Psychology: Physiological Psychology: Relation of eye movements in sleep to dream activity

Psychology: Physiological Psychology: Relation of eye movements in sleep to dream activity

Author: Dement and Kleitman (1957)

Key terms: Sleep and dreaming

Background/Context:

• Scientific studies of sleep and dreaming increased after the invention of physiological techniques to measure brain activity which could indicate dreaming (EEG - Electro-encephalograph) and electrical recording of eye movement (EOG - Electro-oculogram).
• In 1953 Aserinsky identified REM (rapid eye movement) and NREM (non-rapid eye movement). There are 4 stages of NREM, with 1 being the lightest and 4 the deepest. 
• When we sleep, we alternate between REM and NREM sleep. 
• Previous studies show that people woken from REM sleep were more likely to report having a vivid dream than those in NREM. 
• This shows dreams most likely occur in REM sleep. 
• REM is similar to being awake as our brains are most active, and our eyes move behind their eyelids. However we are difficult to wake up during REM sleep and the brain releases signals to keep our muscles paralyzed so we do not move while dreaming.
• EEG (electro-encephalograph) detects and records tiny electrical changes associated with nerve and muscle activity, producing a chart that shows brain waves, which change with frequency and amplitude. 
• Brain waves during REM sleep are low voltage, high amplitude, whereas in NREM sleep they're high voltage and low amplitude.
• Nowadays EEGs are computerized, but in this study the graphs were drawn on continuous running paper. The faster the paper moved, the more detail could be recorded. Paper usually moved at 3mm - 6mm per second. 
• Same electrodes used for the EEG can be used to record eye movements when placed near the eyes. The output is called a EOG (electro-oculagram), which indicates the presence or absence of eye movements, the size and their direction. 

Sleep cycle graph

Aim/hypothesis:

1. Does dreaming occur in REM or NREM sleep?
2. Can participants accurately estimate the length of time they've been dreaming?
3. Do eye movement patterns match dream content?
   Sub-question:
4. Does duration of REM sleep correlate with number of words (description/narrative) in a reported dream?

Method:

• Aim 1, 3 and the sub-question (4) were naturalistic observations because REM and NREM sleep occurs naturally. 
• Aim 2 was a laboratory experiment because participants were woken up to see if they can estimate the time they were dreaming.

Variables:

• Aim 1: IV: REM and NREM sleep. DV: No. of dreams recalled in each state.
• Aim 2: IV: Woken after 5 minutes and woken after 15 minutes. DV: No. of correct dream length estimations. 
• Aim 3: IV: Direction of eye movement. DV: Description of dream (or no dream) when woken.
• Aim 4 (sub-question): DV: Total number of words used to describe a dream correlating with length of dream. 

Design: Repeated measures - All participants slept and woken up at various times.

Participants and sampling technique: 

• 9 adult participants in total: 7 male, 2 female.
• 5 main participants studied in detail
• 4 participants used to confirm data from main 5.
• Main 5 spent 6-17 nights in laboratory and were tested with 50-77 awakenings.
• 4 others spent 1-2 nights in laboratory and tested with 4-10 awakenings.
• Each patient identified with their initials.

Apparatus:

• Bed
• Electrodes connected to EEG
• Door bell
• Tape recorder

Controls:

• All participants asked not to drink alcohol (depressant) or caffeine (stimulant).
• All participants woken with a bell.
• All participants recorded their dream into a tape recorder.

Procedure:

1. During daytime participants ate normally, but didn't drink caffeine or alcohol.
2. Participants arrived at the laboratory before normal bedtime and were fitted with electrodes on the scalp (to record brain waves) and near the eyes (to record eye movements). 
3. They then went to a quiet, dark room and wires from the electrodes were gathered into a 'pony tail' from the person's head so they can move freely.
4. EEG and EOG would run continuously through the night to monitor participant's sleep stages and inform the experimenters when to wake the participant.
5. Participants were woken with a doorbell, so experimenters did not have to enter the room, therefore participants were all treated the same way.
6. Door bell would ring at various times during the night and when participants woke up they would say whether or not they had a dream, and if they did, they would describe their dream into a voice recorder. Descriptions were only considered a dream if it was a understandable, fairly detailed description. 

Participants had different conditions of REM and NREM wakenings, which were chosen by the experimenter:

• PM and KC were woken randomly to prevent possibility of unintentional pattern.
• WD also woken randomly, but told he would only be woken from dream (REM) sleep.
• DN woken in repeating pattern of 3 REM sleeps then 3 NREM sleeps.
• Waking of IR chosen by the experimenter.

Data:

• Quantitative: Instances of dream recall in each stage, dream length estimation and no. of words in dream description.
• Qualitative: Descriptions of dream.

Results

Aim 1: Dream recall in REM and NREM sleepResults from Aim 1's study

• Participants described dreams more often when woken from REM sleep and hardly in NREM sleep. 
• This was accurate for every participant. 
• WD's recalls were no less accurate despite being misled.
• DN was no more accurate even though he might have guessed the pattern of wakings.
• Recall of dreams during NREM sleep was much more likely if participant was awoken soon after end of REM sleep.
• Participants woken in NREM sleep described feelings such as anxiety or pleasantness but not specific dream content.

Aim 2: Estimation of dream length

• Asking participants to guess their dream time was too difficult, so instead they were asked whether they had been dreaming for 5 or 15 minutes, in which participants would respond more accurately.
• For participants who dreamt for 15 minutes, participants guessed right 88% of the time.
• For participants who dreamt for 5 minutes, participants guessed right 78% of the time.
• Only DN consistently underestimated dream length because he could only remember the end of his dream, making his 5 minute dream estimates accurate but not 15.

Aim 4 (sub-question - length of dream description correlating to duration of REM sleep):

• 152 dream narratives collected, but only 146 were able to be accurately transcripted. 
• 15-35 dreams per participant.
• Significant correlation between REM sleep and number of words, although affected by how expressive participant was.
• Dream descriptions for longer REM sleep were not much longer than dream descriptions for 15 minutes.

Aim 3: Eye movement patterns and dream content

• Participants were woken after 1 minute of continuous specific eye movement (vertical, horizontal, both of little movement).
• 3 out of 9 participants who had frequent vertical eye movements reported dreams about vertical movement.
• One participant dreamed about standing under a cliff, looking at various climbers on the cliff and down at his hoist machine. 
• A participant who showed horizontal eye movement had a dream about people throwing tomatoes at each other.
• The 10 times participants were woken after little or no eye movement reported having dreams of watching something in the distance or staring at a fixed object.
• In 2 cases participants had dreams about driving.
• The 21 wakings after mixed eye movement participants reported having dreams about looking at things or people nearby (instead of far away) eg. Fighting or talking to a group.
• Dement and Klietman also recorded participants eye movement when awake and discovered their waking eye movements when looking at nearby objects were similar to when dreaming of looking at nearby objects. 

Conclusions:

Aim 1: Vivid, visual dreams reported only from waking during or shortly after REM sleep. REM sleep is longer later in the night so dreaming is more likely at this time.

Aim 2: Dreams are not instantaneous events but actually happen in 'real time'.

Aim 4 (sub-question): The longer the duration of REM sleep, the more words are in the dream description.

Aim 3: Eye movements during REM sleep correlate to where and what the dreamer is looking at. Therefore eye movements are directly related to dream imagery and are similar to when we are awake.

Strengths:

• Laboratory experiment - High levels of control - No caffeine or alcohol, doorbell sound, EEG monitoring, different set waking for each participant - Easy to replicate to test for validity.
• Laboratory experiment - High levels of control - Easy to conclude IV directly affects DV (dream recall affected by stage of sleep).
• Physiological psychology - Most people are the same physiologically - Can be generalized to an extent.

Weaknesses:

• Laboratory experiment - Lacks mundane realism - Being woken multiple times and asked to verbally record your dream with electrodes on your head is not what usually happens in real life.
• Laboratory experiment - Lacks ecological validity - Sleeping in a laboratory with electrodes on your head is unnatural, therefore results can be invalid since they can be different in real life.
• Generalization - Only 5 participants were studied in detail, therefore the study cannot be generalized to the wider population.
• Reductionism - Findings based around biological reasons for dreams - Does not take into account psychological reasons that could affect dream content.