Promotion of creativity (divergent productions) and

convergent productions by systematic-relaxation

exercises: empirical evidence from five experimental

studies with children, young adults, and elderly

GUÈNTER KRAMPEN*

UniversitaÈt of Trier, Germany

Abstract
Presented are the results of five experimental studies on the effects of systematic-

relaxation exercises on divergent as well as convergent performance. Subjects were 40

elementary school children, 120 college students, and 52 elderly. In Experiments 1, 2, 4,

and 5, subjects were assigned at random to either an experimental group (participating

in a short systematic-relaxation exercise without pre-experience or exercising

autogenic training which subjects had learned at least 6 months before) or to a

control group (with unspecific relaxation instructions). In Experiment 3, subjects with

against without experience of autogenic training were matched for age, gender, years of

study, and creativity pretest scores. Before and after systematic against unsystematic

relaxation, subjects were tested for divergent performances (word fluency,

associational fluency, expressional fluency, and ideational fluency) and convergent

performances (short-term memory, concentration, and general intelligence). Results of

all experiments consistently show significant improvements of divergent andÐto a

lesser extentÐconvergent performances in the experimental groups after systematic-

relaxation exercises. These short-term effects of systematic-relaxation exercises are

discussed with respect to characteristics and problems of divergent problem solving

processes in individuals and groups. & 1997 by John Wiley & Sons, Ltd.

INTRODUCTION

Individual as well as group problem solving and performance sometimes is hindered

or even stopped by exhaustion and tiredness orÐthe oppositeÐtoo much excitement

and task involvement with its dangers for circular thinking and mental

perseveration. This has been known at least since the results of Yerkes and

CCC 0890±2070/97/020083±17$17.50

Received 31 May 1996

& 1997 by John Wiley & Sons, Ltd.

Accepted 6 December 1996

European Journal of Personality, Vol. 11, 83±99 (1997)

*Correspondence concerning this article should be addressed to GuÈnter Krampen, University of Trier,

Department of Psychology, D-54286 Trier, Germany. E-Mail: [e-mail]

Dodson (1908), first obtained in animal experimentation. The `Yerkes±Dodson Law'

became one of the basics of (achievement) motivation theory including the

moderating status of task difficulty in its applications to humans (see e.g.

Atkinson and Raynor, 1974; Heckhausen, 1967). Under-motivation can result in

individuals' passivity, over-motivation can result in over-excitement and distress.

However, both motivational states can refer to feelings of not progressing and of

being in impasse, to circular thinking and compulsive mental perseveration, and to

giving up. These relations were confirmed for different achievement domains (e.g.,

academic achievement in schools and colleges, job performance, experiments on

complex problem solving, etc) in several studies (for an overview, see Atkinson and

Raynor, 1974; Heckhausen, 1967). Therefore, one focus of applied motivation

psychology, educational psychology, and occupational psychology is the

development and evaluation of intervention techniques and teaching methods to

change those under- and over-motivational states (or even habits) in favour of

motivational and emotional states (or habits) that are optimal for performance.

However, most research on the Yerkes±Dodson Law and modern motivation

theory has been done with respect to convergent problem solving; less has been done

with respect to divergent thinking and production. Surely, this partly results from

the unresolved problems in the measurement of creativity (in comparison with the

assessment of convergent performances; see e.g. Hocevar, 1981). Another reason

may be that historical (e.g. from Gestalt Psychology; BuÈhler, 1907; Wertheimer,

1943) as well as modern (see e.g. Csikszentmihalyi, 1990; Olton, 1979)

phenomenological descriptions and analyses of divergent problem solving

processes include phases of motivational±emotional calmness, rest, peace, and

passivity (which sometimes even result in subjective feelings of not progressing and

of being in impasse) as well as phases of motivational±emotional excitement, high

task involvement, and activity (which sometimesÐat bestÐresult in totally

unexpected ideas and solutions; e.g. `aha experiences', BuÈhler, 1907). Indeed,

there is a strong hypothesis that divergent problem solving processes not only

include both forms of motivational and emotional states, but rather the continuous

alternation of the two types of state is one of the main characteristics of divergent

problem solving processes and creativity. Referring to childhood play behaviour,

Heckhausen (1964) described these fluctuations from over- to under-arousal as the

`cycle of activation', in which the individual swings around the optimal motivational

level but never stays at this level for a longer time. Thus, the cycle of activation

describes a continuous ebb and flow of tension and relaxation. Therefore, phases of

subjective passivity and feelings of not progressing and of being in impasse, as well

as phases of high excitement, task involvement, and activity, belong to the creativity

process. However, they must be prevented from resulting in circular thinking and in

ineffective, compulsive perserveration, in total exhaustion, and in giving up. To

assure this, several creativity techniques and creativity training programmes have

been developed and empirically evaluated (see e.g. Osborn, 1963; Rose and Lin,

1984). In short-term applications, those techniques focus mainly on the

optimization of the individual's or group's production and transformation of

ideas (i.e. fluency and redefinition) by the systematic, technically induced activation

of individuals or groups. By comparison, the phases of motivational±emotional

calmness, rest, peace, and passivity have rarely been taken into account, at least not

in empirical studies.

84

G. Krampen

At present, there exist several conceptual proposals to integrate relaxation

techniques (such as progressive relaxation, autogenic training, meditation, or

biofeedback) into creativity promotion programmes or classroom strategies to

initiate and to foster divergent thinking and action (see e.g. Brown and Wolf, 1986;

Councill, 1988; Zelina, 1982), but until now only the effects of long-term relaxation

training programmes on divergent production have been tested empirically. Domino

(1977) found no effects on creativity of a relaxation-response programme and a

transcendental meditation programme which both lasted for six months. Only the

group participating in a psychology-of-creativity course (including exercises with

creativity-enhancing techniques) improved significantly in five creativity measures.

Cowger and Torrance (1982) compared the creativity-enhancing effects of a Zazen

meditation programme and a relaxation training programme. Pretest±posttest

comparisons favoured the meditation training programme, but the relaxation

training improved sensory experiences, synthesis, and unusual visualizations (while

central creativity variables such as verbal and figural fluency as well as originality

dropped). To date, only one study has shown consistent positive effects of a longer

relaxation training programme on creativity: Hershey and Kearns (1979) analysed

the effects of a series of eight half-hour relaxation-guided fantasy sessions (in

comparison to a similar series of arithmetical exercise sessions) in gifted fourth to

sixth graders. Subjects who had participated in the relaxation-guided fantasy

training performed better at posttest on verbal fluency, flexibility, and originality

measures.

This brief overview shows that the effects of long-term relaxation training

programmes on creativity are inconsistent. However, all reported studies focused

mainly on long-term creativity (trait) effects of training programmes in the domain

of creativity (i.e. personality) development. Short-term effects of relaxation

exercisesÐeither trained previously or practised spontaneouslyÐon divergent

productions have not yet been tested empirically. With reference to the cycle of

activation in divergent problem solving processes described above, it can be

hypothesized that relaxation exercises in current problem solving situations enhance

problem solving. Such exercises create motivational±emotional states (i.e. personal

preconditions) that are a prerequisite for the optimal realization of one's own

creative abilities (trait) in the current problem situation. However, this has been

proved up to now only for convergent production (e.g. retention, memory

performance, and concentration; see Budohoska and Orlowska, 1970; Krampen,

1992; Siersch, 1989), and not for divergent production.

There exist many systematic-relaxation methods (mostly applied as additional

treatment methods in clinical psychology) which exceed individual non-systematic

relaxation and resting strategies by their (greater or lesser) training (resulting in their

broader applicability and effectiveness, even when the individual, non-systematic

methods fail) and empirical foundation. There are systematic-relaxation training

methods (e.g. progressive relaxation, Jacobson, 1924, and autogenic training,

Schultz, 1926; see also Luthe, 1969±1973) as well as systematic short-term relaxation

exercises which can be spontaneously realized after instructions without need for

longer training. I therefore investigated, in five experimental studies, the impact of a

systematic-relaxation training technique after training (i.e. autogenic training) and

that of spontaneous systematic-relaxation exercises (not training) on current

divergent productions of children as well as of adults. In addition to creativity

Creativity and relaxation

85

measures, test indicators of convergent productions (i.e., short-term memory

performance, concentration, general intelligence) were included to (i) reduce

subjects' focus on creativity and the transparency of the experiment and (ii) to test

the short-term effects of systematic-relaxation exercises on these variables. The

general hypothesis of the experiments predicts an interaction effect: it is expected

that groups performing systematic relaxation show significantly higher gains in

pretest±posttest comparisons than groups performing unsystematic resting on their

own (i.e., interaction effect of group by pretest±posttest).

EXPERIMENT 1: RELAXATION EXERCISE AGAINST RESTING IN

ELEMENTARY SCHOOL CHILDREN

Method

Subjects
Participants of Experiment 1 were 40 elementary school children aged 8±10 years (22

females and 18 males; age, M=8.8; SD=1.7) of grades 3 and 4. The children had no

pre-experiences with systematic-relaxation exercises and participated in the

experiment within their regular school education. Children were motivated by pre-

information about performing some playful tasks in small groups (while their class-

mates would stay in the classroom). They participated voluntarily and without

refusals.

Procedure
After randomization (controlling for grade and gender) and pretest, subjects in

Group A (n=20) participated in small groups (n=5 each) in systematic-relaxation

exercises. After a short introduction to the exercise, taking a relaxed sitting posture

and closing the eyes, the relaxation was induced by a short non-dynamic relaxation-

guided fantasy story (The Journey of a Leaf ). The story was read aloud with a calm

but not suggestive intonation and long speech pauses. It lasted for 10 min and

included the topics of self-attention, peace, rest, and passivity by repeated insertions

of feelings of peace (referring to passive self-attention and a calm, abdominal

breathing), heaviness (referring to muscle relaxation), and warmth (referring to

vascular dilation in the arms, hands, and legs). Subjects in group B (n=20 in four

small groups) were asked to be quiet for a while (10 min) and to rest in a subjectively

preferred manner. After relaxation/resting children reactivated by flexing their arms

vigorously and breathing deeply.

The whole procedure was implemented by four student teachers. All were

especially trained for test applications and two of them were especially trained in the

systematic-relaxation exercise used in Group A. Each experimenter tested either

Group A or Group B. They knew neither the other experimental condition nor the

hypothesis under study, but all were instructed to carry out fair testings, giving each

individual child good socio-emotional test conditions. Thus, the experiment was

implemented under double-blind conditions making expectancy effects (with respect

to the subjects as well as to the experimenters) unlikely.

86

G. Krampen

Measures
Before and immediately after the small-group exercises (systematic-relaxation

exercise in Group A versus unsystematic resting in Control Group B), all subjects

were tested with parallelized tests of short-term memory performance (the German

adaptation of the subtest `Auditory Digit Span' of the Wechsler Intelligence Scale for

Children, WISC), general intelligence (Subsets A and B of the Standard Progressive

Matrices, SPM), and verbal fluency (ideational and associational fluency, subtests

`Consequences (obvious)' and `Similarities'; TDK 4-6, a German creativity test for

children; Mainberger, 1977). All tests were administered in the small-group settings

with intensive instructions and test materials for each child. The time limit for each

test was 5 min (following test instructions with slight variations depending on task

number within the tests).

Results

Means and standard deviations of all variables for pretest and posttest measures are

reported for both groups in Table 1. Mean comparisons for all pretest measures

indicated that the randomization procedure resulted in comparable groups

(t(38)=0.87, n.s.).

The interaction hypothesis was tested in two-way analyses of variance (ANOVAs)

with the grouping factor groups (A, B) and the repeated measurement factor time

(pretest, posttest). Table 1 shows the F-ratios that indicate differences in

performance gain between the groups and, additionally, the effect sizes d (Cohen,

1977) for simple pretest±posttest comparisons within the groups. Experimental

Group A improved significantly more than Control Group B in short-term memory

performance as well as in the tests of ideational and associational verbal fluency.

Effect sizes (d ) for these variables were large (according to the procedure and criteria

of Cohen (1997)) and, consequently, the effects are of practical relevance. A

corresponding gain from the systematic-relaxation exercise in general intelligence

performance was not observed.

Creativity and relaxation

87

Table 1. Pretest and posttest scores for Experimental Group A and Control Group B in

Study 1 (40 elementary school children), significance of the group6time interaction, and effect

sizes (d) for changes within groups

Pretest

Posttest

Group6time

Variable

Group

M

SD

M

SD

F(1,36)

d

Memory performance

A

1.8

1.2

3.9

1.3

5.12*

1.7

(WISC: digit span)

B

1.9

1.1

2.0

1.2

0.1

General intelligence

A

6.4

2.2

6.6

2.3

1.02

0.1

(SPM)

B

6.3

2.1

6.6

2.0

0.1

Ideational fluency

A

6.8

1.9

9.7

1.8

7.69**

1.6

(TDK: consequences)

B

6.9

1.8

6.8

1.9

0.1

Associational fluency

A

5.7

1.4

9.2

1.6

8.20**

2.3

(TDK: similarities)

B

5.5

1.3

6.0

1.5

0.4

*p50.05, **p50.01.

Additional analyses of variance on possible moderator effects of grade (third

versus fourth graders), age, and gender (included as additional grouping factors in

three-way ANOVAs) did not result in any significant main effect or any further

significant interaction terms. Thus, systematic-relaxation exercises improved short-

term memory performance and verbal fluency independent of grade, age, and gender

of the elementary school children under study.

EXPERIMENT 2: RELAXATION EXERCISE AGAINST RESTING IN

COLLEGE STUDENTS

Method
Subjects
Subjects of Experiment 2 were 48 undergraduate students of psychology (28 females

and 20 males; age, M=22.4, SD=3.4) participating voluntarily without

compensation. The subjects had no pre-experiences with systematic-relaxation

techniques. They were recruited in a lecture by a call for experimental subjects for an

investigation concerning test construction and validation.

Procedure
After randomization (controlling for years of study, age, and gender) and pretest,

subjects in Group A (n=24) participated in small groups (n=12 each) in

systematic-relaxation exercises. After a short introduction to the exercise, taking

a relaxed sitting posture, and closing the eyes, the 10 min relaxation was induced.

The instructions included passive self-attention, peace, heaviness (referring to

muscle relaxation), warmth (referring to vascular dilation in the arms, hands, and

legs), and calm, abdominal breathing. Subjects in Group B (n=24 in two groups)

were asked after pretest to rest for 10 min in a subjectively preferred manner. After

relaxation/resting, subjects reactivated by flexing their arms vigorously and

breathing deeply.

The whole procedure was implemented by four graduate students of psychology

with especial psychodiagnostic training. Again, two of them were especially trained

for the implementation of the systematic-relaxation exercise in Group A. As in

Experiment 1, each experimenter tested either subjects of Group A or Group B. They

knew neither the other experimental condition nor the hypothesis under study, but

all were instructed to carry out fair testings giving each subject good socio-emotional

test conditions. Thus, the experiment was implemented under double-blind

conditions making expectancy effects (with respect to the subjects as well as to the

experimenters) unlikely.

Measures
Before and immediately after the exercises (systematic-relaxation exercise in

Group A versus unsystematic resting in Control Group B), all subjects were tested

with parallelized tests on short-term memory performance (the German

adaptation of subtest `Auditory Digit Span' of the Wechsler Adult Intelligence

88

G. Krampen

Scale, WAIS), concentration (`Test d2', a German concentration test;

Brickenkamp, 1975), and verbal fluency (word and expressional fluency,

subtests `Prefixes' and `Expressional Fluency'; V-K-T, a German creativity test

for adults; Schoppe, 1975). All tests were administered in the small-group settings

with verbal instructions and test materials for each subject. Time limits followed

standard test instructions.

Results

Means and standard deviations of all variables for pretest and posttest measures are

reported for both groups in Table 2. Mean comparisons among all pretest measures

indicated that the randomization procedure resulted in comparable groups

(t(46)=1.22, n.s.).

Data analyses were performed as in Experiment 1. The results of the ANOVA

tests on the interaction hypothesis and the effect sizes d for simple pretest±

posttest comparisons within the groups are presented in Table 2. Whereas there

was no significant pretest±posttest difference in Control Group B, Experimental

Group A improved significantly more in short-term memory and concentration

performance as well as in the two tests of verbal fluency (see Table 2). In

comparison to the results of Experiment 1, the effect sizes are somewhat lower.

According to the criteria of Cohen (1977), however, effect sizes again reach large

values.

Additional analyses of variance on possible moderator effects of years of

university study, age, and gender (used as additional grouping factors in ANOVAs)

did not result in any significant main effect or further significant interaction terms.

Thus systematic-relaxation exercises improved short-term memory performance,

concentration, and verbal fluency independent of years of study, age, and gender of

the college students.

Creativity and relaxation

89

Table 2. Pretest and posttest scores for Experimental Group A and Control Group B in

Study 2 (48 college students), significance of the group6time interaction, and effect sizes (d)

for changes within groups

Pretest

Posttest

Group6time

Variable

Group

M

SD

M

SD

F(1,44)

d

Memory performance

A

3.0

1.9

4.6

2.1

4.63*

0.8

(WAIS: digit span)

B

3.1

2.2

3.3

2.2

0.0

Concentration

A

23.4

5.3

28.0

6.7

4.33*

0.7

(Test d2)

B

24.0

5.7

25.0

6.0

0.2

Word fluency

A

24.0

6.1

29.6

5.9

8.09**

0.9

(V-K-T: prefixes)

B

24.8

6.4

25.9

6.2

0.2

Expressional fluency

A

11.3

2.7

13.8

2.4

8.51**

1.0

(V-K-T)

B

10.9

3.1

11.4

2.9

0.2

*p50.05, **p50.01.

EXPERIMENT 3: AUTOGENIC TRAINING AGAINST RESTING IN

COLLEGE STUDENTS

Method
Subjects
Subjects of Experiment 3 were 42 undergraduate students of psychology (24 females

and 18 males; age, M=23.7, SD=3.7) participating voluntarily without other

benefits. One half of the subjects (Group A) had learned and practised autogenic

training with good success and transfer to everyday life for at least 6 months (known

from follow-up data from a study on learning autogenic training). The other half of

the subjects (Group B) had no pre-experiences with systematic-relaxation techniques

(including autogenic training). Subjects in Group B were matched to those of Group

A for age, gender, years of study, and pretest creativity performance. Pretest

creativity scores had to be considered in the matching procedure because of

increased pretest scores of those students having learned and practised autogenic

training. This may be an effect of self-selection of students learning and practising

autogenic training, or an effect of long-term practice of autogenic training. Data

available do not allow us to test these hypotheses. However, pretest creativity scores

had to be considered additionally in matching. Thus subjects in Group B (n=21)

were selected from a larger sample of college students (N=58) and matched to the

subjects in Group A according to their pretest creativity scores, age, gender, and

years of study. Subjects of Group A were recruited by personal request with

reference to their participation in courses on autogenic training, which had been

finished at least 6 months previously. Subjects of Group B were recruited in a lecture

by a call for experimental subjects for an investigation concerning test construction

and validation (this pre-information was also given to the subjects of Group A).

Procedure
One week after pretests and matching procedure, subjects in Group A (n=21) did

their well known and consolidated autogenic training exercise in (two) small-group

settings. After greeting and information about the experimental session, subjects

practised the simple sitting posture of autogenic training with eyes closed and under

passive self-attention for 10 min. They used the seven mental standard formulas

(auto-suggestions of peace, heaviness, warmth, calm breath, calm heartbeat,

abdominal warmth, and cool forehead), following the autogenic training standard

procedure (see e.g. Schultz and Luthe, 1969, p. 15). After greeting and information

about the experimental session, subjects in Group B (n=21 in two groups) were

asked to rest for 10 min in a subjectively preferred manner. After relaxation/resting

subjects reactivated by flexing their arms vigorously and breathing deeply.

As in Experiment 2, the whole procedure was implemented by four graduate

students of psychology with special psychodiagnostic training. Again, two of them

were especially trained in the implementation of the systematic-relaxation exercise in

Group A. As in Experiments 1 and 2 each experimenter tested either subjects of

Group A or Group B. They knew neither the other experimental condition nor the

hypothesis under study, but all were instructed to carry out fair testing giving each

subject good socio-emotional test conditions. Thus, the experiment was implemented

90

G. Krampen

under double-blind conditions making expectancy effects (with reference to the

experimental subjects as well as to the experimenters) unlikely.

Measures
At pretest and immediately after the exercises (autogenic training in Group A versus

unsystematic resting in Control Group B), all subjects were tested with the same

measures as applied in Experiment 2 (parallelized tests on short-term memory

performance, WAIS-Subtest Digit Span, and concentration test d2, as well as V-K-T

subtests Word and Expressional Fluency).

Results
Means and standard deviations of all variables for pretest and posttest measures are

summarized in Table 3. Mean comparisons between all pretest measures indicated

that the randomization procedure resulted in comparable groups (t(40)40.79,

p40.05). Comparisons of pretest means (see Table 3) with norm data from German

standardization samples as well as with pretest results obtained in Experiment 2

(studying the same college student population as in Experiment 3; see Table 2) show

that the present sample is described by slightly increased short-term memory and

concentration performances as well as increased verbal fluency. Available data do

not allow a decision on whether this is an effect of self-selection of autogenic training

applicators or an effect of long-term practice of autogenic training (and matching for

pretest creativity scores of subjects in Group B).

Data analyses were run as in Experiments 1 and 2. Whereas there was no

significant pretest±posttest difference in Control Group B, Experimental Group A

improved significantly more in short-term memory and concentration performance

as well as in the two tests of verbal fluency (see Table 3). Effect sizes (d ) reached

large values.

Again, additionally computed analyses (ANOVAs) on possible moderator effects

of years of university study, age, gender, and pretest creativity scores (included as

Creativity and relaxation

91

Table 3. Pretest and posttest scores for Experimental Group A and Control Group B in

Study 3 (42 college students), significance of the group6time interaction, and effect sizes (d)

for changes within groups

Pretest

Posttest

Group6time

Variable

Group

M

SD

M

SD

F(1,38)

d

Memory performance

A

4.1

1.8

6.6

1.6

5.27*

1.5

(WAIS: digit span)

B

4.2

1.7

4.3

1.5

0.1

Concentration

A

26.0

5.4

32.1

5.1

4.39*

1.2

(Test d2)

B

26.2

5.3

27.0

5.5

0.1

Word fluency

A

26.3

6.0

33.2

5.6

7.78**

1.2

(V-K-T: prefixes)

B

27.2

5.9

27.0

6.0

0.0

Expressional fluency

A

13.2

2.5

17.9

2.1

8.43**

2.0

(V-K-T)

B

12.9

2.3

13.1

2.5

0.1

*p50.05, **p50.01.

further grouping factors in ANOVAs) did not result in any significant main effect or

any further significant interaction terms. Thus, autogenic training practised in the

current situation improved short-term memory performance, concentration, and

verbal fluency independent of years of study, age, gender, and pretest creativity

performance of the college students.

EXPERIMENT 4: AUTOGENIC TRAINING AGAINST RESTING AND

HEARING MUSIC IN COLLEGE STUDENTS

Method
Subjects
Subjects of Experiment 4 were 30 undergraduate students of psychology (18 females

and 12 males; age, M=23.6, SD=3.5) participating voluntarily without other

benefits. Known from follow-up data of a study on the learning of autogenic

training, all subjects had learned and practised autogenic training with good success

and transfer to everyday life for at least 6 months. All subjects were recruited by

personal request with reference to their participation in the courses on autogenic

training, which had been finished at least 6 months previously. Written invitations to

a reunion meeting of the course included information about data collection within a

study on test construction and validation. Subjects were accustomed to testing

because the courses on autogenic training included pretest, process, and several

posttest measures concerning the indication and effectiveness of autogenic training

(e.g. a symptom checklist, a biographical data sheet, and personality inventories).

These earlier data collections within the autogenic training courses did not include

any tests on divergent or convergent production.

Procedure
All subjects were asked to bring to the experimental session a Walkman or CD

player and some pieces of instrumental music (either pop or classical music with slow

rhythms and no singing) that they preferred in rest and relaxation situations. After

pretest and randomization (controlling for age, gender, and years of study), subjects

in Group A (n=15) were asked to practise their well known and consolidated

autogenic training exercise for 10 min in a group setting (procedure as in Experiment

3). After pretest, subjects in Group B (N=15) were asked to rest in a subjectively

preferred manner for 10 min hearing their most preferred rest and relaxation music

without practising autogenic training or other systematic-relaxation techniques. This

was done because of the significance of music for resting in the everyday life of many

people. After autogenic training, or resting and hearing music, subjects reactivated

by flexing their arms vigorously and breathing deeply. Posttests followed

immediately. In individual post-experimental interviews, all subjects in Group B

confirmed that they had not used a systematic-relaxation technique during the

experiment.

After subjects' greeting by the former instructor of the courses on autogenic

training, the whole experimental procedure was implemented by four graduate

students of psychology with especial psychodiagnostic training. Two of them had

92

G. Krampen

had special training as instructors for autogenic training (Group A). As in the

experiments before, each experimenter tested either subjects of Group A or Group B.

They knew neither the other experimental condition nor the hypothesis under study,

but all were instructed to carry out fair testings giving each subject good socio-

emotional test conditions. Thus, the experiment was implemented under double-

blind conditions making expectancy effects (with reference to the experimental

subjects as well as to the experimenters) unlikely.

Measures
At pretest and posttest all subjects were tested with the same parallelized tests as in

Experiments 2 and 3 following standard test procedures.

Results
Means and standard deviations of all variables for pretest and posttest measures are

summarized in Table 4. Mean comparisons between all pretest measures indicated

that the randomization procedure resulted in comparable groups (t(28) 41.12,

p40.05). Overall, pretest means were in good accordance with those of Group A in

Experiment 3 (see Table 3) which had been sampled from the same population of

college students experienced with and practising autogenic training in their everyday

life. In comparison to norm data from German standardization samples as well as to

pretest scores obtained from subjects without autogenic training experience, the

present sample is characterized by better short-term memory and concentration

performance as well as higher verbal fluency. Again, available data do not allow a

decision on whether this is an effect of self-selection of autogenic training applicators

or an effect of long-term practice of autogenic training.

The hypothesis of a group6time interaction was confirmed for short-term

memory and concentration performance only, and not for the two creativity

measures of verbal fluency (see Table 4). Significant (overall) main effects for the

repeated measurement factor time (pretest±posttest) showed, however, that subjects

Creativity and relaxation

93

Table 4. Pretest and posttest scores for Experimental Group A and Control Group B in

Study 4 (30 college students), significance of the group6time interaction, and effect sizes (d)

for changes within groups

Pretest

Posttest

Group6time

Variable

Group

M

SD

M

SD

F(1,26)

d

Memory performance

A

4.4

1.6

6.9

1.7

6.84*

1.5

(WAIS: digit span)

B

4.3

1.4

4.7

1.5

0.3

Concentration

A

25.7

4.9

31.8

4.7

5.01*

1.3

(Test d2)

B

26.0

5.2

27.3

4.6

0.3

Word fluency

A

27.1

5.6

32.1

5.2

0.89

0.9

(V-K-T: prefixes)

B

26.7

5.5

30.5

6.2

0.6

Expressional fluency

A

12.6

2.1

18.1

2.4

1.15

2.4

(V-K-T)

B

13.0

2.4

16.3

2.1

0.4

*p50.05.

of both groups improved significantly in word fluency (F(1,26)=6.77, p50.05) and

expressional fluency (F(1,26)=8.85, p50.01). Effect sizes of the within-group

changes reached moderate (Group B, word fluency) to large values (Group A, all

dependent variables; Group B, expressional fluency).

Again, ANOVAs on possible moderator effects of years of study, age, and gender

(included as further grouping factors) did not result in any further significant main

effect or any further significant interaction. Thus autogenic training practised in the

current situation improved short-term memory performance, concentration, and

verbal fluency independent of years of study, age, and gender of the college students.

The same was valid for the impact of simple resting and hearing music on verbal

fluency in college students with autogenic training pre-experiences (but not using it

currently), but not for its impact on their short-term memory and concentration

performance.

EXPERIMENT 5: AUTOGENIC TRAINING AGAINST RESTING AND

HEARING MUSIC IN THE AGED

Method
Subjects
Subjects of Experiment 5 were 52 aged persons (32 females and 20 males; age

M=72.1, SD=5.2, range, 67±80 years) participating voluntarily without other

benefits. As known from follow-up data from courses on autogenic training, all

subjects had learned and practised autogenic training with good success and transfer

to everyday life for at least 6 months. They were engaged in the experiment 6 months

after they had regularly finished an introductory course on autogenic training with a

community service programme. Written invitations to reunion meetings of their

course group included information on data collection within a study on test

construction and validation. As in Experiment 4, subjects were accustomed to testing

because the course on autogenic training included pretest, process, and several

posttest measures concerning the indication and effectiveness of autogenic training

(e.g. a symptom checklist, a biographical data sheet, and personality inventories).

These earlier data collections within the autogenic training courses did not include

any test on divergent or convergent production.

Procedure
All subjects were asked to bring to the experimental session a music cassette or a

compact disc with some pieces of instrumental music (either pop or classical music

with slow rhythms and no singing) preferred in rest and relaxation situations.

Walkmen and CD players were put at subjects' disposal. After pretest and

randomization (controlling for age and gender), subjects in Group A (n=26) were

asked to practise their well known and consolidated autogenic training exercise for

10 min in small-group settings (n=13; procedure as in Experiment 3). After pretest,

subjects in Group B (N=26) were asked to rest in a subjectively preferred manner

for 10 min hearing their most preferred rest and relaxation music without practising

autogenic training or other systematic-relaxation techniques. After autogenic

94

G. Krampen

training/resting and hearing music, subjects reactivated by flexing their arms

vigorously and breathing deeply. Posttests followed immediately. In individual post-

experimental interviews, all subjects in Group B confirmed that they had not used a

systematic-relaxation technique during the experiment.

After subjects' greeting by the former instructor of the courses on autogenic

training, the whole experimental procedure was implemented by six graduate

students of psychology with especial psychodiagnosis training. Three of them had

had a special training as instructors for autogenic training. As in the experiments

before, each experimenter tested either subjects of Group A or Group B. They knew

neither the other experimental condition nor the hypothesis under study, but all were

instructed to carry out fair testings giving each subject good socio-emotional test

conditions. Thus, the experiment was implemented under double-blind conditions

making expectancy effects (with reference to the experimental subjects as well as to

the experimenters) unlikely.

Measures
At pretest and posttest all subjects were tested with the same parallelized tests as in

Experiments 2±4, following standard test procedures.

Results
Means and standard deviations of all variables for pretest and posttest measures are

summarized in Table 5. Mean comparisons for all pretest measures indicated that the

randomization procedure resulted in comparable groups (t(50)41.36, p40.05). In

comparison to pretest scores obtained in the college student samples of autogenic

training applicators (Experiments 3 and 4), the present sample of the aged was

characterized by decreased short-term memory and concentration performance as

well as verbal fluency. At first glance, these age differences are consistent with

research results on cognitive development in the elderly. It should be noted, however,

that all pretest means obtained in the present sample of the aged were in relatively

Creativity and relaxation

95

Table 5. Pretest and posttest scores for Experimental Group A and Control Group B in

Study 5 (52 elderly), significance of the group6time interaction, and effect sizes (d) for

changes within groups

Pretest

Posttest

Group6time

Variable

Group

M

SD

M

SD

F(1,48)

d

Memory performance

A

2.9

2.1

4.8

2.3

7.34**

0.9

(WAIS: digit span)

B

3.0

1.9

3.1

2.2

0.1

Concentration

A

23.1

6.1

27.6

6.6

5.83*

0.7

(Test d2)

B

23.5

6.4

24.1

6.7

0.1

Word fluency

A

24.2

5.6

28.7

6.0

1.36

0.8

(V-K-T: prefixes)

B

23.5

5.3

26.8

5.4

0.6

Expressional fluency

A

10.9

1.8

14.2

2.1

0.90

1.7

(V-K-T)

B

11.2

2.0

13.0

2.4

0.8

*p50.05, **p50.01.

good accordance with the pretest means obtained in college students without

experience in systematic-relaxation training (see Table 2). The available data do not

allow a decision on whether this is a positive effect of successful learning and

practising autogenic training in the elderly, because it may also be due to self-

selection of autogenic-training learners and applicators in the aged.

Consistent with the results of Experiment 4, the hypothesis of a group6time

interaction was only confirmed for short-term memory and concentration

performance, and not for the two creativity measures of verbal fluency (see Table

5). Significant (overall) main effects for the repeated measurement factor time

(pretest±posttest) show, again, that subjects of both groups (autogenic training

versus resting and hearing music for 10 min) improved significantly in word fluency

(F(1,48)=7.43, p50.01), and expressional fluency (F(1,48)=8.68, p50.01). Effect

sizes of the within-group changes reached moderate (Group A, concentration;

Group B, word fluency) to large values (Group A, memory and both indicators of

verbal fluency; Group B, expressional fluency).

Again, ANOVAs on possible moderator effects of age and gender (included as

further grouping factors) did not result in any further significant main effect or any

further significant interaction. Thus autogenic training practised in the current

situation improved short-term memory performance, concentration, and verbal

fluency independent of age and gender of the elderly. The same was valid for the

impact of simple resting and hearing music on verbal fluency in elderly subjects

experienced with but not practising autogenic training in the current situation, but

not for its impact on their short-term memory and concentration performance.

DISCUSSION

The results obtained in all five experiments indicate significant short-term effects of

systematic-relaxation exercises on measures of divergent as well as convergent

performance. This is valid for samples of all age groups under study (elementary

school children, college students, and aged persons), pointing at the age

independency of the positive effects of relaxation exercises on short-term memory

performance and verbal fluency. Furthermore, in all five experimental studies

gender, age, and years of education were not significant moderators of the positive

effects of systematic-relaxation exercises. These results point at a broad applicability

of systematic-relaxation exercises to promote convergent as well as divergent

problem solving in different settings. These exercises (that in none of the experiments

exceeded 10 min) have a high feasibility in various educational and occupational

settings. This is especially true for the systematic-relaxation exercises used in

Experiments 1 and 2, because they do not require longer training and can be

practised spontaneously following instructions. Moreover, due to the double-blind

procedures used in the reported experiments, the internal validity of the results is

high because the probability of expectancy effects either from experimental subjects

or from experimenters is low.

The impact of systematic-relaxation exercisesÐeither performed spontaneously

without training and pre-experience of the subjects or practised on the background

of longer personal relaxation trainingÐon psychometric measures of convergent and

divergent performances exceeded the effects of non-systematic resting (in the

96

G. Krampen

subjectively preferred manner) in all experiments. However, the results of

Experiments 4 and 5 show that resting and hearing subjectively preferred relaxing

instrumental music has effects on divergent (but not on convergent) productions

similar to autogenic training exercises, at least in subjects experienced with autogenic

training but not using it in the current situation. Although control group subjects in

both experiments confirmed post-experimentally that they did not exercise autogenic

training during resting and hearing music, this cannot definitely be excluded.

Perhaps resting and hearing preferred slow-rhythm instrumental music has an effect

similar to autogenic training in the autogenic-training-experienced subjects, and an

effect similar to the relaxation-guided fantasy session practised in Experiment 1 in

elementary school children. Resting and hearing music as well as relaxation-guided

fantasies can induce peripheral relaxation while there is ongoing but slow mental

activation (mentally following the fantasy story or the music in a relaxed, passive

and non-compulsive way). These are, however, the central psychophysiological state

effects during autogenic training (i.e. mentally following the autogenic formulas

under peripheral relaxation) which Schultz (1926; Schultz and Luthe, 1969) named

`autogene Umschaltung' (autogenic switch) including `the fusion of physiological,

cognitive, and behavioral elements into what for some becomes a life-long method of

emotional and physical self-control' (Pikoff, 1984, p. 622).

Whereas divergent productivity (i.e. verbal fluency) is improved by systematic-

relaxation exercises in all experimental groups as well as by resting and hearing music

in autogenic-training-experienced subjects in Experiments 4 and 5, the results on

convergent performances are somewhat different. Short-term memory and

concentration performance are improved by relaxation exercises in all experiments,

but not under the treatment condition `resting and hearing music' in Experiments 4

and 5. This may be due to the mental focusing of instrumental music andÐperhapsÐ

the daydreaming induced by the music. This resulted in the reported positive effects on

divergent, but not on convergent performances. However, focusing relaxation-guided

fantasies in Experiment 1 improved short-term memory performance in elementary

school children, but not the performance in a general intelligence measure (SPM). This

is consistent with the trait view of general intelligence as well as with the relatively

higher dependence of short-term memory performance and divergent thinking on

situation factors (e.g. disturbances and emotional states). Moreover, relaxation-guided

fantasy sessions may be a prototype training session for short-term memory,

concentration, self-attention, and ideational fluency under positive motivational and

emotional conditions. Thus, they may improve short-term memory as well as

divergent thinking in a current situation, but not general intelligence.

It must be emphasized that all reported positive effects of systematic-relaxation

exercises are experimentally obtained short-term and not long-term effects.

According to our results, such exercises improve divergent and convergent

performance in the current situation. The reported results do not allow the

interpretation that systematic-relaxation exercises improve divergent and convergent

problem solving ability (i.e. creativity and intelligence). This hypothesis was not

under study. It can only be concluded that the application of systematic-relaxation

exercises promotes the individual's tendency to make full use and to realize his/her

intellectual abilities as far as possible in the particular situation. This is due to

restfulness and capability following systematic-relaxation exercises and is consistent

with the state-centred phenomenological descriptions of current divergent-problem-

Creativity and relaxation

97

solving processes (i.e. the cycle of activation or flow experiences; see e.g.

Csikszentmihalyi, 1990; Heckhausen, 1964). However, future research has to exceed

the narrow experimental approach realized here (featuring psychometric methods of

data gathering and not real life divergent problem situations) in favour of field-

experimental studies analysing coping behaviour with more realistic problems and

including subjects who are in problem solving phases of under- versus over-motivation

with their dangers for problem solving performances of individuals and groups.

Thus, the results reported here are internally valid for short-term effects of

systematic-relaxation exercises and are in need of external validation. They do not

refer to personality development in the domains of creativity and/or intelligence. To

date, results on long-term effects of longer systematic-relaxation-training

programmes on creativity development are inconsistent (Domino, 1977; Cowger

and Torrance, 1982; Hershey and Kearns, 1979). Some of the results of Experiments

3±5 point at the possibility of such long-term developmental effects of autogenic

training on creativity development in college students and the aged. It must be noted,

however, that these results are obtained neither experimentally nor longitudinally.

The observed cross-sectional differences in pretest measures of creativity between

subjects using autogenic training and subjects not mastering a systematic-relaxation

method may be due either to training effects or to a self-selection of subjects who

decided once before in their life to learn such a relaxation technique and decideÐin a

second stepÐto use it in their everyday life. With reference to the available

experimental data the alternative explanation of self-selection can not be rejected

within the presented investigation. Research using controlled longitudinal designs is

needed to answer this question for the long-term effects of systematic-relaxation

training on creativity development. The results presented are limited to short-term

effects of systematic-relaxation exercisesÐeither trained previously or used

spontaneously following instructions. With these limitations, however, they are

very encouraging to promote individuals' competences to realize their own

intellectual abilities in a more optimal way.

REFERENCES

Atkinson, J. W. and Raynor, J. O. (1974). Motivation and Achievement, Winston, Washington,

DC.

Brickenkamp, R. (1975). Test d2: Aufmerksamkeits-Belastungs-Test [Test d2: Attention-

Endurance Test], Hogrefe, GoÈttingen.

Brown, G. W. and Wolf, J. S. (1986). `Development of intuition in the gifted', Journal for the

Education of the Gifted, 9: 157±164.

Budohaska, W. and Orlowska, A. (1970). `E€ect of autogenic training on learning and

retention', Studia Psychologiczne, 10: 55±60.

BuÈhler, K. (1907). `Tatsachen und Probleme zu einer Psychologie der DenkvorgaÈnge: I. UÈber

Gedanken' [`Facts and problems for a psychology of thinking processes: I. About

thoughts'], Archiv fuÈr die gesamte Psychologie, 9: 297±365.

Cohen, J. (1977). Statistical power analysis for the behavioral sciences, 2nd edn, Academic,

New York.

Councill M. (1988). `Creating inspiration', Journal of Creative Behavior, 22: 123±131.

Cowger, E. L. and Torrance, E. P. (1982). `Further examination of the quality of changes in

creative functioning resulting from meditation (Zazen) training', Creative Child and Adult

Quarterly, 7: 211±217.

98

G. Krampen

Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience, Harper and Row,

New York.

Domino, G. (1977). `Transcendental meditation and creativity', Journal of Applied Psychology,

62: 358±362.

Heckhausen, H. (1964). `Entwurf einer Psychologie des Spielens' [Draft of a psychology of

play behavior], Psychologische Forschung, 27: 225±243.

Heckhausen, H. (1967). The anatomy of achievement motivation, Academic, New York.

Hershey, M. and Kearns, P. (1979). `The e€ect of guided fantasy on the creative thinking and

writing ability of gifted students', Gifted Child Quarterly, 23: 71±77.

Hocevar, D. (1981). `Measurement of creativity: review and critique', Journal of Personality

Assessment, 45: 450±464.

Jacobson, E. (1924). `The technique of progressive relaxation', Journal of Nervous and Mental

Disease, 60: 568±578.

Krampen, G. (1992). `E€ekte der GrunduÈbungen des Autogenen Trainings im schulischen

Anwendungskontext' [E€ects of basic exercises of autogenic training in schools],

Psychologie in Erziehung und Unterricht, 39: 33±41.

Luthe, W. (Ed.) (1969±1973). Autogenic therapy, Vols 1±4, Grune and Stratton, New York.

Mainberger, U. (1977). Test zum divergenten Denken: KreativitaÈt [TDK 4-6; Test on Divergent

Thinking: Creativity], Beltz, Weinheim.

Olton, R. M. (1979). `Experimental studies of incubation: Searching for the elusive', Journal of

Creative Behavior, 13: 9±22.

Osborn, A. F. (1963). Applied imagination: Principles and procedure of creative thinking,

Scribner, New York.

Piko€, H. (1984). `A critical review of autogenic training in America', Clinical Psychological

Review, 4: 619±639.

Rose, L. and Lin, H. (1984). `A meta-analysis of long-term creativity training programs',

Journal of Creative Behavior, 18: 11±22.

Schoppe, K.-J. (1975). Verbaler KreativitaÈt-Test [V-K-T; Verbal Creativity Test], Hogrefe,

GoÈttingen.

Schultz, J. H. (1926). `UÈber Narkolyse und autogene OrganuÈbungen' [On narcolysis and

autogenic organ exercises], Medizinische Klinik, 22: 952±954.

Schultz, J. H. and Luthe, W. (1969). `Autogenic methods'. In: Luthe, W. (Ed), Autogenic

Therapy, Vol. 1, Grune and Stratton, New York.

Siersch, K. (1989). `Zum Ein¯uû des Autogenen Trainings auf die Konzentrationsleistung und

auf subjektiv erlebte Beanspruchungsfolgen' [Impact of autogenic training on concentration

and on experienced stress], AÈrztliche Praxis und Psychotherapie, 11: 5±10.

Wertheimer, M. (1943). Productive thinking, Harper and Row, New York.

Yerkes, R. M. and Dodson, J. D. (1908). `The relation of strength of stimulus to rapidity of

habit-formation', Journal of Comparative and Neurological Psychology, 18: 459±482.

Zelina, M. (1982). Program rozvoja tvorivosti ziakov [Students' creativity development

program], Ceskoslovenska Psychologie, 26: 145±155.

Creativity and relaxation

99

Copyright of European Journal of Personality is the property of John Wiley & Sons, Inc. and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.