®Acttdemy of Management Review ¡982. Vot. 7, No. 3. 433-441

Relapse Prevention for Managerial Training:A Model for Maintenance of Behavior Change^

ROBERT D. MARXWestern Washington University

Although organizations invest heavily in training programs to enhancemanagerial effectiveness, little attention is paid to the transfer of suchtraining from the workshop to the workplace. This paper describes acognitive-behavioral model that offers a systematic approach to themaintenance of behavior. Relapse prevention strategies are discussed, andimplications for management training and research are considered.

Corporations spend a great deal of time andmoney on management development training. Pro-grams ranging from leadership training to com-munication skills are conducted in the hope ofenhancing the effectiveness of managerial behavior.There is little evidence, however, of the stayingpower or maintenance of these costly and time con-suming interventions. In a period of increasing ac-countability and diminishing resources, trainingpersonnel will need to be able to justify the efficacyof training and make improvements whennecessary.

One of the most promising approaches to thelong term maintenance of newly trained behaviorsis based on the relapse prevention (RP) model ofMarlatt and Gordon (1980). This model consists ofa set of self-control strategies designed to facilitatethe maintenance of behavior change by teaching in-dividuals to understand and cope with the problemof relapse. This model views maintenance ofbehavior from the novel perspective of identifyingdeterminants of treatment failure. That is, longterm behavior change is predicted to be enhancedby anticipating and monitoring past and presentfailures. Data from these episodes then are utilizedto equip managers with appropriate coping skillsfor dealing with future difficult situations.

'Appreciation is extended to Joseph Litterer, Allen Ivey, andHenry Loehr for their scholarly review of an earlier draft of thismanuscript. The paper was prepared while the author wasVisiting Associate Professor of Management at the University ofMassachusetts at Amherst, 1980-1981.

The Relapse Prevention Model

Marlatt and Gordon (1980) observed the behaviorof individuals who were unable to maintain absti-nence from addictive behaviors following participa-tion in a treatment program. Responses to question-naires by smokers, alcoholics, and drug addictsrevealed similar psychological and environmentalantecedents to relapse episodes. Typical triggers ofrelapse were intrapersonal emotional states such asanxiety, depression, and joy and interpersonal cir-cumstances such as social pressure, arguments, andcelebrations.

Further analysis revealed that the circumstancesof the initial lapse or slip had major implicationsfor the likelihood of further slips leading to a per-manent resumption of the addictive behavior, or arelapse. Anyone who has tried to quit smoking orlose weight is likely to be familiar with the poten-tially damaging effects of a single lapse during aperiod of controlled behavior.

Although originally developed with addictivebehaviors, the RP model has powerful implicationsfor the maintenance of managerial training. Man-agers are spared the physiological component ofbehavior change present in addiction problems, butthey must negotiate an analogous array of disruptivepsychological and environmental influences inorder successfully to maintain long term behaviorchange.

The RP paradigm displayed in Figure I describes

433

the cognitive and behavioral self control strategiesdesigned to reduce the likelihood of relapse. Thefirst step prescribed by the model is to make man-agers aware of the relapse process itself. Trainingprograms typically are designed to stress the likeli-hood of positive results for participants. Thus,awareness of how the training process is vulnerableto breakdown often is neglected. Managers then areasked to pinpoint situations that are likely tosabotage their attempts to maintain new learning.Ability to diagnose such high risk situations pro-vides managers with an early warning system in-dicating when their ability to maintain new learningwill be severely tested. Anticipation of high risk cir-cumstances must be accompanied by adequate cop-ing skills if relapse is to be avoided. If, for example,time pressure for completion of a project is a high

FigureCognitive-Behavioral Model

risk situation, a variety of coping responses, such astime management skills and delegative leadershipstyle to spread the workload, may be required.Whereas the presence of these skills likely will resultin a feeling of mastery and decreased probability ofrelapse, their absence may culminate in unproduc-tive cognitive responses, such as guilt, decreasedself-efficacy, and anxiety. Cognitive strategies de-signed to counter such reactions to imperfect per-formance are included in the implementation of theRP model.

Managers are taught that temporary difficultiesor slips are the predictable outcomes of any trialand error learning paradigm. Errors in implement-ing concepts from a training seminar are perceivedas inadequate coping skills rather than lack of will-power. The goal of this model is to provide man-

of the Relapse Processa

COPINGRESPONSE

INCREASEDSELF-EFFICACY

Monitors high risksituations

Uses time, stress man-agement skills

Awareness of the relapseprocess

Sense of accomplishmentfor having utilized new skillsin high risk situation

HIGH-RISKSITUATION

Time pressureshortly aftertraining

Self-attributionfor lack of control

NOCOPING

RESPONSE

Unaware of high risksituations

Reliance on will-power rather thanskills

Excessive anxiety

DECREASEDSELF-EFFICACY

POSITIVEOUTCOME

EXPECTANCIES(for effects of

the old behavior)

Enhanced expectanciesof other leadershipstyles. Positive effects ofold behavior are ex-aggerated.

Guilt and confusion overhaving violated trainingconcepts. Cognitivedissonance, denies pos-sible effectiveness oftraining.

ABSTINENCEVIOLATION

EFFECT

DECREASEDPROBABILITYOF RELAPSE

Increased likelihood ofattempting new behaviorin other high risk sit-uations

INCREASEDPROBABILITYOF RELAPSE

Gives up on attemptsto incorporate newlearning when situationis not going smoothly

* Adapted from Marlatt (1980) in order to describe the relapse process for managerial training.

434

agers with the necessary cognitive and behavioralskills to keep slips from becoming full blown re-lapses by using information from past failures forfuture prevention.

Recent articles by Leifer and Newstrom (1980)and Michalak (1981) have found transfer of trainingto be enhanced by a number of factors, includingcontinued interest and involvement by superiorsand attention to the maintenance problem before,during, and after training and through direct re-inforcement for maintenance of behavior. The RPmodel adds to these factors the self-control strat-egies of relapse prevention. Even in the most effec-tive programs, supervisory involvement, for exam-ple, may not be immediately available, or it may notbe perceived to be present by trainees. Such circum-stances may leave managers vulnerable to the relapseprocess.

Research demonstrating the validity of the RPmodel thus far has focused on the addictive dis-orders. Empirical evidence comparing RP trainedindividuals with non-RP trained controls showedsignificantly greater maintenance in RP trainedalcoholics after a one year followup (Chaney,O'Leary, & Marlatt, 1978) and RP trained dietersafter a three month followup (Rosenthal & Marx,1979).

The RP model has implications for empirical re-search in the field of management training. In a re-cent review of research pertaining to training inwork organizations, Goldstein reported a lack of"high quality empirical investigations that examinethe usefulness of training techniques" (1980, p.263). He singled out an investigation of the effectsof modeling on supervisory training (Latham &Saari, 1979) as among the few that provide ade-quate controls and significant performance dif-ferences at a one year followup. The validation ofthe RP model for managerial training addressesseveral of Goldstein's concerns. Assessment of thequalitative and quantitative dimensions of relapserequires collection of followup data and clearoperationalization of training outcomes.

Refining definitions of slips and relapses so thatthey might be more consistent with managerialbehavior is an important prerequisite to careful em-pirical testing. If the abstinence criterion of addic-tions is discrete (i.e., no cigarettes), the slip orrelapse can be measured clearly. Lapses in the useof newly trained managerial behavior are likely to

be defined less easily. The response can be partiallyemitted, inappropriately performed, or deliberatelynot used. Such conceptual issues remain to be clar-ified. For the purposes of this paper, a first attemptto define a slip with relation to managerial behaviorwill be the temporary lapse in the use of a newlylearned managerial behavior when its use would beconsidered appropriate. A total relapse is defined asthe permanent extinction of the changed behavior.These definitions will be operationalized further inthe following illustration.

Dimensions of Relapse

Many managerial skills and techniques areacquired through on-the-job training seminars.Programs in strategic planning, management by ob-jectives, communication skills, time management,and leadership skills are typical of such interven-tions. Each of these training programs containsbehavior change strategies that are designed tomake managers more skillful, fiexible, and effective.Even when such programs are internally valid, theproblem of maintenance or transfer of learningfrom the training setting to the work setting re-mains.

Several theories of leadership, although focusingon different variables, agree that effective leader-ship depends on matching specific leader behaviorswith specific human and/or environmental vari-ables, that is, that different situations require dif-ferent leadership actions. Four of the most widelycited contingency theories of leadership are the lifecycle theory (Hersey & Blanchard, 1977), the leader-ship and decision making model (Vroom & Yetton,1973), the path-goal model (House, 1971), and theleadership contingency model (Fiedler, 1967). In acontingency leadership training program, managersare exposed to a variety of leadership behaviors en-compassing the entire autocratic-democratic con-tinuum. The goals of such training include the op-portunity to diagnose one's own predominantleadership style, practice other modes of leadershipbehavior, and implement these styles contingentlywith subordinates.

An example of how the RP model is applied tomanagerial training could begin with a managerwhose feedback during the program indicates pre-dominantly autocratic or task oriented leadershipstyle. On returning to work, the manager decides to

435

experiment with delegating responsibility and parti-cipative decision making. There are a few mild suc-cesses. Responsible workers are given greater au-tonomy, a welcome change from the boss tellingthem how to do something they already understand.The manager even has a few moments free to beginwork on a long range plan. That afternoon, otherstaff are asked for their input on a production deci-sion. They try to hide their amazement and returnto their desks with a budding sense of confidenceand renewed energy, yet not fully understandingwhat has happened to their leader.

Unfortunately, another scenario is unfolding atthe higher levels of the organization that will soonhave profound effects on the manager's attempts toexpand the available leadership options. Rapidlychanging conditions are calling for the sudden con-clusion of an ongoing project. The manager hasbeen made responsible for its prompt completion.In the anxiety of the moment, all attempts to main-tain the recently learned flexible leadership stylesare short-circuited. Under pressure, the managerabandons the contingency based leadership in whichthe autocratic style is sometimes but not always theoptimal method. Instead, the more familiar auto-cratic mode is rigidly adopted regardless of its ap-propriateness to the situation.

The manager experiences conflicting thoughtsregarding this slip. Even though the autocraticleadership style had achieved the desired outcome,giving up on a promising new approach to managingemployees left the manager with a defeated feeling.In light of the day's occurrences, the new leadershipskills appeared useful when things were runningsmoothly but, when push came to shove, one really

needed to "take over." The temporary effectivenessof returning to the autocratic style thus attributedto it greater effectiveness than had been perceivedpreviously. The combination of time pressure, thelack of experience with the new skills, and a result-ing failure to communicate to employees how con-tingent leadership would be implemented had re-sulted in a lapse in their appropriate application.The lapse at this point is temporary and is definedas a slip.

The manager now stands at a critical choice point.The choices are described in Table 1. In order forthis slip to remain a unique and temporary event,the manager must be able to recognize and copewith such potentially dangerous situations in thefuture. Skills must be developed to cope with theseevents. Mastery of difficult management problemsis likely to enhance self-efficacy and decrease theprobability of a permanent lapse in utilizing newtraining. Without RP training, a temporary lapse ismore likely to be defined as a total failure. Such afailure will be attributed to personal inadequacyrather than lack of skills, and the attractiveness ofthe old, less effective style will be exaggerated. Themanager may experience other cognitive-affectivereactions such as guilt and conflict over havingdiscarded a promising strategy so quickly, and thusthe likelihood of abandoning the leadership trainingentirely may further increase.

An RP trained manager understands the impor-tance of the response to an initial slip. A relapse-trained manager expects crises and temporary fail-ures in implementing new skills. These situationsare considered predictable and nearly inevitable andthus are built into the training program. The critical

Table 1Comparative Responses of Managers Trained and Untrained in Relapse Prevention

Manager Who Is Trained m Relapse Prevention Manager H ho Is Untrained in Relapse Prevention

1. Monitors high risk situations—trained to expect a variety ofhigh risk situations that are likely to interfere temporarilywith new learning. Has probably identified time pressure as alikely possibility.

2. Coping response—manager trained in a variety of copingbehaviors with high risk situations or knows where to obtainhelp. Time management, contingency planning.

3. Increased self-efficacy—manager trained to experience asense of accomplishment at attempting to use new copingskills under pressure.

4. Decreased probability of relapse—manager continues toperceive difficulties in managerial behavior as skill deficitsand seeks training whenever environmental circumstancesthreaten to sabotage positive outcomes.

1. Fails to monitor high risk situations—high risk situation arises"unexpectedly." No preparation for dealing with it.

2. Lack of specific coping skills. Reliance on willpower oravoidance strategies. Perceives temporary lapse as failure.

3. Decreased self-efficacy—manager feels a lack of control andattributes failure to own inabilities.

4. Increased probability of relapse. Abstinence violation ef-fect—manager experiences dissonance for having failed to im-plement new training. Attributes greater effectiveness to oldbehaviors. Decreased probability of experimenting with newskills.

436

difference for the manager untrained in RP isvulnerability to the problematic effects of unplannedevents. Without the benefit of relapse "fire-drills,"this manager is less able to deal with potentialfailures.

The scenario described here is not uncommon tomanagerial training programs. The acquisition of anew and complex behavior is easily attenuated orextinguished by the absence of necessary behavioraland cognitive coping skills. What, then, are thespecific RP strategies and how are they implement-ed in the context of managerial training?

Specific Relapse Intervention Strategies

A number of relapse prevention (RP) strategieshave been utilized to assist in the maintenance ofbehavior following training. Marlatt and Gordon(1980) have developed or refined these interventionsfor use with addictive behaviors such as smoking,drinking, and drug abuse. These techniques includea mixture of behavioral coping skills such as anxietyreduction, stress management, and assertion train-ing and a set of cognitive strategies for assisting thelearner in reinterpreting the meaning of negativeoutcomes and heightening awareness of irrationalor nonadaptive thought processes. Marlatt (1980)suggests that an ideal maintenance program shouldinclude this combination of coping strategies inorder to teach an individual new and more adaptiveways of dealing with failure experiences and to in-crease one's expectancy of being able to achievemastery over future problems. RP strategies wereintegrated into a training program in a recent inves-tigation by Rosenthal and Marx (1979). Specificweight loss strategies were presented during the firsthour of weekly training sessions. The RP groupspent the second hour learning RP techniques aim-ed entirely at maintenance of weight loss followingtreatment. This concurrent focus on basic skills andRP skills was designed to equate the importance ofRP with acquiring the skills themselves. The tech-niques of RP and the methodology of implementingthem into the contingency leadership example arepresented below.

Awareness of the Relapse Process

Many training programs seem to adhere closelyto the Pygmalion phenomenon which emphasizesthe importance of expectations on outcomes

(Rosenthal & Jacobson, 1968). Most trainers expecttheir programs to be successful and underemphasizethe possibility of failure and how to handle slips orrelapses when they occur.

Participants in the training procedures proposedhere are encouraged to describe previous slips orrelapses in detail. Reporting these events is re-framed as valued data for prevention of later dif-ficulty rather than an admission of failure. Bydesign, regular discussions of these situations inprecise situational terms heightens awareness of theenvironmental and emotional circumstances sur-rounding slips for those who are prone to avoid or"repress" this awareness. It also serves to desen-sitize those who are supervigilant of their errors sothat they may reduce debilitating anxiety and relymore on coping skills to deal with the problemsituation.

In the context of a contingency leadership pro-gram, the trainer provides information on the topicof relapse and gives specific examples on how theRP model would apply to this form of training.Trainees are asked to describe previous relapses andto monitor any current experiences of this type.Finally, a questionnaire designed to assess the deter-minants of slips or relapses can be administered anddiscussed when appropriate.

Identification of High Risk Situations

The purpose of identifying potentially trouble-some situations in training is to sensitize the man-ager to work environments that are most likely toresult in a temporary or permanent lapse in ex-perimenting with new learning. Groups of managersare likely to have many high risk situations in com-mon, such as time pressure, angry subordinates,and lack of support from superiors. It is critical thateach manager have a detailed description of his ownhigh risk situations. These then can be furtheranalyzed according to problem type (social, emo-tional, work environment, interpersonal, intra-personal, etc.). With such increased preparedness,the manager is less likely to be surprised and/oroverwhelmed by these stressful occurrences.

In the case of the example of the time-pressedmanager, the mention of deadlines (time pressure)would serve as a "red alert" to the increased prob-ability of a slip and set in motion the rest of the RPsystem.

437

Developing Coping Responses

It is one thing to anticipate a slip or relapse duringtraining and quite another to cope with it on thejob. This model assumes that slips or relapses occurbecause of the absence or unavailability of skills. Ifa skill deficit is diagnosed, additional training in,for example, time management, assertiveness train-ing, or strategic planning is prescribed. If the skill ispresent but unavailable to the manager because ofhigh stress or anxiety, appropriate relaxation train-ing or life style intervention would be considered.

The major objective with regard to coping skillsis for managers to perceive most high risk situationsas environmental hurdles that can be cleared by theappropriate training regimen. The failure to clearinitial hurdles must not be interpreted as indicationsof a "bad hurdler," but rather the need for morecoaching.

Thus, the manager who reverts to autocratic con-trol under time pressure may reduce his anxieties bylearning more effective time management pro-cedures that take into account sudden deadlines.

Self-Efficacy

An important concept in the delicate interplaybetween one's ability to handle a situation and one'sfeeling about oneself is the concept of self-efficacy(Bandura, 1977). Coping successfully with difficultsituations appears to increase one's sense of mastery.However, mastery of most complex skills is a trialand error process requiring monitoring and copingskills. Improvement usually is gradual. For manytrainees, the inability to attain rapid mastery resultsimmediately in a decreased expectation in the abilityto master the skill in the future and a correspondingdecreased sense of self-efficacy.

RP training points out successful applications ofthe leadership model before the stressful situation—for example, the time pressure—occurred and ex-amines the situational determinants of these out-comes. Then comparisons can be made to laterlapses in the use of leadership skills. This analysisfocuses heavily on skills, mastery level, and self-efficacy. The manager's tendency to lower self-efficacy because of self-blame for the failure of thetraining to work is discussed. Other likely responsesto perceived inadequacy or helplessness, such asblaming subordinates and anger at consultants, canbe assessed from a self-efficacy perspective. Roleplaying and self-reinforcement for even minor im-

provements are used to train a cognitive set that em-phasizes skills and training rather than slogans andwillpower.

Expectancies of the Effects of the Activity

Closely related to expectancies of mastery arepositive expectancies surrounding the activity to bechanged. Just as an abstinent smoker has positiveexpectancies and fantasies of the relaxing effects ofa cigarette, the manager may overemphasize thepositive aspects of always using autocratic leader-ship (e.g., people obey) and underemphasize thelow morale, lack of loyalty, and high turnover. Thepositive short term consequences of immediate obe-dience are likely to be far more reinforcing to themanager than the less desirable long term conse-quences of low morale. Managers can gain a greaterperspective on this problem by listing the long andshort term advantages and disadvantages of newlylearned contingency leadership styles and of the oldpredominant autocratic style.

Abstinence Violation Effect (AVE)

Marlatt and Gordon (1980) found that individualswho were trying to abstain from smoking, drinking,overeating, drugs, and so on typically made a com-mitment to maintain full control of the problembehavior. They postulated further that individualsperceived their problem behavior along a dichoto-mous rather than continuous dimension (i.e., one iseither a drinker or a teetotaler). Thus when an ab-staining alcoholic takes a single drink, theabstinence commitment is violated and an emo-tional reaction that can lead to additional drinkingis likely to be experienced. This reaction is made upof two subresponses: (1) guilt and conflict (overviolating the rule), and (2) personal attribution(blaming self for the slip). Thus, after the initialslip, additional drinking occurs to deal with theguilt and self-blame.

Although management training differs dramat-ically from the treatment of alcoholism, the emo-tional reaction to giving up on a promising or ex-pensive training program may be quite similar.Managers can be taught to expect such emotionalreactions, and they also can be trained to reexaminesuch ineffective cognitive interpretations and taughtcompeting ones. Once again, the need for appro-priate skills and the opportunity to value the infor-mation from slips are paramount.

438

Apparently Irrelevant Decisions (AIDS)

Marlatt (1980) cites an example of a compulsivegambler who, on a vacation trip between San Fran-cisco and Seattle, "ended up in downtown Reno."Along the way, minor choices were made about theroute to be traveled, presumably in the name ofscenery and convenience. When the gambler"found" himself in downtown Reno, he was unableto cope with the stimuli present that encouragedgambling and went on a binge. Careful analysis intherapy revealed a series of AIDS that culminatedin his total loss of control.

Seemingly inconsequential actions of managerscan have the effect of sabotaging potentially suc-cessful training. Managers who are trying tobroaden their leadership style may inadvertentlybecome more vigilant of employee errors, thus re-quiring more supervision and a corresponding re-luctance to delegate under pressure. Time pressurecontingency planning might be "forgotten" so thatany sudden deadline "requires" autocratic control.

RP training requires managers to monitor minordecisions leading up to a slip or relapse and examinetheir cost effectiveness. Efforts are made to help themanager become aware of the impact of these seem-ingly unimportant actions before he suddenly findshimself in the manager's equivalent of "downtownReno."

Balanced Daily Life Style

Should/ Want Ratio. The potential for a slip orrelapse appears to be greater for persons who areoverburdened with shoulds (activities they mustperform) and deprived of wants (things they like todo).

Leaders who spend a great proportion of theirtime catering to tasks that primarily enhancesuperiors or subordinates without attending to self-enhancing activity of a professional or psycho-emotional nature may set themselves up for a martyrrole. They are likely to feel deprived and deservingof special attention. This pattern of activity leavesmanagers ripe for impulsive reactions, such ashostile or authoritarian leadership, when they arenot necessary. Unexpected crises always seem to oc-cur when managers are in the greatest need ofwants. A balance between wants and shoulds issought, with some of each occurring throughout theday.

Managers are asked to keep daily activity records,specifying every activity they engage in during theday. Each activity is rated on a scale from one toseven, with one being a complete should and sevenbeing a complete want. Many activities are a com-bination of wants and shoulds and would receive anintermediate number. RP theory predicts that slipsare more likely to occur when a manager has asteady string of shoulds uninterrupted by more self-reinforcing wants. In evaluating their lists, managersexamine which want activities can be moved to adifferent part of the day or which shoulds can bedelegated, discontinued, or changed. Managers withfew wants on their list might also consider changesin their jobs.

Life Style Interventions. Managerial and ad-ministrative work has been identified as among themost stressful of occupations. Responsibility hasbeen found to be one of the major causes of stress.Recent emphases in management training havefocused on the physical and emotional componentsof managerial activity in addition to cognitive andintellectual skills. This holistic view assumes thatphysical health and stable emotions contribute tooptimal performance in intellectual and inter-personal behaviors. Life style interventions such asregular exercise, diet and weight loss, and relaxationtraining are examples of methods for increasingphysical resistance to sustained pressure at work.Training in assertiveness, communication skills,and conflict resolution are utilized for improvinginterpersonal and emotional aspects of work.Counseling services for burnout, substance abuse,career development, retirement, and so on are addi-tional life style interventions that may give themanager a greater buffer zone for coping with oc-cupational stress.

Managers are asked to inventory these areas andascertain the extent to which these situations are set-ting events for a slip or relapse. Once the criticaldeficits are identified, proper remedial steps can betaken. In some instances, modifying only onebehavior can powerfully reduce the likelihood ofsuccumbing to work related stress. The managerwho is attempting to maintain a more delegativeleadership style may experience extreme anxietylevels under time pressure. For such an individual,relaxation training, time management training, anda dose of general stress management techniquesmay be the appropriate life style interventions to

439

provide him with a more potent set of coping skills.

Programmed Relapse

A final intervention prescribed for relapse pre-vention necessitates, ironically, practicing therelapse experience. The technique is somewhatparadoxical, consisting of a required return to theless effective behavior. In the scenario describedearlier, the manager would resume autocratic leader-ship in all situations. By programming this eventrather than allowing it to occur haphazardly understressful conditions, the trainer can monitor theparticipant's response to the relapse. Cognitive andattitudinal reactions that might later precipitateslips can be identified, and effective coping re-sponses can be reinforced. Most important, one'srecovery from a slip can be monitored, and effectivestrategies for future occurrences can be established.The requirement of practicing "failure" often elicitspolarity responses from trainees. Many dieters whoare required to binge frequently report difficultybinging as heavily as they used to and are able toend the binge more easily after 24 hours. It is likelythat managers "required" to be autocratic wouldslip into contingent leadership occasionally.

Because most new training is sabotaged tempo-rarily by overwhelming environmental circum-stances, the programmed relapse allows the traineeto accept such occurrences with less likelihood ofsuccumbing to the guilt and conflict of the absti-nence violation effect. It also ensures careful atten-tion to all aspects of relapse, particularly to resump-tion of the trained behavior at the end of the re-quired relapse period. Programmed relapse can firstbe role-played during the course of training and

then carried out by managers in the work environ-ment.

The list of techniques described here is derivedfrom a systematic model of relapse prevention.Specific strategies are designed to enhance each stepin the process. Although all steps are interrelated,some strategies such as high risk identificationshould occur early in training, but programmedrelapse is likely to be more useful after most of theother steps have been introduced. These techniquesare not meant to be exhaustive, but the list includesmost of the intervention strategies utilized byMarlatt and his colleagues in the clinical treatmentof addictive disorders. The effects of these strat-egies on managerial behavior are summarized inTable 2.

Implications for Research

The opportunities for research in the area of RPin management and industry are enormous. Mostcorporations have their share of deserted trainingpackages and forgotten organizational interven-tions. Many were initiated in a flurry of optimismand excitement. They lasted through a brief im-plementation period and then faded in a quiet returnto the status quo. For many organizational inter-ventions, there exists a casualty list of slips andrelapses that can be investigated. Relapsers can becompared with nonrelapsers and such differentialdata recycled into refined treatment programs.

An important first step in this process would bethe determination of the relevance of the Marlatt-Gordon relapse model for nonaddictive behaviors.It is an empirical question whether a model designed

Table 2Functions of Relapse Prevention Strategies for Management Training

Strategy Purpose

1. Awareness of the relapse process2. Identification of high risk situations

3. Developing coping responses4. Enhancing self-efficacy5. Expectancies of the effects of the activity6. Abstinence violation effect (AVE)

7. Apparently irrelevant decisions (AIDS)

8. Should/want ratio9. Life style interventions

10. Programmed relapse

1. Helps manager to understand how self-control strategies can enhance maintenance2. Heightens manager's sensitivity to specific situations that have previously resulted in

difficulty3. Enables manager to learn appropriate skills to overcome environmental hurdles4. Assists manager in maintaining personal worth, despite imperfect performance5. Counters manager's short term positive expectations of the ineffective behavior6. Helps manager to reduce guilt and avoid self-blame for having failed to implement a

newly trained behavior7. Strengthens awareness of seemingly minor decisions that culminate in a slip or

relapse8. Helps manager to maintain a balance of personal satisfaction in daily activities9. Proposes physical and emotional coping skills to buffer manager from stress-

induced relapse10. Provides manager with monitored failure experiences that can be analyzed to avoid

future relapse

440

to explain components of long term maintenance ofaddictive or appetitive behaviors also would be rele-vant for the maintenance of activities such as skillstraining. Questionnaires similar to those utilized foraddictive behaviors can be designed to measuredeterminants of relapse for management behavior,and the results can be compared with those alreadyobtained for addictive behavior.

Definitions of slips and relapses in skills trainingneed to be specified and refined. Whereas smoking,drinking, and drug taking are discrete and measur-able activities, operational definitions of slips andrelapses are less precise when applied to managerialbehavior.

The role of supervisory reinforcement for RP isanother topic of considerable interest. It is likelythat training programs fail to take hold because oflack of administrative support and interest. In such

cases, trainees are likely to rely primarily on theself-reinforcement that is experienced when thetraining makes the work easier or more successful.However, during the acquisition phase of training,when more errors are likely to occur, the behaviormay not have been shaped to the point at which self-reinforcement is occurring with sufficient frequencyto sustain maintenance. In this early stage, teachingsupervisors to reinforce RP behaviors on the part oftrainees may enhance the maintenance of the newskill.

The cognitive-behavioral RP model may providea promising methodology for enhancing the main-tenance of management training and other organi-zational interventions. By focusing attention on thepotential for temporary failure, the likelihood oflong term success may be increased.

References

Bandura, A. Self-efficacy: Toward a unifying theory of behaviorchange. Psychological Review, 1977, 84, 191-215.

Chaney, E. F, O'Leary, M. R., & Marlatt, G. A. Skill trainingwith alcoholics. Journal of Consulting and Clinical Psy-chology, 1978,46, 1092-1104.

Fiedler, F. E. A theory of leadership effectiveness. New York:McGraw-Hill, 1967.

Goldstein, 1. L. Training in work organizations. Annual Reviewof Psychology, 1980, 31, 229-272.

Hersey, P , & Blanchard, K. H. Management of organizationalbehavior: Utilizing human resources. Englewood Cliff, N.J.Prentice Hall, 1977.

House, R. A path-goal model of leader effectiveness. Adminis-trative Science Quarterly, 1971, 16, 321-338.

Latham, G. P., & Saari, L. M. The application of social learningtheory to training supervisors through behavioral modeling.Journal of Applied Psychology, 1979, 64, 239-246.

Leifer, M. S., & Newstrom, J. W. Solving the transfer of trainingproblems. Training and Development Journal, August, 1980,34, 42-46.

Marlatt, G. A. Relapse prevention: A self-control program forthe treatment of addictive behaviors. Invited addresspresented at the International Conference on Behavior Mod-ification, Banff, Alberta, March 1980.

Marlatt, G. A., & Gordon, J. R. Determinants of relapse: Im-plications for the maintenance of behavior change. In P. O.Davidson & S. M. Davidson (Eds.), Behavioral medicine:Changing health lifestyles. New York: Brunner/Mazel, 1980,410-452.

Michalak, D. F. The neglected half of training. Training andDevelopment Journal, May, 1981, 35, 22-28.

Rosenthal, B. S., & Marx, R. D. A comparison of standardbehavioral and relapse prevention weight reduction program.Paper presented at the meeting of the Association for the Ad-vancement of Behavior Therapy, San Francisco, 1979.

Rosenthal, R., & Jacobson, L. Pygmalion in the classroom:Teacher expectations and pupils' intellectual development.New York: Holt, Rinehart, & Winston, 1968.

Vroom, V. H., & Yetton, P. W. Leadership and decision making.Pittsburgh: University of Pittsburgh, 1973.

Robert D. Marx is Associate Professor of Psychology atWestern Washington University.

441

Copyright of Academy of Management Review is the property of Academy of Management and its content may

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

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