Williams et al

• year?

  1992
• aim?

  show that dreams are product of different brain activity which occurs in sleep
• hypothesis?

  bizarre content of dreams would be different from bizarre content of daytime fantasies as they're products of different brain activity
• method?

  natural experiment using self report
• IV?

  whether people were reporting dreams or fantasies
• DV?

  measure of how bizarre dreams and fantasies were
• sample?

  12 biopsychology students from American university aged 23 - 45
• how study was conducted? (1)

  * students were asked to keep dream and fantasy journal for one term

  * using these, researchers selected 60 dreams and 60 fantasies to analyse; ones they picked were more detailed and had more visual content

  * dreams and fantasies scored on 2 scales - one for locus and one for bizarreness per sentence
• how study was conducted? (2)

  * bizarreness density score calculated by dividing number of bizarre items by number of sentences

  * scoring done by 3 separate judges and this was checked for inter-rater reliability

  * judges didn't know if they were scoring dream or fantasy but asked what they thought it was after
• what study found?

  * 3 judges showed good inter-rater reliability

  * significant difference between scores for dreams and fantasies, with dreams being on average 3 times more bizarre

  * most significant difference between dreams and fantasies was on plot discontinuity

  * 7 out of 12 ppts had dreams with much higher bizarreness than fantasies

  * judges were 89% accurate on whether they were testing a dream or fantasy
• conclusions?

  * bizarreness of dreams correlates with brain activity during REM sleep which is why dreams contain more bizarreness

  * only some similarities between content of dreams and fantasies, but this was supposed to be bc of similarities in brain activity during REM sleep and during wake - sleep boundary
• criticisms?

  * sample was too small to generalise and 10 out of 12 ppts were female so could be gender biased

  * self report is inaccurate way of measuring findings

  * results lack construct validity bc qualitative data was turned into quantitative so complexity of dreams was lost when turned into numbers

  * IV was difficult to control bc no certainty that dreams that were described happened in REM sleep

  * differences between scores could be bc of how they were recorded - fantasies recorded not long after but dreams recorded longer after they happened