Smoke notebook

Study influences on the third-person effect (or illusory superiority, or optimism bias) in addictivity. It appears that people think they're less likely to get addicted than the average person. See if you can replicate this. Then study what might influence or ameliorate this bias.

What kind of influences should I examine?

• Don't expect people to think hard about this stuff, of course.
• Chase strong rather than weak effects.
• I could have people write down a reason they thought that they were less vulnerable to a vice than the average person (and later try to taxonomize these reasons), but is lay justification really what I'm interested in?
• One basic social-psychological sort of prediction, similar to the fundamental attribution error, is that people may see others as more affected by outside forces, like an addictive drug, than themselves. (Here the similarity with the third-person effect in media psychology is particularly clear.)
  • The dumbest version of this is thinking "If I was smoking too much, I could just decide to stop smoking, or smoke less." with no elaboration on the difficulty of this, or how well the same idea applies to other people.
• Broader self-enhancement biases, like people believing they're smarter or more competent than the average person, could explain this particular self-enhancement bias. This may include dissonance reduction, but that might only apply if the person already has a bad habit, rather than imagining the situation if they had it.

The effects perceptible for the last and penultimate bullet point here seem unlikely to be strong. For lack of better ideas, let's have people write down some reasons and see what we get.

• https://cancercontrol.cancer.gov/brp/research/constructs/optimistic-bias - a review piece on optimistic bias
• Davison (1983) - The original paper about the third-person effect
• Popova and Halpern-Felsher (2016), Weinstein, Slovic, and Gibson (2004), Twigg and Byrne (2015) - Adolescents believed they were less likely to get addicted to smoking than a comparable average person. Similarly, Strombotne, Sindelar, and Buckell (2021) found this for vaping. Note that in Twigg and Byrne (2015) it appears there is also a pessimistic bias in which nonsmokers believed they would find it more difficult than average to quit (if they took up smoking).
• Arnett (2000): people agreed that smoking was addictive, but smokers tended to say that they "could smoke for a few years and then quit" if they wished
• Morris, Tattan-Birch, Albery, Heather, and Moss (2025) - more "defensive processing" among people with higher AUDIT-C scores
• http://www.jamesdivory.com/Ivory20043pGamesICA.pdf - People agreed less with "I am addicted to video games." than "I think that other people may become addicted to video games." Similarly, Çakır, Çakır, and Gülnar (2015) found that people agreed more with "Watching television causes addiction" than "I am a television addict"
• Heck, Simons, and Chabris (2018) - 65% of American believe they're more intelligent than the average person
• Greening and Chandler (1997) - optimistic bias for risky events, like car accidents, may be caused by people's beliefs that they have better skills than average

How an optimistic bias for addiction was measured:

• Popova and Halpern-Felsher (2016): The subject rated the probability from 0% to 100% for "If you/John/Jill just started smoking 2-3 cigarettes per day" or "continued to smoke 2-3 cigarettes per day for the rest of your/his/her life" (it's not clear how "John" or "Jill" were defined) followed by various outcomes such as "become addicted", "still be smoking in 5 years", "quit if you/he/she want(s)".
• Weinstein et al. (2004): "In your opinion, once someone has been smoking half a pack of cigarettes a day for several years, how easy is it for them to quit and never smoke again: easy, and most people can do it without much trouble; hard, but most people can do it if they really try; very hard, and most cannot do it; almost impossible, and only a few will be able to do it?" vs. "In your opinion, if you were to smoke a pack of cigarettes a day, how easy would it be for you to quit and never smoke again: very easy, you could quit with no trouble; hard, but you could do it if you really tried; very hard, you don’t know that you could do it; almost impossible, you doubt that you could do it?". Moreoever: "Compared to the average smoker, do you think that you could quit smoking more easily, less easily, or about as easily?"
• Twigg and Byrne (2015): "Compared to an average smoker [or "Compared to a friend of yours"], if you smoked at least three cigarettes each day for the next five years, how easy would it be for you to quit smoking?", rated from 1 (much harder than) to 7 (much easier than).
• Strombotne et al. (2021): "If you or the person/group below were to use e-cigarettes (including JUUL-types) with nicotine on a regular basis, how difficult do you think it would be for them to quit?" The types of people were "you"; "your best friend"; "someone else your age at your school, at work, or in your neighborhood"; and "an average U.S. teenager of your age". They were rated from 0 (easy to quit) to 10 (hard to quit).

Ask separately about a few things that people might be addicted to or have a bad habit of use: alcohol, tobacco, some purely behavioral stuff. For each item, you could ask if the subject is already hooked, how readily he could quit if he was hooked, and how readily the average person could quit if he was hooked.

Rather than asking about "addiction" per se, you could ask to the effect of "Supposing you did this thing a lot, and you wanted to quit, how hard to do you think it would be to quit?". This avoids some definitional issues around addiction.

When asking about others, you could use the term "average person", but you could also try to use something a little more well-defined, like "a random person among those who take this survey". This makes it easier to say whether the sample is optimistically biased.