15b subliminal perception

subliminal

below the threshold of awareness

supraliminal

above the threshold of awareness

confabulation

making up plausible theories for why we do things

big picture for experiments studying subliminal perception

seems there is some evidence for subliminal perception but the effects are weak and finicky --not easily replicated

dichotic listening task

-*experiment studying subliminal message* -*studying by diverting attention from the subliminal message* -corteen and wood 1973 -*training phase*: participants listen to audio recordings of words - some of the words are paired w an electric shock -*test phase*: participants perform a dichotic listening task and subliminal messages are presented in the ignored stream --*attended stream*: story they have to pay attention to ----*verbally shadow* the attended stream (repeat what attended stream says) --*ignored stream*: words that had been paired during training as neutral or shock -*result*: when a prior shock word is played, galvanic skin response increases - suggesting subliminal detection of the word ----sweat glands activate (galvanic skin receptors/response increases) even though completely unaware that the words were being said in other stream

Wilson Bryan Keys experiment for subliminal advertising

-*method*: participants view 3 rum ads and choose which one looks most attractive -*result*: pick the one with "U Buy" embedded in it -WB Key interprets this as clear evidence of bias to the subliminal message -still problems with his message (due to hillyard principle) ---> the same ad contains the subliminal message for every participant ------->don't know if ppl choose the ad bc of the subliminal message or bc of other image properties ------>no base rate to see how often ppl picked that ad with NO subliminal images in it

exclusion paradigm results

-*supraliminal (showed cue for 500 ms)*: clearly see word, follow instructions well- either do or don't fill in prime word based on group ----exclusion: 10% ----inclusion: 96% -*subliminal (showed cue for 50 ms)*: ----exclusion: 50% ----inclusion: 63% ---more likely than chance to use/not use primed word -*baseline*: completion rate of word stem without a prime ----exclusion: 36% ----inclusion: 38%

3 methods to study subliminal perception

-*to study if there is any science that supports these claims* 1. present a weak stimulus 2. divert attention from the subliminal message 3. present the subliminal message very quickly and mask it

judas priest lawsuit (1985)

-2 boys attempt suicide on a playground (one succeeds and one fails) -parents play Judas Priest for it -Wilson Bryan Keys claims it is subliminal messaging (saying "Do It") in the song ---shows that subliminal messaging was really serious topic in 80s -the suicide victim's families lose the court case-- not enough evidence blaming music, family lives were also bad

lingerie experiment

-an experiment reviewed by Nisbett and Wilson -400 mall shoppers choose a favorite night gown and list reason for choice -the trick?: position of gown on rack (order counterbalanced) -*result*: participants pick the rightmost gown 4:1 ---ppl think it's bc they like the dress -*no one mentioned position as an influence on their decision*

electric shock experiment

-an experiment reviewed by Nisbett and Wilson -participants (all males) were asked how much more electrical shock they could take? -*key variable*: one grp receives strong reassurance of safety, and the other does not -*result*: makes no difference -participants assert that reassurance (or lack of) strongly affected their response

debate team experiment

-an experiment reviewed by Nisbett and Wilson -participants write an essay abt controversial issue (ex. racial integration) -two weeks later, they debate with persuasive confederate for the opposite opinion who changes their opinion --->confederate participant in on the experiment -**confederate tried to convince ppl to change ideas --->succeeded most of the time ---then when asked to recall original position, participants falsely recalled original position as closer to new opinion

exclusion paradigm experiment

-debner and jacoby 1994 -same priming technique as word-stem completion ---inclusion: complete word-stem with prime ---exclusion: do not complete word-stem with primed word -*competing predictions in exclusion condition*: ---if subjects are *aware of the prime*, this should *decrease* probability of completing the word stem with the primed word ---if subjects are *unaware of the prime*, this should *decrease* the probability of completing the word stem with the primed word

how do we study what influences our decisions?

-don't always know what influences our decisions -instead, do experiments that measure how ppl react to different things -study subliminal perception

subliminal self-help tapes

-greenwald et al 1991 -subjects listen to subliminal self-help tapes for memory or self esteem -but the labels to the tapes were randomly assigned -*result*: ----tape labels had strong effect on self-reported improvement ----actual tape identity had absolutely no effect -*likely a placebo effect*

hillyard principle

-he claims that when assessing attentional effects in visual perception, you must control the "stimulus energy" bw conditions -this allows you to conclude the effect was attentional, not due to differences in the images themselves -DOES NOT FOLLOW IN WB KEYS EXPERIMENT -Hillyard's way: would have conditions alternate bw ALL three of the ads that have the subliminal message ---->more grps that see subliminal images in the other ads too so you know its not just bc of the properties of that one picture

problems with semantic priming

-if the stimulus priming is too long, the participant might explicitly be aware of it (supraliminal) -if the stimulus is presented too briefly, it might be too weak to be subliminally perceived -*big picture*: supraliminal and subliminal perception would cause the same response pattern ---hard to make sure it's subliminal and not too supraliminal

james vicary

-in 1957, james vicary inserts advertisement into movie to encourage concession sales --ex. "Eat Popcorn" -claimed concession sales skyrocketed -later admitted it was a hoax -ppl still believed it *could have* been true -subliminal perception has a *suspect history*

wilson bryan key/subliminal seduction

-in 1973, wilson bryan key publishes subliminal seduction -he argues media and advertising are full of subliminal messages -goes on lecture circuit to many US universities -said it says sex in ice in the *whiskey* campaign with makes ppl want to buy it -said in *camel cigarettes* logo, there is a little man with phallus in the image which makes ppl want to buy it -released other books: media sexploitation, subliminal adventures in erotic art, clam-plate orgy -subliminal perception has a *suspect history*

choice blindness/ attractiveness ratings

-johannson et al. 2005 (newer study that goes along w nisbett and wilson) -have participant choose bw two faces -experimenter hands participant the "unchosen" card using slight of hand -participant is asked why they choose that face as more attractive (even though you didn't really pick that one) -*problem*: they didn't pick that face -*result*: participants often gave elaborate responses for why they picked that (wrong) face

backmasking experiment vokey and keys

-method: participants dicriminate various properties of backmasked audio -their performance is compared to chance probabilities (random guessing) -*results*: ----really good at detecting things about the backmasked voice (*voice 1 or voice 2 and what language*) ----bad at detecting semantic info for backmasking (*what the message says*) --makes sense bc can tell someone's voice without recognizing what it's saying

subliminal messaging in advertising

-moore (1982) does an extensive lit review of all studies on subliminal perception in advertising -*conclusion*: the literature is *full of flawed experiments* and nonreplicable results -strong evidence that *supraliminal* messages infl consumer behavior ---->don't need to use hidden message, up front messages are just/more useful

introspective access

-nisbett and wilson -we have little introspective access to our cognitive processes --ex. don't know how certain mental processes work -introspection just involves making up plausible theories (*confabulation*) to answer questions -from one of the most famous papers in psychology --->this article has been cited over 6000 times

top down perception

-perception driven by context, learning and expectation -sometimes allows us to see/hear things that are not present in the physical stimulus ---ex. we can read the statement perfectly even though the A is a misshapen H

subliminal perception

-perception without conscious awareness -unaware of stimulus, but detected or identified it in some manner -these stimuli could affect our subsequent behaviors or decisions

backmasking

-supposed subliminal messaging in pop music -when you play a song backwards and it contains a hidden message or sound -in the 1970s, various rock and pop artists are accused of doing this -promoting satanism, suicide, murder, drug use and sex

problem with backmasked messages

-the messages are not easy for listeners to recognize unless the supposed lyrics are in front of them ---*totally suggestible* -*top down perception*: perception driven by context, learning, and expectation -can make us see or hear things that are not present in the physical stimulus

presenting a weak stimulus (studying subliminal perception)

-way to study subliminal perception -present a stimulus that is below the absolute threshold for detection and test whether it affects subsequent behavior -*problem*: detection thresholds are so low that this is impractical for auditory and visual stimuli --ex. can detect candle from really far away, hear clock ticking from far away

semantic priming experiment

-way to study subliminal perception by presenting the subliminal message very quickly and mask it -*prime*: word (ex salt), brief flash, should still be unable to report it -*mask*: ### after prime word disappears -*word stem*: s _ _ _ ----have them try to complete it, say if it was a word or a nonword -*compare baseline* likelihood of filling in the word bw when *primed v unprimed* -*result*: people can't explicitly report the prime, but they are more likely to fill in the word stem with the prime