Key Study APFCC- Dement and Kleitman (sleep and dreams)- 1957

IV:
Aim 1) Whether awoken from REM (rapid eye movement) or nREM (non-rapid eye movement)
sleep
Aim 2) covariable: whether woken after 5 mins or 15 mins of REM
Aim 3) Type of eye movement
DV:
Aim 1) Whether ppts recalled dream or not and the detailed description of dream content
Aim 2) covariable: whether ppts estimated they were dreaming for 5 or 15 mins
Aim 3) content of dream

Aims
Overall: To investigate dreaming in an objective way by looking for relationships between eye
movements in sleep and dream recall
 To investigate whether dream recall differs between nREM sleep and REM sleep.
 To investigate whether subjective estimate of dream duration is correlated with length of REM
period before awakening
 To investigate whether eye movement patterns are related to dream content -> whether represent
visual material in dream or just random movements because active CNS)

Hypotheses (can be inferred but not explicitly given)

 There will be a difference between dream recall in REM and nREM
 There will be a significant positive correlation between length of REM and estimated dream
length
 Eye movement patterns will represent a visual experience of dream content

Background
 Sleep and dreams hard to investigate as ppt can’t communicate with researcher + only subjective
self-report data on dream content which may have low validity
 Electroencephalogram (EEG) – Measures electrical activity of brain produced when brain cells
communicate in order to measure brain activity
o Differences in signals detected between about 25 diff. electrodes attached to cap and fitted to
scalp w. conductive gel provide info on which brain regions most active. Each electrode will
display a series of lines (brain waves) which represent the sum of activation of brain cortex
just under that electrode.
o Amplitude (height of waves) shows brain wave intensity. Frequency (distance between each
wave) shows activation speed.
 Electrooculogram (EOG) – electrical recording of eye movement patterns (presence/ absence,
size and vertical/horizontal direction)
 Electromyogram (EMG) – records electrical charges associated w. nerve + muscle activity
 Circadian body rhythm repeats every 24hrs.
 REM – theta waves (low voltage, high frequency)
 nREM- delta waves (high voltage, low frequency) interspersed with K complexes (brief, high
amplitude spike in electrical activity) which are followed by sleep spindles (short-lived burst of
high amplitude, high frequency waves 1-2 second long)
 Psychology being investigated
 Aserinsky and Kleitman, 1955 – ppts awakened during periods of rapid eye movements had high
incidence of dream recall (based on subjective recall) and low incidence when awoken other
times. Eye movements at regular intervals relating to cyclical changes in depth of sleep. Found in
normal and schizophrenics.
 Sleep either nREM or REM
o nREM sleep has 3 stages
 Repair and growth
o REM (stage 5 sleep) occurs ~ every 90 mins
 Eyes move rapidly under eyelids
 Brain waves resembling awake stage – brain active
 Dreams (endogenous sensory, motor, emotional, and other experiences that seem like
actual events occurring)
 Difficult to be woken up from, paralysed, insensitive to stimuli

Procedures:
Sample:
 7M + 2F (adults) to begin with
 5 studied intensively while data from other 4 used to confirm the results. Self-selected
sample. Repeated measures design as all ppts put through all conditions of IV

Method:
1) Day prior to study ppts instructed to eat normally but not to not drink alcohol or caffeine on day
of experiment.
2) They arrived just before bedtime at lab then 2+ electrodes attached near eyes -> register electric
potential changes as eyes moved, 2/3 electrodes on scalp -> record brain waves
3) Slept in dark, quiet sleep lab w. wires gathered into single cord at back of head-> freedom-of-
movement while asleep, sending data to EEG (paper speed 3 or 6mm/s for eye movement and
3cm/s for brain waves
4) Ppts woken up by a loud doorbell placed near bed (because if people woke up the ppts ->
experimenter effects). Took < 5 mins to fall back asleep. Awoken either during REM/ time
increments after REM stopped (nREM) but ppts not told whether eyes had been moving.
 Awakenings fell pretty evenly in first 2, second 2, third 2 and fourth 2 hours.
 Total awakenings = 351 times/ 61 nights (average 5.7/night), w. breaks in between so not
constant nights
5) They were asked to speak into recording device whether or not dreaming and relate content of
their dream. Some ppts had to describe the contents of their dreams, which was recorded by a
tape recorder. ‘Dream’= if could give coherent, detailed description. Vague recall= ‘no dream’.
Experiment occasionally entered to question ppt further on particular point of dream but didn’t
do this most of the time (avoid demand characteristics)
Hypothesis 1:
 Participants PM and KC awoken acc. to random number table
 DN: 3 REM then 3 nREM
 WD: told he’d be awoken only when he would be dreaming but acc. awoken in REM/nREM
(random order)
 IR: acc. to whim of experimenter
 Hypothesis 2:
Asked how long they thought they were sleeping for (too difficult) so a series then done where
awoken either 5 or 15 mins after REM onset and asked to estimate if they were dreaming for 5 or
15 mins, then asked to say dream content and length of words of their narrative counted
Hypothesis 3:
Ppts awakened when 1 of main eye movement patterns (mainly vertical mainly horizontal,
vertical and horizontal, little/no eye movement) detected by EOG persisted for 1 min then to
describe dream content but also woken 10 times after 1 min little/no. To confirm relation
between eye movement and dream content, 20 naïve and 5 experimental ppts observed
distant/close-up activity awake w. electrodes around eyes

Findings
Hypothesis 1
 All ppts: high dream recall after REM (79%) awakening and low recall after nREM (7%)
awakening esp if nREM awakening in later part of night-> hypothesis 1 accepted
 WD not less accurate (but misled) + DN not more accurate (but pattern) -> weren’t
responding to demand characteristics
 Significantly more nREM dreams recalled within 8 minutes of REM than after more than 8
mins -> could be recalling dream from previous REM sleep
 When ‘not dreaming’ in nREM: EEG showed sleep spindles. When awoken during high
voltage, low frequency waves, bewildered, felt like dreaming, and reported pleasantness,
anxiety etc

Hypothesis 2
 Significant correlation between estimated length of dreams and number of words in dream
narrative and ppts mostly accurate in estimation of dream length -> hypothesis 2 accepted.
 88% ppts guessed 5 mins correctly and 78% guessed 15 mins correctly
 DN estimated his dreams as shorter -> responsible for large proportion of incorrect estimates

Hypothesis 3

 Eye movements related to dream content as if they represent visual imagery-> hypothesis 3
accepted
 Bursts of activity then inactivity
 Eye movement potentials measured in awake ppts similar to those when dreaming

How often and length of REM/nREM

 All 9 ppts showed periods of rapid eye movements every night in lab after initial onset sleep
low voltage, high frequency EEG
 1 REM period every 92 minutes for whole group. Variations: 70 -104 mins. Comparable to
previous study of uninterrupted sleep
  REM 3 - 50 mins (mean 20 min) long. Increase in length as night progressed.

Conclusions
 Dreams probably only occur during REM sleep as dreams from nREM from previous REM states
 Dream length increased proportionally as REM period increases -> dreams progress in ‘real
time’, not instantaneous events and ppts can estimate their dream length
 Eye movements during REM sleep correspond to where and what the dreamer is looking at in the
dream so they are not random events caused by activation of CNS.

Criticisms
Strengths
 Conducted under carefully-controlled laboratory conditions -> standardised procedure where
situational variables e.g light level, noise in which the ppts sleeping in to be controlled and wires
gathered in ponytail so they would not cause the participant to wake up/disrupt sleep -> increased
reliability of the study by allowing the full length of REM/nREM stages of sleep to be measured
in an uninterrupted manner to estimate the average length of these stages across the different
participants
 EEG and the EOG provide empirical data (changes in electric potential around eyes -> eye
movement patterns, brain waves during REM/nREM) -> minimises the effect of researcher bias
as there is not any room for subjectivity in interpreting the length of REM -> correlational
analysis of REM length and estimated dream duration to be calculated to allow objective
conclusions to be drawn.
 Supports nature (nature-nurture) debate as it provides evidence that the underlying biological
processes causing dreaming are particular voltages and frequencies of cortical activation ->
experimental method involving EEG can be used as an objective measure of dream duration and
REM/nREM -> can be used to diagnose and help people with sleep apnea, insomnia and
narcolepsy (real-world application)
 High generalisability -> study measuring eye movements and brain waves, which are generally
consistent across all humans -> despite small sample, link between REM length and dream length
has high population validity.

Weaknesses
 Self-report measures to gather qualitative data about the contents of participants’ dreams ->
confounding variable of the level of expressiveness of each participant influencing the number of
words they use to describe their dream making it appear longer/shorter or more detailed than in
actuality. This reduces the reliability of the conclusion that there was a significant positive
correlation between REM duration and the length of dreams.
 Potential beta bias -> unrepresentative sample of 7 males and 2 females where females are very
underrepresented -> sleep is mostly similar across members of the same species but the small
sample size may make results less generalisable to the overall population, so the research could
have low external validity and may not be representative of the way all female brains recall
dreams.
 Low ecological validity -> ppts who are used to drinking coffee/alcohol could have experienced
dreams/slept in a way that would be unusual for them under normal circumstances and sleeping
in sleep lab connected to EEG and under observation has low mundane realism -> the
conclusions drawn from this data on ppt brain waves and dream content may be less externally
valid