PROGRESSIVE MUSCLE RELAXATION: EFFECTS OF EXPECTANCY · Bernstein and Borkovec (1973). These authors recommend the following These authors recommend the following sequence of events

PROGRESSIVE MUSCLE RELAXATION: EFFECTS OF EXPECTANCY

AND TYPE OF TRAINING ON MEASURES OF ANXIETY

by

Michael E. Stefanek

Thesis submitted to the Faculty of the

Virginia Polyteclmic Institute and State University

in partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE

APPROVED:

R. Hodes

in

Psychology

Richard Eisler, Chairman

May, 1982

Blacksburg, Virginia

A. Schulman

ACKNOWLEDGEMENTS

I would like to thank my committee for its support and suggestions

throughout this project. In particular, I would like to express my ap-

preciation to Robert Hodes, without whose expertise in psychophysiology

this project would not have been undertaken. A hearty thank you is also

extended to Mike Patsfall for his assistance with the data analyses,

~nd to Jean Sales and Dan Fones, research assistants, whose dependabili-

ty through a sometimes tedious task is much appreciated.

ii

TABLE. OF CONTENTS

ACKNOWLEDGEMENTS • • • • • • • • • • • • • • • • . • . • • • • • • • • . ii

LIST OF TABLES. • • • • • • • • • • • • • • •

LIST OF FIGURES

INTRODUCTION .•

Background and Description of Progressive Muscle Relaxation Efficacy of Progressive Muscle Relaxation. • • • • • ••• Psychophysiological and Self Report Assessment of

Relaxation Training • • • • • • • . • • • • • • • Procedural Variables of Progressive Muscle Relaxation •••• Live Versus Taped Relaxation Investigations •••••• Methodological Smnmary: Live versus Taped Instructions Treatment Expectancy and other Control Procedures. Present Investigation ••••••••••

METHOD ••.

Subjects •• Design ••• Apparatus. Dependent Variables Procedure •.••••

Live Relaxation Groups •• Taped Relaxation Groups ••••• Self Relaxation Groups. • • • • • • High Expectancy Conditions. Low Expectancy Conditions •••••

RESULTS

Pretreatment Measures •••• Analysis of Treatment Effects

·. . .

DISCUSSION . . . . . . . . . . ' . . The Role of Expectancy and Other Active Multidimensional Nature of Anxiety •.• Caveats and Future Directions ••••• Concluding Remarks •••••••••

Therapeutic Elements •••

REFERENCES. • • • • • • • • • • • • • . . . . . . . . . . . . . . .

V

vi

1

1 3

6 ll 14 18 24 29

32

32 34 34 36 37 39 41 42 43 43

44

44 49

61

62 68 71 74

76

APPENDICES. •

Anxiety Differential. • • •• A. B. c. D. E. F. G.

Inventory of General Trait Anxiousness •••••••••••• Consent Form and Self Monitoring Contract.

82

83 86 90 92 94 96

Expectancy Questionnaire. • • • • •••••• Post-Session 1 Questionnaire. • • • • • .• Procedural Outline: Session 1. Procedural Outline: Session 2.

H. L

• •• 98 • •• 100

• • 102 J.

Self Monitoring Sheets ••• ~ •• Pre-Relaxation Instructions--Taped. Live and Taped Relaxation Instructions. Self Relaxation Self-Monitoring Form ••

• ••• 104 K. L. Tables. . . . . . . . . . . . . . . . . . . . .

1. Analysis of Variance for SR-GTA. • ••• 2. Analysis of Variance for Baseline Period Measure. 3. Session 1 Baseline Heart Rate ••• 4. 5. 6. 7. 8. 9.

10.

Session 2 Baseline Heart Rate • • • • • • • • • t-test for Heart Rate Median Split •••••••• Analysis of Variance for Experimenter Differences. Analysis of Variance for Self-Monitoring Sheets •• Analysis of Variance for Relaxation Ratings •••• Multivariate Analysis of Variance for the Overall Effects of Treatment, Expectation, and Sessions. . • . • • . • • • • •.• Univariate Analyses of Variance for Dependent. Variables . . . . . . . . • • • •

VITA • ••

ABSTRACT

iv

• 108 • 110 • 111 • 112

• •• 114 • 115

• • 116 • 117 • 118

119

• •• 120

• 121

•• 123

LIST OF TABLES

Table

1. Live versus Taped Relaxation Training.

2. Mean Scores for SR-GTA.

3. Relaxation Rating Means ••

4. Mean Change Scores for Dependent Variables.

22

45

51

56

5. Correlation Matrix for Self Report Dependent Variables. 58

6. Correlation Matrix for Self Report/Physiological Dependent Variables .. 59

V

Figure

1.

2.

3.

LIST OF FIGURES

Baseline Heart Rate~Session 1.

Baseline Heart Rate~Session 2.

Heart Rate Change Scores ••••

4. Skin Fluctuation Change Scores ••

5. Anxiety Differential Change Scores .•

6. Finger Pulse Change Scores.

vi

Page

46

48

52

53

54

55

Introduction

The therapeutic technique known as progressive muscle relaxation

has been used as a treatment technique for a variety of psychological

disorders (e.g., tension headaches, test anxiety) and as a component

in other behavioral strategies (e.g., systematic desensitization,

participant modeling). Despite rather widespread use, a variety of

issues pertaining to progressive muscle relaxation remain clouded.

More specifically, psychophysiological effects in stress and nonstress

situations, and critical procedural variables involved in its effec-

tive implementation, have not been definitely delineated. Th~ present

investigation examines the latter area of confusion, specifically, the

modality used in training progressive muscle relaxation (i.e., live or

taped training) and the role of expectancy factors in examining the

efficacy of progressive muscle relaxation.

Background and Description of

Progressive Muscle Relaxation

At this point in time, progressive muscle relaxation is considered

the most prevalent technique of inducing relaxation (King, 1980). The

procedure itself involves systematically tensing and relaxing various

gross muscle groups throughout the body. The individual attends to the

feelings of tension and relaxation concomitant with the tensing and

relaxing of the muscle groups. The goal of the procedure is to in-

crease the individual's ability to discern and eliminate tension.

Despite the early work of Jacobsen (1929) with progressive relaxa-

tion, it was several decades before the technique was used extensively

1

2

in applied settings. As King (1980) noted, the major reason may well

have been due to the central role psychoanalysis played as a model of

psychopathology. Psychoanalytic treatment is based on the unconscious

role played by the id, ego, and superego and the dynamics created by

the interplay of these inner forces. In addition., great emphasis is

given the role of childhood development as the individual progresses

through a series of psychosexual stages. Psychopathology results when

progression through these stages is prevented. Relaxation training

does not purport to deal with unconscious forces or psychosexual con-

flict. Hence, based on this view, progressive relaxation W'Ould be seen

as a very superficial treatment strategy. In addition, the substantial

time investment needed for the full training as proposed by Jacobsen

served as a further deterrent. As developed by Jacobsen, progressive

relaxation required a total of 50 hourly sessions of systematic

training.

As alternatives to psychoanalysis have developed, particularly

behavior therapy, the use of progressive muscle relaxation has in-

creased extensively. Wolpe (1958) served to promote the procedure

when he selected muscle relaxation as the main anxiety inhibitor in

his procedure of systematic desensitization. In addition, he estab-

lished an abbreviated form of progressive muscle relaxation, incorpo-

rating six 20 minute sessions with two 15 minute practice sessions

daily between training sessions. This shortened version of progres-

sive relaxation has increased the popularity of the technique. un-

doubtedly, the most thorough and frequently cited description of

3

abbreviated progressive muscle relaxation (Israel & Beinan, 1977;

Russell,Sipich, & Knipe, 1976, among many others) is the manual by

Bernstein and Borkovec (1973). These authors recommend the following

sequence of events with each of the major muscle groups: (1) the

attention of the subject is focused on a particular muscle group; (2)

the muscle group is tensed upon a signal from the therapist; (3) ten-

sion is maintained for 5-7 seconds; (4) the muscle group is then re-

laxed, upon signal from the therapist; and (5) the subject's attention

is focused upon the muscle group as it relaxes. Sixteen muscle groups

are tensed and relaxed in the following order: dominant hand and

forearm; dominant biceps; nondominant hand and forearm; nondominant bi-

ceps; forehead, upper cheeks, and nose; lower cheeks and jaw; neck and

throat; chest, shoulders, and upper back; abdominal or stomach region;

dominant thigh; dominant calf; dominant foot; nondominant thigh; non-

dominant calf; nondominant foot.

Efficacy of Progressive Muscle Relaxation

Since Wolpe's initial work, progressive relaxation has been ex-

tensively used as a treatment technique in its owri right in addition

to its use as a component of the desensitization paradigm. Russell and

Sipich (1974) found significant decreases in the Test Anxiety Scale

(Sarason, 1957), the State-Trait Anxiety Inventory {Spielberger &

Gorsuch, 1966) and the Anxiety Differential {Husek & Alexander, 1962)

following progressive muscle relaxation with a test anxious population.

Russell, Miller, and June (1976) also found changes due to progressive

muscle relaxation with this same population on the Test Anxiety Scale

4

and State Trait Anxiety Inventory. Borkovec and Fowles (1973) found

decreases in the number of minutes required to fall asleep, the number

of minutes awake during the night, and the degree of restfulness upon

awakening with insomniacs with training in progressive muscle relaxa-

tion. Matthews and Gelder (1969) investigated progressive muscle re-

laxation with individuals reporting phobic reactions or generalized

anxiety and found decreases with relaxation on EMG, skin conductance

activity, and skin conductance level measures. Borkovec, Grayson, and

Cooper (1978),working with a generalized anxiety population, found

significant effects attributable to progressive muscle relaxation in

daily tension percentage and a respiration measure, inspiration/

expiration amplitude ratio. Tasto and Chesney (1974) have also used

progressive muscle relaxation for treatment of dysmenorrhea finding

decreases in the Symptom Rating Scale (Muller, 1971) and the Menstrual

Activities Scale. Finally, Shoemaker and Tasto (1975) found that

progressive muscle relaxation training had a significant effect upon

lowering systolic and diastolic blood pressure in hypertensives. In

general, then, researchers using abbreviated progressive relaxation

as a single model therapy have reported positive results across re-

sponse systems (somatic, cognitive, and behavioral), particularly in

the areas of insomnia and anxiety.

Progressive relaxation also compares favorably with other modes

of relaxation training. In the area of insomnia, Nicasso and Bootzin

(1975) found progressive relaxation superior to control groups and

equivalent to autogenic relaxation. This latter type of relaxation

5

training (Schultz & Luthe, 1959) involves no direct tensing of muscle

groups. Rather, phrases are introduced to the trainee (e.g., "My arm

is heavy and warm."), and the trainee is instructed to passively con-

centrate on these phrases. These instructions are shifted to different

limbs, and also to cardiac activity ("heartbeat calm and regular"),

respiration, the abdominal region, and the forehead. Similar results

supporting progressive relaxation exist with other problem areas in-

cluding dental fears (Miller, Murphy, & Miiler, 1978) and headaches

(Blanchard, Theobald, Williamson, Silver, & Brown, 1978; Cox, Freund-

lich, & Meyer, 1975). King (1980) noted that despite investigations

supporting the efficacy of progressive muscle relaxation as a single-

model therapy (Deffenbacher, 1976; Russell et al., 1976), a number of

methodological problems prevent a definitive statement of the effec-

tiveness of the single model therapy approach with progressive muscle

relaxation. Among these problems are inadequate controls for demand

characteristics or expectation and lack of reliability and validity of

the dependent measures. More specifically, studies attesting to the

efficacy of progressive relaxation have implemented control strategies

involving subjects discussing a variety of topics with the therapist in

lieu of progressive relaxation training (Goldfried & Trier, 1974;

Haynes, Woodward, Moran, & Alexander, 1974), a rather weak control for

demand characteristics. In addition, dependent measures frequently

involve only self report, a measure fraught with inherent reliability

and validity problems. As an example of the prevalence of this

measurement problem, Luiselli, Steinman and Steinman (1979) reviewed

6

70 articles from four major psychological journals regularly publish-

ing research concerned with relaxation training and noted that 70% of

the articles gave no indication of how the effects of relaxation were

assessed. Moreover, only five studies involving physiological

measures reported the use of more than one physiological measure, and

only five studies reported data involving more than one category of

measurement. Finally, a large number of studies have used college

students with problems of less than clinical intensity implementing

very lenient selection criteria. This choice of subjects severely

limits the ability to generalize to clinical populations. Thus, al-

though results are very promising, a conservative view of the efficacy

of progressive relaxation 'WOUld fall short of enthusiastic endorsement.

In sum, widespread use of progressive relaxation should not be

seen as proof of the efficacy of progressive muscle relaxation in the

problem areas discussed in this section. Despite great promise and a

preponderance of favorable research results, the efficacy of this ap-

proach as a single model therapy and facilitator of other interven-

tions can still be questioned. In addition to the methodological

problems noted, the infrequent measurement of or inconsistent results

garnered from self report measures or psychophysiological assessment

contributes to this question of efficacy.

Psychophysiological and Self Report

Assessment of Relaxation Training

As might be gleaned from the above discussion of physiological

measures, results of the numerous investigations of relaxation on

7

physiological responding are somewhat inconsistent. In his early

research, Jacobsen (1929, 1934, 1940) found that training in progres-

sive relaxation produced dramatic decreases in muscle activity,

blood pressure, and heart rate. However, as noted previously, Jacob-

sen relied upon a large number of lengthy training sessions and ne-

glected to utilize statistical procedures and appropriate control pre-

cautions (Mathews, 1971). Abbreviated progressive relaxation has been

shown to effect changes in electrodermal measures (Beiman, Israel, &

Johnson, 1978; Brandt, 1973), muscle tension (Mathews & Gelder, 1969;

Paul, 1969), respiration (Delman & Johnson, 1976; Paul, 1969), heart

rate (Ollendick & Murphy, 1977; Paul & Trimble, 1970), and blood

pressure (Deabler, Fidel, Dilenkoffer, & Elder, 1973; Shoemaker &

Tasto, 1975). However, Brokovec and Sides (1979), investigating 25

studies directly related to this issue, found 15 reporting progressive

relaxation superiority, while 10 indicated equivalence. In addition,

King (1980) suummarized studies of the psychophysiological effects of

relaxation training noting: "A clear picture fails to emerge with

respect to the psychophysiological effects of progressive muscle re-

laxation in stress and nonstress situations" (p. 155). In an attempt

to provide some explanation of these results, Borkovec and Sides

(1979) posit that the number of sessions, population differences

(anxious v. normal subjects), and live vs. taped presentation of pro-

gressive relaxation instructions act as differences between those

studies showing and not showing physiological effects.

8

Historically, heart rate is by far the most frequently assessed

cardiovascular output parameter, and, as noted by Borkovec, Weerts,

and Bernstein (1977), has received the greatest empirical investiga-

tion with respect to anxiety assessment. A host of investigators

have utilized this measure in assessing anxiety (Edelman, 1970; Mat-

thews & Gelder, 1969; Paul, 1969; Schandler & Grings, 1970, among

others). The majority of investigators have noted decreases in heart

rate with progressive muscle relaxation over control conditions

(Brandt, 1973; Edelman, 1970; Ollendick & Murphy, 1977; Paul, 1969)

although these results are not unanimous (Beiman et al., 1978; Borko-

vec et al., 1978). In sum, heart rate has consistently been the car-

diovascular measurement of choice in the physiological assessment of

anxiety.

Finger pulse volume amplitude (FPV) has also been used as a

measure of cardiovascular functioning. As Borkovec et al. (1977) note,

different measures of cardiovascular functioning are all interrelated,

but not in a simple, positive, linear fashion, and thus, cannot be

assumed interchangeable. Blood volume represents the absolute level

of blood in the tissue with blood volume pulse representing the blood

flow through the tissue with each cardiac contraction. Bloom, Houston,

and Burish (1976) explored the utility of FPV as a measure of anxiety.

In this investigation, subjects were exposed to either a threatening

(threat of electrical shock) or nonthreatening situation, while in-

dices of physiological arousal (pulse rate and finger pulse volume)

and self reports of anxiety (Affect Adjective Checklist) were

9

collected. Results indicated significant correlations of FPV with

pulse rate and AACL measures and that FPV was differentially respon-

sive to two levels of experimentally manipulated anxiety. It should

be noted, however, that the strengths of the relationships between FPV

and pulse rate and AACL were not substantial (r .41 and .24, respec-

tively), although typical of correlations between physiological

measures and other physiological and self report measures of anxiety

(Martin, 1961). A further investigation (Bloom & Trautt, 1977) served

to substantiate the above findings.

Measures of electrodermal activity (EDA) have also been investi-

gated with relation to changes in skin conductance with progressive

relaxation training. Beiman et al. (1978), Brandt (1973, Lehrer

(1977, and Schandler and Grings (1976) have all found significant

changes on GSR frequency with progressive muscle relaxation over a

variety of control conditions, while Davidson and Hiebert (1971) and

Mathews and Gelder (1969) found no differences in skin conductance

measures between treatment and control conditions. Other than the

Beirnan et al. (1978) investigation, those studies examining live versus

taped relaxation have not selected a physiological measure from the

electrodermal response system.

vec et al. (1977), note: "

This is despite the fact that, as Borke-

• of all the bodily systems available :·for

measurement, the skin's electrical properties have most often been

elected as a convenient measure of the physiology of the anxiety state"

(p. 411).

10

In the present investigation, spontaneous fluctuations in skin

resistance, heart rate, and finger pulse volume amplitude were all

used as a physiological assessment of anxiety.

As opposed to the literature's inconsistent physiological results,

self report measures consistently support the use of progressive re-

laxation training as a means of reducing subjective tension and anxiety.

The Affect Adjective Checklist (Zuckerman, 1960), State-Trait Anxiety

Inventory (Spielberger, Gorsuch, & Lushene, 1968), and Anxiety Differ-

ential (Husek & Alexander, 1962) have been used extensively in re-

search with progressive relaxation (Beiman & Johnson, 1978; Edelman,

1970; Paul, 1969, 1970; Schandler & Grings, 1976) with positive re-

sults. The latter measure was used in the present investigation. This

instrument is a self report semantic differential state anxiety measure

(Appendix A). The AD was developed as a verbal response measure of

situational anxiety, with ratings of concepts simply involving checking

bipolar scales of adjectives (e.g., loose-tight, deep-shallow}. Anx-

ious subjects instructed to "fake good" have been found to display

higher anxiety scores than a nonanxious control group. In a subsequent

investigation, knowing the purpose of the AD was unrelated to anxiety

scores. Internal consistency was adequate in several studies (Alex-

ander & Husek, 1962; Husek & Alexander, 1963}. In addition, Paul

(1966) reported a moderate 7 week test-retest correlation (r = .54)

for a speech anxious sample, while a 3 week test-retest relationship

among 47 nontreated socially anxious subjects has been found to be

fairly high (r = .78) in a study by Borkovec, Stone, O'Brien,

11

and Kaloupek (1978). Finally, all investigations dealing with the

issue of live versus taped relaxation have used this measure (Beiman

et al., 1978; Israel & Beiman, 1977; Paul & Trimble, 1970; Russell,

Sipich, & Knipe, 1976), thus facilitating comparison of results.

Typically, correlations between subjective ratings of anxiety

and physiological variables have been low and nonsignificant. A num-

ber of factors may account for ,:the inconsistent findings including

population differences, techniques of assessing anxiety level, inten-

sity of training, the complexity of the "anxiety" construct itself,

and/or emotional desynchrony (Hodgson & Rachman, 1974; Lang, 1978;

Lehrer, 1978). The latter possibility refers to different rates of

change for different emotional behaviors observed in the same subject.

Procedural Variables of

Progressive Muscle Relaxation

A host of issues related to progressive relaxation training re-

main unsolved. Borkovec and Sides (1979), in a recent review of pro-

cedural variables related to the physiological effects of progressive

relaxation referred to earlier in this paper, found a series of dif-

ferences between studies dem::>nstrating progressive relaxation superiority

and those not demonstrating superiority. Among the former, the mean

number of sessions was 4. 5.7, 73% employed live administration of the

procedures, and 47% involved patient samples. A:nK>ng studies finding

equivalence between progressive relaxation and control conditions,

the average number of sessions was 2.30, 80% involved normal subjects,

and.70% employed standardized taped instructions. It is this latter

12

issue, i.e., the issue of live versus taped training, which will be of

concern in the present study •. In addition to the authors' disclaimer

stating that considerable overlap existed between groups of the studies

providing different results on the procedural aspects noted above,

methodological problems in ·the studies discussed serve to question

their conclusions. More specifically, the seven studies employing

taped relaxation instruction which produced findings indicating pro-

gressive muscle relaxation to be equivalent to or inferior to appro-

priate control conditions include a high frequency of methodological

weaknesses. The large number of methodological problems serves to

dissolve the significant difference found between live and taped

instructions with regard to physiological effects. A series of these

experiments (Edelman, 1970; Haynes, Moseley, & McGowan, 1975; Lader &

Mathews, 1970) included only one relaxation session. Also, Davidson

and Hiebert (1971) employed, in addition to one experimental session,

two relaxation training sessions, one the day prior to the experimental

session and one immediately prior to the experimental session. Hence,

the degree of training of relaxation skills may have been inadequate

in these studies reporting negative results for taped relaxation.

Further, Davidson and Hiebert used only skin conduction as their

measure of assessing physiological effects, while Haynes et al. (1975)

restricted their physiological measures to frontalis EMG recordings.

This poses a problem since different individuals may respond to pro-

gressive relaxation by a large decrease in heart rate and a small or

insignificant decrease in skin conductance, while another individual

13

may respond with the opposite pattern (Lacey, Bateman, & Van Lehn,

1953; Martin, 1961). Thus, the use of a single measure of physio-

logical response is severely limiting the prospects for finding

changes with relaxation training. Tasto and Huebner (1976), another

study cited by Borkovec and Sides (1979), found no differences in

measures of systolic and diastolic blood pressure with normotensives,

using taped instructions. However, Shoemaker and Tasto (1975) found

that muscle relaxation has a significant effect upon lowering systolic

and diastolic blood pressure using taped progressive muscle relaxation

instructions with a hypertensive population. It is conceivable, as

Tasto and Huebner point out, that there may be a lower limit below

which blood pressure will not go regardless of how relaxed a person

becomes. If this is true, Tasto and Huebner's work can hardly be said

to be an indictment of the efficacy of progressive muscle relaxation.

Finally, these investigations, including Paul and Trimble (1970), suf-

fer from weaknesses in the area of subject selection (i.e., use of

normal population) and lack of controls for treatment expectancies.

Based on all seven studies reviewed, the issue regarding the ef-

fectiveness of taped instructions has not been isolated from the

effects of normal versus clinical populations, number of sessions

(amount of practice), or other factors as a determinant in not achiev-

ing physiological effects with progressive muscle relaxation. For-

tunately, a series of studies have specifically investigated the ef-

fects of live versus taped relaxation instructions (Bieman et al.,

1978; Israel & Beiman, 1977; Paul & Trimble, 1970; Russell et al., 1976).

14

Live Versus Taped Relaxation Investigations

The first investigation directly comparing the effects of recorded

versus live relaxation training was that of Paul and Trimble (1970).

In this study, three groups (N = 10) of undergraduate females par-

ticipated individually for two sessions, 1 week apart, receiving

either: (1) abbreviated progressive muscle relaxation via recorded

tape; (2) hypnotic induction emphasizing direct suggestions of heavi-

ness, warmth, etc. via recorded tape; (3) self relaxation control

procedure. The experimenter was not present during the sessions in

all three conditions. Among other results, the authors found recorded

relaxation to be inferior to live, with live data based on results ob-

tained previously. That is, recorded relaxation training was com-

pared to live relaxation training, while taped self relaxation and

hypnotic induction groups were compared to live self relaxation and

live hypnotic induction groups, respectively. With regard to the pro-

gressive relaxation condition, significant differences were found on

measures of heart rate and tonic muscle tension (forearm), favoring

the live condition. No differences were found among groups on the

Anxiety Differential (Husek & Alexander, 1962), a self report measure

of state anxiety.

Obviously, the retrospective data comparison must be considered

a weakness in the investigation under discussion. The fact that the

data from the live groups across conditions was drawn from a prior

investigation (Paul, 1969) poses problems of internal validity.

Other weaknesses include: (1) the use of different tapes for each of

15

the two sessions; (2) the lack of close monitoring of between session

practice ("requested" of the subjects by the authors); (3) the use of

a population consisting of female undergraduates participating as part

of a research requirement, unselected except for absence of drug use

and presence of good physical health; (4) use of one experimenter

across groups; and (5) lack of controls for expectancy factors.

Russell et al. (1976) also contrasted live versus taped instruc-

tion,in a four-group design involving two sessions 1 week apart; the

groups included: (1) live presentation, 16 muscle groups; (2) taped

presentation (experimenter absent), 16 muscle group; (3) live presen-

tation, four muscle groups; and (4) self relaxation control group.

Dependent measures included EMG recordings (location not noted), and

the Anxiety Differential. The population for the study consisted of

34 undergraduate females. Significant reductions in both EMG and

Anxiety Differential measures occurred only in the 16 live condition,

while the 16 taped condition showed significant reductions on the

Anxiety Differential, but not with EMG measure. The four live condi-

tion showed no changes in either dependent measure, while the control

condition demonstrated significant pre~post EMG changes. This inves-

tigation suffers from a host of weaknesses including: (1) small and

unequal number of subjects across conditions (16 live= 11, 16 taped=

6, 4 live = 8, self relaxation= 9), and no explanation given regarding

these unequal cell assignments; (2) lack of practice between sessions;

(3) use of nonanxious female undergraduates as subjects with no infor-

mation provided regarding screening decisions; (4) no information

16

provided regarding assignment of subjects to groups or assignment of

the three therapists to groups; (5) no control for client expectan-

cies; (6) no specification of electrode placement for the EMG measure;

and (7) the use of only one physiological measure (EMG) and no mention

of between group comparisons with only pre-post within-group changes

provided.

Israel and Beiman (1977) compared the effects of live and taped

progressive relaxation training and self relaxation using frontalis

EMG and the Anxiety Differential as dependent measures. Subjects were

individuals who responded to advertisements for tension relief (14

males, 11 females). Treatment involved three sessions, and results

indicated no difference between live, taped, and self relaxation con-

ditions on heart rate, respiration, or EMG (frontalis) measures, with

the live relaxation group superior to both taped and self relaxation

groups on the Anxiety Differential. All subjects showed significant

reductions on all measures during the sessions with no difference be-

tween treatment conditions or across sessions. No mention is made of

experimenter presence/absence during taped relaxation, and practice

between sessions was not monitored. Weaknesses in this particular

study include possible subject selection problems since no mention is

made of exclusion due to previous relaxation training or medication.

Also, considerable ambiguity exists regarding the criterion of anxiety

measures used, i.e., "above average" scores on the STAI (Spielberger

et al., 1968) and the MCAL (Zuckerman, 1960). In addition to the above

weaknesses, this investigation unfortunately involved no controls for

17

treatment expectancy variables, while use of the same therapist across

all three groups is also questionable.

Finally, Beiroan et al. (1978} compared live and taped relaxation

training, self relaxation, and electromyogram biofeedback on measures

of autonomic and somatic arousal and subjective tension. Male and fe-

male respondents (N = 40} to an advertisement soliciting tense indi-

viduals to participate in a psychology study were assigned to one of

the groups above and evaluated in five sessions of decreasing duration

(35, 35, 15, 15, and 5 minutes). This decreasing duration coincided

with training involving 16 muscle groups, then four, then training in

relaxation by recall, spread over a total of 24 days. Due to this de-

crease in session length for live and taped groups, statistical com-

parisons were limited to live versus taped and self versus biofeedback

group comparisons. Dependent measures included the Trait Scale of the

STAI, the MACL, the skin resistance response (SRR), heart rate, respira-

tion rate, and muscle tension measures. The experimenter was not pres-

ent during taped instruction. Among other findings, live progressive

relaxation was significantly superior to taped training on three of the

four physiological measures (no significant difference on respiration

rate), while no significant difference was found between the live and

taped groups with regard to the Anxiety Differential or other self

report measures implemented. A posttreatment 10-minute assessment

(i.e., a sixth session) also favored the live relaxation condition.

During this session, subjects were told to relax as much as possible

using the relaxation skills developed in the previous five training

18

sessions. Specifically, the live relaxation condition was superior

on the SRR measure. The self and live relaxation condition were

equivalent and superior to biofeedback and taped groups with regard

to the respiration rate measure, while EMG and subjective tension

measures were equivalent across groups. The lack of objective measures

of anxiety as a screening device, lack of control for experimenter in-

fluence,and expectancy effects, and lack of monitoring of requested

practice between sessions limit internal validity and generalizability

of this investigation.

Methodological Swnn1ary:

Live Versus Taped Investigations

Briefly reviewing the four studies dealing directly with the issue

of live versus taped relaxation (Israel & Beiman, 1977; Israel, et al.,

1978; Paul & Trimble, 1970; Russell et al., 1976), a series of common

methodological weaknesses are found.

1. Subject selection. While Israel and Beiman and Beiman et al.

purport to use "anxious" populations, the criteria for selection are

either inadequate (i.e., no objective measure of anxiety implemented,

no assessment of prior use of relaxation) or ambiguous (use of objec-

tive measures with little delineation of participant's scoring rela-

tive to normative data). In addition, Paul and Trimble (1970) and

Russell et al. (1976) used unselected female undergraduates, with no

measures of anxiety, either subjective or objective, used for screen-

ing purposes. Despite these weaknesses, all investigators attempt to

19

generalize findings to clinical settings with only Paul and Trimble

adding a cautionary note.

2. Group assignment: Therapists and subjects. While Russell

et al. (1976) make no mention of how subjects were assigned to differ-

ent groups, random assignment was used by the other three investiga-

tors under discussion. Based on the relatively small number of sub-

jects used in each investigation (Beiman et al.~10/group; Israel &

Beiman~8/group; Paul & Trimble~lO/group), a better option might

have involved random assignment based on anxiety level~ With the small

number of subjects in each group, placement of two to three of the

most anxious individuals in a particular group may bias findings in

that anxious individuals have shown more physiological response to

progressive relaxation training (Borkovec & Sides, 1979; Lehrer, 1978).

In addition, Russell et al. (1976) provide no information regarding

assignment of therapists to groups, while Beiman et al. (1978), Israel

and Beiman (1977) and Paul and Trimble (1970) all used the same thera-

pist across groups with the resulting possible confounds of therapist

expectancy factors.

3. Expectancy factors. None of the four investigations atte~ted

to control for subject expectancy variables. Hence, it may well be

that expectancies for improvement are higher for live versus taped or

self relaxation. This point is especially critical in Paul and Trim-

ble's (1970) study in which different instructions are given among

groups which clearly favors the live relaxation condition in terms of

purported efficacy.

20

4. Practice effects. Israel and Beiman (1977) and Russell et

al. (1976) do not mention between session practice, while Paul and

Trimble (1970) requested subjects to practice, twice daily, but pro-

vided no check on whether the practice was in fact completed. Finally,

Beiman et al. (1978) improve in this area somewhat.by making the re-

quest and discussing the practice at the second session. Unfortunately,

no monitoring of actual task completion was attempted. This failure

to either request or monitor relaxation practice seems critical in

terms of generalizing to clinical settings and possible between-group

differences (considering the small number of subjects per group across

studies) in practice leading to differential changes across groups in

the dependent measures.

In sum, based on the above considerations, the rather accepted

notion that live relaxation is superior to taped relaxation training

appears somewhat premature. To add to the confusion, consider that

self reported anxiety has been found not to differ between taped and

live instruction (Paul & Trimble, 1970; Russell et al., 1976), to be

significantly lowered with live instruction (Israel & Beiman, 1977),

and to be equivalent across types of instructions but showing defini-

tive downward trends favoring live relaxation (Beiman et ai., 1978).

With regard to physiological measures, Russell et al. (1976) found

significant EMG decrease with live and self relaxation (16 muscle

group condition), while Israel and Beiman (1977) found lower EMG re-

cordings for all groups (taped, live, and self). Further, Paul and

Trimble (1970) found taped relaxation inferior to live on EMG and

21

heart rate measures, and no significant difference with respiration

measures. Finally, Beiman et al. (1978) found decreases in SRR,

heart rate, and muscle tension with live instruction, superior to

changes with taped relaxation instruction within training sessions

(see Table 1) •

Those studies finding a superiority of live versus taped instruc-

tions have postulated either experimenter presence/absence or the is-

sue of response versus program contingent progression as a causal

factor in this superiority. The latter refers to the fact that with

live relaxation training the experimenter does not proceed to tension

release of -successive muscle groups until the current group is com-

pletely relaxed, a procedure at best impractical with taped relaxation.

Beiman et al. (1978) and Paul and Trimble (1970) note the need for

further research to delineate the basis for the superiority of live

versus taped relaxation training. Three of the four studies noted

that have dealt directly with the issue of taped versus live instruc-

tion (Bieman et al., 1978; Paul & Trimble, 1970; Russell et al., 1976)

have had the experimenter absent during taped instructions, while the

fourth (Israel & Beiman, 1977) does not specifically detail this pro-

cedural aspect.

A recent study by Borkovec et al. (1978) with anxious undergradu-

ates (based on questionnaire data relating to percent of time "anxious"

daily, an anxiety level moderately severe or greater, and a desire to

receive treatment for tension problems) assigned subjects randomly

within blocks of percent tense to three treatment conditions. These

Authors

B0iman, Israel, & Johnson (1978)

Israel &

Beiman (1977)

Paul & Trimble (1970)

Russell, Sipich, & Knipe (1976)

Table 1

Live versus Taped Relaxation Training

Type of Group, N

Live, 10 Tape, 10 Self, 10 EMG biofeedback, 10

Live Taped Self *Total N = 25

Taped, 10 Hypnotic Induction, 10 Self, 10

Live, 16 muscle group, 11 Tape, 16 muscle group, 6 Live, 4 muscle group, 8 Self, 9

Number of Sessions

5

1 (post)

3

2

2

Dependent Variables

EMG

Skin Conductance

Heart rate

Respiration

MIIACL

STAI-Trait

Anxiety Differential

EMG

Respiration

Heart Rate

Anxiety Differential

EMG (forearm)

Heart rate

Respiration

Anxiety D.lfferential

EMG

Anxiety Differential

Results

Live> Taped Self= Biofeedback (Bf)

Posttraining Session: Live=Taped=Self=Bf

Live> Taped Self= Biofeedback

Posttraining Session: Live > Self=Taped=Bf

Live> Taped Self > Biofeedback

Posttraining Session: No pre-post ma.in effects

Live=Taped Self and Biofeedback=<no pre-post reductions

Posttraining Session: Live=Self) Bf, Taped

Live=Taped=Self=Bf (Significant pre-post main effects)

Live=Taped=Self=Bf (Significant pre-post main effects)

Live=Taped=Self=Bf {Significant pre-post main effects)

Live=Taped=Self

Live=Taped=Self

Live=Taped=Self

Live> Taped, Self

Live>Taped

Live >Taped

Live=Taped

Live=Taped

Pre-post analyses: significant reductions with Live, 16 muscle group and self

Pre-post analyses: significant reductions with Taped and Live, 16 muscle group

*No information provided regarding distribution of subjects across groups.

N N

23

groups included: (1) progressive relaxation training with tension

release; (2) relaxation without tension release; and (3) no treatment.

The instructions were taped, but, by means of two cassettes and sub-

ject control over alternation between tapes, subjects controlled

their progress in treatment establishing a response-contingent progres-

sion procedure. The measure of percent tense indicated superiority

for treatment conditions, while subjects in the tension release group

required significantly fewer training cycles to produce subjective

reports of complete relaxation. It is interesting to note that the

literature would suggest that the study under discussion would maxi-

mize the probability of obtaining significant physiological reduction

effects: several sessions, subject controlled treatment progression,

and a P!~senting:problem of anxiety or the use of an anxious popula-

tion. The absence of treatment effects on physiological measures

(heart rate, respiration, frontalis EMG), then, is somewhat surpris-

ing. The main difference between the authors' procedure and previous

studies finding significant relaxation effects with live instruction

was therapist presence. The authors conclude that the therapist factor

may be the critical factor in promoting physiological reduction during

relaxation training. The components of the therapist present factor

have yet to be addressed. It seems feasible that the support of the

therapist, i.e., verbal encouragement, may increase motivation on the

part of the subject. Other therapist-subject factors may prove of

critical importance such as therapist-subject relationship, the sub-

ject's perception of the therapist's involvement as a result of

24

therapist presence/absence, or increased client expectancy with thera-

pist present. This latter postulation, i.e., expectancy explanations

for the superiority of live versus taped relaxation training, has not

been addressed in the studies noted in this paper investigating live

versus·taped relaxation training.

Treatment Expectancy and

Other Control Prodcedures

In addition to the taped versus live issue, several studies have

noted relaxation effects (both in terms of subjective reports and

physiological measures) with a control or self-relaxation condition.

Israel and Beirnan (1977) found no difference in EMG measures among

live, taped, and self relaxation controls, while Miller and Bornstein

(1977) reported that the self relaxation group used in their investi-

gation relaxed as effectively as any of the formal training groups

based on EMG (forearm) measures and self report. Other investigators

(Borkovec & Fowles, 1973; Lader & Mathews, 1970) have noted similar re-

sults but attributed these results to experimental confounding. Fin-

ally, several investigators (e.g., Paul & Trimble, 1970; Schandler &

Grings, 1970) have not found self relaxation to be efficacious with

respect to physiological or self report measures. In addition, these

differences cannot be related to different population samples (i.e.,

"normal" versus "anxious"). That is, across population samples, self

relaxation has been, on oocasion, as efficacious as live or taped

strategies. In all of the above investigations, self relaxation has

25

involved instructing the subject to relax in any manner that they

choose,without falling asleep.

Israel and Beiman (1977), interpreting their positive results

with a self relaxation condition, note that this condition was pre-

sented as a potentially effective treatment for tension in their in-

vestigation, a presentation not given in previous research. This is

especially evident in the research conducted by Paul and Trimble (1970)

in which different instructions across groups clearly favored the ac-

tive treatment condition. Reinking and Kohl (1975), reporting their

finding of no differences in subjective tension scores between a self

. relaxation control group and treatment groups (EMG plus progressive

relaxation, EMG only, progressive relaxation only, EMG plus money),

note that if the goal is subjective calm, any procedure should work

equally well if presented as a relaxation procedure.

Within a more general framework, this issue of expectancy or

treatment credibility has become a major issue in treatment outcome

studies (Borkovec & Nau, 19721 Kazdin, 19791 McGlynn & McDonnell,

1974). The question has been raised as to whether typically used

placebo conditions in outcome research do, in fact, control for client

expectations or experimental demand. A number of authors (Baker &

Kohn, 19721 Rosenthal & Frank, 1956) have noted that a placebo condi-

tion should be theoretically inert and capable of generating a positive

expectancy equivalent to its comparison treatment condition. Kazdin

(1980) and Kazdin and Wilcoxon (1976) note several strategies for

developing adequate control groups in which the investigator wishes to

26

rule out differential expectancies for changes across groups as a ri-

val explanation of results. The attention placebo control strategy

consists of any procedure the experimenter designates as a control for

nonspecific treatment effects. This strategy may or may not fulfill

its intended purposes, however, since client expectancies for change

may or may not be equal across control and treatment conditions. The

treatment element control strategy involves a control group resembling

the actual treatment as closely as possible, with as few procedural

changes as possible. Again, this strategy is limited since the credi-

bility of treatment and client's expectancy are not necessarily con-

trolled. That is, the inclusion of treatment elements in a control

group does not necessarily equalize credibility across groups. For

example, as Borkovec and Nau (1972) noted, groups receiving a single

component of desensitization do not generate expectancies for success

equivalent to groups receiving the complete desensitization procedure.

Finally, the empirically derived control strategy involves an empirical

demonstration that the expectancy for change is equivalent across

treatment and control groups. One method is to devise any procedure,

followed by assessment of this procedure's credibility relative to the

treatment groups. A potential weakness in this procedure is that any

procedure equivalent in credibility can be used as a control group.

Kazdin (1976) noted that the greater the procedural difference, the

more likely that some difference in group procedures contributes to

outcome differences, thus reflecting other nonspecific treatment ef-

fects or other specific therapeutic ingredients (e.g., different

27

behavior change mechanisms). In order to explain the mechanism of the

treatment being evaluated, this possibility must be ruled out •. Based

on the above, Kazdin recommends a synthesis of nonspecific treatment

control strategies. A combination of the empirically derived control

strategy and the treatment element control strategy control effectively

for equal credibility across treatment and control groups and the simi-

larity between treatment,and control conditions. This combination of

strategies, incorporated into the present investigation, most effec-

tively rules out differential,nonspecific treatment effects across

treatment and control conditions.

Historically, despite the recommendations of Kazdin (1980) and

Kazdin and Wilcoxon (1976) noted above, the assumption that the place-

bo condition is credible has rarely been empirically tested, while

placebo and therapy credibility comparisons are rare. With regard to

the latter, a series of investigations (Borkovec & Nau, 1972; Boudewyn

& Borkovec, 1974; McGlynn & McDonnell, 1974) indicate that treatment

conditions often may not be equivalent in credibility. Borkovec and

Nau, using a subject population of 450 introductory psychology .stu-

dents, found the rationale and procedural description of systematic

desensitization more er.edible than Davidson's (1968) relaxation-recall

control, Borkovec and Nau's (1972) avoidance response placebo, Paul's

(1966) attention-placebo, and ,Marcia, Rubin, and Effran' s. (1969)

tachistoscope placebo, while implosive therapy was also superior to

the control groups. Boudewyn and Borkovec (1974) conducted a study

with 120 psychiatric inpatients as subjects and found drug treatment

28

and psychoanalytic rationales significantly more credible than an

attention-placebo rationale. Finally, McGlynn and McDonnell (1974)

demonstrated that snake phobic college students exposed to samples

from a desensitization treatment and a frequently employed pseudo-

therapy control procedure, rated the desensitization treatment as

significantly more credible than the latter.

All of the investigators above used the Borkovec and Nau (1972)

scale, a five question; ~0-point rating scale for credibility/expec-

tancy for any improvement. The scale inquires as to the logic of the

type of treatment, confidence in recommending the treatment to a

friend, willingness to undergo such a treatment, and generality of

effects predicted from the treatment. Hence, these results indicate

that certain treatment conditions are more credible than others, at

least as measured on the aforementioned rating scale. These results

are strengthened in a study by Nau, Caputo, and Borkovec (1974) using

Ornes' (1965) simulation procedure. In this investigation, snake

phobic subjects, following pretesting on a behavioral avoidance task,

were exposed to a ratiom~le and procedural descriptions of: (1) sys-

tematic desensitization (2). attention placebo, (3) implosive therapy,

(4) tachistoscope placebo, (5) relaxation plus recall strategies.

Subsequent to the rationale and description, subjects were asked to

simulate on the posttest the effects they would expect to occur had

they undergone five sessions of the described technique. Based on

three investigations, the hypothesis that variability in self reported

confidence in different treatment conditions is related to different

29

demand characteristics for improvement in the treatment procedures

themselves was supported. That is, in all three studies credibility

ratings of the rationales correlated significantly with simulated

treatment outcome. It seems apparent that efforts to guarantee equi-

valence of credibility among therapy and control conditions are criti-

cal. It should be noted that these findings do not show that treatment

effects are due to credibility of treatment procedures. The different

treatments might be equally effective for different reasons. However,

they do indicate that claims attributing therapeutic effects to spe-

cific treatment ingredients may not be valid. Within the context of

the present discussion, this may mean that live relaxation may indeed

be more credible than taped or self relaxation, particularly consid-

ering the traditional procedure of experimenter presence for live re-

laxation and experimenter absence for taped relaxation training.

Hence, physiological and subjective changes in measures of anxiety

favoring live relaxation may be due to expectancy factors.

Present Investigation

The purpose of the present investigation is to compare, on self

report and physiological indices, live, taped, and self relaxation.

Changes from the previous studies investigating live versus taped

relaxation include: (1) the use of an "anxious" population assessed

objectively, based on the recommendations of Borkovec and Sides (1979)

and Shapiro and Lehrer (1980); (2) the use of multiple physiological

measures as dependent variables, as a means of assessing changes in

more than one physiological response system (i.e., cardiac and somatic);

30

(3) assessing the role of expectancy across treatment modalities

(live, taped, and self); (4) the use of between session self monitor-

ing across groups to ensure equivalent practice effects; (5) control~

ling for the experimenter absent/present by having the experimenter

present across groups for training; and (6) soliciting ratings of

relaxation across taped and live groups to assess differences in sub-

jective relaxation as movement is made through the 16 muscle groups.

The latter two inclusions are based on hypotheses related to the

superiority of live relaxation training. More specifically, Paul and

Trimble (1970) support the notion that the lack of response contingent

progress in the recorded mode may be critical. That is, while subjects

in live relaxation training move to the next muscle group contingent

upon reported relaxation of the current muscle group, subjects in taped

conditions progress with the recorded instructions. Borkovec et al.

(1978), however, postulated, based on their results, that the presence

of the experimenter may be the critical component. A review of the

four studies investigating taped versus live progressive relaxation

training reveals that the experimenter is absent for all taped condi-

tions,and present for all live instructions. Thus, the present inves-

tigation will elicit subjective ratings of each muscle group prior to

movement into the next group with the aim of comparing these ratings

across groups. This will provide a test of the response-progression

hypothesis.

Contrasting these inclusions with prior investigations, two of the

studies have used an "anxious" population (Bieman et al., 1978; Israel

31

& Beiman, 1977). However, the weaknesses of screening procedures in

these studies have been discussed previously. The use of multiple

physiological measures has been employed on only two of the four

studies (Beiman et al., 1978; Paul & Trimble, 1970), while the assess-

ment or control of experimenter present/absent effects, between ses-

sion self-monitoring, expectancy factors, and subjective ratings of

successive muscle group relaxation have been completely neglected in

prior research.

In sum, the chief issues this proposal will explore are: (1) Can

taped progressive muscle relaxation and/or self relaxation be as ef-

fective as measured by physiological and self-report measures of re-

laxation as live progressive muscle relaxation? (2) Is the degree of

subjective and/or physiological relaxation contingent upon client

expectancy factors regardless of relaxation instruction provided?

In addition, the effects of experimenter absence/presence and

the differences in subjective relaxation ratings across treatment

conditions will be investigated. This latter factor will help to de-

termine the influence of response progression or therapist presence/

absence influences on progressive muscle relaxation training.

Method

Subjects

The 54 participants in this study were selected from the under-

graduate psychology mass testing session at Virginia Polytechnic Ins-

titutute and State University. All students involved in this testing

(N = 262) were administered the S-R Inventory of General Trait Anxious-

ness (Endler & Okada, 1975), a multidimensional test of trait anxiety.

This instrument employs a sample of four general situations (interac-

tions with others, physical danger, novel situations, daily routine);

and nine modes of response to each of the four situations. These re-

sponse modes include: Seeks experiences like this, perspire, have an

"uneasy" feeling, feel exhilarated and thrilled, get fluttering feel-

ing in stomach, feel tense, enjoy these situations, heart beats faster,

and feel anxious. Subjects are asked to respond to this 36 item

instrument on a 5-point scale ranging from "not at all" to "very much"

in terms of the intensity of the particular response (see Appendix B}.

Over a series of investigations (Endler & Hunt, 1966; Endler & Okada,

1975), reliability for the situations have been consistently high,

while strong evidence for the validity of the inventory as a multidi-

mensional measure of trait anxiety exists.

Those students scoring highest on the SR-GTA, representing a high-

er anxiety level, were offered a position in the investigation in re-

turn for receiving three introductory psychology credits. Other in-

clusion criteria included: (1) no prior participation in previous

experiments dealing with relaxation or experience with relaxation

32

33

procedures; (2) no prior or current history of heart disease; (3) no

current ingestion of tranquilizers or prescribed medication which

might interfere with relaxation procedures. The 54 subjects included

in the present study were among the top 66 scorers on the SR-GTA, with

10 participants in the mass testing sessions solicited unable to par-

ticipate in the study for a variety of reasons, while two prospective

subjects were excluded for prior experience with relaxation procedures.

The 54 subjects participating in the present investigation repre-

sented the top 25.2% of scores among the total subject pool of 262

students. The overall mean of students participating in the mass test-

ing session was 98.90 with a standard deviation of 14.48. This close-

ly parallels Endler and Okada's (1975) mean across sexes of 98.75. In

contrast, the range of scores among subject participating in the

present study was 106-141, with a mean of 116.38.

The students selected as subjects were stratified according to

their scores on the S-R GTA (top, middle, bottom) and randomly assigned

to one of the three groups constituting the live, taped, self component

of the design. For assignment to high or low expectancy groups, a high

or low slip was randomly selected followed by selection of one slip

from each of the top, middle, and lower SR-GTA scores within the live

group. This procedure was followed for the taped and self conditions

until three of the six slots within each category (live-top 1/3, live-

mid 1/3, live-low 1/3, etc.) were filled with either high or low expec-

tancy conditions. At this point, selection of high or low expectancy

34

conditions were based on equalizing expectancy conditions within each

of these categories.

Design

This investigation involved a (3 X 2 X 2) mixed design, with the

factors represented by treatment modality (Le., live, taped, self

relaxation training), expectancy conditions (high, low),and number of

sessions. Thus, across both sessions, all 54 subjects were assigned to

the following conditions: (1) live training/high expectancy; (2) live

training/low expectancy; (3) taped training/high expectancy; (4) taped

training/low expectancy; (5) self training/ high expectancy; and (6)

self training/ low expectancy. Each group was balanced according to

initial level of anxiety,based on SR-GTA scores.

Apparatus

The heart rate and finger pulse volume in this investigation were

recorded from the subject's right thumb on a Lafayette model 76604 re-

flectance photoelectric plethysmograph coupled with a Lafayette model

76406 amplifier and model 76102-10 Data Graph Systems polygraph. The

skin resistance response was measured on the right palmar site, spe-

cifically, one 76602 model silver-silver chloride electrode on the

thenar eminence area and one on the hypothemar eminence. Subjects

were seated in a comfortable reclining chair throughout each session.

Therapists included a second year Clinical Psychology graduate

student, who was familiar with the rationale and implementation of pro-

gressive muscle relaxation training, and two upper level undergraduate

students (1 male, 1 female), who volunteered for the study as part of

35

an independent study project. The students were trained in relaxation

training via written material delineating the muscle groups to be re-

laxed and relaxation instructions to be used during the study. The

total time invested in training the two therapists consisted of ap-

proximately 4 hours and, in addition to the material noted above, in-

volved familiarization with the physiological recording equipment and

procedural aspects of both scheduled sessions. A step-by-step pro-

cedural manual was provided for each session and kept in the equipment

room for review as needed. The senior investigator was available dur-

ing the first session for each of the junior investigators to provide

assistance as needed. In addition, 12 procedural folders, represent-

ing all groups across both sessions, were available with questionnaires

and instructions (e.g., expectancy conditions, pre-relaxation instruc-

tions) in sequential order to provide procedural structure for the re-

searchers involved. In terms of group assignment, therapists were

assigned randomly and counterbalanced across all groups. Each thera-

pist was randomly assigned to three subjects within each treatment

condition (live/high, live/low, taped/high, taped/low, self/high,

self/low) and initial scoring level on the SR-GTA. The latter was done

as a means of balancing the degree of subject's anxiety across thera-

pists. Finally, experimenters were assigned to subjects across both

treatment sessions. That is, an experimenter who ran a particular

subject in Session 1 remained with that particular subject for Session

2.

36

Dependent Variables

Dependent variables in this investigation directly assessing anxi-

ety reduction consisted of both physiological and self report measures.

The self report measures included: (l) subjective relaxation ratings on

a 1 (very relaxed) to 10 (very tense) scale during the relaxation training

itself~ and (2) the Anxiety Differential (Husek &Alexander, 1962). Physi-

ological measures included heart rate, finger pulse volume, and spon-

taneous fluctuations in skin resistance (SF). In addition to the meas-

ures noted above ( self report ratings, Anxiety Differential, heart rate, SF,

fingerpulsevolume) directly assessing the anxiety construct, subjects

were also administered the Borkovec and Nau (1972) Expectancy Scale which

served as a manipulation check to ensure that subjects assigned to high

and low expectancy conditions did, in fact, differ on this dimension.

All physiological data were scored during a 1 minute baseline peri-

od at the termination of the relaxation training. Heart rate was mea-

sured by beats/minute, and finger pulse volume amplitude measures were

. derived by measuring (in mm) every fourth heartbeat and dividing by the

number of heartbeats per period, thus deriving a mean amplitude over the

two sampling periods. Skin fluctuations involved a frequency count for

each period requiring at least a 1 mm deflection to be scored as a

fluctuation. Change scores were used to analyze the physiological mea-

sures (presession-postsession values) with the exception of finger pulse

volume amplitude. To correct for the wide range of amplifer sensitivi-

ties employed in recording finger pulse volume, the following formula

was used: post-presession/pre-session values.

37

Procedure

Following assignment of subjects to groups (N = 9), each subject

was scheduled for two relaxation sessions of approximately 1 hour at a

1 week interval. During the first session, the designated therapist

greeted the subject. The subject was then taken to the relaxation

chamber. A consent form and self monitoring contract were signed by

the subject (see Appendix C), and the subject received a brief descrip-

tion of the relaxation procedure appropriate for his/her group assign-

ment and expectancy condition rationale. The physiological measures

and equipment were then explained by walking the subjects to the equip-

ment room and describing the measures to be taken and the form of the

data to be gathered from the physiological monitoring equipment (i.e.,·

SF electrodes and finger-wrap for finger pulse volume and heart rate).

Following attachment of the physiological monitoring equipment, an ex-

pectancy questionnaire was administered consisting of five items rated

on a 10-point credibility scale. The Borkovec and Nau (1972) expect-

tancy assessment instrument, which emphasizes the elimination of

speech anxiety, was altered to emphasize learning the ability to relax

away general tension and anxiety (see Appendix D). At the same time,

the self report measure of anxiety chosen for the present investigation,

the Anxiety Differential, was administered. Following the completion

of these instruments, a 5 minute adaptation period was begun with the

last minute used as the pretreatment baseline for physiological meas-

ures (Period 1). During this time, the subject was asked to sit

quietly with eyes closed for instrument calibration. Following this

38

post-adaptation period, the pre-relaxation instructions were given for

each group followed by the beginning of relaxation proper. During the

relaxation training, subjects were intermittently requested to rate

the level of relaxation in the particular muscle group being tensed

and relaxed at the time of request. In the live and taped condition,

this occurred following the second tense-release cycle of the muscle

group, while in the self relaxation group this occurred every other

minute (i.e., 1, 3, 5, ..• 29). Following relaxation proper, the

posttreatment Anxiety Differentialwasadministered, electrodes detached,

and a discussion of the training (maximum duration= 5 minutes) ini-

tiated. For the latter procedural aspect, three identical open-ended

questions were given across groups (see Appendix E). At this point,

one self monitoring form was distributed and discussed. An example was

provided on the sheet demonstrating appropriate completion of the

form. Requirements included practicing once daily for 20 minutes and

a signature (pledge) for each daily entry confirming the fact that re-

laxation practice had occurred. These sheets were deposited daily

(not including weekends) in a box in the lobby of the fifth floor of

Derring Hall,where the subjects also picked up a new sheet assigned

for that day's practice. For a summary outline of the experimental

procedures for Session 1, see Appendix F.

Session 2 was identical to Session l1but without the introduction/

rationale, administration of the expectancy questionnaire, introduc-

tion to the physiological monitoring equipment, and disbursement and

explanation of the self monitoring forms. A 5 minute pre-session

39

discussion occurred to review self monitoring forms. Again, three

open-ended questions were used with the therapist inquiring: "How did

you find the relaxation during the week compared to the first session

of training?", "How did the relaxation sessions go during the week?",

"What questions do you have either about the forms you have been using

or the relaxation you have been practicing?". An addition to Session 1

included a debriefing for low expectancy subjects in the live and

taped conditions. Subjects in these groups were informed that progres-

sive relaxation training has been shown to be a rather effective pro-

cedure in learning to reduce anxiety and that the information provided

to them regarding its ineffectiveness was to evaluate the influence of

expectancy factors. For a summary outline of the experimental pro-

cedures for Session 2 across groups, see Appendix G.

Live relaxation groups. Each subject in the live relaxation con-

ditions, either high (N = 9) or low (N = 9) expectancy, was given the

following general relaxation instructions/rationale.

For this study, we are interested in examining the process of progressive muscle relaxation. Basically, progressive muscle relaxation training consists of learning to sequentially tense and then relax various groups of muscles all through the body while at the same time paying very close and careful attention to the feelings associated with both tension and relaxation. The tensing of the muscle prior to letting them relax is like giving ourselves a "running start" toward deep relaxation through the momentum created by the tension re-lease. Another important advantage to creating and releasing tension is that it will give you a good chance to focus your attention upon and become clearly aware of what tension really feels like in each of the various groups of muscles we will be dealing with. In addition, the tensing procedure will make a vivid contrast between tension and relaxation and will give you an excellent opportunity to directly compare the two and appreciate the difference in feeling associated with each of these states. While you are relaxing, I will be

40

monitoring your heart rate, skin conductance, and finger pulse volume. Finally, after the relaxation session, in about 45 minutes, we will get a chance to talk about how you felt during the relaxation exercise. Do you have any questions?

In addition to the above general rationale provided to the sub-

jects in the live condition, pre-relaxation instructions (i.e., follow-

ing completion of the physiological adaptation period and self-report

measures and irmnediately prior to relaxation proper) were:

As I described before, the procedure we will be using is called progressive relaxation training which consists of learning to tense and release various muscle groups through-out the body. I will be asking you to tense a particular muscle group for about 5 seconds and then to relax that muscle group for 30-45 seconds. We will go through e.ach muscle group twice. It is important to remember to release the muscle tension immediately rather than gradually, when I say the word, "relax." Also, once a group of muscles is relaxed, do not move it unnecessarily (except to make your-self more comfortable). Finally, after each muscle group, I will ask you to rate on a scale of 1-10 how relaxed that muscle group is. A rating of 1 will indicate that the mus-cle group is completely relaxed, while a rating of 10 will indicate a great deal of tension in that muscle group. Re-member, a 1 means that you are thoroughly and completely relaxed, while a 10 means that you are very tense. Do you have any questions?

The relaxation training in the live groups was response contingent.

That is, prior to progression to the next muscle group, the subject was

required to verbally report his level of relaxation on a l(very re-

laxed) to lO(very tense) scale. A rating of at least 3 was required to

progress. Ratings of 4 or above after the second tension release cycle

prompted a third presentation of the tension-release cycle for the

particular muscle group involved. If, following a third presentation,

self report indicated a tension level of 4 or above, training

progressed to the next muscle group. The relaxation procdur~ itself

41

involved 5-10 seconds of tensing for each muscle group followed by .30-

45 seconds of relaxation within.that muscle group. Subjects received

a list of muscle groups to relax included on their self monitoring

sheets (see Appendix H).

Taped relaxation groups. As in the live condition, subjects in the

taped relaxation conditions (both high expectancy [N = 9] and low ex-

pectancy [N=9] conditions) were presented with a general instructions/

rationale regarding progressive muscle relaxation and group specific

pre-relaxation instructions. These were identical to the live condition

in every respect other than the mode of presentation involved. That is,

subjects were told that they would be listening to a tape directing them

to tense and relax the particular muscle groups, and, wit:h regard to. the

pre-relaxation instructions, informed that the tape would request the

subjective relaxation ratings (see Appendix I). The other procedural

difference between this group and the live relaxation group was the

lack of response contingent progression. Progression to muscle groups

was program contingent in that the tape dictated when new muscle groups

were introduced for relaxation. Subjects in this group progressed to

the next muscle group regardless of these ratings. The tapes them-

selves consisted of 5-10 seconds of tensing instructions for each

muscle group followed by 30-45 seconds of relaxation and a prompt to

rate the relaxation level of the particular muscle group. The relaxa-

tion "patter" was identical across live and taped training conditions.

That is, the content of the relaxation instructions on the tape matched

matched those on the instructional sheet used during live relaxation.

42

A series of six different phrases were used systematically across

muscle groups in both groups to control for differential relaxation in-

structions across groups and amount of relaxation instructions (see Ap-

pendix J). As in the live relaxation group, subjects received a list of

muscle groups to relax included on their self rronitoring sheets.

Self relaxation groups. In this group, subjects were told to relax

in any way they chose but not to fall asleep. Verbal self reports of

relaxation level were requested every 2 minutes beginning with the

first minute (i.e., 1, 3, 5, ••• 29) to control for attending to and

rating relaxation levels as done in the live and taped groups. Self

monitoring forms in this group did not list muscle groups to attend to

in relaxing their bodies. The forms were more global in nature rather

than referring to particular muscle groups (see Appendix K). For the

self/high and self/low groups, the following general rationale was

provided:

For this study, we are interested in examining the process of relaxation. Since different people relax in different ways, you will be asked to relax in any way that you would like without falling asleep. In other words, any thoughts or images which help you to relax can be used during the relaxation exercises. While you are relaxing, I will be monitoring your heart rate, skin conductance, and finger pulse volume. Finally, after the relaxation session, in about 45 minutes, we will get a chance to talk about how you felt during the relaxation exercise. Do you have any questions?

In addition to the foregoing, the self relaxation groups also re-

ceived pre-relaxation instructions immediately prior to relaxation

proper:

As mentioned before, we are interested in the process of relaxation. When I ask you to begin, just relax yourself

43

in any way you feel will be effective. Please do not move around unnecessarily (except to make yourself comfortable), or fall asleep during relaxation. Finally, at varying in-tervals I will ask you to rate how relaxed you feel, on a scale of 1 to 10. A rating of 1 will indicate that you are completely relaxed, while a rating of 10 will indi-cate that you are very tense. Remember, a 1 means that you are thoroughly, completely relaxed, while a 10 means that you are very tense. Do you have any questions?

High expectancy conditions. Across live, taped, and self relaxa-

tion instructions, 27 subjects received the following high expectancy

instructions immediately following the general relaxation instructions/

rationale.

The relaxation training you will undergo has been proven to be a very effective procedure in a lot of problem areas. People using relaxation have been able to sleep better, elimi-nate headaches and, more generally, report feeling much more relaxed on a day to day basis. Since the evidence for the beneficial effects of relaxation is abundant, there is little doubt that you will get some benefit out of this relaxation practice. You will also very likely notice some beneficial effects very quickly and will feel very relaxed and calm dur-ing your relaxation exercises.

Low expectancy conditions. Across all training conditions, 27 sub-

jects assigned to this ·condition received the following low expetancy

instructions immediately following the general relaxation instructions/

rationale:

It will very likely be the case that this relaxation pro-cedure will be of little benefit to you. The evidence demonstrating a positive effect for the relaxation train-ing you will be undergoing is very weak. You have been assigned to what is called a "control" group for this study. Basically, this means that the relaxation training you will be given is not of proven effectiveness but does control for the time you will spend in the relaxation room, practice in relaxation during the week, etc. In sum, the relaxation procedure you will be undergoing simply has not been shown to be an effective means of learning relaxation, but we are interested in studying this particular method of relaxation.

Results

Pretreatment Measures

The S-R Inventory of General Trait Anxiousness was used in this

investigation as a screening instrument to select high anxious subjects.

To assess whether subjects, across conditions, were equivalent on the

measure of anxiety, a 3 X 2 (treatment X expectancy) ANOVA was per-

formed. No significant differences were observed (see Appendix L, Table

1), indicating that subjects were indeed equivalent on this measure

across treatment and expectancy conditions. Mean scores for each ex-

perimental condition are presented in Table 2.

To determine if subjects in the two expectancy conditions, high

and low, did differ on the Expectancy Questionnaire, at-test was per-

formed on subjects' scores between the two conditions. Those subjects

in the High Expectancy condition did score higher on the Expectancy

Questionnaire (X = 37.37) than those subjects in the Low Expectancy

condition (X = 29.18). Thus, the high and low expectancy instructions

did establish different expectancies regarding the effectiveness of re-

laxation between the two conditions (t = 3.82, R < .003).

Differences in baseline measures for heart rate (Session 1 = HRll,

Session 2 = HR21), finger pulse volume (FPll, FP21), spontaneous fluc-

tuations in skin resistance (SFll, SF21), and the Anxiety Differential

(ADl, AD3) were tested across Treatment and Expectancy conditions.

Figure 1 summarizes the results of the 3 X 2 (treatment X expectancy)

ANOVA, indicating a significant difference among groups for the Session

1 baseline heart rate data (HRll), with the Live High and Taped High

44

45

Table 2

Mean Scores' for SR-GTA

Treatment Expectancy Group Condition Score

Live High 119.11

Tape High 113.66

Self High 117.77

Live Low 116.22

Tape Low 115.22

Self Low 116. 33

46

84

83

82

81

80

79

78

77 Tape High

76 r Q) 76.33 -g 75 -~ 74 ~

[/J 73 .µ m Q) 72 ~

(!) 71 .µ

~ 70 X

.µ 69 70.22 lo-I Live Low m Q) 68 ::i::

67

66

65

64

63 63.55 62 Tape Low

61-

60-

Live Tape Self Treatment Levels

Figure 1. Baseline heart rate~Session 1.

47

conditions exhibiting a significantly higher baseline rate. The analy-

sis demonstrated a significant effect on HRll for Expectancy, F (1, 48)

= 7.34, E. ).009, and a Treatment X Expectancy interaction,! (2, 48) = 3.44 £_).04 {see Appendix L, Tables 2 and 3). In addition, differences

existed for the heart rate baseline data in Session 2 {HR21). Spe-

cifically, the Taped High condition had a significantly higher baseline

heart rate than the Taped Low condition. Figure 2 illustrates this

Treatment X Expectancy interaction for the HR21 data,! (2, 48) = 5.39,

p).007 {see Appendix L, Table 4). Finally, there were a greater number

of skin fluctuation responses in the High Expectancy condition (X = 4.59) than in the Low Expectancy condition {X = 1. 79) for Session 2

baseline data (SF21), indicating a main effect for Expectancy,! (1, 48)

= 7.08, p) .01 (See Appendix L, Table 2). No other significant differ-

ences were found among baseline measures. Borkovec and Sides (1979), as

previously noted, postulated that progressive muscle relaxation is

superior to control conditions when high anxious subjects are used.

Also noting the potential importance of baseline differences on subse-

quent physiological changes, Wilder's_(l953) Law of Initial Values

would predict greater changes on physiological measures when baseline

values are elevated. Therefore, to assess the influence of these base-

line differences on subsequent physiological and self report changes,

subjects were split on baseline heart rate (HRll), and change scores

across dependent measures were analyzed via at-test procedure. Sub-

jects were divided into a high heart rate category (HR11)71) and low

heart rate category (HRll 4 71) based on the median heart rate values

48

84

83

82 X

81 82.00 Self Low

80

79

78 77

...... 76 76.44 Q) .µ

75 Live High ::I -~ 74 ~ ....... rn 73 .µ

Self High Ill Q) IXl 72 ...... Q) 71 .µ ;2 70 X

70.88 .µ Live Low M 69

Ill Q)

68 :::i:::

67

66 65 64

63 62 62. 77 61 Tape Low

60

Live Tape Self Treatment Levels

Figure 2. Baseline heart rate~Session 2.

49

of the Session 1 baseline period. No significant differences were found

on any measure due to initial heart rate (see Appendix L, Table 5). To

further control for the baseline differences noted above, change scores

were used to analyze changes in the main dependent variables in this

investigation (heart rate, skin fluctuation responses, finger pulse vol-

ume a.zt\Plitude, Anxiety Differential). It should be noted that finger

pulse change scores were derived by the following formula rather than

post-pretreatment values (post-pretreatment/pretreatment values). This

was done to correct for the wide range of amplifier sensitivities em-

ployed in recording finger pulse volume.

Analysis of Treatment Effects

Prior to investigating differences across dependent measures due

to training and expectancy variables, differences in change scores due

to the experimenter variable were tested. A one-way ANOVA was completed

.for both Sessions 1 and 2. As summarized in Appendix L, Table 6, finger

pulse change scores in Session 1 were significant,! (2, 51) = 7.56, £

<. 001. Nb .other change scores attained significance due to the experi-

menter variable. Due to the large number of comparisons made in this

investigation, this difference is seen as a chance finding and will not

be considered in further analyses.

To determine if differences among groups on the dependent measures

could be attributed to differential practice effects, differences among

groups on the return of the daily self monitoring sheets were investi~

gated. Testing for these differences·via a 3 X 2 (treatment X expec-

tancy) ANOVA found no differences among groups (see Appendix L, Table 7),

50

The number of sheets returned per subject ranged from 5.44 (Taped Low)

to 6.22 (Self Low) out of a possible 7 sheets.

One of the aims of the present investigation was to assess dif-

ferences across types of relaxation training in within session relaxa-

tion ratings. A 3 X 2 X 2 (treatment X expectancy X session) ANOVA was

performed on· the in-session relaxation ratings, with the dependent

variable consisting of the number of ratings above 3 on a 1 (relaxed) to

10 (tense) rating scale. Results indicated that subjects gave fewer re-

ports of residual tension during Session 2 than during Session 1, !. ( 1, 48) = 8. 90, £ ( . 004 ( see Appendix L, Table 8) • Further, within

session relaxation ratings were equivalent across treatment groups.. In

particular, relaxation ratings did not differ between the live and taped

groups. Table 3 presents the means for each of the Treatment X Expec-

tancy cells for these ratings.

The main empirical question in the present investigation was

whether live relaxation was superior to taped and self relaxation and to

assess the role of expectancy across the relaxation training conditions.

For a graphic. summary of change scores for each dependent variable

across sessions, see Figures 3, 4, 5, and 6. Mean values for each de-

pendent measure are included in Table 4.

Due to the number of dependent variables selected, a multivariate

analysis of variance (~OVA) including treatment, expectancy, and ses-

sion variables was completed on change scores for the physiological mea-

sures and the main self report measure in the investigation, the Anxiety

Differential. Three scores for the SF measure were missing, resulting

51

Table 3

Relaxation Rating Means

Treatment Expectancy Session N *Relaxation

Group Condition Ratings

Live High 1 9 2.55

Live High 2 9 1.44

Live Low 1 9 4.77

Live Low 2 9 2.88

Tape High 1 9 6.22

Tape High 2 9 4.44

Tape Low 1 9 6.00

Tape Low 2 9 3.66

Self High 1 9 4.33

Self High 2 9 3.33

Self Low 1 9 6.44

Self Low 2 9 5.88

*Number of relaxation ratings greater .than 3 on 1 (very relaxed) to 10 (very tense) scale.

Session 1 X = 5.05 Session 2 X = 3.60

9-

8-

. 7-i:: ·.-1 0 ~ 6.66 ' 6-Ill 6.22 .l,J m ;< Q) 5-l:!l ,5. 77 ......

/ Q) 4- / t,'I i:: / ~ 3- / CJ / Q) / .jJ 2 / &! X .jJ 1 1.55 J..l m Q) ::i:: 0

-1

-2

L Live Live

Session 1 Session 2

5.~

X

2.44'

'

·Tape Session 1

' ' '

4.55

' ' ' 'x -1.44

'!'ape· Session 2

Figure 3. Heart Rate Change Scores

6.88

/ x"'

3.00

; /

.,,,/'

,,x ;5. 77

o = High Expectancy----

x = Low Expectancy - -

Self Session 1

Self Session 2

U1 r,.,

4.2 4.0 3.8 3.6 3.4

II) 3.2 (I) I,.! 3.0 0 0 2.8 Cl) ....... 2.6 (I) . g' -~ 2.4 Ill ::E: 2.2 tj ~ 2.0 ~ § 1.8 0 ·rl

.... +l 1.6 +l Ill ~ .B

1.4~ 1.50 ... ........

+l 0 1.2 ~

8 .:1 1.0 1.22 r-1 r.. r.. ..... .8 -~ .6 ~ .4 Cl)

.2

.0 -.2 -.4 -.6

I I Live Live

Session 1 Session 2

2.55

o = High Expectancy----x = Low Expectancy - - -

~ 2.33, 2.22 ', ....

',, ' 'x

1.37

2.22 X ' 2.00

\

Self

\ \

'

Tape Session 1

Tape Session 2 Session 1

Figure 4. Skin fluctuation change scores.

' U1 \ w ' \ ' \ \

\ \

\ \

\ 'x

-.so Self

Session 2

10

(I) 9 °' X 8.55 (I)

H 9.33 ,9.11 0 '\x,/ ~55 u 8 U) x---------iC (I) 8.00 8.00 tn 7 7.66 §

,, ,, , ..c: , CJ 6

, ,, .--l ,, Cll X

·.-l 5 5.88 5.88 .µ ~ 5.33 (I) H 4 (I)

'1-1 X '1-1 ,, ·.-l

3 ., ., 3. 88

Q ., U1 ,, ~ >, ., o = High Expectancy~ .j.) .,

(I) 2 ,, ,, ·.-l ,, x = Low Expectancy - ~ >< x" ~ 1

1.11

Live Live Tape Tape Self Self Session 1 Session 2 Session 1 Session 2 Session 1 Session 2

Figure 5. Anxiety Differential Change Scores

Cll Q) 1-1 ...... 8 ~ tJl

i:: Q) ·ri t:11 i:: "d rt! Q) .c: :::s u .µ

·ri Q) ..... ~, p..

1-1 § ~~ i:: ....... .... r:r.i

.25

• 20.

.15

.10

.OS

.oo

-.OS~ -.10

-.15

-.20

-.25

-.30

I

u' -.OS I

X -.09

Live Session 1

I I

I I

X 119 r

I I

I I

Live Session 2

X .. .10,

' ' ' ' ' ' ' ' ' ' ' ' X -.11

o,:. _______ .,..,

-.22 -?24

Tape Session 1

Tape Session 2

o = High Expectancy--

x = Low Expectancy - - -

0... ... ..x -.oa ... ~, ,, ' -.oa

X' -.17

-.21

Self Session 1

Self Session 2

Figure 6. Finger pulse chanqe scores. (Positive scores represent vasodilation.)

U1 U1

Table 4

Mean Change Scores for Dependent Variables

Treatment Expectancy Session N HR SF FP AD Group Condition

Live High 1 9 6.66 1.50 -o.os 9.33

Live High 2 9 6.22 3.44 0.05 5.88

Tape High 1 9 5.66 2.55 -0.22 7.66

Tape High 2 9 4.55 2.22 -0.24 8.55

Self High 1 9 6.88 2.22 -0.08 7.11

Self High 2 9 8.11 3.88 -0.21 5.33 U1 O'I

Live Low 1 9 1.55 1. 77 -0.09 5.88

Live Low 2 9 5.77 1.22 0.19 9.11

Tape Low 1 9 2.44 2.33 0.10 1.11

Tape Low 2 9 -1.44 1.37 -0.11 3.'88

Self Low 1 9 3.00 2.00 -0.17 8.00

Self Low 2 9 5.77 -a.so -0.08 8.00

57

in unequal cell size. Since each missing value came from different

groups (i.e., Live-High, Taped-Low, Self-Low), values derived from the

mean for each group were substituted for the missing values. The MANOVA

results indicated an expectancy main effect (See Appendix L, Tjible 9).

That is, regardless of the type of training (live, taped, self), those

subjects in the high expectancy condition demonstrated greater change

toward a relaxed state than those subjects in the low expectancy condi-

tion. This main effect is further examined by looking at the univari-

ate analyses of variance for each dependent measure. As shown in

Appendix L, Table 10, there was a significant Expectancy effect for

heart rate with significantly higher heart rate change scores for the

High Expectancy condition, !_ (1, 48) = 10.94, g,( .001. In addition,

the SF change scores were significantly higher for the High Expectancy

condition, demonstrating a significant expectancy effect for this de-

pendent variable, !_ (1, 48) = 5.43, g_ ( .024. Although it should be

viewed very conservatively based on the fact that the MANOVA was sig-

nificant only for the Expectancy manipulation, the heart rate univari-

ate analysis of variance showed a trend toward a significant treatment

effect, F (2, 48) = 3.12, £( .053, with the Duncans Multiple Range Test

indicating a superiority of Live and Self conditions over the Taped

relaxation condition.

Finally, Tables 5 and 6 demonstrate a series of correlations among

self. report measures and between self-report and physioloqical measures,

respectively. More specifically,. correlations were m:,derately high and

very significant between the relaxation ratings (RRl = Session 1, RR2 =

58

Table 5

Correlation Matrix for Self Report Dependent Variables

Dependent Variables ADl AD2 AD3 AD4 ADCHl ADCH2 RRl RR2

ADI .68+ .61+ .40** -. 57+ · -.29* -.36** -.34**

AD2 .59+ .64+ .19 .09 -.56+ -.46+

AD3 • 76+ -.16 -.31* -.28* -.28* ' AD4 .17 .36** -.40** -.32**

ADCHl .50+ -.14 -.05

ADCH2 -.18 -.07

RRl .73+

*£ ( .05 **.2. < .01 +£<. .001

Dependent Variables

ADl

AD2

AD3

AD4

ADCHl

ADCH2

RRl

RR2

HRll

HR15

HR21

HR25

HRCHl

*£ <.. 05 **12. <.. 01

59

Table' 6

Correlation Matrix for Self ·Report/

Physiological Dependent Variables

HRll HR15 HR21 HR25 .HRCHl

-.29* -.25 "".'.00 .08 -.13

- .. 02 -.05 -.09 -.04 .06

-.13 -.02 .12 .26* -.29*

.07 .10 .23 .29* -.04

.36** .29* -.10 -.15 .24

.30* .18 .15 .05 • 34**

-.02 .07 .03 .03 -.22

.07 .13 .11 .10 -.12

.91** .41** .33* • 38**

.47** .43** -.OB

.85** .04

-.15

HRCH2

.19

-.10

-.23

-.08

.OB

.21

.oo

.03

.19

.12

.38**

-.15

.20

60

Session 2) and the Anxiety Differential within sessions. That i.s, the

fewer relaxation ratings above 3 on the within session ratings given

after each tense/relax cycle for each muscle group, the higher the Anx-

iety Differential scores (high scores indicating lower anxiety). Thus,

a relationship existed between self reports of muscle tension and anx-

iety. In addition, the pre-relaxation AD measures in seach session (ADl

= Session 1, AD3 =Session 2) were significantly and negatively correlated

with the Anxiety Differential change scores within that same session.

Thus, a relationship existed between scores indicating high anxiety

prior to relaxation (i.e., low scores) and change on this measure in

the relaxed direction. The more anxiety reported prior to relaxation,

the greater decrease across the session. Finally, with regard to cor-

relations between physiological and self report measures, only heart

rate correlated with the latter. More specifically, Table 6 shows a

moderately high and significant negative correlation between Session 1

baseline heart rate (HRll) and the pretreatment Session 1 Anxiety Differ-

ential (ADl). Thus, higher baseline heart rate in Session 1 was related

to lower Anxiety Differential scores (denoting hiaher anxiety). Also a

positive correlation between HR-11 and Session l Anxiety Differential

change scores (ADCH 1) was found. There were no significant correlations

between the in-sesison relaxation ratings and heart rate measures.

Finallv, none of the other physiological measures (SF, FPV) correlated

significantly with any self report measure or with heart rate measures.

However, SF change scores in Session 1 did correlate with finger pulse

change scores in Session 1 (r = • 36; J2. (. 007) •

Discussion

The chief concerns of the present investigation were with the

relative efficacy of live versus taped and self relaxation training pro-

cedures as measured by physiological and self report measures of relaxa-

tion and the assessment of the role of expectancy on these measures.

The present study found no differences among live, taped, and self re-

laxation procedures on any of the physiological measures used (heart

rate, finger pulse volume amplitude, skin fluctuation responses) or the

self report measure of anxiety, the Anxiety Differential. These results

are in marked contrast to those of Beiroan et al. (1978), Paul and Trim-

ble (1970), and Russell et al. (1976), who found live relaxation to be

superior to taped on both physiological and self report measures. How-

ever, the results concord ·with the Israel and Beiman (1977) finding that

live, taped, and self relaxation conditions all resulted in significant

heart rate decreases with no differences across groups. Although these

same authors found a significant difference favoring the live relaxa-

tion condition on the same self report measure of anxiety used in the

present investigation, the Anxiety Differential, this finding was not

replicated by these authors in a later investigation (Beiman et al.,

1978). Interestingly, it was Israel and Beiman (1977) who initially

presented their control or self relaxation condition as a potentially

effective treatment for tension and posited expectancy as a possible

explanation for the equivalent effects across groups. Miller and Born-

stein (1977) also found a self relaxation treatment group to be

61

62

equivalent to several other relaxation conditions on EMG and self re-

port measures, including a taped progressive muscle relaxation group.

Borkovec and Sides (1979) had posited, in their meta-analytic

review of progressive muscle relaxation training, that certain critical

procedural variables were related to effecting physiological changes

with progressive muscle relaxation. More specifically, studies demon-

strating progressive muscle relaxation training superiority over control

conditions had involved live relaxation procedures, with high anxious

populations, over several sessions, although the latter condition showed

a great deal of variability (X = 4.57, SD= 3.02). The present investi-

gation quite clearly challenges these findings. In the present investi-

gation, a high anxious population was used along with live relaxation

training. In addition, although only two supervised training sessions

occurred, subjects attested to the fact that practice occurred between

sessions, the frequency of which was equivalent across all treatment

conditions. Despite the presence of these conditions, the presumed

superiority of progressive muscle relaxation training over the self

relaxation condition was not demonstrated in this investigation.

Rather, the subjects' expectancy for change based on information pro-

vided by the therapists regarding the efficacy of the given relaxation

procedure was the critical factor in decreasing anxiety.

The Role of Expectancy and Other

Active Therapeutic Elements

In essence, the chief finding in this investigation involved the

critical role of expectancy in achieving decreases in physiological and

63

self report indices of anxiety. That is, regardless of the type of

training or whether explicit training existed at all, those subjects

who were informed, and based on the expectancy measure employed, be-

lieved that the particular relaxation procedure was an effective means

of achieving relaxation, did indeed show significantly larger decreases

on heart rate and spontaneous skin fluctuation measures than did low

expectancy groups.

The four investigations comparing live versus taped and self re-

laxation (Beirnan, Israel & Johnson, 1978; Israel & Beiman, 1977; Paul &

Trimble, 1970; Russell et al., 1976) did not investigate the role of

differing expectancies across treatment conditions. It seems very

plausible that a live relaxation condition could be viewed as a more

powerful intervention than either taped or self relaxation training.

As noted previously, Paul and Trimble 1 s (1970) instructions to subjects

certainly favored the live relaxation condition in terms of expectancy

for change. Perhaps the critical procedural variable of progressive

muscle relaxation is an expectancy for change rather than those varia-

bles postulated by King (1981) and Borkovec and Sides (1979). Investi-

gations by Borkovec and Nau (1978), McGlynn and McDonnell (1974) and

others in which active treatment conditions were indeed seen as more

credible and as fostering a higher expectation for improvement than

control conditions certainly support this position. Moreover, these in-

vestigations included as active treatment components the therapeutic

procedures of desensitization, of which progressive muscle relaxation is

an integral part. Thus, the increased credibility of these procedures

64

may have been due, in part, to the increased credibility of the progres-

sive muscle relaxation procedure.

In addition to the reasons noted above, perhaps all of the train-

ing conditions shared elements producing a relaxed state. Benson,

Beary and Carol (1974) delineated four basic elements of the relaxation

response which are necessary and sufficient to induce a relaxed state.

Specifically, these elements include a shift away from logical extern-

ally oriented thought, a passive attitude with attention directed toward

the relaxation technique, a comfortable position, and a quiet environ-

ment. These elements were included in all of the relaxation groups used

in this investigation. Finally, perhaps the choice of relaxation pro-

cedure is not critical with high anxious populations. That is, high

anxious subjects may relax so infrequently, either subjectively or

physiologically, that any technique producing a state of rest may induce

physiological and self report changes in anxiety level. This postula-

tion contradicts the evidence reviewed by Borkovec and Sides (1979),

but has some support in the work of Miller and Bornstein (1977) and

Israel and Beiman (1977), who found self relaxation with anxious indi-

viduals to be as effective as progressive muscle relaxation. In addi-

tion, Smith (1976), using subjects responding to advertisements for an

anxiety reduction program, found a transcendental meditation treatment

condition to be no more effective in reducing trait anxiety than a con-

trol treatment consisting of sitting without meditation. Unfortunately,

no physiological measures were implemented. Finally, in one of a series

of investigations, Cuthbert, Kristeller, Simons, Hodes, and Lang (1981)

65

found that subjects given only instructions to try to relax were as suc-

cessful at achieving heart rate reductions as either of two training

groups (heart rate feedback and meditation).

An interesting finding involves the Expectancy and Treatment X

Expectancy effects for heart rate baseline data, and a main effect for

expectancy for skin fluctuation response on Session 2 baseline data.

These baseline data were taken following the instructions,,and it seems

plausible that the expectancy manipulation influenced basal physio-

logical measures. Although this possibility should certainly be viewed

with caution, perhaps high expectancy instructions are detrimental ini-

tially by instilling a strong demand among subjects to change. In addi-

tion, high expectancy instructions given during a pretreatment phase may

prompt decreased effort. That is, if a treatment is proposed as an ex-

tremely efficacious one, the perception by the subject may be that ex-

tensive effort is not needed for change. In this investigation, high

heart rate and a greater number of skin fluctuation responses, indi-

cants of anxiety, were higher for those individuasl given high expec-

tancy instructions. Obviously, an alternative explanation to those

noted above is that the differences were in fact real basal level dif-

ferences, uninfluenced by expectancy manipulations. This question cer-

tainly deserves further investigation.

It is noteworthy that self report measures of anxiety .and muscle

tension did not differ among groups regardless of training or expectancy

manipulations. Reinking and Kohl (1975) had noted that if the desired

effect is subjective calm, then" ••• any procedure should work

66

equally well if 'sold' as a relaxation procedure" (p. 599). The present

investigation found changes in the relaxed direction across all condi-

.tions on the Anxiety Differential (see Figure 5 and Table 4). Thus,

perhaps no "selling" is necessary to facilitate subjective calm. It

may be that merely providing instructions to try to relax may be the

critical variable in producing decreases in subjective anxiety.

As noted, the within session relaxation ratings, while indicating

increased relaxation among muscle groups across sessions, did not differ

across live, taped, or self relaxation training conditions. These rat-

ings were included to test the response contingent hypothesis (Paul &

Trimble, 1970) as an explanation for previous findings of live training

superiority over taped training. In addition to this hypothesis, Borko-

vec et al. (1978) had postulated that experimenter presence in live

relaxation training and absence in taped training was the critical

variable leading to live relaxation training superiority. In the pres-

ent investigation, experimenter presence was equivalent across groups

with the experimenter present for all treatment conditions. As noted

in the present study, relaxation ratings did not differ between live

and taped conditions. The lack of superiority for live versus taped

relaxation, combined with the equivalence of the relaxation ratings

across live and taped conditions thus favors the Borkovec et al. (1978)

conceptualization stressing experimenter presence. That is, some aspect

of therapist presence may be critical in relaxation training. Albeit

this study does not allow for the specification of this aspect, it

should be noted that the therapists in the present investigation were

67

active in terms of explaining the relaxation process, physiological

monitoring equipment, self-monitoring homework, etc. and served a func-

tion during relaxation proper by recording the relaxation ratings given

after the second tension-release cycle of each muscle group. Cuthbert

et al. (1981) addressed this issue by actively manipulating the subject-

experimenter relationship across two training conditions designed to

decrease heart rate. The high-involvement procedure included considera-

ble general conversation between subject and experimenter, positive

support and information about performance, instr.uctions to practice at

home, and literature about the relationship between the experimental

task and relaxation. The low-involvement condition was defined as the

experimenter interacting no more than necessary to answer procedural

questions. Although the relationship was a complex one, results strong-

ly indicated that relationship factors and knowledge of results inter-

acted significantly, lending some support in favor of the experimenter

presence argument. Obviously, more definitive answers to this issue of

therapist presence would have been possible within the present investiga-

tion if live relaxation training had indeed proved superior to taped

relaxation. With the equivalence of these training modalities, the

above discussion should be considered cautiously.

This finding may, however, have implications for self-help thera-

pies, if indeed therapist presence proves to be a critical variable in

effecting changes in anxiety level. Without therapist involvement,

self-help strategies,particularly those dealing with tension and anxiety

may be lacking a critical component necessary for symptom reduction.

68

Unfortunately, as Foreyt and Goodrick (1979) note, there is a glaring

absence of data based evaluations of self-help treatment strategies,

thus precluding conclusions regarding this postulation.

Multidimensional Nature of Anxiety

As noted previously, the different response systems that define

anxiety have been found to be frequently discordant (Lang, 1978; Martin,

1961). The present investigation provides little challenge to these

consistent findings. More specifically, in the present investigation,

while heart rate and skin fluctuation measures showed significant dif-

ferences due to the expectancy manipulation, self reports of anxiety

did not. There were, however, some rather interesting relationships

among the dependent measures. Correlations between the anxiety self

report measure in this investigation (Anxiety Differential) and the re-

laxation ratings were significant and negative. More specifically,

posttreatment anxiety reports in Sessions 1 and 2 (AD2, AD4) were very

significantly correlated with RR! and RR2 respectively (r = -. 56, E_(.001;

r = -.32, 12.(.0l). That is, high scores on the AD indicating low

anxiety, were correlated with fewer relaxation ratings greater than 3

on the 1 (relaxed) to 10 (tense) scale used to rate muscle relaxation.

This has clinical relevance, in that, with replication, within session

relaxation ratings may be an accurate indicant of subjective anxiety

level. This is particularly encouraging when considering the covert

nature of the Anxiety Differential, as substantiated by the finding

previously noted that anxious subjects instructed to "fake good" have

69

been found to display higher anxiety scores than a non-anxious control

group.

In addition, baseline heart rate was significantly and positively

correlated with the Session 1 Anxiety Differential Change Score (ADCH 1).

The fact that heart rate did indeed.correlate with. self report, along

with significant decreases in this dependent variable due to the expec-

tancy manipulation and a trend toward training modality differences

(i.e., live and self> tape) indicates the sensitivity of this measure

in the present investigation. The sensitivity of the heart rate measure

in studies of anxiety has been noted previously (Agras, 19811 Bo~kovec

et al., 1977). Indeed, Agras (1981) in ranking a variety of measures

in order of increasing sensitivity to change in anxiety-arousing situa-

tions, found heart rate to be the most sensitive physiological measure.

Other correlations between heart rate measures and self report measures

were consistently higher than other physiological-self report pairings,

although,when significant, were low (.20's). This finding, added to the

lack of correlation among physiological measures and the significant

expectancy effects with heart rate and skin fluctuation response but

not finger pulse volume amplitude, is typical of research findings both

across (i.e., self report, physiological,behavioral) and within (e.g.,

different physiological measures) response systems. Martin (1961)

notes that one reason for the lack of intercorrelations among physio-

logical measures stems from Lacey, Bateman and Van Lehn's (1953) finding

of individual patterns of autonomic responses. Certain individuals may,

for example, respond to stress by increased heart rate and a small

70

change in skin conductance while another may respond with the opposite

pattern. Lang (1978) concurs noting:

Despite the ubiquity of relevant physiological measures, empirical investigations have also shown that the group intercorrelations among these measures are remarkably low, and that the shared variance within the physiological system seldom exceeds 10 or 15%. (p. 383)

The lack of change in the finger pulse volume amplitude measure

warrants some discussion, particularly considering the significant

change scores on the heart rate and skin fluctuation response measures

related to the expectancy factor. In addition to the response stereo-

typy noted above, several other factors may be related to this lack of

effect. First, room temperature and humidity must be closely regulated

in the measurement of peripheral blood flow, regulation not possible

in the laboratory in which this study was conducted. Second, position

of the transducer is critical when sites such as the finger are em-

ployed (Jennings, Tahmoush, & Redmond, 1980). Thus rather minute varia-

tions in position placement among subjects may have contributed to non-

significant effects. Third, none of the four relaxation investigations

contrasting live versus taped relaxation training used this cardiovascu-

lar measurement. Moreover, only three studies investigating progressive

muscle relaxation have used peripheral blood flow, with two investiga-

tions using forearm blood flow as the physiological measure of choice

(Lader & Mathews, 1970; Mathews & Gelder, 1969), with neither study

finding differences between progressive muscle relaxation and control

groups. In the third investigation, Van Egeren, Feather, and Hein

(1971) found that relaxed subjects showed less decrease in skin

71

resistance to phobic stimuli than those who were not relaxed but no

differences in finger pulse amplitude between relaxed and unrelaxed sub-

jects. Finally, Kallman and Feuerstein (1977) note that autonomic in-

nervation of the blood vessles is poorly understood. To illustrate this

confusion, Lader and Mathews (1970) reported a subject who showed in-

consistency within the cardiovascular system, with a rise in forearm

blood flow and a drop in heart rate.

The failure of the self report measures (~nxiety Differential,

relaxation ratings) to correlate with the physiological measures in the

present study is also no surprise. Indeed, Parloff, Waskau and Wolfe

(1978) note that a repeated finding in psychotherapy is that different

measures of even the same criterion fail to show high correlations.

This oft-repeated finding may reflect the role of individual differ-

ences in patterns of anxiety responses and also the complexity of the

construct of anxiety. The latter is illustrated by the fact that

anxiety has been u.ewed as a behavior ("doesn't he look anxious"), a

trait ("he's an anxious individual"), an explanation of the behavior ("he

smokes because he is anxious"), and a state ("I feel anxious doing

this"), the latter influenced by a host of situational variables. It is

obvious that anxiety is a multidimensional construct, and the definition

or conception of anxiety as unitary (e.g., an emotional state) is not

sufficient.

Caveats and Future Directions

Finally, there are several procedural variables to be considered

prior to generalizing the present results beyond this investigation.

72

First, volunteers were solicited for this investigation with subjects

receiving extra points in an introductory psychology course for parti-

cipation. However, these participants did score substantially above

the norm for normal male and female samples on the anxiety measure used

for screening purposes, the S-R Inventory of General Trait Anxiousness

(Endler & Okada, 1975), indicating that the subjects used in this in-

vestigation were indeed high anxious subjects.

A number of investigations have examined the effects of progres-

sive muscle relaxation on psychophysiological responding to stress in-

ducing stimuli with inconsistent results (King, 1981). This question

was not a focus of the current investigation. Therefore, results should

not be generalized to situations involving the express presentation of

anxiety-eliciting stimuli during the relaxation procedure.

Therapists actively interacted with the subjects before and after

each session and were active during relaxation proper, eliciting relaxa-

tion ratings. Since therapist presence may indeed be a critical varia-

ble in the efficacy of the relaxation training procedure, any or all of

these components may be critical. Relaxation practice between sessions

was stressed, with participants filling out daily relaxation rating

sheets requesting them to rate their ability to relax. This practice

and the emphasis given it by the therapists involved may also be an

important factor in acquiring relaxation skills.

The expectancy conditions instructions (i.e., high and low) in

the present investigation were extremely disparate limiting generaliza-

tion to investigations implementing expectancy instructions of equal

73

disparity. That is, low expectancy subjects were told that they were in

a control group and that the evidence did not support the use of the

particular procedure, etc., while high expectancy subjects were told

that the relaxation procedure was of proven effectiveness and that they

would undoubtedly receive benefit from the relaxation training.· The

question remains as to whether the same effects would be found if sub-

jects were given moderately vs. extremely high expectation or counter

demand instructions (i.e., informed that progress would not occur for

several weeks). As Kazdin (1979) suggests, assessing expectancy at

different times within treatment (e.g., post-instructions, post-session

1) may also provide valuable information regarding the role of this

construct in the relaxation process by noting changes in expectancy ~nd

concurrent changes in self report and physiological indices. There is

no question that more data are needed to confirm the role of expectancy

in relaxation training. However, the present investigation does cer-

tainly corroborate the findings of others who have suqgested that the

role of expectancy in a host of psychotherapeutic procedures may be a

critical one (Borkovec & Nau, 1972; Israel & Beirnan, 1977).

It is possible that extending the number of sessions beyond the

two used in this study would result in differential results among

training conditions. This possibility is supported by the findings of

Borkovec and Sides (1979), although the present investigation found no

changes due to the sessions variable on physiological or self report

measures of anxiety. Finally,_ the present investigation did not in-

volve a follow~up session. Such a session may well produce different

74

results across treatment conditions. Indeed, Beiman et al. (1978) did

find some differences across physiological measures during a oost-

treatment session. Replications of the present research should certain-

ly investigate this issue.

Concluding Remarks

In sum, this investigation, using high anxious subjects under-

going two relaxation sessions under either live, taped, or self relaxa-

tion training found that information provided to the subject regarding

the efficacy of the procedure was the critical independent variable in

two physiological change scores, heart rate and skin fluctuation re-

sponses. There were no differences between groups in within session

relaxation ratings or with1he self report of anxiety as measured by the

Anxiety Differential. However, relaxation ratings did decrease across

sessions, and the Anxiety Differential change scores did indicate a

decrease in anxiety in all treatment groups.

As noted above, self report measures (relaxation ratings, Anxiety

Differential) were equivalent across groups while physiological mea-

sures (heart rate, skin fluctuation response) showed a significantly

greater decrease under high expectancy conditions. Thus, the possi-

bility exists that the active elements needed to reduce subjective com-

ponents of anxiety may be different than that for physiological compon-

ents. This response fractionation certainly supports the multidimension-

al view of anxiety.

Further investigations are needed to delineate the necessary com-

ponents for the influential role of therapist presence in relaxation

75

training and to further substantiate the chief finding of this investi-

gation, i.e., the critical influence of expectancy in the relaxation

process. Investigations should also include incorporation of a measure

of tonic muscle tension (EMG). This measure may prove to differentiate

live versus taped and self relaxation since progressive muscle relaxa-

tion direcly deals with the tensing and relaxing of different muscle

groups, albeit some evidence exists (Israel & Beirnan, 1977; Beiman et

al., 1976) incidating that self relaxation training can be as effective

on this measure of anxiety.

References

Agras, S., & Jacob, R. G. Phobia nature and measurement. In U. Mavis-· sakalian, & D. Barlow, Phobia. New York: Guilford Press, 1981.

Baker, B. L., & Kahn, M.A. A reply to "Critique of 'Treatment of in-somnia by relaxation training': Relaxation training, Rogerian therapy, or demand characteristics." Journal of Abnormal Psychology, 1972, J..:i, 94-96.

Beiman, I., Israel, E., & Johnson, S. During and post-training effects of live and taped extended progressive relaxation, self relaxation and electromyogram feedback. Journal of Consulting and Clinical Psychology, 1978, 46, 314-321.

Benson, H. , Beary, J. F ~ , & Carol, M. F. The relaxation response. Psychiatry, 1974, 37, 37-46.

Bloom, L. J., Houston, B. K., & Burish, T. G. Psychophysiology, 1976, 13(1), 40-42.

Bloom, L. J., & Trautt, G. M. iety: Further Evaluation.

Finger pulse volume as a measure of anx-Psychophysiology, 1977, 14(6), 541-544.

Borkovec, T. D., & Fowles, D. C. A controlled investigation of the ef-fects of progressive and hypnotic relaxation on insomnia. Journal of Abnormal Psychology, 1973, 82, 153-158.

Borkovec, T. D., Grayson, J.B., & Cooper, K. M. Treatment of general tension: Subjective and physiological effects of progressive relaxa-tion. Journal of Consulting and Clinical Psychology, 1978, 46(3), 518-528.

Borkovec, T. D., & Nau, s. D. Credibility of analogue therapy ration-ales. Journal of Behavior Therapy and Experimental Psychiatry, 1972, l, 257-260.

Borkovec, T. D., & Sides, J. K. Critical procedural variables related to the physiological effects of progressive relaxation: A review. Behaviour Research and Therapy, 1979, 17, 119-125.

Borkovec, T. D., & O'Brien, G. T. Methodological and target behavior issues in analogue therapy outcome research. In M. Hersen, R. M. Eisler, & P. M. Miller (Eds.), Progress in behavior modification. New York: Academic Press, 1976.

Borkovec, T. D., Stone, N. M., O'Brien, G. T.·, & Kaloupek, D. G. Evaluation of a clinically relevant target behavior for analogue outcome research. Behavior Therapy, 1974, ~, 504-514.

76

77

Borkovec, T. D., Weerts, T. c., & Bernstein, D. A. Behavioral assess-ment of anxiety. In Ciminero,A., K. Calhoun, & H. E. Adams (Eds.), Handbook of behavioral assessment. New York: Wiley, 1977.

Brandt, K. on basal selected 11, 242.

The effects of relaxation training with analog HR feedback levels of arousal and response to aversive tones in groups according to Fear survey scores. Psychophysiology, 1973,

Boudewynns, P.A., & Borkovec, T. D. Credibility of psychotherapy and placebo therapy rationales. V. A. Newsletter for Research in Mental Health and Behavioral Sciences, 1974, 16, 15-18.

CUthbert, B., Kristeller, J., Simons, R., Hodes, R., & Lang, P. J. Strategies of arousal control: Biofeedback, meditation, and motiva-tion. Journal of Experimental Psychology~ General, 1981~ 11(4), 518-546.

Davidson, P. o., & Hiebert, s. F. Relaxation training, relaxation instruction, and repeated exposure to a stressor film. Journal of Abnormal Psychology, 1971, 78, 154-159.

Deabler, H., Fidel, E., Dillenkoffer, R., & Elder, S. T. The use of relaxation and hypnosis in lowering blood pressure. American Journal of Clinical Hypnosis, 1973, 16, 75-83.

Deffenbacher, J. L. Relaxation in vivo in the treatment of test anxiety. Journal of Behavior Therapy and Experimental Psychiatry, 1976, 2, 389-292.

Edelman, R. I. Effects of progressive relaxation on autonomic pro-cesses. Journal of Clinical Psychology, 1970, 26, 421-425.

Endler, N. S., & Hunt, J. McV. sured by the S-R Inventory 1966, 65, 336-346.

Sources of behavioral variance as mea-of Anxiousness. Psychological Bulletin,

Endler, N. s., & Okada, M.A. Multidimensional measure of trait anx-eity: The S-R Inventory of General Trait Anxiousness. Journal of Consulting and Clinical Psychology, 1975, 43{3), 319-329.

Epstein, L. H., Hersen, M., & Hemphill, D. P. Muscle feedback in the treatment of tension and headache: An experimental case study. Journal of Behavior Therapy and Experimental Psychiatry, 1974, .§., 59-63.

Foreyt, J., & Goodrick, G. K. Book review of R. Stuart and B. Davis. Slim chance in a fat world: Behavior control of obesity. The Behavior Therapist, 1979, ±._{3), 22-23.

78

Haynes, s., Moseley, D., & McGowan, W. Relaxation training and biofeed-back in the reduction of frontalis muscle tension. Psychophysiology, 1975, 12, 547-552.

Haynes, s. N., Woodward, s., Moran, R., & Alexander, D. Relaxation treatment of insomnia. Behavior Therapy, 1974, 2,, 555-558.

Hodgson, B., & Rachman, s. Desynchrony in measures of fear. Behaviour Research and Therapy, 1974, 12, 319-326.

Husek, T. R., & Alexander, s. The effectiveness of the anxiety differ-ential in examination and stressful situations. Educational and Psychological Measurement, 1963, ~, 309-318.

Israel, E., & Beiman, I. Live versus recorded relaxation training: A controlled investigation. Behavior Therapy, 1977, ~, 251-254.

Jacobsen, E. Progressive relaxation. Chicago, Ill.: University of Chicago Press, 1929.

Jacobsen, E. Electrical measurements concerning muscular contraction (tonus) and the cultivation of relaxation in man. American Journal of Physiology, 1934, 107, 230-248.

Jacobsen, E. Variation in blood pressure with skeletal muscle tension and relaxation. Annals of Internal Medicine, 1939, g, 1194-1212.

Jacobsen, E. Progressive Relaxation. Chicago, Ill.: University of Chicago Press, 1969

Kallman, W. M., & Feuerstein, M. Psychophysiological procedures. In Ciminera, A., Calhoun, K., ioral assessment. New York:

& Adams, H. (Eds.}, Handbook of behav-Wiley & Sons, 1977.

Kazdin, A. Therapy outcome questions requiring control of credibility and treatment generated expectancies. Behavior Therapy, 1979, 10(1), 81-93.

Kazdin, A. E., & Wilcoxon, L.A. Systematic desensitization and non-specific treatment effects: A methodological evaluation. Psycho-logical Bulletin, 1976, 83, 729-758.

Jennings, J. R., Tahmaush, A. J., & Redmand, D. P. surement of peripheral vascular activity. In P. Martin (Eds.), Techniques in Psychophysiology. Sons, 1980.

Non-invasive rnea-H. Venables & I.

New York: Wiley and

79

King, N. J. The therapeutic utility of abbreviated progressive relaxa-tion: A critical review with implications for clinical practice. In M. Hersen, R. Eisler, and P. Miller {Eds.), Progress in behavior modification. New York: Academic Press, 1980.

Lacey, J. I., Batemen, D. E., & Van Lehn, R. Autonomic response speci-ficity: An experimental study. Psychosomatic Medicine, 1953, 15, 8-21.

Lader, M. H., & Mathews, A. M. ing physiological measures. 331-338.

Comparison of methods of relaxation us-Behaviour Research and Therapy, 1970, .§!.,

Lang, P. J. Anxiety: Toward a psychophysiological definition. In H. Akiskal (Ed.), Psychiatric diagnosis. N.Y.: Spectrum, 1978.

Lehrer, P. M. Psychophysiological effects of progressive relaxation in anxiety neurotic patients and of progressive relaxation and alpha feedback in non patients. Journal of Consulting and Clinical Psy-chology, 1978, 46, 389-404.

Luiselli, J. K., Marholin, D., Steinman, D., & Steinman, w. Assessing the effects of relaxation training. Behavior Therapy, 1979, 10(5), 663-668.

Marcia, J.E., Rubin, B. M., & Efran, J. S. Systematic desensitization: Expectancy change as counter-conditioning. Journal of Abnormal Psy-chology, 1969, 74, 382-387.

Martin, B. The assessment of anxiety by physiological behavioral mea-sures. Psychological Bulletin, 1961, 58(3), 234-255.

Mathews, A. M., & Gelder, M. G. Psychophysiological investigations of brief relaxation training. Journal of Psychosomatic Research, 1969, .!l, 1-12.

McGlynn, F., & McDonnel, R. Subjective ratings of credibility following brief exposure to desensitization and pseudotherapy. Behaviour Re-search and Therapy, 1974, ~, 141-146.

Miller, M. P., Murphy, P. J., & Miller, T. P. Comparison of electromyo-graphic feedback and progressive relaxation training in treating circumscribed anxiety stress reactions. Journal of Consulting and Clinical Psychology, 1978, 46, 1291-1298.

Miller, R. K., & Bornstein, P.H. Thirty minute relaxation: A compari-son of some methods. Journal of Behavior Therapy and Experimental Psychiatry, 1977, .§., 291-294.

80

Nau, S. D., Caputo, J. A., & Borkovec, T. D. The relationship between credibility of therapy rationale and the reduction of simulated anx-iety. Journal of Behavior Therapy and Experimental Psychiatry, 1974, .§., 129-133.

Nicasso, P., & Bootzin, R. A. A comparison of progressive relaxation and autogenic training as treatments for insomnia. Journal of Ab-normal Psychology, 1974, 83, 253-260.

Ollendick, T., & Murphy, M. J. Differential effectiveness of muscular and cognitive relaxation as functions of locus of control. Journal of Behavioral Therapy and Experimental Psychiatry, 1977, ~, 223-228.

Orne, M. T. On the social psychology of the psychological experiment: With particular reference to demand characteristics and their impli-cations. American Psychologist, 1962, 17, 776-783.

Parloff, M. B., Waskow, I.E., & Wolfe, B. E. Research on therapist variables in relation to process and outcome. Ins. Garfield & A. Bergin, Handbook of psychotherapy and behavior change. New York: Wiley and Sons, 1978.

Paul, G. L. Insight vs. desensitization in psychotherapy: An experi-ment in anxiety reduction. Stanford: Stanford University Press, 1966.

Paul, G. L. Physiological effects of relaxation training and hypnotic suggestion. Journal of Abnormal Psychology, 1969, 74, 425-437.

Paul, G. L., & Trimble, R. W. Recorded vs. "live" relaxation training and hypnotic suggestions: Comparative effectiveness for reducing physiological arousal and inhibiting stress responses. Behavior Therapy, 1970, .!_, 285-302.

Reinking, P.H., & Kohl, M. L. Effects of various forms of relaxation training on physiological and self report measures of relaxation. Journal of Consulting and Clinical Psychology, 1975, 43(5), 595-600.

Rosenthal, D., & Frank, J. Psychotherapy and the placebo effect. Psy-chological Bulletin, 1956, 53, 294-302.

Russell, R. K., Miller, D. E., & June, L. M. laxation in the treatment of test anxiety. .§., 572-573.

Group cue-controlled re-Behavior Therapy, 1974,

Russell, R. K., & Sipich, J. F. Treatment of test anxiety by cue-controlled relaxation. Behavior Therapy, 1974, .§_, 672-676.

81

Russell, R. K., Sipich, J. F., & Knipe, J. Progressive relaxation training: A procedural note.· Behavior Therapy, 197.6, 2(4}, 566-567.

Shandler, S. L., & Grings, W.W. An examination of methods for produc-ing relaxation during short-term laboratory sessions. Behaviour Re-search and Therapy, 1976, 14, 419-426.

Shapiro, S., & Lehrer, P. M. Psychophysiological effects of autogenic training and progressive relaxation. Biofeedback and Self Regula-tion, 1980, 5(2), 249-255.

Shoemaker, J. S., & Tasto, D. L. The effects of muscle relaxation on blood pressure of essential hypertensives. Behaviour Research and Therapy, 1975, 13, 29-43.

Smith, J.C. Psychotherapeutic effects of transcendental meditiation with controls for expectations of relief and daily sitting. Journal of Consulting and Clinical Psychology, 1976, 44(4), 630-637.

Spielberger, D., Gorsuch, R., & Lushene, R. Manual for the state-trait anxiety inventory. Palo Alto, Calif.: Consulting Psycholo-gists Press, 1968.

Tasto, D. L., & Chesney, M.A. Muscle relaxation treatment for primary dysmenorrhea. Behavior Therapy, 1974, ~, 668-672.

Tasto, D. L. & Huebner, L.A. The effects of muscle relaxation and stress on the blood pressure levels of normotensives. Behaviour Research and Therapy, 1976, _!!, 89-91.

Van Egeren, phobias: 1971, ~.

L. F., Feather, B. W., & Hein, P. L. Desensitization of Some psychophysiological propositions. Psychophysiology,

213.

W'ld J Sti'mulus and Response· The law of 1'n1't1'al value. Br1'st·o1: i er, . . Wright, 1967.

Wolpe, J. Psychotherapy by reciprocal inhibition. Stanford, Calif.: Stanford University Press, 19.58.

Zuckerman, M., & Lubin, B. Manual for the multiple affect adjective checklist. San Diego, Calif.: Educational and Industrial Testing Service, 1965.

APPENDICES

82

APPENDIX A

ANXIETY DIFFERENTIAL

83

84

INSTRUCTIONS

The purpose of this form is to measure the meanings of certain things to various people by having them judge them against a series of de-scriptive scales. In completing this form, please make your judgments on the basis of what these things mean to you. Below are different concepts to be judged and beneath each a set of scales. You are to rate the concept on each of these scales in order. Here is how to use these scales:

If you feel that the concept at the top of each scale is very closely related to one end of the scale, you should place your check-mark as follows:

Fair X Unfair ---or

Fair X Unfair ---If you feel that the concept is quite closely related to one or the other end of the scale (but not extremely), you should place your check mark as .follows:

Strong _____ x ________________ Weak

or Strong ___ ------ ______ __ x _ ___ Weak

If the concept seems only slightly related to one side as opposed to the other side, then you should check as follows:

Active X Passive ---or

Active X Passive --- ---If you consider the concept to be neutral on the scale, both sides of the scale being equally associated with the concept, or if the scale is completely irrelevant, unrelated to the concept, then you should place your check mark in the middle space:

Safe -~---- _____ x __________ Dangerous

IMPORTANT: (1) Place your check marks in the middle of spaces, not on the boundaries.

(2) Be sure you check every scale for every concept~ do not omit any.

(3) Never put more than one check mark on a single scale.

85

ME

Frightened Fearless

DREAMS Loose Tight ---

MY MIND Loose Tight

LITTLE BOYS Safe Dangerous

ME Jittery Calm

BREATHING Careful Carefree

FINGERS Loose Tight

SCREW Strong Weak

GERMS Deep Shallow

ME Helpless Secure

SCREW Nice Awful

FINGERS Stiff Relaxed

MOVIES Loose Tight

SCREW Loose Tight

HANDS Good Bad

DREAMS Near Far

HANDS Wet Dry

APPENDIX B

S-R INVENTORY OF GENERAL TRAIT ANXIOUSNESS

86

87

INVENTORY OF ATTITUDES TOWARD GENERAL SITUATIONS

NAME:

AGE: SS#:

SEX: PHONE:

This inventory represents a means of studying people's reactions to and attitudes towards various types of General situations. On the fol-lowing pages are represented four general kinds of situations which most people have encountered. For each of these general kinds of situa-tions certain conunon types of personal reactions and feelings are listed.

You Are in Situations Involving Interaction with Other People

(We are primarily interested in your reactions in General to those situations that involved interacting with other people. This includes situations that involve frineds, family, acquaintances, strangers, etc.

l 2 3 4 5 1. Seek experiences Very much Not at all

like this 1 2 3 4 5

2. Perspire Not at all Perspire much

1 2 3 4 5 3. Have an "uneasy Not at all Very much

feeling: 1 2 3 4 5

4. Feel exhilarated Very much Not at all and thrilled

1 2 3 4 5 5. Get fluttering Not at all Very much

feeling in stomach 1 2 3 4 5

6. Feel tense Not at all Very tense

l 2 3 4 5 7. Enjoy these situations Very much Not at all

1 2 3 4 5 8. Heart beats faster Not at all Much faster

1 2 3 4 5 9. Feel anxious Not at all Very anxious

88

You Are in Situations Where You Are About to Or May Encounter Physical Danger

(We are primarily interested in your reactions in General to those situations involving Physical Danger.

10. Seek experiences like this

11. Perspire

12. Have an "uneasy feeling"

13. Feel exhilarated and thrilled

14. Get fluttering feel-ing in stomach

15. Feel tense

16. Enjoy these situations

17. Heart beats faster

18. Feel anxious

1 Very much

1 Not at all

1 Not at all

1 Very much

1 Not at all

1 Not at all

1 Very much

1 Not at all

1 Not at all

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

2 3 4

You Are in a New Or Strange Situation

5 Not at all

5 Perspire much

5 Very much

5 Not at all

5 Very much

5 Very tense

5 Not at all

5 Much faster

5 Very anxious

(We are primarily interested in your reactions in General to~~ strange situations.)

19. Seek experiences like this

20. Perspire

21. Have an "uneasy feeling"

22. Feel exhilarated and thrilled

1 Very much

1 Not at all

1 Not at all

1 Very much

2

2

2

2

3 4 5 Not at all

3 4 5 Perspire much

3 4 5 Very much

3 4 5 Not at all

89

1 2 3 4 5 23. Get fluttering Not at all Very much

feeling in stomach 1 2 3 4 5

24. Feel tense Not at all Very tense

1 2 3 4 5 25. Enjoy these Very much Not at all

situations 1 2 3 4 5

26. Heart beats faster Not at all Much faster

1 2 3 4 5 27. Feel anxious Not at all Very anxious

You Are Involved in Your Daily Routines

(We are primarily interested in yoru reactions in General to Routine Situations.)

1 2 3 4 5 28. Seek experiences Very much Not at all

like this 1 2 3 4 5

29. Perspire Not at all Perspire much

1 2 3 4 5 30. Have an "uneasy Not at all Very much

feeling" 1 2 3 4 5

31. Feel exhilarated Very much Not at all and thrilled

1 2 3 4 5 32. Get fluttering Not at all Very much

feeling in stomach 1 2 3 4 5

33. Feel tense Not at all Very tense

1 2 3 4 5 34. Enjoy these situations Very much Not at all

1 2 3 4 5 35. Heart beats faster Not at all Much faster

1 2 3 4 5 36. Feel anxious Not at all Very anxious

APPENDIX C

CONSENT FORM

SELF MONITORING CONTRACT

90

91

CONSENT FORM

I, , have been informed of my responsibility as a volunteer in this research project and of the time commitment in-volved. More specifically, I am aware that the study will involve two sessions of approximately 60 minutes duration, with sessions scheduled one week apart. I am also aware that: (1) the study of the relaxation process is the focus of the project; (2) the relaxation training is not meant to serve as a substitute for psychological counseling; and (3) that physiological measures of relaxation will be taken during the two sessions, the results.of which will be unavailable to me until the com-pletion of the study. These measures will involve the attachment of monitoring equipment to the forehead and one finger and will cause no discomfort or pain. In addition, I realize that my commitment also in-volves practicing relaxation for 20 minutes daily during the interval between session 1 and session 2. Finally, I understand that I will re-ceive 3 academic credits for my participation in this research project and that my participation in this study will remain completely confi-dential other than the identification necessary for academic credit assignment.

Signature

Date

SELF MONITORING CONTRACT

I, , agree to:

1. Practice relaxation once daily for a minimum of 20 minutes;

2. Pick up a relaxation practice sheet daily from the 5th floor lobby at which time the completed sheet for the previous day will be returned;

3. Sign each relaxation practice sheet as a pledge that relaxa-tion practice did indeed occur.

Signature

Date

Witness

APPENDIX D

EXPECTANCY QUESTIONNAIRE

92

93

QUESTIONNAIRE

Please read the following questions and circle the number which matches how you feel about the type of .relaxation that you will be learning. Please do not hesitate to make use of the entire scale.

1. How logical does this type of treatment seem to you?

2 1 totally

illogical; makes no sense

3 4 5 6 7 8 9 10 extremely

logical and sensible

2. How confident are you that this treatment will be successful in making you less tense and anxious?

_J, __ ~~--...._ __ _,_ __ _,_ __ _._ __ _._ __ -"' ___ ~----

1 2 not confident

at all

3 4 5 6 7 8 9 10 extremely

confident; certain

3. How confident would you be in recommending this treatment to a friend who was tense and anxious?

1 2 not confident

at all

3 4 5 6 7 8 9 10 extremely confident

4. How successful do you feel this treatment would be in decreasing fears; for example, being very nervous before taking tests?

1 2 certain it would be unsuccessful

3 4 5 6 7 8 9 10 certain it would work

5. If you were extremely nervous in certain situations, for example, taking tests, would you be willing to undergo the relaxation training you will be practicing?

_l __ _.._ __ --J __ ___. __ __,_ __ __._ _______ ___._ __ __,_ ______ .__

1 definitely

not

2 3 4 5 6 7 8 9 10 yes,

definitely

APPENDIX E

POST-SESSION 1 QUESTIONNAIRE

94

95

POST-SESSION l QUESTIONNAIRE

1. Tell me a little bit about how you felt during the relaxation practice.

2. What difficulties, if any, did you have in getting relaxed?

3. What questions do you have about the relaxation that you have just practiced?

APPENDIX F

PROCEDURAL OUTLINE: SESSION 1

96

97

PROCEDURAL OUTLINE: SESSION 1

1. Greeting of subject a. Consent form signed b. Self monitoring contract signed

2. Statement of relaxation instruction and rationale and expectancy condition.

3. Physiological monitoring explained and instruments attached.

4. Expectancy measure and Anxiety Differential administered.

5. Subject asked to remove glasses, contacts, and rings.

6. 5 minute adaptation period with final 1 minute serving as pretreat-ment baseline period for physiological measures (begun following completion of Anxiety Differential and expectancy measure)

7. Pre-relaxation instructions (includes modeling tensing the face, relaxing "all at once" and instructions to recline and close eyes).

8. Begin relaxation proper. a. Record self ratings following relaxation of each muscle group. b. Press event marker in live and taped conditions (once for

tensing, twice for relaxation). c. Press event marker signalling final 30 seconds of relaxation

session.

9. Anxiety Differential administered.

10. Subject detached from physiological equipment.

11. Discussion of post-session 1 questions.

12. Self monitoring forms explained and distributed with statement that a discussion of practice will occur next session.

APPENDIX G

PROCEDURAL OUTLINE: SESSION 2

98

99

PROCEDURAL OUTLINE: SESSION 2

1. Greeting of subject. a. Collection of self monitoring sheet from previous day's

practice. 2. Pre-session 2 questions discussed.

3. Apparatus attached.

4. Anxiety Differential administered.

5. 5 minute adaptation period with final 1 minute serving as pretreat-ment baseline period for physiological measures.

6. Begin relaxation proper. a. Record self ratings following relaxation of each muscle group. b. Press event marker in live and taped conditions (once for

tensing, twice for relaxation). c. Press event marker signalling final 30 seconds of relaxation

session.

7. Anxiety Differential administered.

8. Subject detached from physiological monitoring equipment.

9. Debrief subject in live/low, taped/low conditions.

APPENDIX H

SELF MONITORING SHEETS

100

101

NAME:

1. Fist on nondominant hand; relax 2. Elbow pressed down agains arm of chair, nondominant arm; relax 3. Fist on dominant hand; relax 4. Elbow pressed down against arm of chair, dominant arm; relax 5. Lift eyebrows; relax 6. Squint and wrinkle nose; relax 7. Bite teeth together, corners of mouth back; relax 8. Chin down; relax 9. Deep breathmd hold, shoulder blades back; relax

10. Stomach hard; relax 11. Tense right upper leg; relax 12. Right foot, toes up; relax 13. Right foot, point toe, foot inward, curl toes; relax 14. Tense left upper leg; relax 15. Left foot, toes up; relax 16. Left foot, point toe, foot inward, curl toes; relax

Relaxation Level Relaxation Level Date Time Before Practice After Practice

(1 = Extremely Relaxed) (1 = Extremely Relaxed) (10 = Extremely Tense) (10 = Extremely Tense)

PLEDGE:

APPENDIX I

PRE-RELAXATION INSTRUCTIONS~TAPED

102

103

PRE-RELAXATION INSTRUCTIONS~TAPED

As I described before, the procedure we will be using is called pro-gressive relaxation training, which consists of learning to tense and release various muscle groups throughout the body. You will be listen-ing to a tape, directing you to tense a particular muscle group for about 5 seconds and then to relax for 30-45 seconds. The tape will go through each muscle group twice. It is important to remember to re-lease the muscle tension immediately rather than gradually when the tape directs you to relax. Also, once a group of muscles is relaxed, do not move it unnecessarily (except to make yourself comfortable). Finally, after each muscle group, the tape will ask you to rate on a 1 to 10 scale how relaxed that muscle group is. A rating of 1 will in-dicate that the muscle group is completely relaxed, while a rating of 10 will indicate a great deal of tension in that muscle group. Remeber, a 1 means that you are thoroughly, completely relaxed, while a 10 means that you are very tense. Do you have any questions?

~Begin Relaxation Exercise~

APPENDIX J

LIVE AND TAPED RELAXATION INSTRUCTIONS

104

105

LIVE RELAXATION INSTRUCTIONS

Muscle Groups

1. Dominant hand and forearm

*la. Repeat *lb. Please rate level of

relaxation in your hand and forearm.

Exercise

Cup hand and spread fingers

(Repeat if rating is greater than 2)

2. Dominant biceps

3. Nondominant hand and forearm

4. Nondominant biceps

5. Forehead

6. Central face

7. Lower face and jaw

8. Neck

9. Chest, shoulders, upper back

10. Abdomen

11. Dominant upper leg

12. Dominant calf

Elbow down against chair

Cup hands; spread fingers

Elbow down against chair

Lift eyebrows

Squint and wrinkle nose

Bite hard; pull back corners of mouth

Chin to chest and keep it from touching

Deep breath and hold; pull blades back

Stomach hard (punch)

Lift leg slightly

Point toes toward head (brief)

Area of Tension Patter**

Hand, knuckles, lower arm

Biceps

Hand, knuckles, lower arm

Biceps

Forehead and scalp

Central part of face; upper cheeks and through eyes

Lower face and jaw

Neck

Chest, shoulders, upper back

Stomach

Upper leg

Calf

A

B

C D

E F

A B

C D

E F

A B

c· D

E F

A B

C,D

E F

13.

14.

15.

106

LIVE RELAXATION INSTRUCTIONS (Cont'd.)

Muscle Groups

Dominant Foot

Nondominant upper leg

Nondominant calf

Exercise

Curl toes; foot inward

(see 11)

(see 12)

Area of Tension Pattern**

Arch ball of A of foot B

C D

E F

16. Nondominant foot (see 13) A B

*Same for each muscle group.

**A= and relax, letting all the tension go, focusing on these muscles as they just relax completely, noticing what it feels like as the muscle becomes more and more relaxed, focusing all your attention on the feelings associated with relaxation flowing into these muscles.

B = just enjoying the pleasant feelings of relaxation, as the muscles go on relaxing more and more deeply, more and more completely. There is nothing for you to do but focus your attention on the very pleasant feelings of relaxation flowing into this area. Just noticing what it is like as the muscles become more and more deeply relaxed.

C = just enjoying the feelings in the muscles as they loosen up, smooth out, unwind,and relax more and more deeply. Just experiencing the sensations of deep, complete relaxation flowing into these muscles more and more deeply and completely relaxed. Just letting the muscles go, thinking about nothing but the very pleasant feelings of relaxation.

D = just let those muscles go and notice how they feel now as compared to before. Notice how those muscles feel when so completely re-laxed. Pay attention only to the sensations of relaxation as the relaxation process takes place. Calm, peaceful and relaxed. Now that you have relaxed as much as you have.

E = attending to the difference in how your feels now as com-pared to just a moment ago. Let your become very relaxed all over. Let the tension flow away as your muscles relax more and more completely as the tension and tightness dissolve.

107

LIVE RELAXATION INSTRUCTIONS (Cont 1 d.)

F = letting all the tension go, enjoying the feeling of relaxation as the muscles loosen up and unwind completely. Pay attention only to your muscles, noting the difference in feeling as you relax it. There is nothing for you to do but pay attention to these relaxed sensations.

APPENDIX K

SELF RELAXATION SELF-MONITORING FORM

108

109

NAME:

Relaxation Level Relaxation Level

Time Before Practice After Practice Date (1 ~ Extremely Relaxed) (1 = Extremely Relaxed) (10 = Extremely Tense) (10 = Extremely Tense)

PLEDGE:

APPENDIX L

110

111

Table l

Analysis of Variance for SR-GTA

Source

Treatment

Expectancy

Treatment X Expectancy

df

2

1

2

ss

105.44

11.57

46.25

F

0.99

0.22

0.44

112

Table 2

Analysis of Variance for Baseline Period Measure

Source df ss F

Heart Rate (HRll) Treatment 2 276.70 1.22 Expectancy 1 832.29 7.34** Treatment X Expectancy 2 780.03 3.44* Error 48 5442.88

Finger Pulse (FPll) Treatment 2 2.60 3.00 Expectancy 1 0.35 0.82 Treatment X Expectancy 2 o.oo 0.00 Error 48 20.83

Anxiety Differential (AD1) Treatment 2 107.11 0.52 Expectancy 1 21.40 0.21 Treatment X Expectancy 2 651.25 3.16 Error 48 4949.55

Skin Fluctuation Res:12onses (SFll) Treatment 2 9.30 0.44 Expectancy 1 8.28 0.78 Treatment X Expectancy 2 3.32 0.16 Error 48 498.22

Heart Rate (HR21) Treatment 2 483.44 2~02 Expectancy 1 280.16 2.34 Treatment X Expectancy 2 1290.77 5.39** Error 48 5749.11

Finger Pulse (FP21) Treatment 2 0.51 0.98 Expectancy 1 0.00 0.00 Treatment X Expectancy 2 0.07 0.15 Error 48 12.53

113

Table 2 (Cont'd.)

Source df ss F

Anxiety Differential (AD3) Treatment 2 24.48 0.13 Expectancy 1 109.79 1.14 Treatment X Expectancy 2 77.37 0.40 Error 48 4614.66

Skin Fluctuation Res~nses (SF21) Treatment 2 2.40 0.08 Expectancy 1 101.22 7.08** Treatment X Expectancy 2 24.03 0.84 Error 49 658.08

**E. <. 01 *p <. .OS

114

Table 3

Session 1 Baseline Heart Rate

Simple Main Effects for Treatment X Expectancy Interaction

Source df ss MS F

Expectancy at 1 624.22 624.22 5~'50* Live Condition

Expectancy at 1 953.39 953.39 8.40** Tape Condition

Expectancy at 1 34. 72 34.72 0.30 Self Condition

Error 50 113. 39

*E. ~. 025 **£. ~.01

115

Table 4

Session 2 Baseline Heart Rate

Simple Main Effects for Treatment X Expectancy Interaction

Source df ss MS F

Expectancy at 1 138.88 138.88 1.07 Live Condition

Expectancy at 1 1152.00 1152.00 9.62* Tape Condition

Expectancy at 1 280.06 280.06 2.33 Self Condition

Error 50 119. 77

*12.<-0l

116

Table 5

t-Test for Heart Rate Median Split

Group N df t

Heart Rate Change Session 1 (HRCHl)

HR high 26

HR low 28 52 1. 751

Finger Pulse Change Session l (FPCHl)

HR high 26

HR low 28 52 0.131

Anxietl Differential Change Session 1 (ADCHl)

HR high 26

HR low 28 52 1.935

Skin Fluctuation Change Session 1 (SFCHl)

HR high 26

HR low 28 52 0.020

117

Table 6

Analysis of Variance for Experimenter Differences

Source df ss F

Heart Rate Change Session 1 (HRCHl)

Experimenter 2 26.70 0.61 Error 51 1121.88

Finger Pulse Change Session 1 (FPCHl)

Experimenter 2 6.55 5.93 Error 51 22.08

Anxiety Differential Change Session 1 (ADCHl)

Experimenter 2 314.48 2.82 Error 51 2847.00

Skin Fluctuation Change Session 1 (SFCHl)

Experimenter 2 32.25 Error 50 285.44

Heart Rate Change Session 2 (HRCH2)

Experimenter 2 24.77 0.29 Error 51 2166. 72

Finger Pulse Change Session 2 (FPCH2)

Experimenter 2 1.37 2.92 Error 51 16.79

Anxiety Differential Change Session 2 (ADCH2}

Experimenter 2 78.37 0.90 Error 51 2208.38

Skin Fluctuation Change Session 2 (SFCH2)

Experimenter 2 26.11 1.31 Error 49 585.88

E.<: .005

118

Table 7

Analysis of Variance for Self-Monitoring Sheets

Source

Treatment

Expectancy

Treatment X Expectancy

Error

df

2

1

2

48

ss

1.03

0.46

1.92

35.33

F

0.70

0.63

1.31

119

Table 8

Analysis of Variance for Relaxation Ratings

Source df ss F

Treatment 2 108.50 1.49

Expectancy 1 40.33 1.11

Treatment X Expectancy 2 41.16 0.56

Error 48 1750.00

Session 1 46.33 8.90*

Treatment X Session 2 7.38 0.58

Expectancy X Session 1 o.59 0.09

Treatment X Expectancy

X Session 2 1.90 0.15

Error 48 303.77

*E_<.005

120

Table 9

Multivariate Analysis of Variance for the Overall

Effects of Treatment, Expectation, and Sessions

Pillai's Trace

Effect F value p<

Treatment 1.09 0.374

Expectancy 5.05 0.001*

Treatment X Expectancy 0.66 o. 718

Sessions 0.11 o. 977

Treatment X Sessions 1.10 0.371

Expectancy X Session 1.85 0.135

Treatment X Expectancy X Sessions 0.91 0.515

*Significant main effect.

121

Table 10

Univariate Analyses of Variance for Dependent Variables

Source df ss F

Heart Rate Treatment 2 188.46 3.12 Expectancy 1 330.75 10.94* Treatment X Expectancy 2 17.16 0.28 Error 48 1451.00

Sessions 1 5.78 0.24 Treatment X Session 2 118.57 2.45 Expectancy X Session 1 8.98 0.37 Treatment X Expectancy X Sessions 2 62. 90 1.30 Error 48 1162.33

Skin Fluctuation Responses Treatment 2 0.88 0.06 Expectancy 1 43.37 5.43** Treatment X Expectancy 2 15.33 0.96 Error 48 385.23

Sessions 1 0.40 0.05 Treatment X Sessions 2 9.23 0.52 Expectancy X Sessions 1 39.84 4.47** Treatment X Expectancy X Sessions 2 14.15 0.79 Error 48 428.19

Finger Pulse Volume Treatment 2 0.49 0.88 Expectancy 1 0.34 1.24 Treatment X Expectancy 2 0.18 0.33 Error 48 13.50

Sessions 1 0.03 0.18 Treatment X Sessions 2 0.46 1.41 Expectancy X Sessions 1 0.06 0.34 Treatment X Expectancy X Sessions 2 0.30 0.93 Error 48 1228.88

122

Table 10 (Cont'd.)

Univariate Analyses of Variance for Dependent Variables

Source df ss F

Anxiety Differential Treatment 2 102.24 0.67 Expectancy 1 46.67 0.61 Treatment X Expectancy 2 265.24 1. 74 Error 48 3654.66

Sessions 1 2.08 0.08 Treatment X Sessions 2 35.38 0.69 Expectancy X Sessions 1 80.08 3.,13 Treatment X Expectancy X Sessions 2 35.05 0.68 Error 48 1228.88

*R. c:: • 05 **R.< .005

The vita has been removed from the scanned document

PROGRESSIVE MUSCLE RELAXATION: EFFECTS OF EXPECTANCY

AND TYPE OF TRAINING ON MEASURES OF ANXIETY

by

Michael E. Stefanek

(ABSTRACT)

The present study compared live versus taped and self relaxation

training conditions under high and low expectancy conditions. Anxious

college students (N = 54) were stratified according to scores on the

screening instrument employed (S-R Inventory of General Trait Anxious-

ness) and randomly assigned to one of six training/expectancy conditions

for two relaxation sessions: live-high, taped-high, self-high, live-

low, taped-low, or self-low. Physiological measures of anxiety in-

cluded heart rate, spontaneous skin fluctuation responses, and finger

pulse volume amplitude. The Anxiety Differential was used to assess

self report of anxiety. Finally, within-session relaxation ratings were

elicited to evaluate differences between live and taped training condi-

tions. Results indicated that subjects in the high expectancy condition

showedasignificantly greater decrease in heart rate and spontaneous

skin fluctuation responses than low expectancy subjects, but no differ-

ences were found with the self report measure of anxiety. There were no

differences due to type of training (live, taped, self). Within-session

relaxation ratings indicated increased relaxation across sessions, but

no differences across type of training. Results were discussed in terms

of the multidimensionality of the anxiety construct and. the role of ex-

pectancy factors in progressive muscle relaxation procedures.