Week 1 - Classics of Attitude Change

Acquiring conviction through active participation (Hovland et al, 1953)¹

Design/Method Participants gave persuasive talks based either on improvised arguments or prepared scripts, while controls only read silently.

Findings Improvised speakers showed more attitude change than script-readers or silent readers.

Theoretical Contributions Demonstrated that persuasion depends on active self-generation of arguments, not just exposure or repetition.

Associationism vs Propositionalism Supports propositionalism: improvisation requires creating structured, inferential propositions, while mere association (oral reading) did not produce additional persuasion.

Notes

• This article is dated and shows by the way the methodology is not totally clear.
• “The second study was the most important” - Mandelbaum

Questions

• This reminds me of learning science where active engagement with the study material is more effective than passive reading and re-reading. Is there a similar effect happening here where generating your own argument aligns with learning science’s idea of elaboration and higher-order learning?
• Is it effort that improves self-persuasion? Reading is passive and less engaged and effortful compared to improvisational justification.

Cognitive Consequences of Forced Compliance (Festinger & Carlsmith, 1959)²

Design/Method
Participants performed a boring task, then were asked to tell another participant it was enjoyable. They were paid either $1(lowincentive)or$20 (high incentive). Afterwards, they rated how enjoyable the task actually was.

Findings
Paradoxically, those paid $1ratedthetaskasmoreenjoyablethanthosepaid$20. The $1groupchangedtheirattitudestoresolvedissonance;the$20 group did not need to because they had sufficient external justification.

Theoretical Contributions
Established cognitive dissonance theory: when external justification is insufficient, people change their internal attitudes to resolve the inconsistency. This undermined behaviorist reinforcement models and showed that beliefs are actively regulated for coherence.

Associationism vs Propositionalism
Supports propositionalism: the shift results from reconciling inconsistent propositions (“the task was boring” vs. “I said it was fun for only $1”). Associationism cannot easily explain why smaller reinforcement produced larger attitude change.

Notes

• Mandelbaum: “This is the worst named studied in the history of psychology.” Why? It’s not really forced.
  • Nothing in the design forced participants to comply. They were induced: offered either $1or$20 to tell someone the task was enjoyable. They could have refused.

Questions

• Does dissonance theory point to an innate need for propositional coherence in the mind?
• What is the mechanism by which coherence is sought and why is this necessary? Is it not possible to hold onto two conflicting beliefs?
• What exactly is dissonance? A emotional state? An incommensurate set of beliefs? Something else?

Is Choice-Induced Preference Change Long Lasting? (Sharot et al., 2012)³

Design/Method
Participants rated vacation destinations, made hypothetical choices between equally liked options, and later re-rated them. Follow-ups tracked changes over time.

Findings
Choices reshaped preferences: people rated chosen destinations more positively and rejected ones more negatively, and these changes persisted for years—but only when choices were freely made.

Theoretical Contributions
Showed that self-generated choices restructure attitudes in a durable way. Choice is not just a behavioral output but feeds back into belief and preference architecture.

Associationism vs Propositionalism
Supports propositionalism: the effect depends on the propositional act of endorsing one option over another, not mere co-occurrence. Associationism would predict weaker or transient effects.

Notes

• “People want to justify every stupid little behavior” - Mandelbaum

Questions

• Why do free choices uniquely generate durable attitude change—does the propositional act of commitment explain this?
• Can AI/LLMs simulate this phenomenon, or does it require genuinely propositional belief states?

What Doesn’t Kill Me Makes Me Stronger (Tormala & Petty, 2002)⁴

Design/Method
Participants resisted counterarguments of varying strength against their attitudes. Researchers measured subsequent certainty in the original attitudes.

Findings
Resisting strong counterarguments increased certainty in one’s original attitude, but resisting weak ones had little effect.

Theoretical Contributions
Introduced metacognitive inference as a mechanism of persuasion. Attitude strength can change without valence change: people infer from their own resistance that their beliefs are valid, thereby increasing certainty.

Associationism vs Propositionalism
Supports propositionalism: persuasion depends on propositional inference about one’s own cognitive performance (“I resisted strong arguments, so my belief must be true”). Associationism cannot easily capture this metacognitive, context-sensitive process.

Notes

• We see this in implicit attitudes too. When they encounter strong arguments, they’re implicit attitudes get stronger.
• The main idea is to push the “what” here—that argument strength matters.

Questions

• How does this finding complicate the idea of effort justification?
• Could failed persuasion attempts actually entrench attitudes more than leaving them untouched?
• Could this be explained by PP, where priors are updated after resisting strong arguments?

Asymmetry in Belief Revision (Yang, Stone, & Marsh, 2022)⁵

Design/Method
Participants learned facts and later received corrections. Researchers compared ease of affirming vs. negating beliefs.

Findings
Updating was asymmetric: affirmations were readily integrated, but negations were harder to encode and less durable.

Theoretical Contributions
Demonstrated that belief revision is constrained by the architecture of memory and representation, not just motivation. The asymmetry highlights structural limits in how propositions are encoded and negated.

Associationism vs Propositionalism
Supports propositionalism: asymmetry arises from the difficulty of attaching and retrieving “false” tags to propositions (Spinozan model). Associationism cannot explain why negations, though equally frequent pairings, fail to persist.

Questions

• Does this asymmetry imply that belief perseverance is an inevitable cognitive bias?
• How does the Spinozan model of automatic acceptance and effortful rejection best explain these results?
  • Automatic acceptance + effortful rejection predicts exactly this asymmetry. Affirmations require only automatic encoding; negations require attaching and maintaining a “false” tag. Forget the tag, and the original belief reasserts itself.

Footnotes

1. Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communication and persuasion: Psychological studies of opinion change. Yale University Press. (Chapter 7: Acquiring conviction through active participation) ↩

2. Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58(2), 203–210. https://doi.org/10.1037/h0041593 ↩

3. Sharot, T., Fleming, S. M., Yu, X., Koster, R., & Dolan, R. J. (2012). Is choice-induced preference change long lasting? Psychological Science, 23(10), 1123–1129. https://doi.org/10.1177/0956797612438733 ↩

4. Tormala, Z. L., & Petty, R. E. (2002). What doesn’t kill me makes me stronger: The effects of resisting persuasion on attitude certainty. Journal of Personality and Social Psychology, 83(6), 1298–1313. https://doi.org/10.1037/0022-3514.83.6.1298 ↩

5. Yang, B. W., Stone, A. R., & Marsh, E. J. (2022). Asymmetry in belief revision. Applied Cognitive Psychology, 36(5), 1072–1082. https://doi.org/10.1002/acp.3991 ↩