THERAPEUTIC RECREATION JOURNAL Vol. 36, No. 1, 35-47, 2002

Effects of Attribution Retraining During Therapeutic Recreation on Attributions and Explanatory Styles of Adolescents with Depression

Rodney B. Dieser and Edward Ruddell

The purpose of this study was to examine the effects of attribution retraining on attributions and during therapeutic recreation activity. Research participants were sixteen adolescents who were in-patients at a residential treatment facility and from a psychiatric hospital. Results from this study indicated that after attribution retraining (a) the experimental group scored significantly higher than the control group on personal control and stability attributions; (b) the experimental group did not score significantly higher than the control group on locus of attributions and explanatory styles; (c) the experimental group did not score significantly lower than the control group on external control attributions. Implications for practice and directions for future research are addressed.

KEY WORDS: Attribution Retraining, Attribution Style Questionnaire, Attribution Retraining, Causal Dimension Scale II, Verbal .

Many studies support the existence of a pessimistic explanatory styles, and depression relationship among unhealthy attributions, (Buchanan, 1995; Oettingen, 1995; Peterson &

Rodney B. Dieser—Assistant professor in the School of Health, Physical Education, and Leisure Services, University of Northern Iowa. Edward Ruddell—Associate professor in the Department of Parks, Recreation, and Tourism, University of Utah.

We wish to express our thanks to Dr. Gary Ellis, Dr. Cathy Morris, and Dr. Mark Widmer for their comments and suggestions.

First Quarter 2002 35 Seligman, 1984; Sweeney, Anderson, & the study by Ellis et al. (1993) demonstrated Bailey, 1986). For example, in a longitudinal that attribution based verbal persuasion during study that consisted of 508 third grade chil- therapeutic recreation experiences can have a dren, Nolen-Hoeksema, Girgus, and Seligman positive significant effect on self-efficacy (1992) measured explanatory styles and de- judgments, outcome judgments, generality of pression every six months. They found that efficacy judgments, and task persistence. They later in childhood pessimistic explanatory concluded that "attribution based verbal per- styles were significant predictors of depres- suasion can be a potent force in impacting on sion. Likewise, Raps, Peterson, Reinhard, clients' efficacy judgments" (p. 94). Abramson, and Seligman (1982) examined 30 Integrating attribution retraining with ther- clients with unipolar depression, 15 clients apeutic recreation interventions has important with nondepressed schizophrenia, and 61 cli- implications for improving mental health (see ents with nondepressed medical and surgical Dieser & Voight, 1998; Ellis et al., 1993; conditions and found that clients with depres- Iso-Ahola, 1980; Voight, 1983). Therapeutic sion explained negative life events with more recreation interventions provide a modality to pessimistic attributions than clients in the change pessimistic attributions because recre- other groups. ation can be designed to present an optimal Scholars in psychology have posited that and challenging activity while providing a attribution retraining, or attribution therapy, high probability of success (Widmer, 1996). can improve the health of individuals with Combining successful recreation experiences depression (Beck & Weishaar, 1995; DeRu- with healthy attributions can lead clients to beis & Holland, 1995; Forsterling, 1988; greater perceived control (Iso-Ahola, 1980). Peterson, 1982; Seligman, Reivich, Jaycox, & Also, therapeutic recreation settings, as op- Gillham, 1995). Attribution retraining inter- posed to clinical settings, may be perceived as ventions have consisted of therapists targeting less threatening to clients. Consequently, cli- automatic thoughts rooted in unhealthy attri- ents may be less defensive and resistant to butions and challenging these pessimistic at- having their pessimistic attributions chal- tributions via offering alternative explanations lenged. premised upon healthy attributions (see For- Due to scant empirical evidence regarding sterling, 1988; Seligman et al., 1995). In the the integration of attribution retraining with past, attribution retraining usually lasted 8-12 therapeutic recreation intervention, it is un- weeks and occurred in clinical (DeRubeis & known whether attribution retraining can be Holland, 1995), group counseling (Green-Em- utilized successfully in therapeutic recreation rich & Altmaier, 1991), psycho-educational practice. The purpose of the study presented in (Seligman et al., 1995), and therapeutic recre- this paper was to build on the research of Ellis ation (Ellis, Maughan-Pritchett, & Ruddell, et al. (1993) by examining the effects of verbal 1993) settings. persuasion based attribution retraining during Although many therapeutic recreation therapeutic recreation on attributions and ex- scholars and practitioners have advocated to planatory styles of adolescents with a clinical integrate attribution retraining with therapeutic diagnosis of major depression. recreation practice to ameliorate the well being of individuals with psychological disorders Literature Review (Dieser & Voight, 1998; Dixon, 1979; Ellis et al., 1993; Iso-Ahola, 1980; Maughan & Ellis, Attribution Theory and 1991; Voight, 1983; Widmer, 1996), little em- Depression pirical evidence exists to validate using attri- Attribution theory describes the systematic bution retraining as a theoretical intervention patterns people use when ascribing cause to a in therapeutic recreation practice. However, human event (Peterson & Seligman, 1984;

36 Therapeutic Recreation Journal Weiner, 1985). Inquiry into the manner in 1989; Morris & Ellis, 1993), attribution based which people assign cause to human events verbal persuasion (Ellis et al., 1993), and lei- began in the 1940s. In an early study, during sure education intervention (Dieser & Voight, which Heider (1958) presented pictures of 1998). Investigation of cross-cultural attribu- meaningless drawings to individuals, he found tions became prevalent in the latter 1990s (see that people regularly and predictably assigned Choi, Nisbett, & Norenzayan, 1999; Matsu- patterns of causality where such causality did moto, 1996; Menon, Morris, Chiu, & Hong, not exist. 1999; Oettingen, 1995). In the early 1970s, Kelley's (1971) work continued the exploration and advancement of Causal Dimensions attribution theory. The underlying premise of Central to research regarding attribution the- his inquiry was that causal thoughts play a ory during the 1980s and 1990s was a movement central role in determining individuals' feel- aimed toward clustering causal explanations into ings and behaviors. Furthermore, Kelley con- sets of dimensions. Causal dimensions are cluded that individuals seek a causal explana- groupings of causal attributions—four frequently tion for events that occur in physical and social studied causal dimensions are , environments because explanations function in locus of causality, stability, and globality (Kent ways that help individuals attain personal & Martinko, 1995). goals. In this regard, Kelley (1971) posited Locus of control. Locus of control refers to that "the attributor is not simply an attributor, internal and external attributions. Internal a seeker after knowledge; his latent goal in causes refer to factors within the person and attaining knowledge is that of effective man- external causes relate to the situation or cir- agement of himself and his environment" (p. cumstance (Abramson et al., 1978). For exam- 220). ple, a university student attributing an excel- In the latter half of the 1970s, using attri- lent mark on a test to his/her effort in studying bution theory as a theoretical base, Seligman would exemplify an internal attribution be- (1975) developed a model to explain learned cause the success is attributed to something helplessness which three years later was about the individual (his/her studying). The changed into a reformulated model (see opposite of this, an external attribution, would Abramson, Seligman, & Teasdale, 1978). be demonstrated by a student attributing his/ , a psychological phe- her mark to an easy exam. nomenon in which experiences with uncon- Locus of causality. Locus of causality is a trollable events create passive behaviors to- causal dimension closely related to locus of ward maintaining well-being, emphasizes that control. Locus of causality refers to whether people learn to become helpless in controlling the cause resides within or outside the attribu- their lives because of their pessimistic attribu- tor (McAuley, Duncan, & Russell, 1992). A tions (Abramson et al., 1978; Seligman, 1990). chief difference between locus of control and According to Gilbert (1992), learned helpless- locus of causality is the way in which the ness theory has become one of the most im- external causal attribution is defined. Within portant theories used to explain depression. the causal dimension of locus of causality, Attribution-based models of intervention external refers to causes that reside outside of began to be introduced across psychology in the attributor. Within the casual dimension of the 1980s and 1990s (see DeRubeis & Hol- locus of control, external refers to something land, 1995; Forsterling, 1988; Seligman et al., about the situation or circumstances. That is, 1995). Likewise, attribution theory served as an external attribution within the locus of cau- the conceptual and theoretical basis for many sality dimension refers to causes that reside aspects of therapeutic recreation, such as the outside the control of the attributor (e.g., the Leisure Diagnostic Battery (Witt & Ellis, weather) however, an external attribution

First Quarter 2002 37 within the locus of control dimension refers to attribute—she can cause her grip to remain causes regarding the situation or circum- constant. stances which may refer to something outside Globality. Globality refers to breadth of the individuals (e.g., the weather) but also may attributions. Global attributions are those that be part of the individual (e.g., genetics). An affect a variety of outcomes; conversely, spe- example of an external attribution within the cific attributions are limited to narrow and locus of control dimension could be an indi- specific outcomes (Abramson et al., 1978). vidual who does not have the genetic makeup The remark "I am stupid" by a student who to become a famous bodybuilder. Although fails a math exam exemplifies a global at- the individual may eat well and train hard, his tribute because such a remark implies a per- physiological disposition may not enhance vasive variety of outcome—the person is stu- muscle cell expansion. This situation is out- pid in many things. However, the remark "I'm side the individual's own responses and does terrible in math" implies a particular area of not relate to environmental factors, but still is weakness and hence is a specific attribution. considered external.1 Hence, the external attri- bution that is conceptualized within the locus Explanatory Style of control dimension may have more breadth An explanatory style is an individual's ten- in regard to ascribing cause to a human event dency to offer similar attributions for different than the external dimension framed within the events (Peterson, Buchanan, & Seligman, locus of causality dimension. 1995). Within a Western context, an optimistic Stability. Stability alludes to consistency of (healthy) explanatory style is one in which an attributions. Stable attributions remain con- individual applies internal, stable, and global stant, whereas unstable fluctuate attributions to positive outcomes, or external, (Weiner et al., 1971). That is, stable refers to unstable, and specific attributions to negative whether an individual perceives that an out- outcomes (Seligman, 1990). In contrast, a pes- come is, over time, unchangeable, while un- simistic (unhealthy) explanatory style in West- stable refers to whether an individual per- ern culture is one through which an individual ceives that an outcome is, over time, applies internal, stable, and global attributions changeable. For example, if an individual ex- to negative outcomes, or external, unstable, and specific attributions to positive outcomes plained that she played an excellent round of (Seligman, 1990). Bandura (1997) highlighted golf due to luck, this would imply that her that optimistic attributions can bolster positive success is changeable—luck is not controlled changes in self if the appraisal is based upon by individuals. If she attributed golfing well to realistic and accurate criteria. a new way of gripping her golf clubs, then her attribution would imply that her success may Table 1, adapted from Peterson & Selig- be permanent, and thus would be a stable man (1984), presents a clear conceptualization of optimistic and pessimistic attribution styles regarding the negative event of playing a poor 1 This is one of the subtle yet paramount differ- round of recreational golf. The internal, stable, ences between locus of control and locus of causal- and global explanation "I can't do anything ity. Furthermore, Weiner (1985) argued that locus right" is the most pessimistic attribution be- and control are separate because an individual can cause the negative event is perceived as caused control the amount of effort he/she puts into an event. This is why Weiner (1985) has argued for an by the individual, unchangeable, and affecting additional attribution dimension of intentionality. many aspects of the person's life. The exter- The intentionality dimension has received scant sup- nal, unstable, and specific explanation "I'm port. Kent and Martinko (1995) provided a convinc- surprised—this golf course has never been in ing argument that insufficient effort and improper strategy are related to internal and unstable attribu- such poor shape" is the most optimistic attri- tions and not related to intentionality. bution because the negative event is perceived

38 Therapeutic Recreation Journal Table 1. Examples of Causal Attributions for the Negative Event of Playing a Poor Round of Golf

Internal External

Stable Global I can't do anything right. All golf courses have poorly designed fairways and greens. Specific I always have trouble with my This particular golf course has putting. terrible greens. Unstable Global I've not felt well for a couple of This time of year demands that days and I've let my whole life one concentrate on other fall apart, including my golf game. things than golf. Specific The one time I forget my golf shoes I'm surprised—this golf course is the one time that my partner has never been in such poor plays his/her best round of golf. shape.

Adapted from: Peterson, C, & Seligman, M. E. P. (1984). Causal explanations as a risk factor for depression: Theory and evidence. Psychological Review, 91, 347—3"r4. as caused by situational factors, changeable, study one percent of people born prior to and affecting only golf. World War I experienced depression and four Pessimistic attributions and explanatory percent of people born around 1925 experi- styles are at the core of some types of depres- enced depression. However, for people born in sions (Peterson & Seligman, 1984; Seligman, the 1950s, seven percent of the research par- 1990, 1993). Sweeney et al. (1986) conducted ticipants had experienced severe depression by a meta-analysis review of studies which in- their early twenties. Likewise, Klerman et al. volved nearly 15,000 participants from 104 (1985) studied 2,289 close relatives of 523 reported studies and concluded that for nega- people who had been hospitalized for severe tive events pessimistic explanatory styles depression. In this study 60 percent of women (composite scores of internal, stable, and born in the early 1950s had experienced severe global attributions) had a positive, linear asso- depression by the time they were in their mid ciation with depression. They also concluded thirties, but in contrast, only three percent of that for positive events optimistic explanatory women who were born around 1910 had ex- styles (composite scores of internal, stable, perienced depression. In this regard, Wenegrat and global attributions) had a negative, linear (1995) argued that depression among Ameri- association with depression. can women has reached epidemic propor- The relationship between pessimistic ex- tions—depression among American women is planatory styles and depression is paramount twice that of men (Seligman, 1993). This same because over the course of a century, depres- escalating trend has been observed among sion has increased at an alarming rate (Selig- American teenagers (Lewinsohn, Rohde, See- man, 1993). Robins et al. (1984) cross-sec- ley, & Fischer, 1993) and among people of tional study that randomly interviewed 9,500 differing cultures, such as American Indians people found that depression is increasing con- (May, 1987), Alaskan Natives (Herring, siderably over the course of the century. In this 1992), and African-Americans (Dinges &

First Quarter 2002 39 Cherry, 1995). In sum, Americans born in the People who are persuaded verbally that second half of the century are ten times likelier they possess the capabilities to master to suffer from depression than Americans born given tasks are likely to mobilize in the first half of the century (Seligman, greater effort... to try hard enough to 1993). Hence, depression has been termed 'the succeed ... [these] self-affirming be- common cold of psychopathology' (Seligman, liefs promote development of skills and 1975). a sense of personal efficacy (p. 101). Pessimistic explanatory styles, unhealthy attributions, and depression can be decreased Furthermore, by convincing individuals to be- if individuals develop or maintain a sense of lieve they possess new capabilities, pessimis- personal capacity and control (Peterson, tic attributions and explanatory styles can be Maier, & Seligman, 1993; Seligman, 1993). challenged because the resulting new behav- One way that personal control can be in- iors provide visible, contradictory evidence to creased is by attribution retraining (Forster- dispute pessimistic attributions. ling, 1988; Peterson, 1982; Seligman, 1990; The Ellis et al. (1993) study which con- Seligman et al., 1995). sisted of attribution based verbal persuasion in a therapeutic recreation setting highlighted the Attribution Retraining in positive effect attribution retraining could Cognitive Therapy have on the mental health of adolescents with Attribution retraining is a form of cognitive major depression. After playing video games therapy (Beck & Weishaar, 1995; Seligman, the research participants who received attribu- 1993). Cognitive therapy is based on attempts tion based verbal persuasion messages scored to reduce excessive emotional reaction and higher on self-efficacy related variables than self-defeating behavior by modifying faulty or the research participants who received ambig- erroneous thinking and maladaptive beliefs uous verbal messages. In this study, the exper- that underlie these reactions (Beck, Rush, imental group received attribution based ver- Shaw, & Emery, 1979). Within a cognitive bal statements that indicated personal success therapy framework, attribution retraining usu- was due to internal and stable factors (e.g., ally involves targeting pessimistic attributions "You know how to use great strategy") while and changing them into optimistic attribution the control group received vague verbal state- by disputing the accuracy of pessimistic ments (e.g., "Nice job"). thoughts (Peterson, 1982; Seligman et al., 1995). Forsterling (1988) advocated attribu- Hypotheses tion retraining via verbal persuasion because Based upon the preceding rationale of the verbal persuasion can convince people that relationship between attributions, explanatory they possess capabilities that will enable them styles, and attribution retraining, the following to achieve personal goals. In this regard, Ban- hypotheses were tested: dura (1997) suggested that verbal persuasory HI: The mean locus of causality score of efficacy attributions serve as a means of an attribution retraining group will be signifi- cantly higher than the mean locus of causality ... Strengthening people's beliefs that score of a minimal verbal response group. they possess the capabilities to achieve H2: The mean stability attribution score of what they seek. It is easier to sustain a an attribution retraining group will be signifi- sense of efficacy, especially when cantly higher than the mean stability attribu- struggling with difficulties, if signifi- tion score of a minimal verbal response group. cant others express faith in one's capa- H3: The mean personal control attribution bilities than if they convey doubts ... score of an attribution retraining group will be

40 Therapeutic Recreation Journal significantly higher than the mean personal planatory styles. The ASQ measures explana- control attribution score of a minimal verbal tory style (based upon the causal dimensions response group. of locus of control, stability, and globality) and H4: The mean external attribution score of does not measure situational attributions an attribution retraining group will be signifi- (Henry & Campbell, 1995; Tennen & Herz- cantly lower than the mean external attribution berger, 1985). The scale's internal consistency score of a minimal verbal response group. has been reviewed (Peterson et al., 1993)— H5: The mean composition explanatory alpha coefficients for the individual dimen- style score of an attribution retraining group sions ranged from .40 to .70 while the test- will be significantly higher than the mean retest correlation was reported as .60 and composition explanatory score of a minimal above. A higher score on the ASQ is desirable. verbal response group. Procedure Method After informed consent was obtained, par- Participants ticipants completed the CDS II and the ASQ. Participants then interacted in their daily ther- Eighteen Caucasian adolescents (aged 14- apeutic recreation activities that the treatment 17) who were medically diagnosed with major facilities provided. Twice a week the re- depression participated in this study. Eight searcher met individually with participants and participants resided at a residential treatment asked each participant the following open- facility and the other 10 participants resided in ended question: "Tell me one successful and a psychiatric hospital. All research participants unsuccessful therapeutic recreation activity and their parents were informed of the pro- you were involved in during the past week." posed study and informed consent was se- After a participant disclosed a successful and cured. After all participants were randomly unsuccessful therapeutic recreation activity, assigned to either the experimental group or optimistic verbal messages (e.g., internal, sta- the control group two participants from the ble and global attributions for successful control group withdrew from the study due to events and external, unstable, and specific at- early discharge. Thus, the experimental group tributions for unsuccessful events) were pro- consisted of nine participants and the control vided by the researcher (see Table 2 and Table group included seven participants. 3). For example, the verbal persuasion state- ment "It sounds like the activity was success- Instrumentation ful because you put much effort into it" was stated by the researcher after disclosure of a Attributions were measured by the Causal successful therapeutic recreation activity be- Dimension Scale II (CDS II) (McAuley et al., cause such a sentence is premised upon an 1992). The CDS II measures four attributions: internal locus of control. locus of causality, stability, personal control, and external control. Internal consistencies for When each client disclosed a successful the four causal dimensions have been reported and unsuccessful therapeutic recreation activ- as follows: locus of causality, .67; stability, ity the researcher was able to present attribu- .67; personal control, .79; external, .82 tions linked to the event, and thus attempted to (McAuley et al., 1992). In regard to the CDS change attributions toward an optimistic II, higher scores in locus of causality, stability, framework via verbal persuasion statements. and person control are desirable and a lower In other words, attribution retraining verbal score on external control is desirable. persuasion intervention was used to convince The Attribution Style Questionnaire (ASQ) people that they possessed more control than (see Peterson & Seligman, 1984; Tennen & they thought (Bandura, 1997). The control Herzberger, 1985) was used to measure ex- group received minimal verbal responses such

First Quarter 2002 41 Table 2. Verbal Persuasion Statements for Successful Events

Attribution Verbal Persuasion Statements

Internal It sounds like the activity was successful because you put much effort into it. Internal What I'm hearing you say is that the activity was successful because you worked hard at it. Stable It sounds like (the name of the therapeutic recreation activity) will continue to be successful for you. Stable I'm guessing you will continue to succeed at (the name of the therapeutic recreation activity) . Global It seems to me that if you concentrated on other tasks like you have on (the name of the therapeutic recreation activity) that you will become successful in it too. Global You will do well in many activities if you put forth the amount of effort you did during (the name of the therapeutic recreation activity) .

as the following phrases "good," "uh-huh," was purposely designed to be a brief interven- "sure," "super," and "I see." tion. Overall, 12 verbal statements premised To reflect contemporary health care prac- upon optimistic-oriented attributions were im- tice that is premised upon brief treatment plemented over four weeks. That is, the inter- (Carter, Van Andel, & Robb, 1995; Witman & vention lasted four weeks, with a total of six Batchelder, 1992) the attribution retraining sessions. verbal persuasion intervention in this study To understand if the research participants

Table 3. Verbal Persuasion Statements for Unsuccessful Events

Attribution Verbal Persuasion Statements

External It sounds like the activity was unsuccessful because there wasn't much effort put into it. External It sounds like the activity was unsuccessful because it simply wasn't your day. Stable Unlucky break, but I bet you can do well in (the name of the therapeutic recreation activity) if you try again. Stable I'm guessing you will succeed at (the name of the therapeutic recreation activity) next time. Specific It seems to me that the only activity you had difficulty with was (the name of the therapeutic recreation activity) . Global What I'm hearing you say is that (the name of the therapeutic recreation activity) was one of the few activities you felt was unsuccessful.

42 Therapeutic Recreation Journal Table 4. higher in the attribution retraining verbal per- Posttest Scores of Dependent Variables suasion group, the difference of the locus of causality attribution means were not statisti- Dependent Variables M SD F cally significant between the two groups. The second hypothesis was supported. The differ- Locus of Causality ence of the stability attribution means were Control Group 21.14 3.53 statistically significant between the two 0.47 Experimental Group 22.56 5.36 groups. The third hypothesis was also sup- Stability ported—the difference of the personal control Control Group 12.43 4.12 attribution means were statistically significant 7.40* Experimental Group 18.33 4.21 between the two groups. The fourth hypothesis Personal Control was not supported. Although the posttest score was lower in the attribution retraining verbal Control Group 20.00 5.45 7.00* persuasion group, the difference of the exter- Experimental Group 25.56 2.35 nal control attribution means were not statisti- External Control cally significant between the two groups. The Control Group 14.29 6.37 0.00 fifth hypothesis was not supported. Although Experimental Group 13.11 7.46 the posttest score were higher in the attribution Explanatory Styles retraining verbal persuasion group, the differ- Control Group 1.87 2.35 1.86 ence of explanatory style means were not sta- Experimental Group 3.80 2.76 tistically significant between the two groups. Hence, the results of this study suggest that the *p < .05. research participants who experienced attribu- tion retraining verbal persuasion intervention believed that successful therapeutic recreation in the experimental group heard the verbal activities were caused by factors within them- statements, which were provided by the re- selves (personal control) and that these suc- searcher, a manipulation check was included cessful activities will, over time, remain con- at the end of each session. Thus, after each stant (stability). session participants in the experimental group received a paper that had three written state- The manipulation check indicated that the ments and were asked to indicate which of research participants in the experimental group three statements accurately described the feed- were aware of and could identify the optimis- back provided by the researcher. tic verbal persuasion messages provided by the researcher. Of the possible 162 manipulation Data Analysis check questions (the number of questions [3] multiplied by the number of sessions [6] mul- Data were analyzed through analysis of tiplied by the number of research participants covariance (ANCOVA). An ANCOVA in- in the experimental group [9]) only seven creases precision in measurement by removing questions were answered incorrectly. pre-existing differences among groups (Wildt To gain a more precise measure of whether & Ahtola, 1978). In this study, a priori alpha the independent variable had an effect on the was set at .05 and the pretest means served as dependent variables of attributions and explan- the covariates. atory style, two additional post hoc statistical procedures were implemented. First, a power Results analysis was conducted to check for the pos- Results of the hypotheses tests are pre- sibility of a type II error. The power analysis sented in Table 4. The first hypothesis was not suggested that 46.8 cases would have needed supported. Although the posttest score was to be studied to conclude that a type II error

First Quarter 2002 43 was not made. Furthermore, the power analy- tions, findings of this study point to the need sis score for the 16 cases in this study was for continued investigation of the effects of .031. The underpower of this statistical proce- attribution retraining verbal persuasion on re- dure increases the probability that true differ- lated variables. Future attribution retraining ences between the control group and experi- research can be built upon the design of this mental group would not be discovered. study. For example, in this study attribution Second, reliability scores for the ASQ and the retraining lasted only four weeks—future at- CDS II were conducted so that competing tribution retraining verbal persuasion could explanations of the effect of the experimental examine the effects of a longer intervention, stimulus on the dependent variable could be perhaps six to eight weeks. Additionally, fu- lessened. The alpha coefficients for the four ture research lasting four weeks could provide subscales of the CDS II were stability .43; more verbal persuasion statements (such as external control .75; locus of causality .64; and providing optimistic attribution statements on personal control .37. The alpha coefficient for a daily basis). Likewise, future attribution re- the entire CDS II scale was 0.50. The alpha training verbal persuasion research can repli- score of the ASQ was .085. Due to low reli- cate this study with a larger sample size. The ability scores of the three subscales of the power analysis in this study suggested that CDSII and the low power analysis score, re- 46.8 cases would need to be studied to con- sults from this study need to be interpreted clude that a type II error would not occur— with caution. replication of this study could follow with 50 or more research participants. Second, this Discussion study aligns itself with other studies that dem- Results from this study present unclear onstrated the therapeutic effects of attribution findings. On the one hand, the results suggest retraining verbal persuasion, both within ther- that attribution retraining verbal persuasion in apeutic recreation practice (Ellis et al., 1993) a therapeutic recreation setting can have a and other social and educational settings (see significant effect on stability and personal con- Craske, 1988; Miller, Brickman, & Bolten, trol attributions. Additionally, the manipula- 1975). tion check indicated that participants were Furthermore, this study, combined with aware of the researcher's optimistic verbal other studies (e.g., Ellis et al., 1993), has persuasion messages. On the other hand, the considerable implications for therapeutic rec- power analysis suggests that detection of true reation practice. First, therapeutic recreation differences between the control group and ex- specialists can use similar verbal persuasion perimental group would be virtually impossi- attributions for clients after successful events ble with the current sample size. Furthermore, (see Table 2) and after unsuccessful events the reliability scores of the stability and per- (see Table 3). From a pragmatic perspective, sonal control subscales of the CDS II were this study underscores that what therapeutic low. In light of this study's shortcomings (e.g., recreation specialists say can have positive or underpower and low reliability values), the negative impacts on client attributions. Sec- two positive, significant findings should be ond, attribution theory could be used as a considered impressive. Furthermore, another theoretical base in therapeutic recreation inter- salient finding of this study is that all of the vention to ameliorate certain psychological posttest score means of the dependent vari- disorders. For example, Dieser and Voight ables changed in the predicted relationship. (1998) presented a case study of the manner in In this regard, the findings of this study are which attribution retraining leisure education important for two reasons. First, attribution can be used with people who are substance retraining involved a brief intervention—if a dependent. Furthermore, having therapeutic brief intervention can cause change in attribu- recreation specialists ascribing cause to opti-

44 Therapeutic Recreation Journal mistic attributions can increase perceived free- Dieser, R. B., & Voight, A. (1998). Therapeutic dom among clients, which is paramount to recreation and relapse prevention intervention. Parks and Recreation, 33(5), 78-83. experience leisure and antithetical to learned helplessness (Mannell & Kleiber, 1997). Dixon, J. T. (1979). The implications of attribu- tion therapy for therapeutic recreation services. Therapeutic Recreation Journal, 13(1), 3-11. Ellis, G. D., Maughan-Pritchett, M., & Ruddell, E. (1993). Effects of efficacy information during References recreation activity on efficacy judgments of adoles- cent psychiatric patients. Therapeutic Recreation Abramson, L., Seligman, M. E. P., & Teasdale, Journal, 27, 83-97. J. (1978). Learned helplessness in humans: Critique and reformulation. Journal of Abnormal Psychol- Forsterling, F. (1988). Attribution theory in clin- ogy, 87(1), 49-74. ical psychology. New York: John Wiley & Sons. Bandura, A. (1997). Self-efficacy: The exercise Gilbert, P. (1992). Depression: The evolution of of control. New York: W. H. Freeman and Com- powerlessness. New York: Guilford Press. pany. Green-Emrich, A., & Altmaier, E. (1991). Attri- Beck, A. T., & Weishaar, M. (1995). Cognitive bution retraining as a structured group counseling therapy. In R. J. Corsini & D. Weddings (Eds.), intervention. Journal of Counseling and Develop- Current psychotherapies (5th ed.) (pp. 21-36). ment, 69, 351-355. Itasca, IL: F. E. Peacock. Heider, F. (1958). The psychology of interper- sonal relationships. New York: John Wiley & Sons. Beck, A. T., Rush, A. J., Shaw, B. F., & Emery, G. (1979). Cognitive therapy of depression. New Henry, W. J., & Campbell, C. (1995). A com- York: Guilford Press. parison of the validity, predictiveness, and consis- tency of a trait versus situational measure of attri- Buchanan, G. M. (1995). Explanatory styles and butions. In M. J. Martinko (Ed.), Attribution theory: coronary heart disease. In G. M. Buchanan & An organizational perspective (pp. 35-52). Delray M. E. P. Seligman (Eds.), Explanatory styles (pp. Beach, FL: St. Lucie Press. 225-232). Hillsdale, NJ: Lawrence Erlbaum. Herring, D. R. (1992). Seeking a new paradigm: Carter, M. J., Van Andel, G. E., & Robb, G. M. Counseling Native Americans. Journal of Multicul- (1995). Therapeutic recreation: A practical ap- tural Counseling, 20, 35—43. proach (2nd ed.). Prospect Heights, IL: Waveland Press, Inc. Iso-Ahola, S. E. (1980). The of leisure and recreation. Dubuque, LA: William C. Choi, I., Nisbett, R. E., & Norenzayan, A. Brown. (1999). Causal attributions across cultures: Varia- Kelley, H. H. (1971). Attribution in social inter- tion and universality. Psychological Bulletin, 125, 47-63. action. Morristown, NJ: General Learning Press. Kent, R., & Martinko, M. (1995). The measure- Craske, M. (1988). Learned helplessness, self- ment of attributions in organizational research. In worth motivation and attribution retraining for pri- M. J. Martinko (Ed.), Attribution theory: An orga- mary school children. British Journal of Educa- nizational perspective (pp. 17—34). Delray Beach, tional Psychology, 58, 152-163. FL: St. Lucie Press. DeRubeis, R. J., & Holland, S. D. (1995). Ex- Klerman, G., Lavori, P., Rice, J., Reich, T., planatory styles in the treatment of depression. In Endicott, J., Andreasen, N., Keller, M., & Hirsch- G. M. Buchanan & M. E. P. Seligman (Eds.), Ex- feld, R. (1985). Birth cohort trends in rates of major planatory styles (pp. 99-112). Hillsdale, NJ: Law- depressive disorder among relatives of patients with rence Erlbaum Associates. affective disorder. Archives of General Psychiatry, 42, 689-693. Dinges, G. N., & Cherry, D. (1995). Symptom expression and the use of mental health services Lewinsohn, P., Rohde, P., Seeley, J., & Fi- among American ethnic minorities. In J. F. Aponte, scher, S. (1993). Age-cohort changes in the life- R. Y. Rivers, & J. Wohl (Eds.), Psychological in- times occurrences of depression and other mental terventions and cultural diversity (pp. 1-18). Bos- disorders. Journal of Abnormal Psychology, 102, ton, MA: Allyn and Bacon. 110-120.

First Quarter 2002 45 Mannell, R. C, & Kleiber, D. A. (1997). A Peterson, C, Maier, S., & Seligman, M. E. P. social psychology of leisure. State College, PA: (1993). Learned helplessness: A theory for the age Venture. of personal control. New York: Oxford University Press. Matsumoto, D. (1996). Culture and psychology. Pacific Grove, CA: Brooks/Cole. Peterson, C, & Seligman, M. E. P. (1984). Causal explanations as a risk factor for depression: Maughan, M, & Ellis, G. (1991). Effects of Theory and evidence. Psychological Review, 91, efficacy information during recreation participation 347-374. on efficacy judgments of depressed adolescents. Therapeutic Recreation Journal, 25, 50-59. Raps, C. S., Peterson, C, Reinhard, K. E., Abramson, L., & Seligman, M. E. P. (1982). Attri- May, P. A. (1987). Suicide and self destruction bution style among depressed patients. Journal of among American Indian youths. Journal of the Na- Abnormal Psychology, 91, 102-108. tional Center for American Indians and Alaskan/ Native Mental Health Research, 1(1), 244-289. Robins, L., Helzer, J., Weissman, M., Orvaschel, H., Gruenberg, E., Burke, J., & Regier, D. (1984). McAuley, E., Duncan, T. E., & Russell, D. W. Lifetime prevalence of specific psychiatric disorders (1992). Measuring causal attributions: The revised in three sites. Archives of General Psychiatry, 41, causal dimension scale (CDSII). Personality and 949-958. Social Psychology Bulletin, 18, 566-573. Seligman, M. E. P. (1975). Helplessness of de- Menon, T., Morris, M. W., Chiu, C. Y., & Hong, pression, development and death. San Francisco, Y. Y. (1999). Culture and the construal of agency: CA: W. H. Freeman. Attributions to individual versus group dispositions. Journal of Personality and Social Psychology, 76, Seligman, M. E. P. (1990). Learned optimism. 701-717. New York: Pocket Books.

Miller, R., Brickman, P., & Bolten, D. (1975). Seligman, M. E. P. (1993). What you can Attribution versus persuasion as a means for modi- change . . . and what you can't. New York: Fawcett fying behavior. Journal of Personality and Social Columbine. Psychology, 31, 430-441. Seligman, M. E. P., Reivich, K., Jaycox, L., & Morris, C, & Ellis, G. (1993). Attributional Gillham, J. (1995). The optimistic child: A revolu- basis of perceived freedom of leisure. Therapeutic tionary program that safeguards children against Recreation Journal, 27, 172-185. depression and builds lifelong resilience. Boston, MA: Houghton Mifflin Company. Nolen-Hoeksema, S., Girgus, S., & Seligman, M. E. P. (1992). Predictors and consequences of Sweeney, P., Anderson, K., & Bailey, S. (1986). childhood depressive symptoms: A five year longi- Attribution style in depression: A meta-analysis re- tudinal study. Journal of Abnormal Psychology, view. Journal of Personality and Social Psychology, 101, 405-422. 50, 974-991.

Oettingen, G. (1995). Explanatory style in the Tennen, H., & Herzberger, S. (1985). Attribu- context of culture. In G. M. Buchanan & M. E. P. tional style questionnaire. Test Critiques, 4, 20—30. Seligman (Eds.), Explanatory styles (pp. 209-224). Hillsdale, NJ: Lawrence Erlbaum. Voight, A. (1983). Attribution theory in thera- peutic recreation. In L. L. Neal & C. R. Edginton Peterson, C. (1982). Learned helplessness and (Eds.), Extrae perspectives: Concepts in therapeutic attribution intervention in depression. In C. Antaki recreation (pp. 133-144). Eugene, OR: University & C. Brewin (Eds.), Attributions and psychological of Oregon Press. change: Application of attribution theory to clinical and educational practice (pp. 97—115). London: Weiner, B. (1985). An attribution theory of Academic Press. achievement motivation and emotion. Psychological Review, 92, 548-573. Peterson, C, Buchanan, G. M., & Seligman, M. E. P. (1995). Explanatory style: History and Weiner, B., Frieze, I. H., Kukla, A., Reed, L., evolution of the field. In G. M. Buchanan & M. E. P. Rest, S., & Rosenbaum, R. M. (1971). Perceiving Seligman (Eds.), Explanatory styles (pp. 1-20). Hills- the cause of success and failure. Morristown, NJ: dale, NJ: Lawrence Erlbaum. General Learning Press.

46 Therapeutic Recreation Journal Wenegrat, B. (1995). Illness and power: Worn- Witman, J., & Batchelder, K. (1992). Implica- en's mental disorders and the battle between the tions of brief treatment in therapeutic recreation. In sexes. New York: New York University Press. B. H. Hitzhusen & L. T. Jackson (Eds.), Expanding Horizons in Therapeutic Recreation, Volume 14. Widmer, M. (1996). Book reviews: Learned op- (pp. 273-287). Columbia, MI: University of Mis- timism. Therapeutic Recreation Journal, 30, 79 - 83. souri.

Wildt, A. R., & Ahtola, O. T. (1978). Analysis of Witt, P. A., & Ellis, G. D. (1989). The leisure covariance: Series 12, quantitative applications in the diagnostic battery users manual. State College, PA: social sciences. Beverly Hills, CA: Sage Publication. Venture.

First Quarter 2002 47