Content uploaded by Torsten Norlander
Author content
All content in this area was uploaded by Torsten Norlander on Feb 21, 2018
Content may be subject to copyright.
Norlander, T., Bergman, H., & Archer, T. (1998). Effects of flotation REST on creative
problem solving and originality. Journal of Environmental Psychology, 18, 399-408.
Effects of Flotation REST on Creative Problem Solving and
Originality
Torsten Norlander, Karlstad University, Sweden,
Henrik Bergman, Örebro University, Sweden,
Trevor Archer, Göteborg University, Sweden
Running head: Floating and Creativity
--------------------------------------------------------------------------------------------------------------
Authors note: This study was supported by grants from Adlerbert Research
Foundation (ARF), Göteborg and from Sport Development Center (IUC), Örebro.
Correspondence should be addressed to Dr. Torsten Norlander, Department of
Psychology, Karlstad University, 651 88 Karlstad, Sweden. Email:
at.norlander@mailbox.swipnet.se
Floating and Creativity
2
Abstract
The purpose of the present study was to explore whether or not flotation REST
facilitates the creative problem-solving ability and originality. Sample 1 had 40 subjects, 20
men and 20 women, whom were randomly assigned in equal numbers to either a non-REST
group (armchair-sitting) or to a floating REST group. Both groups worked on a “Chain
Puzzle“ for 5 minutes and were then interrupted with 45 minutes sitting or floating. Then the
subjects were given the task of continuing with the creative problem-solving test. Sample 2
had 54 subjects, 27 men and 27 women, whom were randomly assigned in equal numbers to
either a “non-REST“ (armchair-sitting), a “dry-REST“ group (lying on a couch in a dark
room) and a “flotation-REST“ group. Then the groups had to fill in a couple of paper-and-pen
tests giving scores on fluency, obvious answers, original answers, elegance and deductive
thinking. The results (impaired creative solving ability and higher originality for the floating
group) were interpreted as a indication of a cognitive function where the primary process still
dominates over the secondary process.
Key words: Flotation, REST, Creativity, Problem-solving, Originality, Primary process,
Secondary process, Alcohol
Floating and Creativity
3
Introduction
The flotation-isolation technique, which is the purpose of this investigation to apply,
is a form of sensory deprivation, or as expressed by the more recent concept “Restricted
Environmental Stimulation Technique“ (REST). Earlier conducted experiments with REST
often reported of negative effects (Zubeck, 1973) such as problems with logical thinking,
concentration problems, higher levels of anxiety and even hallucinations. Subsequent
research has demonstrated that even positive effects may occur, such as improved relaxation
(Jacobs, Heilbronner, & Stanley, 1984), less anxiety (Fine & Turner, 1982), increased
receptivity to information (Aquino, 1982), pain reduction (Turner & Fine, 1984), heart rate
reduction (Jacobs, Heilbronner, & Stanley, 1984), and muscular tension reduction (Stanley &
Francis, 1984). Corresponding negative and positive effects are reported for the floating form
of REST (e.g. Best & Suedfeld, 1982; Forgays & Belinson, 1986; Hutchison, 1984; Suedfeld
& Bruno, 1990; Suedfeld, Metcalfe & Bluck, 1987). Several studies (e.g. Lee & Hewitt,
1987; McAleney, Barabasz & Barabasz, 1990; Suedfeld & Bruno, 1990; Wagaman, Barabasz
& Barabasz, 1991) offer support for the premise that flotation-isolation may reinforce aspects
of ‘imagery’, possibly a prerequisite to the creative process.
Flotation REST has been shown to affect creative performance. For instance,
Suedfeld, Metcalfe & Bluck (1987) and of Forgays & Forgays (1992) are two studies directly
dedicated to creativity. Unfortunately, both these studies suffer from concomitant
methodological problems, as described by the authors. Concerning the study of Suedfeld et al.
(1987) there were only seven subjects, including two of the authors of the report. The result
however indicated that subjects, according to their own reports, experienced "better" new
ideas directly following floating than simply after sitting in their offices. Forgays & Forgays
(1992) lost 25% of their subject population due to a procedural problem in connection with
data collection. This proportion of “missing data” is disconcertingly large in view of the
small, though significant, differences obtained (i.e. the floating group succeeded better than
the control group) that could be measured with a divergent test that measured fluency.
Floating and Creativity
4
Another problem with the latter study pertains to the fact that the authors allowed their
subjects to, amongst other things work with a difficult convergent problem, a so-called
"brain-teaser", during the ongoing isolation. The risk with this procedure is that these
subjects, due to this mental activity, were preoccupied with the largely irrelevant "analytical"
stimuli, thereby being retained within the secondary process; this situation presupposes a
radically different experimental set-up to that ordinarily considered during floatation REST.
The capacity of alcohol intake to alter the balance between the primary and secondary
processes (Norlander, 1997) may result in both positive or negative effects for the creative
process, as for example an heightened originality during intoxication (Hajcak, 1976;
Norlander & Gustafson, 1998) or directly after intoxication (Norlander & Gustafson, 1996),
improved restitution (Gustafson & Norlander, 1995), decreased deductive ability (Gustafson
& Norlander, 1994), less persistent effort (Gustafson & Norlander, 1994), a deterioration in
flexibility (Norlander & Gustafson, 1998) and thereby a deterioration in problem-solving
ability (Hajcak, 1976). Further, a decrease in handicraft-efficiency have been observed
(Norlander & Gustafson, 1997). One interesting question in this context would then be
whether or not there may exist other and more suitable techniques, than alcohol use, in order
to induce the "shifts of balance", or other combinations, between the primary and secondary
processes. These processes may be described as follows: primary process involves autistic,
free-associative and analogical thinking, referring to fantasy, reverie and “day-dreaming”,
whereas secondary process involves the abstract, reality-oriented thought of daily
consciousness (Martindale & Dailey, 1996). Several established techniques for inducing
‘shifts’ (Larsson, 1987a; Larsson & Starrin, 1988; Setterlind, 1990) are available: hypnosis,
neuromuscular relaxation, autogenic training, meditation, biofeedback-training, etc. Another
effective technique, given the right circumstances, has been found to be long-distance
running/jogging (Larsson, 1987b). One problem with the different forms of relaxation
exercises remains the complication that those individuals most in need of this training have
the most serious problems in relaxing. Consequently, these individuals have problems
initiating and/or maintaining the exercises (Maslach, 1988). The “floating-technique” may
show potential to be a successful application since it is claimed to produce an immediate and
Floating and Creativity
5
deep relaxation.
The purpose of the present study is to explore whether or not flotation REST
facilitates the creative problem-solving ability and originality. The expected ability of this
procedure to weaken the secondary process thereby allowing the primary process to remain
undisturbed should produce a similar result to that observed under the influence of alcohol,
i.e. the creative problem-solving ability (mainly secondary process-oriented) should
deteriorate whereas originality (mainly primary process-oriented) should be reinforced.
Further, as a control factor the Logical Deductive Test (secondary process-orientation) should
also deteriorate with flotation. “The Cheap Necklace Problem“(Silveira, 1971; Best, 1995)
was applied as a measure of creative problem-solving ability. The type of creativity that
different versions of “Chain Puzzles“ (where the idea is cutting and welding a minimum
number of links in order to create different patterns) measure is usually defined as “adaptive
flexibility“ (Guilford, 1967). Such tests involve a large amount of “trial and error“: “Failing
in one attempt, E must revise his tactics and attempt another approach. If he is too firmly set
on one approach, he is handicapped in doing the test.“ (Guilford, 1967, p. 152). In order to
solve a Chain Puzzle a large proportion of logical thought is required; This would not be
sufficient if one lacked the ability to view the problem from alternative perspective, i.e. an
“intuitive leap” (Guilford, 1967). Thus, different versions of Chain Puzzles creative problem-
solving tests offer different estimations regarding the dimensions of ability. As a measure of
originality a technique involving the production, by each subject, of as many associations
pertaining to various series of events, was assigned. Originality is defined as the “ability to
produce responses that are statistically rare in the population“ (Guilford, 1967, p. 154). As a
control condition a Deductive Test, derived from Holmquist (1974), was applied as a measure
of logical-deductive thought.
Method
Experiment I
Subjects. 40 subjects, 20 men and 20 women were recruited at the University of
Örebro. The mean age was 22.03 years (SD = 3.23 years, range = 19 to 31). They were all
Floating and Creativity
6
Swedish born. Subjects were randomly assigned in equal numbers (10 men and 10 women)
to either a “non-REST“ (group 1) or to a “flotation-REST“ (group 2). There were no
(ANOVA) significant differences (ps > 0.1) between groups with regard to age, weight or
resting-pulse (taken at home by the subjects themselves, both in sitting and lying down
positions). There were no significant differences (Mann-Whitney U Test) between groups
with regard to civil status, consumption of alcohol, breakfast and luncheon habits, amount of
training, health status, cultural background, use of tobacco, interest in sports or level of
education (ps > 0.07). There was a significant difference (Mann-Whitney U Test) between
groups with regard to their background growing up conditions (p < 0.01). The subjects in the
control group had a more provincial background as opposed to the floating group.
Design. The control group (group 1) worked on a creative problem for 5 minutes and
then they reclined in an armchair for 45 minutes during which time they were allowed to read
magazines which were laid out for them (sitting). Finally they were allowed to continue
working on the creative problem either until they had solved the problem or until 30 minutes
(5 minutes before sitting and 25 minutes after sitting) had passed. Subjects in the control
group had a “pre-sitting“ with the pulse measuring instrument about one week before the
designated time for the experiment. No psychological testing was conducted at this point of
time.
The floating group (group 2) worked with the same creative problem as the control
group but after 5 minutes received a 45-minute floating session. Then they were allowed to
continue the creative problem-solving (Chain Puzzle) either until a total of 30 minutes (5
minutes before floating and 25 minutes after floating) had passed or until they had completed
the task. To allow the subjects to become accustomed to the floating situation and the pulse
measuring instrument each was called upon to take a floating session about one week before
the designated time for the experiment. No psychological testing was conducted at this point
in time.
Subjects who managed to solve the creative problem before sitting-time or floating-
time (5 men, 1 women) were consequently removed from the experiment, having solved the
problem before the experimental manipulation. For group 1 (the non-REST group) this
Floating and Creativity
7
caused a subject loss of two subjects and from group 2 (the floating group) four subjects. For
the statistical analysis there remained a total of 34 subjects, 18 subjects in the control group
(8 men and 10 women) and 16 in the floating group (7 men and 9 women). A personality test
measuring creative attitude FS (Holmquist 1986), was administered to all subjects. Test
results were converted to stanine scores using a norm table (Holmquist 1986) created from
industrial employees. This gave an opportunity to compare the 34 subjects with a larger
population (M = 5.15; SD = 1.16). A two-way ANOVA showed no significant differences
between groups [F(1, 30) = 0.58, p = 0.45] or between sexes [F(1, 30) = 3.56, p = 0.07].
There was no interaction between groups and sexes [F(1, 30) = 0.03, p = 0.87].
Instruments. (1) A floating tank (Pantarei Flytexperten, Stockholm) with
measurements 2400 mm x 1260 mm x 950 mm was used. Water depth varied between 200-
300 mm, due to evaporation, volume of water 1 cubic meter. The floating tank was insulated
on the inside so as to maintain a constant temperature and to isolate the subject from sound
and sight. The water temperature was maintained at 34.2oC. Ambient (air) temperature was
the same as the water temperature minimizing sense sensation. The water was saturated with
magnesium sulfate in order to maintain a salt concentration of 1.3 g/cm3.
(2) Pulse measuring instrument (polar Sport tester) was used throughout the entire
duration of the experiment to measure subjects’ heart rate. It consisted of an electrode belt
fastened over the chest by a wide rubber band. It functioned as a radio-transmitter. The
receiver was a clock which received and stored the information. Both the receiver and the
transmitter were water tight to a depth of 1 atmosphere. To reduce disturbance between the
receiver and the transmitter in the salt water the receiver was fastened to the electrode belt.
The information stored in the clock/receiver was then entered into a PC computer. The
subjects’ pulse was registered every 60 seconds.
(3) FS - change and stability (Holmquist, 1986). A test used to measure attitude to
change and stability and which correlates high with several creativity tests, was used and
administered to all the subjects at the beginning of the experiment, i.e. before the
manipulation.
(4) Silviera´s "cheap necklace problem" (Silveira, 1971; Best, 1995) was
Floating and Creativity
8
administered to test the creative problem-solving ability. The links in the chains can be
opened and closed. From 4 small chains with 3 links in each chain, the subject is required, by
opening and closing the links, to build one circle of chains which costs no more than 15
crowns. To open a link costs 2 crowns and to close a link costs 3 crowns. In order to help the
subjects to further visualize the problem, the subjects had the possibility to borrow a box
containing one-crown coins. The subject had two separate opportunities to work on the
problem. At the first opportunity the subject was allowed to work 5 minutes only, whereby
the subject was then interrupted. At the second opportunity the subjects were allowed to
work, if needed, in 25 minutes at the conclusion of which the experiment was terminated.
Procedure. Directly on arrival subjects were informed that they could terminate the
experiment at any time, for any reason. Thereafter, the subjects were instructions to put on the
pulse measuring instrument. Subjects belonging to the floating group were instructed to
change into there swimming costumes, and a bathrobe before the experiment was started.
Then, the subjects were informed of the proceedings for the next 100 minutes, i.e. first they
were required to try to solve a problem for five minutes, then sit for 45 minutes (non-REST
group) or float for 45 minutes (flotation-REST group), and finally to try and solve a further
problem within a short time period.
Then the instructions were given for the cheap necklace problem as follows.
“In front of you are 4 small chains each consisting of 3 links. It costs 2 crowns to open a link
and 3 crowns to close a link. All the links are closed at the beginning of the test. Your task is
to connect all 12 links into one continuous circle, or if you so wish into a necklace, at a toal
price of no more than 15 crowns. If you think it may help, you may deduct and add coins
from the box while you work with the task.“
After five minutes, the subjects were interrupted with the problem- solving task.
Subjects belonging to the non-REST group were instructed to sit in an armchair and to read
the laid out magazines or any literature they may have brought with them for a duration of 45
minutes. Subjects belonging to the flotation-REST group were instructed to visit the
bathroom and then take a shower. Furthermore, they were instructed to thoroughly dry their
faces so as to avoid irritating droplets of water that might have disturbed their relaxation in
Floating and Creativity
9
the tank. To avoid the possibility that the subjects may be thinking about the creative
problem-solving assignment while floating, they were given relaxation instructions according
to Benson (Benson, 1975 p. 159-163). The subject was then informed of how to remove the
cover and how to climb in and out of the tank. So as not to be disturbed by the water and to
further eliminate sounds, wax plugs were inserted into the subjects ears. Thereafter, the
subject entered the tank.
After 45 minutes sitting (in the armchair) and/or floating (in the tank) subjects were
allowed to leave the armchair/tank, respectively, and given the task of continuing with
Silveira´s cheap necklace problem. If the subject had not solved the problem within 25
minutes (after the break) the experiment was terminated. The time was noted. A manipulation
check after the completion of the experiment was conducted to control whether or not the
subject had consciously thought about the experiment while sitting or floating. Subjects were
asked if consideration of the problem had occurred in their thoughts, and if so, how many
times. The answer was written down.
Experiment II
Subjects. 54 subjects (27 men and 27 women) were recruited at the University of
Örebro. The mean age was 23.24 years (SD = 3.50 years, range = 19 to 33). They were all
Swedish born. Subjects were randomly assigned in equal numbers (9 men and 9 women) to
either a “non-REST“ group (group 1), a “dry-REST“ group (group 2) and a “flotation-REST“
group (group 3). There were no significant differences (one-way ANOVAs) between groups
with regard to age, height or weight (ps > 0.18), but there was a significant groups difference
with regard to how many academic terms the subjects had attended at the university [F(2,53)
= 6.62, p < 0.01]. A post hoc test (Scheffé’s) showed that group 3 (M = 4.61, SD = 2.12)
differed from both group 2 (M = 2.22, SD = 2.65) and group 1 (M = 2.17, SD = 2.09). There
were no significant differences (Kruskal-Wallis) between groups with regard to consumption
of alcohol, breakfast and luncheon habits, amount of training, health status, cultural
background, growing up conditions, use of tobacco, interest in sports or education (ps >
0.26), but there was a significant difference with regard to civil status (p = 0.02) where group
Floating and Creativity
10
3 (Mean Rank = 20.5) differed from group 2 (Mean Rank = 31) and group 1 (Mean Rank =
31). Group 3 had more singles compared to the other two groups. A test was administered to
all subjects to provide more background information. The test measured prevailing attitude-
to-creativity with respect to change and stability, namely the FS (change and stability) - test
(Holmquist, 1986). A two-way ANOVA indicated no differences between groups (p = 0.51)
and no differences between sexes (p = 0.42). The FS scoring results were transformed into
stanine following a norm from industrial employees, both workers and officials (Holmquist,
1986), thereby making possible a comparison to a broader population (M = 4.7, SD = 1.59).
Design. The flotation-REST group (group 3) received a 45-minute floating session,
thereafter they had the assignment to conclude three psychological paper-and-pen tests. To
accustom the subjects to the floating situation they were instructed to partake of four floating
sessions before the designated time for the experiment.
The dry-REST group (group 2) received exactly the same treatment as group 3 before
and during the experiment, but instead of floating they spent 45 minutes resting on a couch in
a light- and sound-proofed chamber. The non-REST group (group 1) received exactly the
same treatment as group 3 before and during the experiment, but instead of floating they sat
in an armchair for 45 minutes and were allowed to read the magazines which were laid out
for them (sitting).
Instruments. (1) A floating tank (Pantarei Flytexperten, Stockholm) with
measurements 2400 x 1260 x 950 mm was used. Water depth varied between 200-300 mm
according to evaporation, volume of water was 1 cubic meter. The floating tank was insulated
on the inside so as to maintain a constant temperature and to isolate the subject from sound
and sight (see above). Water and air temperature were maintained at 34.2oC, as for
Experiment I. Magnesium sulfate saturation maintained the concentration, 1.3g/ cm3, as
above.
(2) FS - change and stability. A test (Holmquist, 1986) used to measure attitude to
change and stability and which correlates high with several creativity tests, was used and
administered to all subjects at the beginning of the experiment, i.e. before the flotation
manipulation.
Floating and Creativity
11
(3) Syllogisms I. The test (Holmquist, 1974) measures the ability of logical and
deductive thinking. The Syllogism I test consists of 21 items demanding a quantitative type of
deduction reasoning (e.g. Tom is taller than John. John is smaller than Bill. Is John taller than
Tom?). Time was limited to 5.5 minutes and the number of correct answers were registered.
(4) FREGO. The test (Holmquist, 1973) is a questionnaire which has been constructed
for measuring the variable ’preconscious activity’. It is a translation from the North American
Preconscious Activity Scale (Holland & Barid, 1968). The reasoning behind the test is that
high ’preconscious’ activity stimulates new and unusual associations. The FREGO test
consisted of 36 items assigning the subject yes or no answers to statements (e.g. I would
rather be a politician than a scientist).
(5) Finally, each subject received the assignment of producing as many consequences
as possible to six dramatic events, all with some connection to the Nordic situation (e.g.
“What would happen if there suddenly was a new ice age in Northeurope?”). Time was
limited to two minutes per item, which gave a total of 12 minutes. The fluency score was
obtained through a simple summation of all the responses, excluding all iterations. A panel of
judges (consisting of two high-school teachers) was constituted and was required to assess
responses as being either obvious or original on the basis of quality. In addition, each judge
was required to search out from the responses from each subject that which was considered
most elegant and to adjudge this on a ten-point scale. In the event of a lack of any response
being considered elegant the subject was assigned no points. “Obvious“ was defined as high
frequency response and “original“ as a rare response in large populations and “elegance“ as
"an object/enterprise achieves elegance when two or more apparently opposing components
are combined to one harmonious unit. The greater the initial opposition between the
components involved the more elegant the solution or 'end-point' is experienced" (Norlander,
1997, p. 9).
Procedure. Directly on arrival at the laboratory the subject was informed that s/he
could terminate the experiment at any time, for any reason. If the subject belonged to the
floating group s/he was instructed to change into her/his swimming costume, and a bathrobe
before the experiment was started. Then the subject completed the FS-test and filled out a
Floating and Creativity
12
questionnaire about her/his background. If the subject belonged to the non-REST group s/he
was instructed to sit in an armchair and to read the layed out magazines or any literature they
s/he have brought with for a duration of 45 minutes. If the subject belonged to the dry-REST
group s/he was given relaxation instructions according to Benson (Benson, 1975 p. 159-163)
and the instruction to practice these exercises while relaxing on the pallet. Each subject was
given wax ear plugs to place in their ears. After this, the door to the closed with the subject
left alone in the darkened chamber. Finally, if the subject belonged to the flotation-REST
group s/he was instructed to visit the bathroom and then take a shower. Furthermore s/he was
instructed to thoroughly dry her/his face so as to avoid irritating droplets of water that might
have disturbed the relaxation in the tank. Thereafter the subject was given relaxation
instructions according to Benson (Benson, 1975 p. 159-163) and the instruction to practice
these exercises during floating. So as not to be disturbed by the water and to further eliminate
sounds, wax plugs were inserted into the subjects’ ears. Thereafter the subject entered the
tank.
After the 45 minutes of non-REST, dry-REST or flotation-REST the subject had to
complete the FREGO and Syllogisms I tests and the six items with the dramatic events. The
order of these tests were randomized.
Results
First sample
(a) Test time. A two-way ANOVA showed a significant difference between groups
[F(1,30) = 4.75, p = 0.04], where the floating group used more of the available test time (M =
27.07, SD = 6.08) than the control group (M = 21.00, SD = 10.16). There were no differences
between sexes (p > 0.53), neither was there any interaction between group and sex (p = 0.28,
for means and standard deviations, see Table 1).
___________________________________
insert Table 1 about here
____________________________________
Floating and Creativity
13
(b) Hits and misses. The number of correct reported solutions within the given test
time (hits) was noted as well as the number of incorrect solutions (misses). Results showed
that group 1 (control group), had 9 hits and 9 misses. Group 2 (floating group) had 6 hits and
10 misses. Mann-Whitney U Test showed no significant differences between groups (p =
0.47) or between sexes (p = 0.47).
(c) Pulse measurements. During the entire experiment pulses were registered from
subjects both in the control and floating groups. To simplify further statistical analysis the
registered pulse values were reduced to the following eight categories: (1)-(4) mean pulse
rate, standard deviation for pulse rate values, highest pulse rate value, and lowest pulse rate
value, respectively, for sitting and floating, (2) during sitting and floating, (3) during
sitting/floating, (4) during sitting respective floating, (5)-(8) mean pulse rate, standard
deviation for the mean pulse rate, highest pulse rate value, and lowest pulse rate value
during the resumed work with the cheap necklace problem. A MANOVA (Pillais) where the
eight categories were the dependent variables and group and sex the independent variables,
showed no significant differences for the eight categories with regard to group (p > 0.09), sex
(p = 0.5), neither was there any interaction between groups or sex (p = 0.78). There was a
significant difference (Wilcoxon Matched-Pairs) between the pulse while sitting/floating and
the pulse having resumed the work with the problem (p = 0.02). Further analysis (Wilcoxon
Matched-Pairs) showed no difference for group 1 (p = 0.81) but there was a difference for
group 2 (p < 0.01). The pulse mean for group 1 during sitting was 67.79 (SD = 8.81) and for
group 2 during floating was 66.44 (SD = 10.26). The pulse mean during the resumed work in
group 1 was 69.34 (SD = 9.44) and in group 2 it was 73.68 (SD = 9.44).
Correlation analysis (Pearson’s r) showed a negative correlation (r = - 0.50, p = 0.008)
between test time and the subjects’ standard deviation for mean pulse rate during the resumed
work on the problem, indicating that the less time used the higher the standard deviation.
Further analysis (Pearson’s r) showed a stronger negative correlation between test time and
the standard deviation for the mean pulse rate during work in group 1 (r = - 0.72, p = 0.005),
but no significant correlation for group 2 (r = - 0.38, p = 0.33).
Manipulation control. Subjects were instructed to report how often during
Floating and Creativity
14
floating/sitting they had consciously thought about the necklace problem. Of the subjects in
the control group 10 reported that they had not thought about the problem whatsoever, while
8 reported having once thought about the problem. Of the subjects in the floating group 6
reported not having thought about the problem while 10 reported having once thought about
the problem. A Mann-Whitney U Test showed no differences between groups (p > 0.29)
regarding the amount of conscious thought on the problem during floating/sitting.
Second sample.
Interjudge Reliabilities. The correlation statistics (Pearsons’ r) showed in regard to
fluency, obvious, original and elegance significant (p < 0.05) values between the scores given
by the two judges: fluency (r = 0.99, p < 0.01), obvious (r = 0.85, p < 0.01), original (r = 0.40,
p < 0.01), elegance (r = 0.57, p < 0.01). With this background analysis it was determined as
meaningful to summate the judges scoring on the four variables for further statistical analysis.
Dependent variables. A Pillais’ MANOVA (3 x 2 factorial design) with group and sex
as independent variables and the scores from the FREGO and Syllogisms I tests together with
the averaged scores from the two judges on “fluency“, “obvious“ “original“ and “elegance“ as
dependent variables were used. The analysis showed no interaction between group and sex (p
= 0.5) ) and no significant difference in regard to sex (p = 0.61), but there was a significant
group difference (p = 0.04). An ANCOVA controlling for the differences of the amount of
terms the subjects had attended at the university yielded no other significant indications.
Described below are: for each dependent variable the results from the univariate F-tests with
regard to group and sex (there was no interaction effect between group and sex, for means
and standard deviations see Table 2).
(a) Fluency. Univariate F-test showed no significant difference between groups (p =
0.63) and no difference with regard to sex (p = 0.86).
(b) Obvious. Univariate F-test showed no significant difference between groups (p =
0.92) and no difference in regard to sex (p = 0.96).
(c) Original. Univariate F-test showed a significant difference between groups
[F(2,48) = 4.11, p = 0.02] but no difference with regard to sex (p = 0.33). A post hoc test
Floating and Creativity
15
(Scheffé’s) showed that group 3 (M = 8.53, SD = 3.48) had higher scores on original
compared to group 1 (M = 5.64, SD = 2.98) but not to group 2 (M = 7.33, SD = 2.57).
(d) Elegance. Univariate F-test showed no significant difference between groups (p =
0.48) and no difference in regard to sex (p = 0.70).
(e) FREGO. Univariate F-test showed no significant difference between groups (p =
0.92) and no difference in regard to sex (p = 0.84).
(f) Syllogisms I. Univariate F-test showed no significant difference between groups
[F(2,48) = 2.43, p = 0.09] and no difference in regard to sex (p = 0.86). Even though there
was no significant group difference a slight tendency to a trend could be observed: group 1
(M = 14.67, SD = 5.60), ), group 2 (M = 13, SD = 4.86) and group 3 (M = 11, SD = 4.43).
____________________________________
insert Table 2 about here
____________________________________
Discussion
The present study produced three main results:
1. The subjects in the floating group in the first experiment used more test time (latency) than
the non-REST group.
2. The shorter the test time latency by the subjects in the first experiment required for the
creative problem solving test, the greater the heart rate variability obtained.
3. In the second experiment, subjects in the floating group were scored as more original in
comparison with the dry-REST and non-REST control groups.
The finding that the results of the floating group from the first experiment were not as
high as the control group on the creative problem-solving test seems to indicate that a residual
floating effect must be considered, even after the subject has left the immersion tank. This
effect may possibly be characterized by cognitive functioning wherein the primary process
still dominates over the secondary process and where the ability of making shifts between the
two processes appears impaired/retarded. One support for this suspicion is provided by the
second main result of this study which coincides with certain other experiments, as for
Floating and Creativity
16
instance Bowers & Keeling (1971), who found that subjects scoring high on creative tests had
greater variance of heart rate compared to those who produced low results. With support of
Lacey (1967) Bowers and Keeling concluded that the greater heart rate variability, observed
in more creative subjects, indicated that they had rapid cognitive shifts between reality-
oriented cognition and more imaginal inner-directed ideation. Following a comprehensive
perusal through studies with physiological correlates of creativity, Suler (1980), drew the
conclusion that: “Creative people should possess an above average ability for shifting
between states of low arousal associated with primary process and higher states of arousal
characterized by secondary process.“ (Suler, 1980, p, 158). It is within reason to suggest that
adaptive flexibility, which may be measured with different forms of chain puzzles, demands a
higher propensity to cognitive shifts. It seems relevant that the control group, i.e. that
producing the best results on the test, provided the highest correlation between shortest test
time and greater variance in heart rate.
A third support for the notion of a primary process dominated floating effect directly
after the actual floating is indicated by the third main result in this study, i.e. the subjects in
the floating group of the second experiment were scored as more original compared to the
dry-REST and non-REST groups. To reiterate, it seems as if originality may be connected to
the primary process in much the same manner as problem-solving may be connected to the
secondary process. On the other hand, the second experiment poses a complex result insofar
as deductive ability (secondary process-oriented) was not influenced negatively by the
floating procedure, even though one may discern a weak, non-significant trend. One possible
explanation focuses upon the facilitation of visual imagination through floating. A more
abstractly constructed deductive test may have provided a more explicit deterioration. In
general, it appears that the balance/combination between primary and secondary processes,
during a creative problem-solving test or a originality test, may be influenced by floatation
REST in much the same way as was obtained for alcohol intoxication (Norlander 1997).
An important aspect for future studies concerning flotation REST and tests of
creativity pertains to the more stringent control of expectancy factors. In this regard, it may be
argued that both positive and negative expectations of floating may have induced some
Floating and Creativity
17
impact upon the results. In none of the three studies directly concerned with floating and
creativity, i.e. Suedfeld et al. (1987), Forgays & Forgays (1992) and the present study, has
sufficient consideration been made for the putative role of expectancy factors.
Generally, flotation REST experiments maintain only one control group, i.e. an
armchair group or a couch group. In the present circumstance we have developed a double-
control group design (the DCG-design) maintaining both a NONREST and a DRYREST
condition. This type of design resembles the condition which has been considered in terms of
“double-placebo” and/or “balanced-placebo” designs now often applied in studies on the
effects of alcohol and other drugs on states of consciousness (cf. Marlatt et al., 1973).
References
Arieti, S. (1966). Creativity and its cultivation: relation to psychopatology and metal health.
In S. Arieti (ed.) American Handbook of Psychiatry, vol 3. New York: Basic Books.
Arieti, S. (1967). The Intrapsychic Self. New York: Basic Books.
Aquino, C.C. (1982). Relationships between stimulus deprevation theory and creative
communications. Journal of Creative Behavior, 16, 123-131.
Benson, H. (1975). The Relaxation Response. New York: William Morrow and company.
Best, J.B. (1995). Cognitive Psychology. Minneapolis/St. Paul: West Publishing Company.
Best, J.A. & Suedfeld, P. (1982). Restricted Environmental Stimulation therapy and
Behavioral self-management in smoking cessation. Journal of Applied Social Psychology, 12,
408-419.
Bowers, K. & Keeling, K. R. (1971). Heart rate variability in creative functioning.
Psychological Reports, 29, 160-162.
Fine, T.H. & Turner, J.W., Jr. (1982). The effect of brief restricted environmental stimulation
therapy in the treatment of essential hypertension. Behavioral Research Therapy, 20, 567-570.
Forgays, D.G. & Belinson, M.J. (1986). Is flotation isolation a relaxing environment? Journal
of Environmental Psychology, 6, 19-34.
Forgays, D.G. & Forgays, F.K. (1992). Creativity enhancement through flotation isolation.
Journal of Environmental Psychology, 12, 329-335.
Floating and Creativity
18
Freud, S. (1923), The ego and the id. In: Complete Psychological Works, standard edition,
vol 19, London: Hogarth Press, 1966.
Goldberger, L. (1961). Reactions to perceptual isolation and Rorschach manifestations of the
primary process. Journal of project. Tech., 25, 287-302.
Guilford, J.P. (1967). The nature of human intelligence, New York: McGraw-Hill.
Gustafson, R., & Norlander, T. (1994). Effects of alcohol on persistent effort and deductive
thinking during the preparation phase of the creative process. Journal of Creative Behavior,
28, 124-132.
Gustafson, R., & Norlander, T. (1995). Effects of creative and non-creative work on the
tendency to drink alcohol during the restitution phase of the creative process. Journal of
Creative Behavior, 29, 25-35.
Hajcak, F. (1976). The effects of alcohol on creativity. Dissertation Abstracts International,
36, 6380B-6381B.
Holland, J. L., & Barid, L. L. (1968). The Preconscious Activity Scale: The Development
And Validity of an Orikginality Measure. The Journal of Creative Behavior, 2, 18-25.
Holmquist, R. (1973). Testbeskrivning för FREGO [Manual for the FREGO test]. Stockholm:
Psykologiförlaget.
Holmquist, R. (1974). Testbeskrivning för Syllogismer I och II [Manual for the Syllogismer I
and II Test]. Malmö: Meddelande från Metodoteket.
Holmquist, R. (1986). Manual till FS: Förändring och stabilitet (Manual for the FS Test:
Change and stability). Stockholm: Psykologiförlaget.
Hutchison, M. (1984). The book of floating: exploring the private sea. New York.
Jacobs, G.D., Heilbronner, R.L. & Stanley, J.M. (1984). The effects of short-term flotation
REST on relaxation: a controlled study. Healthy Psychology, 3, 99-111.
Klein, G. S. (1961). On inhibition, disinhibition and “primary process“ in thinking. Proc. 14
th Int. Congr. Appl. Psychol. Copenhagen: Munksgaard.
Kris, E. (1952). Psychoanalytic explorations in art. New York: International Universities
Press.
Lacey, J. I. (1967). Somatic response patterning and stress: Some revisions of activation
Floating and Creativity
19
theory. In M. H. Appley & R. Trumbull (Eds.). Psychological stress: Issues in research. New
York: Appleton-Century-Crofts.
Larsson, G. (1987a). Routinization of mental training in organizations: Effects on
performance and well-being. Journal of Applied Psychology, 72, 88-96.
Larsson, G. (1987b). Jogging och kreativitet: En explorativ fallstudie [Jogging and creativity.
An explorative case study]. In G. Patriksson (Ed.), Aktuell beteendevetenskaplig
idrottsforskning: SVEBIs årsbok (pp. 106-121). Lund, Sweden: SVEBI.
Larsson, G. (1991). Mentala träningstekniker i preventivt och behandlande syfte. Karlstad:
Centrum för folkhälsoforskning.
Larsson, G., & Starrin, B. (1988). Effects of relaxation training on verbal ability. Journal of
the Society for Accelerative Teaching and Learning, 13, 147-163.
Lee, A. B. & Hewitt, J. (1987). Using visual imagery in a flotation tank to improve gymnastic
performance and reduce physical symptoms. International Journal of Sport Psychology, 18,
223-230.
Marlatt, G.A., Demming, B., & Reid, J. (1973). Loss of control drinking in alcoholics.
Journal of Abnormal Psychology, 81, 233-241.
Maslach, C. (1988). Utbränd [Burnout - the cost of caring]. Trans. P. Wiking, Stockholm:
Natur och Kultur.
Maslow, A. H. (19689. Toward a psychology of being. New York: D. Van Nostrand
Company.
McAleney, P.J., Barabasz, A. & Barabasz, M. (1990). Effects of flotation restricted
environmental stimulation on inter-collegiate tennis performance. Perceptual and Motor
Skills, 71, 1023-1028.
Norlander, T. (1995). Alcohol and the Creative Process. Formulation of a research model
through experiments with alcohol and integrative creativity. Örebro: REPRO.
Norlander, T. (1997). Alcohol and the Creative Process. Frameworks of influence by alcohol
upon creative performance. Örebro: Tryckverkstaden. Doctorial dissertation.
Norlander, T., & Gustafson, R. (1996). Effects of Alcohol on Scientific Thought during the
Incubation Phase of the Creative Process. Journal of Creative Behavior, 30, 231-248.
Floating and Creativity
20
Norlander, T., & Gustafson, R. (1997a). Effects of alcohol on picture drawing during the
verification phase of the creative process. Creativity Research Journal. In press.
Norlander, T., Gustafson, R. (1997b). Effects of Alcohol on a Divergent Figural Fluency test
during the Illumination Phase of the Creative Process. Creativity Research Journal. In press.
Noy, P. (1969). A revision of the psychoanalytic theory of the primary process. International
Journal of Psychoanalysis, 50, 155-178.
Setterlind, S. (1990). Från Hypnos och Suggestion till Avslappning och Meditation. En metod
och forskningsöversikt [From Hypnosis and Suggestion to Relaxation and Meditation. A
review]. Karlstad: Stress Management Center.
Silveira, J. (1971). The effect of interruption timing and length on problem solution and
quality of problem processing. Eugene Oregon: University of Oregon.
Stanley, J.M. and Francis, W.D. (1984). The effects of REST and REST enhanced self-
regulation on essential hypertension. Presented at the XXIII International Congress of
Psychology, Alcapulco, Mexico, 3 September 1984.
Suedfeld, P. & Bruno, T. (1990). Flotation REST and imagery in the improvement of athletic
performance. Journal of Sport & Exercise Psychology, 12, 82-85.
Suedfeld, P., Metcalfe, J. & Bluck, S. (1987). Enhancement of scientific creativity by
flotation REST (Restricted Environmental Stimulation Technique). Journal of Environmental
Psychology, 7, 219-231.
Suler, J. R. (1980). Primary process thinking and creativity, Psychological Bulletin, 88, 144-
165.
Turner, J.W., Jr. and Fine, T.H. (1984). REST-assisted relaxation and chronic pain. Presented
at the XXIII International Congress of Psychology, Alcapulco, Mexico, 3 September 1984.
Wagaman, J., Barabasz, A. & Barabasz, M. (1991). Flotation rest and imagery in the
improvement of collegiate basketball performance. Perceptual and Motor Skills, 72, 119-122.
Zubeck, J.P. (1973). Behavioral and physiological effects of prolonged sensory and
perceptual deprivation: A review. In J.E. Rasmussen (ed.), Man in Isolation and Confinement
(pp. 9-83). Chicago: Aldine.
Floating and Creativity
21
Table 1
Group 1
Non-REST
Group 1
Non-REST
Group 2
Floating
Group 2
Floating
Men
Women
Men
Women
Test time
Mean
20.26
21.60
29.93
24.85
SD
10.69
10.26
0.19
7.53
Floating and Creativity
22
Table 2
Type of
scores
Group 1
Non-REST
Group 1
Non-REST
Group 2
Dry-REST
Group 2
Dry-REST
Group 3
Floating
Group 3
Floating
Men
Women
Men
Women
Men
Women
Fluency
Mean
29.95
30.44
32.37
31.90
33.78
32.39
SD
10.57
5.88
7.09
7.50
11.64
10.18
Obvious
Mean
23.95
25.25
25.75
25.10
25.00
24.04
SD
8.26
4.17
6.77
7.02
9.18
6.29
Original
Mean
6.00
5.19
8.06
6.75
8.72
8.33
SD
3.21
2.80
2.98
2.16
2.79
4.23
Elegance
Mean
1.30
0.44
0.94
2.45
1.44
1.44
SD
2.35
1.24
1.74
1.99
2.51
2.17
FREGO
Mean
25.90
24.50
24.75
24.90
24.44
26.56
SD
2.96
5.07
5.63
5.26
3.97
6.88
Syllogisms
Mean
14.00
15.50
13.38
12.70
11.78
10.22
SD
6.65
4.24
5.58
4.50
4.89
4.06
