experiment showed that as the size of the group 
witnessing an emergency increased, helping behavior decreased. 
(Source: Darley & Latané, 1968.)
feL82795_ch02_030-057.indd Page 46 8/17/10 9:18 PM user-f464
feL82795_ch02_030-057.indd Page 46 8/17/10 9:18 PM user-f464

---

Applying Psychology 
in the 21st Century
Testing the Value 
of Self-Affi
rmations: 
Am I Lovable Because 
I Tell Myself I’m Lovable?
If you’ve ever felt down about yourself, 
you might have sought comfort from a 
friend or family member who spent some 
time telling you what a great and wonder-
ful person you really are. Perhaps you’ve 
done the same for a friend of yours who 
was having a crisis of confi dence. Pop psy-
chology wisdom tells us that these affi rma-
tive statements, or affi rmations, help us feel 
better about ourselves. We’re sometimes 
even advised to use self-affi rmations to 
help boost our sense of self worth, and it 
makes good common sense that if you keep 
repeating to yourself statements such as 
“I’m a lovable person” or “I will succeed,” 
you’ll believe it and like yourself more. But 
is that really the case? (Gordon, 2001)
Relying on intuition alone to answer this 
question is not enough. Something that 
sounds like it should be true may not be. 
The only way to tell is to put the hypothesis 
to the test—and that’s exactly what a group 
of psychology researchers did with affi rma-
tions. They brought several dozen college 
students into their laboratory, where they 
fi rst asked students to assess their self- 
esteem level and mood using question-
naires. Then the researchers instructed the 
students to write down their thoughts and 
feelings for several minutes. Half the stu-
dents were given an additional instruction: 
Whenever a certain tone sounded, they 
were to give themselves a self-affi rmation, 
telling themselves “I’m a lovable person.” 
(The group providing themselves the self-
affi rmation was the experimental group. 
The other group, which did not receive this 
additional instruction and did not engage 
in any self-affi rmations, was the control 
group (Wood, Perunovic, & Lee, 2009).
After several minutes (and 16 repeti-
tions of the self-affi rmation for the experi-
mental group), both groups were then 
given new tests of their current level of 
self esteem and mood to see if changes 
occurred. The results clearly showed that 
a change did occur for students who had 
low self-esteem and who engaged in self-
affi rmations: their mood and self-esteem 
went down. This is, of course, the exact 
opposite of the result that proponents of 
self-affi rmation would intuitively expect. 
Instead, the study showed that not only 
were self-affi rmations not helpful but that 
they actually seemed to backfi re.
Why might this happen? Interpreting 
these surprising fi ndings is easier if we con-
sider the fi ndings of other, related research. 
If you consider these self-affi rmations to be 
a form of self-persuasion, we can look at 
research on persuasion and fi nd that people 
tend to be more persuaded by messages 
that are in line with their pre- existing atti-
tudes. When a message is strongly contra-
dictory to what a person already believes, 
he or she may counter that message with an 
argument.
So when people who have low self- 
esteem tell themselves a contradictory 
message, such as “I’m a lovable person,” 
instead of being able to accept the message, 
they might think of times when they were 
not lovable because they acted unpleas-
antly or were rejected by others. The self-
affi rmation might therefore backfi re by 
causing people with low-self-esteem to lin-
ger on negative thoughts that just confi rm 
their poor self-image.
These 
fi ndings, which run counter to 
common-sense explanations, underscore 
the need to submit hypotheses to scientifi c 
testing. Even when the outcome seems 
“obvious,” we need to do research, because 
our intuitions are often simply wrong 
(Eagly & Chaiken, 1993).
• Are you convinced by the experiment results that self-affi rmations are not effective? 
Why or why not?
• Could you design another experiment to test the same hypothesis?