How we change what others think, feel, believe and do
People decide on their own attitudes and feelings from watching themselves behave in various situations. This is particularly true when internal cues are so weak or confusing they effectively put the person in the same position as an external observer.
Self-Perception Theory provides an alternative explanation for cognitive dissonance effects. For example Festinger and Carlsmith's experiment where people were paid $1 or $20 to lie. Cognitive dissonance says that people felt bad about lying for $1 because they could not justify the act. Self-perception takes an 'observer's view, concluding that those who were paid $1 must have really enjoyed it (because $1 does not justify the act) whilst those who were paid $20 were just doing it for the money.
Note that this indicates how changing people's attitudes happens only when two factors are present:
Zanna and Cooper gave people a placebo pill and asked them to perform a counter-attitudinal activity. Control people who were told the pill was a placebo did as expected, becoming more supportive of the attitude (because they had enacted it). Others, who were told that the pill would make them tense, did not change their attitude, as they could attribute their dissonance to the effects of the pill.
If you hear a lot of rock music and do not particularly dislike it, you will probably conclude that you do like it.
If you want someone to believe or feel something about themselves, first get them to do it. This works best when they have no particular view about the area in question. If they already have a strong view, you will need to call the view into doubt, for example by giving disconfirming examples.
When people ask you to do things about which you have no clear view, ask yourself what they could gain by your believing something about yourself in this matter.
And the big