|
|
|
|
Paper presented at meeting of the Society for Psychotherapy Research, Amelia Island, Florida, June 1996. Measuring Session Effects in Process-Experiential and Cognitive-Behavioral Therapies of PTSD: The Revised Session Reaction ScaleJulie Reeker, Robert Elliott, and David Ensing Address for Correspondence:
We acknowledge the contributions of David Rennie, Sharon Young, Nicole Taylor, Laili Radpour-Markert, Patsy Suter, Chris McCullen, Carla Gibson, Nancy Solis, Lynn Simmer-Dvonch, Karen Yurko, Carol Layman, Janie Manford, Michele Knox, and Ken Davis. Address correspondence to Robert Elliott, Department of Psychology, University of Toledo, Toledo, OH 43606.
Table 1: Means, Standard Deviations, and Factor Analysis of the Revised Session Reaction Scale Table 4: Predictors of Revised Session Reaction Scale Indexes ABSTRACTThe Revised Session Reaction Scale (RSRS) is a brief client-report measure of the experienced effects of therapy sessions. Data were collected from clients seen in short-term process-experiential and cognitive-behavioral treatment for crime-related PTSD. Factor analyses were consistent with the expected hierarchical structure of the measure (Helpful Reactions divided into Task Reactions and Relationship Reactions, and Hindering Reactions). Internal reliabilities for all of the subscales were excellent. Helpful reaction scales were most strongly correlated with the client initial experiences, client global helpfulness ratings, and the Session Evaluation Questionnaire Depth and Positivity scales. Clients gave more positive ratings to cognitive-behavioral sessions than to process-experiential sessions, the reasons for which are discussed. Measuring Session Effects in Process-Experiential and Cognitive-Behavioral Therapies of PTSD: The Revised Session Reaction Scale Session-level measures that assess therapeutic reactions from the client's perspective are relevant to practicing therapists and counselors because they are more precise than overall client satisfaction measures and vastly more efficient than videotape-assisted recall methods. The session reactions measure described in this paper was derived from earlier cluster- and content-analytic studies (Elliott, 1985; Elliott, James, Reimschuessel, Cislo, & Sack, 1985) of clients' open-ended descriptions of significant therapy events. In these studies, therapeutic reactions were found to fall into two broad groups: helpful and hindering. In addition, two kinds of helpful reactions were found: (a) Task reactions, involve clients' experienced progress on their presenting problems; These reactions typically consist of clients' reports of either changes in cognitive-emotional schemas (e.g. increased awareness of self) or changes in behavioral-action processes (problem clarification or solution) (b) Relationship reactions, involve clients’ reports of some form of beneficial interpersonal contact (e.g. closer to the therapist) with the therapist, often accompanied by positive emotions (e.g. feeling supported). In contrast to these two types of helpful reactions, hindering reactions involve the clients' negative experiences, such as feeling misunderstood or impatience with the lack of progress. On the basis of this earlier research, Elliott & Wexler (1994), devised a session-level questionnaire to provide a quantitative measure of therapeutic impacts. In this study, the authors confirmed the tripartite nested structure of the measure, obtained good reliabilities for the scales (except for hindering reactions), and reported good convergent validity with similar measures such as Stiles' (1980) Session Evaluation Questionnaire (SEQ). In addition, Elliott, Clark and Kemeny (1986; see also Elliott et al., 1990), carried out a content analysis of clients’ postsession descriptions of significant therapy events, developing a more differentiated account of the task domain of client postsession effects. In particular, they distinguished between client reports of four different kinds of schema change: insight (cognitive connection among schematic elements, e.g., seeing why an event caused an emotional reaction), awareness (enhanced salience of particular schematic elements, e.g., becoming aware that one is angry), positive reevaluation (e.g., coming to see the self in a more positive - or less negative - light), and meta-perception (looking at self or other from a different perspective, e.g., as someone else might see it). Furthermore, for most of these schema changes, they distinguished between changes in self or other schemas. In light of these developments, it became clear that it would be useful to undertake a revision of the Session Impacts Scale. First, we added hindering items to improve the reliability of the hindering effects scale. Second, we added items tapping the broader range of task effects which had been distinguished. A final issue was the need to change the linguistic frame of the construct of session effects, from “session impacts” (implying the client as a passive object of therapeutic activity) to “session reactions” (implying the client as an active agent interacting with the therapist). In this article, we report psychometric data on the Revised Session Reaction Scale (RSRS) using data from a study of short-term process-experiential and cognitive-behavioral treatment of crime-related PTSD. The main research questions that are addressed are as follows: First, what is the psychometric structure of this instrument, including the underlying dimensions and internal reliability? Is this structure consistent with the model found in previous research? Second, what pattern of session reactions are typically present in process-experiential and cognitive-behavioral therapy? Are these consistent with the treatment models? Third, what is the evidence for the convergent and predictive validity of the RSRS? That is, how well does it correspond to ratings on conceptually similar measures, such as the Depth, Smoothness, and postsession positivity scales of the SEQ? How well does it predict week-to-week change in the client's problem ratings? Fourth, how much of the variance in RSRS ratings is accounted for by possible predictor factors, such as therapist gender, treatment condition, and session number? In other words, which of these factors make a difference for RSRS ratings? MethodParticipants and TreatmentClients. Clients experiencing crime-related PTSD were primarily recruited through advertisements in local newspapers and on a local time and weather channel. After an initial telephone interview to determine appropriateness and willingness to participate, prospective clients were given a screening battery, which included the Clinician-Administered PTSD Scale (CAPS; Blake et al., 1990), Millon Clinical Multiaxial Inventory (MCMI), Structured Clinical Interview for DSM-III-R Disorders (SCID; Spitzer, Williams, Gibbon, & First, 1990), Impact of Events Scale (IES; Horowitz, Wilner, & Alvarez, 1979, Symptom Checklist-90-Revised (SCL-90-R; Derogatis, 1983, PTSD Scale from the MMPI (MMPI-PTSD; Keane, Malory & Fairbank, 1984, Inventory of Interpersonal Problems (IIP; Horowitz, Rosenberg, Baer, Ureno, & Villasenor, 1988, and Toronto Alexithymia Scale (TAS; Taylor, Ryan, & Bagby, 1985). All of the clients included in the study fit the criteria for PTSD as a primary diagnosis. Clients were excluded for a variety of reasons (previous psychiatric hospitalization, personality disorders, recent substance abuse, or recent therapy). The sample reported here consisted of 16 clients, with data available from 204 sessions. All of the clients included in the study were female except for one male. Therapists. 13 therapists were involved in the study: Two were licensed clinical psychologists, and the rest were graduate students in clinical psychology. All of the therapists were of European-American ethnic background. Most of the therapists saw only one client: 4 therapists saw 2 clients. Therapists underwent at least 3 months of training, involving lectures, reading, role playing, and training clients. Process-experiential treatment. Clients were typically seen for 16 sessions of process-experiential therapy. The process-experiential approach to therapy synthesizes client-centered and Gestalt approaches (Greenberg, Rice, & Elliott, 1993). The process-experiential approach organizes treatment according to a set of treatment principles, response intentions, and therapeutic tasks. The six treatment principals provide the overall guidelines for the treatment. The response intentions define what the therapist does, and predominately include empathic understanding responses, empathic exploration responses, and process directives. A key aspect of the Process-Experiential approach is its focus on a set of therapeutic tasks, signalled by client "markers" (signs that the client is ready to work on a particular problematic experience). A total of nine therapeutic tasks have been described, but with PTSD, the major tasks appear to be (a) Empathic Exploration/Re-experiencing of the victimization (helping the client to re-experience the traumatic event or trauma-related dreams); (b) Creation of Meaning (helping the client to symbolize the cherished belief which was shattered by the victimization experience; Clarke, 1993); (c) Empathic Prizing (providing the client the experience that the therapist cares and accepts them in their most vulnerable moments); (d) Two Chair Dialogue (in which the client enacts the internal conflict between aspects of self which want to re-experience vs. to avoid the trauma); and (e) Finishing Unfinished Business (in which the client imagines a perpetrator or unhelpful other in an empty chair and speaks to him or her in order to complete blocked emotional experience). Cognitive-behavioral treatment. Clients were typically seen for 16 sessions of cognitive-behavioral therapy. The cognitive-behavioral treatment used in this study combines Veronen and Kilpatrick's (1983) Stress Inoculation Therapy, Resick and Schnicke's (1993) Cognitive Processing Therapy as well as Foa's (1991) prolonged exposure treatment. The treatment was designed to specifically address symptoms of victimization-related PTSD. The treatment is divided into eight modules each which could, for the most part, be addressed in any order. Flexibility was built into the treatment in order to allow clients to address their most problematic issues first. MeasuresRSRS. As noted, the RSRS consists of three subscales derived from previous research on the reactions of significant therapy events (Elliott, 1985; Elliott, Clark, & Kemeny, 1988; Elliott, James et al., 1985). The Task Reactions subscale consists of 10 items, while the Relationship Reactions subscale consists of four items. These two subscales together are referred to as helpful reactions and were combined to create a 14 item Helpful Reactions scale. The third scale, Hindering Reactions, consists of eight items. Finally, an "other reaction" item was included as a check on scale completeness. All items are rated on 5-point adjective-anchored scales (1 = not at all, 2 = slightly, 3 = somewhat, 4 = pretty much, 5 = very much). For each item, clients were asked to rate the item on the basis of the descriptor that best fits their experience. SEQ. The SEQ consists of twenty-four 7-point bipolar semantic differential style items (Stiles, 1980). The most commonly used SEQ indexes are the Depth and Smoothness scales, derived by factor analysis, each which consists of five items that begin with the stem, "This session was." In this study, we also used two less commonly studied SEQ indexes, Positivity and Arousal. The former provides an additional measure of session reaction, and the latter provides one for discriminant validity (that is, RSRS scales were expected to correlate with Positivity but not with Arousal). These scales consist of five items each, which begin with the stem, "Right now I feel." Stiles (1980) and Stiles and Snow (1984) reported high internal reliabilities for all of the SEQ scales. Helpfulness ratings. The Helpfulness Scale has been used in several studies as a tape-assisted recall measure of the general effect of therapist reactions (e.g. Elliott, 1985; Hill, Helms, Tichenor et al., 1988). However, it can also be used to rate the most helpful event in the session and the session as a whole. These helpfulness ratings complement the specific reaction items of the RSRS. In this study, three one-item global helpfulness ratings were used for comparison purposes: client session helpfulness, therapist session helpfulness, and client event helpfulness (clients were asked to describe and rate the most helpful event in the session). The Helpfulness Scale is a 9-point bipolar (i.e. hindering vs. helpful), adjective-anchored (i.e. neutral, slightly, moderately, greatly, and extremely) scale. Initial Experiences Scale. This eight-item measure has been constructed for use in a larger study comparing two types of treatment for crime-related PTSD. This measure was created in order to control for the immediate therapeutic resources with which the client enters the session (differential client motivation, task focusedness, felt readiness, and attitude toward treatment model; see Stiles, 1988). Both client and therapist complete the measure. In the present sample, excellent internal reliabilities (alphas of .81 and .87) for the client and therapist versions of the measure were obtained. The client version correlated moderately (r = .41) with the therapist version and was substantially associated with client ratings of session Depth and Positivity. Impact of Events Scale (IES). This is a 15-item self-report measure assessing the severity of current reexperiencing and avoidance symptoms related to the victimization. Horowitz et al. (1979) reported good internal reliabilities for the IES (.79 for intrusion and .82 for avoidance subscales). Zilberg, Weiss & Horowitz (1982) reported good internal reliability, validity and sensitivity to change for the IES. Murphy et al. (1988) found that the IES was useful for screening and tracking treatment progress in childhood sexual victimization survivors. The IES was completed by the client before each session so that it was used as a weekly change measure. ProcedureClients completed the IES before each session and 1-week after the last session. Clients completed the RSRS, the session helpfulness and event helpfulness ratings, and the SEQ after each session, and therapists filled out the session helpfulness rating after each session. ResultsScale StructureWe subjected the first 22 items (excluding "other reactions") of the RSRS to exploratory factor analyses. Because the RSRS is intended for use as a session-by-session measure, the session was used as a unit of analysis for the factor analyses. (Exploratory factor analysis is a descriptive rather than an inferential statistical method; therefore, the nonindependence of session within cases is not critical.) In keeping with the original cluster-analytic research (Elliott, 1985) on which the measure was based, we predicted a hierarchical structure of two main factors and three subfactors. For this reason, we expected that both the two- factor and three-factor solutions would be valid and interpretable, with the two-factor solution consisting of helpful and hindering items and the three-factor solution made up of task, relationship, and hindering items. (Rather than carry out the confirmatory factor analyses, we used the more conservative strategy of attempting to rediscover the hypothesized factor structure through exploratory principal axis factor analysis). An unconstrained, unrotated principal-axis factor analysis revealed four factors with eigenvalues greater than 1, but when a varimax rotation was used, no items loaded on it substantially (>.30); therefore, we used the three factor solution instead. Before rotation, the largest factor was a general Helpful Reactions factor (eigenvalues = 7.35, 2.37, 1.58, respectively), while the other two factors were much smaller and overlapped almost entirely with the first factor, indicating the presence of a substantial general factor. The rotation of these factors (which together accounted for 51% of the variance) resulted in a structure nearly identical to a predicted three-component subfactor structure (see Table 1). The one exception was the Relieved item which was found to load on the Task Reactions factor rather than the Relationship Reactions factor as previously found (Elliott & Wexler, 1994). Next, the data were forced into a two-factor solution, which yielded the predicted higher-order clustering of task and relationship items into a single Helpful Reactions factor. We then calculated interitem reliabilities for the items that loaded above .35 and were the highest on a particular factor (represented in boldface in Table 1). Internal reliability (alpha) for all scales was excellent (Hindering Reactions, .83; Task Reactions, .91; Relationship Reactions, .89; higher-order combined Helpful Reactions, .92). These results indicate that the attempt to improve Hindering Reactions scale reliability was successful. Descriptive AnalysesMeans and standard deviations for the RSRS items and scales are presented in Tables 1 and 2. The Relationship Reactions scale and its constituent items received uniformly high ratings, clustering around scale point “4” (“pretty much”). Of the Task Reactions items, highest ratings were given for self-awareness, self-insight and problem definition. Task items involving insight or positive beliefs about others, as well as problem progress, received the lowest ratings on this scale. On the other hand, the Hindering Reactions scale was given the lowest ratings overall; only distressed and more distanced received mean ratings that were substantially above the bottom of the scale (closer to “slightly” than to “not at all.” All Hindering Reactions had lower mean ratings than any of the Helpful Reactions. Components of Variance AnalysisAn important aspect of understanding what the RSRS measures is learning what classes of variables appear to correlate with it. An initial approach to this question may be gained by an analysis of components of variance, such as has been done by Howard, Orlinsky, and Perilstein (1976) for the TSR and by Stiles and Snow (1984) for the SEQ. Thus, the variance due to case and session was analyzed for each of the RSRS scales by one-way analyses of variance. The analysis gave an overall estimate of variance due to client-therapist pairs (referred to in Table 2 as “case variance”). In addition, variance due to sessions was estimated by the residual variance not accounted for by overall case variance. The summarized results of this analysis are presented in Table 2. Case variance accounted for about half of the variance for Task Reactions and the combined Helpful reactions, 63% of the variance for Relationship reactions, but only for 40% of the variance for Hindering Reactions. Conversely, 37-51% of the variance in RSRS Task Reactions, Relationship reactions, and the combined Helpful Reactions appears to be due to fluctuations between sessions within cases. For Hindering Reactions, this figure is 60%, which suggested that negative reactions were predominately a session-level phenomenon. Convergence With Measures of Session ReactionAssociations with the SEQ. In comparing the RSRS with the SEQ, we expected strong associations between RSRS scales and the most conceptually similar SEQ scales, Depth, Smoothness, and postsession Positivity. We found that the RSRS Helpful Reactions was strongly correlated with client Smoothness and postsession Positivity, while it was moderately correlated with client Depth and Arousal (see Table 3). Hindering reactions were negatively correlated with client session Positivity, Smoothness, and Arousal, which indicated that sessions that were rated higher in hindering reactions were experienced as more emotionally negative and rough, and slightly less arousing. Associations with other measures of session reaction. Client helpful reactions were also strongly positively correlated with client general session helpfulness ratings. Helpfulness reactions were moderately correlated with client event helpfulness ratings and therapist session helpfulness ratings (see Table 3). Nevertheless, both client and therapist initial experience ratings were moderately positively correlated with client helpful reactions, suggesting therapists ratings of their clients’ readiness to engage in therapeutic work do predict client reaction ratings. In addition, there appeared to be no relationship between the RSRS indexes and the IES change score. Finally, there were significant negative correlations between client hindering reactions and client general session helpfulness ratings and client event helpfulness ratings which suggests that in sessions and events that clients rated as more helpful, there were less hindering reactions. Predictors of Session ReactionFinally, three simple descriptive variables were examined as possible correlates of session reactions. First, treatment condition, coded in the direction of cognitive-behavioral treatment, was positively correlated with Task, Relationship, and overall Helpful Reactions scales. On the other hand, cognitive-behavioral therapy received slightly lower ratings for Hindering Reactions, as indicated by the small negative correlation in Table 4. Second, therapist gender, coded in the direction of male therapists, was a positive predictor of Task Reactions (note, however, that therapist gender was confounded with therapist experience, since the two male therapists were the only post-Ph.D. therapists in the sample). Third, session number predicted Task and Helpful Reactions scores, indicating that the levels of these reactions increased over the course of therapy. DiscussionThe most interesting substantive finding in this study was the higher ratings for cognitive-behavioral versus process-experiential sessions on Helpful reactions, including Relationship reactions, together with the lower ratings on Hindering Reactions. We had expected Relationship Reactions to be higher for the process-experiential therapy and many of the Task Reactions to be higher for the cognitive-behavioral therapy sessions. Our results were not in the expected direction; in fact, they were significant in the opposite direction than what was expected. Several different explanations may account for these surprising findings. First, all of the graduate student therapists, including the cognitive-behavioral therapists, received basic training in process-experiential therapy, emphasizing the empathic elements of the treatment prior to their training in cognitive-behavioral training. Study of the session videotapes (Young, Ensing & Elliott, 1995) suggests that the student cognitive- behavioral therapists incorporated the process-experiential empathic skills as a strong component of their cognitive-behavioral work; thus they enacted cognitive-behavioral tasks within the context of a strong, supportive empathic relationship. In addition, it is important to note that cognitive-behavioral therapy may be easier for inexperienced therapists to learn because the tasks are more straightforward and do not involve constant adjustment from the therapist. Process-experiential therapy tasks appear to be harder to implement, because they must be continuously adapted to the client’s immediate state within each session. As a result, the process-experiential student therapists may have been more anxious and self-conscious during the therapy. The cognitive-behavioral student therapists on the other hand, may have been comfortable and consequently more empathic, yet task-oriented at the same time. Also, theoretical rationales were provided in great detail to cognitive-behavioral clients, and the therapeutic tasks were carefully articulated in relationship to the client’s presenting problems. In the process-experiential therapy, however, much less time and emphasis was placed on explaining the rationale and tasks of the therapy. As a result, it appears that the task and relationship aspects of the therapeutic alliance were stronger in the cognitive-behavioral cases. It is also the case that the process-experiential approach to PTSD at this point is, much less developed than the cognitive-behavioral approach. For example, the process-experiential therapists in this study spent very little time briefing clients about the nature of the treatment or providing rationales for particular therapeutic tasks, and gave very little information about how PTSD is understood from a process-experiential perspective. Perhaps the process-experiential approach is more difficult to explain to clients because it is still in the developing stages as a therapy model. As a result, clients may have felt that their concerns were not fully understood by the process-experiential therapists, and may have had trouble engaging in and benefiting from process-experiential tasks. Clearly, further research is needed on this finding. Currently, we are investigating the process-experiential treatment in the larger PTSD treatment study in order to determine what the sources of difficulty are, using a combination of quantitative and qualitative methods. The failure of RSRS ratings to predict weekly symptom ratings, reported by Elliott and Wexler (1994) was replicated in this study, and raises the question of what measures like the RSRS or the SEQ are measuring. We suspect that these scales tap therapeutic processes also assessed by measures of client satisfaction and therapeutic alliance. Furthermore, our clinical experience is that client week-to-week symptom change typically reflects external life events more than therapeutic processes. It is important to look at trends in weekly symptom measures over time to see their general relationship to treatment. In other words, we suspect that the session may be too small of a unit to trace relationships between client-experienced therapeutic reactions and symptom change. Methodologically, it appears that the changes made in revising the earlier measure (Elliott & Wexler, 1994) have resulted in a measure with substantially improved internal reliability, especially for the Hindering Reactions scale. Even though negative treatment processes are relatively rare, it is important to have a good measure of negative processes in therapy. In addition to further research on cognitive-behavioral and process-experiential therapy differences, further investigation of the factors which influence RSRS scores is needed. For example, we were unable to separate client and therapist variance with the present sample, as most therapists in this sample only saw one client. A larger sample of therapists seeing multiple clients will be needed for this type of analysis. Carry-over (week-to-week autocorrelations) and other nonindependence effects with the RSRS also need further study. Practically, the Hindering Reactions scale may be useful for identifying treatments (or sessions within treatments) in which problems are occurring. The Hindering Reactions scale may help to facilitate the improvement of certain treatments. On the other hand, the replicated finding of a lack of relationship between RSRS and week-to-week symptom change suggests that the RSRS cannot be used as a substitute for a measure of client symptomatic status or change. As a result, we would recommend that both therapeutic reactions and weekly symptom measures be used for gaging the unfolding effects of psychotherapy. However, the major clinical implication of this study has been to bring about a re-examination of the process-experiential therapy approach with PTSD. We are currently developing procedures for providing more information for clients, referred to as “experiential teaching”; this involves giving clients the process-experiential explanation of post-trauma difficulties, the theoretical rationale for the treatment and specific therapeutic tasks. This is done early in treatment and at later points, as the need arises (e.g., when introducing a new task or when the client questions a task). In addition, we are adapting therapeutic tasks, such as two chair work, specifically for trauma work. For example, traumatized clients often present conflicts in the form of “anxiety splits,” which are characterized by a client marker in the form of “X (e.g., driving on freeways) makes me anxious.” Process-experiential therapists are now being trained to recognize and work with this kind of conflict split. Developments such as these will result in a revised treatment manual for the use of process-experiential therapy for the treatment of crime-related PTSD.
ReferencesBlake, D., Weathers, F., Nagy, L., Kaloupek, D., Klauminzer, G., Charney, D., & Keane, T. (1990). Clinician-administered PTSD scales (CAPS). Boston, MA: National Center for Posttraumatic Stress Disorder. Clarke, K.M. (1993). Creation of meaning in incest survivors. Journal of Cognitive Psychotherapy, 7, 195-203. Derogatis, L.R. (1983). SCL-90-R administration, scoring and procedures manual - II. Towson, MD: Clinical Psychometric Research. Elliott, R. (1985). Helpful and nonhelpful events in brief counseling interviews: An empirical taxonomy. Journal of Counseling Psychology, 32, 307-322. Elliott, R., Clark, C., Wexler, M., Kemeny, V., Brinkerhoff, J., & Mack, C. (1990). The impact of experiential therapy of depression: Initial results. In G. Lietaer, J. Rombauts, & R. Van Balen (Eds.), Client-centered and experiential psychotherapy towards the nineties (pp.549-577). Leuven, Belgium: Leuven University Press. Elliott, R., James, E., Reimschuessel, C., Cislo, D., Sack, N. (1985). Significant events and the analysis of immediate therapeutic impacts. Psychotherapy, 22, 620-630. Elliott, R. & Wexler, M.M. (1994). Measuring the Impact of Treatment Session: The Session Impacts Scale. Journal of Counseling Psychology, 41, 166-174. Foa, E.B., Olasov-Rothbaum, B., Riggs, D.S., & Murdock, T.B. (1991). Treatment of posttraumatic stress disorder in rape victims: a comparison between cognitive-behavioral procedures and counseling. Journal of Consulting and Clinical Psychology, 59, 715-723. Greenberg, L.S., rice, L.N., & Elliott, R. (1993). Facilitating emotional change: The moment-by moment process. New York: Guilford Press. Hill, C.E., Helms, J.E., Spiegel, S.B., & Tichenor, V. (1988). Development of a system for categorizing client reactions to therapist interventions. Journal of Counseling Psychology, 35, 27-36. Horowitz, L.M., Rosenberg, S.E., Baer, B.A., Ureno, G., Villasenor, V.S. (1988). Inventory of interpersonal problems: psychometric properties and clinical applications. Journal of Consulting and Clinical Psychology, 56, 885-892. Horowitz, M.J., Wilner, N., & Alvarez, W. (1979). Impact of event scale: A measure of subjective stress. Psychometric Medicine, 41, 209-218. Keane, T.M., Mallory, P.F., Fairbank, J.A. (1984). Empirical development of an MMPI subscale for the assessment of combat-related posttraumatic stress disorder. Journal of Consulting and Clinical Psychology, 52, 888-891. Murphy, S.M., Kilpatrick, D.G., Amick-McMullan, A., Veronen, L.J., Paduhovich, J., Best, C.L., Villeponteaux, L.A., & Saunders, B.E. (1988). Current psychological functioning of child sexual assault survivors. Journal of Interpersonal Violence, 3, 55- 79. Resick, P.A., & Schnicke, M.K. (1993). Cognitive processing therapy for rape victims. Newberry Park: Sage. Spitzer, R.L., Williams, J.B.W., Gibbon, M., & First, M.B. (1990). User's Guide for the Structured Clinical Interview for DSM-III-R. Washington, D.C.: American Psychiatric Association. Stiles, W.B. (1980). Measurement of the impact of psychotherapy sessions. Journal of Consulting and Clinical Psychology, 48, 176-185. Stiles, W.B., & Snow, J.S. (1984). Counseling session impact as viewed by novice counselors and their clients. Journal of Consulting and Clinical Psychology, 31, 3-12. Vernonen, L.J., & Kilpatrick, D.G. (1983). Stress management for rape victims, In D. Meichenbaum & M. Jaremko (Eds.), Stress reduction and prevention (pp. 341- 374). New York: Plenum. Young, S.L., Ensing, D., & Elliott, R. (June,1995). The investigation of cognitive-behavioral therapy processes during client-identified peak moments. Paper presented at meeting of Society for Psychotherapy Research, Vancouver, BC. Zilberg, N., Weiss, D., & Horowitz, M. (1982). Impact of event scale: A cross-validation study and some empirical evidence supporting a conceptual model of stress response syndromes. Journal of Consulting and Clinical Psychology, 50, 407-414.
Table 1 Factor Item
M SD
1 2 3
Communality Positive beliefs (self)
3.10 1.31 .75 -.23
.10 .63 Close to
therapist 3.94
1.16 .20 -.06 .87
.80 Stuck/Lack of progress
1.44 .98 -.13 .76 -.20
.63 Note. n = 204 sessions for M and SD; n = 196 for factor analysis. The three-factor solution accounted for 51% of the total variance. A two-factor solution, in which Factors 1 and 2 were combined, accounted for 43% of the variance. Numerals in boldface indicate substantial factor loadings and item membership on each of the three scales. The other reactions item was not included in the factor analysis. Factor 1 = Task Reactions; Factor 2 = Hindering Reactions; Factor 3 = Relationship Reactions.
Table 2 Scale
M
SD
Case Session Task
3.03
.91
49 51
Table 3 Measure
Task Relationship
Helpful Hindering df Client SEQ Helpfulness ratings Initial experiences
Table 4 Predictor
Task Relationship Helpful Hindering
df Treatment condition .24*
.21*
.26*
-.18* 204
|
|
Send mail to experiential-researchers@focusing.org
with questions or comments about this web site.
|