The Activation Synthesis Theory of Dreaming Research (OCR GCSE Psychology): Revision Note

Activation synthesis core study: Williams et al. (1992)

Background

• The study was based on the assumption that the strange, illogical nature of dreams is linked to the neurobiology of REM sleep

• Hobson & McCarley (1977) proposed that during REM:

  • the brainstem (pons) sends random, disconnected neural signals to the cortex

  • the cortex then attempts to synthesise these into a coherent story

  • this produces the bizarre content typical of dreams

• However, critics (e.g., Foulkes, 1985; Reinsel et al., 1992) argued that:

  • not all REM dreams are bizarre

  • REM dreams may be no more strange than waking fantasies

• Williams et al. set out to test whether dreams are genuinely more bizarre than fantasies, as activation–synthesis theory predicts

Aim

• To determine whether dreams contain more bizarre elements than waking fantasies

• To evaluate whether dream bizarreness arises from the random neuronal activation associated with REM sleep

Method

• Type of study: Natural experiment using a self-report method

  • Independent variable (IV):

    • Whether the report was a dream or a fantasy

  • Dependent variable (DV):

    • The bizarreness density score assigned to each report

• Design: Independent measures – dreams and fantasies were treated as separate conditions

• Sample:

  • 12 university students enrolled in a biopsychology course at Harvard University, USA

  • Two males and ten females aged 23–45 years

• Materials:

  • Participant dream journals

  • A detailed two-stage scoring system for bizarreness

  • Judges trained in the scoring method

• Procedure:

  • Over one academic term, participants kept written records of

    • all remembered dreams (morning/night recall)

    • spontaneous fantasies that were unrelated to external events

  • From all entries, researchers selected:

    • 60 dream reports

    • 60 fantasy reports

  • Each report was broken into one-sentence units

  • Units were scored using two stages:

    • Stage 1: Type of content

      • plot

      • thoughts

      • emotions

      • ad hoc (odd/unclassifiable content)

    • Stage 2: Type of bizarreness

      • Discontinuity

      • Incongruity

      • Uncertainty

      • Not bizarre

  • Three judges rated all 120 reports independently and blindly

    • Three independent judges scored all 120 reports blindly (unaware of conditions) to test inter-rater reliability

Results

• Inter-rater reliability was high (about 80% agreement) for bizarre and non-bizarre scoring

• Dreams had significantly higher bizarreness density than fantasies

• Dreams were more bizarre in terms of:

  • discontinuities

  • unusual characters

  • remote or shifting times/places

• Fantasies tended to involve:

  • familiar environments

  • first-person perspective

  • fewer bizarre elements overall

Conclusions

• The findings support the activation–synthesis theory

  • Dreams contain more bizarre features because they arise from random neural activation during REM sleep

• Fantasies were less bizarre because they occur while awake, when cognition is more organised and grounded in reality

• However, the researchers noted some cognitive overlap between dreaming and mind-wandering

  • This means that the boundaries between the states may not be entirely separate

• Overall, the findings suggest that:

  • dreaming and fantasising involve different cognitive processes

  • bizarreness in dreams reflects REM-related brain activity

Criticisms

• Self-report issues

  • Dreams and fantasies rely on memory recall, which may be inaccurate

  • Social desirability may lead participants to leave out embarrassing content

• Lack of control over the IV

  • Researchers could not control when dreams were recorded

  • Some fantasies may have been written while sleepy, blurring conditions

• Construct validity concerns

  • Dreams and fantasies are complex, yet researchers reduced them to numerical categories

  • Critics say this oversimplifies dream content

• Sample problems

  • The study included a very small sample

  • Mostly female, mostly psychology students, which means it is hard to generalise findings to wider populations