Predictors - Hill et al (2001) ASD Journal Dreaming 11(2)

Predictors of outcome of dream interpretation sessions:

Volunteer client characteristics, dream characteristics, and type of interpretation

Clara E. Hill, Frances A. Kelley, Timothy L. Davis, Rachel E. Crook, Leslie E. Maldonado, Maria A. Turkson, Teresa L. Wonnell, V. Suthakaran, Jason S. Zack, Aaron B. Rochlen, Misty R. Kolchakian, and Jamila N. Codrington

Abstract

105 volunteer clients completed single sessions of dream interpretation using the Hill (1996) model, with half randomly assigned to waking life interpretation and the other half to parts of self interpretation in the insight stage of the Hill model. No differences were found between waking life and parts of self interpretations, suggesting that therapists can use either type of dream interpretation. Volunteer clients who had positive attitudes toward dreams and presented pleasant dreams had better session outcome; in addition, volunteer clients who had pleasant dreams gained more insight into their dreams. Results suggest that therapists doing single sessions of dream interpretation need to be cautious about working with dreams when volunteer clients have negative attitudes toward dreams and present unpleasant dreams.

Key Words: dream interpretation, attitudes toward dreams, dream valence, waking life, parts of self

Predictors of outcome of dream interpretation sessions:

Volunteer client characteristics, dream characteristics, and type of interpretation

Although many theories of dream interpretation have been developed, minimal empirical research has been conducted on dream interpretation until recently. Several studies have now investigated the Hill (1996) cognitive-experiential dream interpretation model in single sessions as well as within brief therapy. One of the most impressive and consistent outcomes is that volunteer clients have reported higher session quality and gained more insight from dream interpretation sessions than from regular therapy sessions (Cogar & Hill, 1992; Diemer, Lobell, Vivino, & Hill, 1996; Heaton, Hill, Hess, Hoffman, & Leotta, 1998; Heaton, Hill, Petersen, Rochlen, & Zack, 1998; Hill, Diemer, & Heaton, 1997; Hill, Diemer, Hess, Hillyer, & Seeman, 1993; Hill et al., 2000; Rochlen, Ligiero, Hill, & Heaton, 1999; Wonnell & Hill, 2000; Zack & Hill, 1998). For example, in Hill et al.’s (2000) study where clients with recent losses and troubling dreams received brief therapy (8-10 sessions) involving at least five sessions of dream interpretation or at least five sessions of a focus on their loss, clients in the dream condition rated the session process (depth, working alliance, exploration-insight gains, and action gains) significantly higher throughout the therapy than did clients in the loss condition.

Given that dream interpretation seems to be effective in terms of positive session quality and gains in insight, a next step is to identify predictors of session outcome. Determining the predictors of outcome in dream interpretation sessions is important so that therapists can make informed decisions about with whom and how to do dream interpretation. In this study, we were interested in investigating client characteristics (need for cognition and attitudes toward dreams) and dream characteristics (valence, distortion, vividness, and recency) in relation to session outcome. In addition, we were interested in investigating the effects of waking life versus parts of self interpretations on session outcome.

Client Characteristics

In terms of client characteristics, we were first interested in need for cognition. Therapists have reported that it is easier to conduct dream interpretation sessions with clients who enjoy thinking about their dreams and are psychologically minded (e.g., Wonnell & Hill, 2000), but previous attempts to relate the client-rated outcome of dream interpretation to client self-report measures of psychological mindedness have not yielded significant results (Cogar & Hill, 1992; Diemer et al., 1996; Hill et al., 1997). We wondered if perhaps a measure of need for cognition would be more likely to tap the relevant client characteristic than measures of psychological mindedness. Hence, for the present study, we used the Need for Cognition Scale (Cacioppo, Petty, & Kao, 1984), which assesses one’s tendency to engage in and enjoy thinking.

A second client variable that might influence the session outcome of dream interpretation is attitudes toward dreams. Using a sample of undergraduate students who had completed a number of psychological measures for course credit, Hill et al. (1997) found that students who volunteered for a dream interpretation session (not for course credit) had more positive attitudes toward dreams than those who did not volunteer. However, among those volunteers who actually participated in dream interpretation sessions, attitudes toward dreams did not correlate with session outcome, perhaps because of limited variability of attitudes in the sample. Moreover, Zack and Hill (1998) found that volunteer clients who had moderate attitudes toward dreams had better session outcomes than those with very positive attitudes, who in turn had better session outcomes than those with very negative attitudes. Because of the somewhat conflicting findings between the previous studies and because both of these studies used small samples, we wanted to study the relationship between attitudes and session outcome again with a larger sample.

Dream Characteristics

We expected that characteristics of dreams would influence the outcome of dream interpretation sessions. The first dream characteristic of interest was valence, or the pleasantness of the dream. Zack and Hill (1998) found better session outcome for moderately unpleasant and extremely pleasant dreams than for neutral, moderately pleasant, or extremely unpleasant dreams. In this study, we wanted to see if we could replicate Zack and Hill’s findings, especially since they had a small sample size.

We were also interested in distortion, or how discrepant the dream is from what one might commonly experience in waking life. We speculated that dreams that were less distorted, or closer to waking life, might be easier for clients who have a low need for cognition to understand because they could be clearly linked to waking life. If a client is worried about an exam and dreams that she is taking an exam, it may be easy for the client to understand the meaning of the dream. On the other hand, we speculated that clients who have a high need for cognition would profit more from working with distorted dreams because they would find them to be puzzling and intriguing. A dream about cows dancing on a suburban front porch would be more discrepant with common experience and might make clients with a high need for cognition want to understand it.

Finally, vividness and recency of dreams made sense as predictors of the outcome of dream interpretation sessions. In our clinical experience, clients have an easier time working with vivid dreams because it is easier to remember the details and reexperience the emotions. Similarly, it is often easier to remember the details of recent dreams. The only study (Wonnell & Hill, 2000) that included recency of dreams as a predictor of session outcome found that it was easier for clients to develop plans for waking life changes with more recent dreams.

Type of Interpretation

Hill (1996) proposed that during the insight stage of her three-stage model of dream interpretation (exploration, insight, action), therapists and clients can work collaboratively to construct an interpretation of dreams from one of several perspectives. In a previous study (Hill et al., 1997), clients most often reported gaining waking life and parts of self insights from dream interpretation, so we focused on these two interpretation types for this study. A focus on waking life assumes that dreams reflect current waking life issues and is supported by research indicating that dreams are related to waking life (see review in Hill, 1996). For example, if a client dreams of being attacked, the therapist might focus on whether the client felt attacked in waking life. A focus on parts of self is based on Jungian theory (Jung, 1974) and Gestalt theory (Perls, 1970), which suggest that people incorporate representations of significant others into themselves but that these parts of self are often not integrated into a cohesive sense of self. Hence, the dreamer projects onto others what are actually parts of his or her self. For example, if a client dreams of a lazy person being berated by a tyrannical boss, the assumption is that both the lazy person and the tyrannical boss are parts of the client. Hence, we wanted to examine the effects of the type of interpretation on the outcome of dream interpretation sessions.

We thought that type of interpretation might interact with need for cognition. Our clinical experience in conducting dream interpretation is that clients who have low need for cognition prefer waking life to parts of self interpretations because they can readily understand how dreams reflect waking life but have a hard time understanding how dream images reflect parts of themselves. On the other hand, clients who have a high need for cognition seem to like and profit from both waking life and parts of self interpretations.

Purposes of the Present Study

The first purpose of this study was to examine the relationship of volunteer client characteristics (need for cognition, attitudes toward dreams) and dream characteristics (valence, distortion, vividness, recency) with the outcome of dream interpretation sessions. We hypothesized that need for cognition, attitudes toward dreams, dream valence, dream vividness, and dream recency would be positively related to session outcome. Furthermore, we hypothesized an interaction between distortion and need for cognition, such that the best session outcomes would be obtained when volunteer clients with a low need for cognition presented undistorted dreams and when clients with a high need for cognition presented distorted dreams.

The second purpose of this study was to examine the effects of type of interpretation (waking life or parts of self) on session outcome. We hypothesized an interaction between type of interpretation and need for cognition, such that waking life interpretations would result in better session outcome than parts of self interpretations for clients low in need for cognition but that there would be no differences between conditions for clients high in need for cognition.

We assessed session outcome in terms of volunteer client perceptions of depth (quality) of the session, overall insight gained in sessions, and changes in insight into the presented dream from pre to post-session, all of which have been used in previous studies of dream interpretation. We used single sessions of dream interpretation because possible confounding variables (e.g., stage of therapy, length and type of treatment) could be controlled. In addition, we did not want to use volunteers for dream interpretation because Hill et al. (1997) found that volunteers for dream interpretations sessions had higher attitudes toward dreams than nonvolunteers. Hence, we used introductory psychology students who participated for research credit because we thought that they would have a larger range of attitudes toward dreams and type of dreams than volunteers.

Method

Therapists

Twelve therapists (7 female, 5 male; 9 Caucasian, 1 African American, 1 Hispanic American, 1 Asian American; 11 doctoral students in counseling psychology, 1 counseling psychology professor), ranging in age from 22 to 50 (M = 31.25, SD = 8.14), participated in this study. The graduate students were all enrolled in a dream interpretation seminar with the professor, and all were authors of the study. All had received previous training in the Hill model of dream interpretation and had completed at least two semesters of practica in a counseling psychology doctoral program. Experience since beginning training as a therapist averaged 5.42 (SD = 7.66) years, and number of individual dream interpretation sessions conducted previously averaged 18.92 (SD = 26.67). Using three 5-point scales (1 = low, 5 = high), therapists indicated that they believed in and adhered to the techniques of humanistic/experiential therapy 3.92 (SD = .67), psychoanalytic/psychodynamic 3.50 (SD = 1.24), behavioral/cognitive-behavioral 2.92 (SD = 1.08), and the Hill model of dream interpretation 4.75 (SD = .45).

Volunteer clients

105 undergraduate students (62 female, 43 male; 57 White American, 20 African American, 15 Asian American, 6 Hispanic American, 7 other), ranging in age from 18 to 26 (M = 19.27, SD = 1.59), served as volunteer clients (52 in the waking life condition and 53 in the parts of self condition). Participants were not aware of the hypotheses of the study and received course credit for their participation.

Judges

Seven judges (5 undergraduates and 2 graduate students; 6 Caucasian, 1 Asian-American, ranging in age from 21 to 26) rated insight in dream interpretation. Three other judges (2 undergraduates and 1 graduate student; 1 female, 2 male; 1 Caucasian, 1 African American, 1 Asian-American; ranging in age from 21 to 34) read the dreams and rated level of distortion in the dreams. Five other judges (4 undergraduate, 1 graduate student; all female; 2 Caucasian, 1 African-American, 1 Asian-American, 1 Latina; ranging in age from 21 to 40 yrs) listened to excerpts of the insight stage and judged which of the two conditions was being delivered. The four graduate student judges were team leaders of their respective rating teams and were also therapists and authors of the study.

Measures for Predictor Variables

Demographics. Therapists indicated their age, race, gender, experience as a therapist in general and with dream interpretation, and theoretical orientation. Clients indicated their age, race, and gender.

The Need for Cognition Scale (NCS; Cacioppo, Petty, & Kao, 1984) is a trait measure of an individual’s tendency to engage in and enjoy thinking. The NCS has 18 items using 5-point Likert scales (0 = extremely uncharacteristic; 4 = extremely characteristic). Sample items are, “The notion of thinking abstractly is appealing to me, “ and “I like to have the responsibility of handling a situation that requires a lot of thinking.” Items are summed for a total score. The reliability and validity are well established (Cacioppo & Petty, 1982). Item-total correlations ranged between .52 and .80 (Cacioppo & Petty, 1982) and the theta coefficient (a maximized alpha coefficient) was .90 (Cacioppo et al., 1984). College students scoring high on the NCS reported enjoying a complex mental task more than a simple task, In addition, university faculty scored higher on the NCS than assembly-line workers. Internal consistency (Cronbach’s alpha) for the present study was .90.

Attitudes Toward Dreams-Revised (ATD-R) is an 9-item self-report measure of how a person feels about dreams. Participants respond to all items on a 5-point Likert scale (1 = not at all, 5 = very). The ATD was originally constructed by Hill et al. (1997) by combining 7 true-false items from Cernovsky (1984), 3 items (two using 4-point Likert scales, 1 using a yes-no format) from Robbins and Tanck (1988), and 1 item using a 5-point Likert scale that the authors created. A factor analysis on the 11 items yielded a single factor accounting for 36% of the variance with all items loading above .40 and an internal consistency alpha of .79. In the Hill et al. (1997) study, ATD was related positively to the NEO Openness Scale (Costa & McCrae, 1992), estimated dream recall, and diary dream recall. The ATD also predicted whether participants volunteered for dream interpretation sessions, providing some evidence of validity.

For the present study, we revised the ATD so that all items used 5-point scales, reversed the wording of half of the items so that all items were not scored in the same direction, and added one new item (“I value my dreams”) that seemed important for the measure. We administered the original ATD and the ATD-R to a sample of 26 upper level undergraduates (not the participants from the main part of the present study) and readministered the ATD-R to them two weeks later. We found a high correlation, r(24) = .91, between the ATD and ATD-R. Preliminary examination indicated three items that did not correlate as highly as other items with the total score; consequently, these three items were dropped. Internal consistency (Cronbach’s alpha) for the remaining 9 items was .93. A 2-week test-retest reliability of .92 was found for the 9-item ATD-R with the same 26 students. We conducted a factor analysis on the 9-item measure for the 105 participants in the present study and found that a single factor (eigenvalue = 4.60) accounted for 51% of the variance, with all items loading .50 or higher on the factor. Internal consistency (Cronbach’s alpha) on the ATD-R for the 105 participants in the present study was .88. Scores were determined by summing the ratings (after reversing the negatively worded items) and dividing by the total number of items.

Valence of the dream was assessed through the Semantic Differential Measure of Emotional States (Mehrabian & Russell, 1974). This measure includes 18 pairs of bipolar adjectives on 9-point scales assessing valence (happy-unhappy, pleased-annoyed, satisfied-unsatisfied, contented-melancholic, hopeful-despairing, and relaxed-bored), arousal (stimulated-relaxed, excited-calm, frenzied-sluggish, jittery-dull, wide-awake-sleepy, and aroused-unaroused), and dominance (controlling-controlled, influential-influenced, in control-cared for, important-awed, dominant-submissive, and autonomous-guided). Half of the items in each factor have the order of the scoring reversed and all of the items are presented in a random order. Responses are averaged across the six items of each dimension. High scores indicate pleasant dreams, and low scores indicate unpleasant dreams. In a factor analysis of the 18 items, Mehrabian and Russell reported three factors, accounting for 62% of the variance. Factor loadings ranged from .79 to .92 for valence, .75 to .82 for arousal, and .46 to .79 for dominance. We dropped the dominance score as suggested by Lang, Greenwald, Bradley, and Hamm (1993) and Russell (1978) because the factor loadings were low and participants find these items hard to rate. We also dropped the arousal scores, because Zack and Hill (1998) found that arousal was not significantly related to outcome of dream interpretation sessions. The internal consistency (Cronbach’s alpha) for valence was .90 in both Zack and Hill (1998) and the present study.

Distortion of dreams refers to the degree to which people or objects in the dream appeared or behaved in ways that would seem uncommon in waking life. This scale was created for the present study because we did not find an appropriate scale in the literature. Distortion was rated by trained judges using a 9-point Likert scale (1 = not at all distorted, e.g., “I dreamed about the test I actually took yesterday,” 9 = very distorted, e.g., “Aliens were coming out of the space ship, and then a purple cow flew over the moon.”) Interrater reliability (Cronbach’s alpha) for the three judges in the present study was .72.

Vividness of the dream was rated by volunteer clients on a 9-point Likert scale (1 = not at all vivid, vague, unclear, indistinct; 9 = vivid, clear, distinct, strong, intense). This scale was created for the present study because we did not find an appropriate scale in the literature. Recency of the dream was also assessed by volunteer clients on a 5-point Likert scale (1 = last night, 2 = within the last week, 3 = within the last month, 4 = within the last year, and 5 = several years ago).

Client-Rated Session Outcome

Session outcome was assessed with three measures. The Depth Scale of the Session Evaluation Questionnaire-Form 4 (Stiles & Snow, 1984) is a widely-used, valid, 5-item, bipolar, adjective-anchored, self-report measure used by volunteer clients to evaluate the quality of counseling. High scores indicate high levels of perceived quality of the session. Stiles and Snow reported high internal consistency (alpha = .87). Internal consistency (alpha) for the present study was .89. The Mastery-Insight Scale of the Therapeutic Realizations Scale (Kolden, 1991) assesses the level of mastery and insight gained by clients in sessions and can be completed by clients or therapists. Kolden formed the Therapeutic Realizations Scale from the Therapy Session Report (Orlinsky & Howard, 1975), using face validity and internal consistency estimates. The 5 items of the Mastery-Insight Scale (e.g., "I feel I got more understanding of the reasons behind my thoughts and feelings") are rated by clients or therapists using a 3-point Likert scale (0 = no, 2 = a lot). Kolden (1991) reported internal consistency for the total scale of the Therapeutic Realizations Scale (including two other scales of Unburdening and Encouragement) of .86 to .87 and test-retest reliability estimates of .65 between sessions 1 and 3. Hill et al. (1993) found an internal consistency (alpha) of .79 for the client-rated Mastery-Insight Scale. Diemer et al. (1996) presented evidence of concurrent validity in that the Mastery-Insight Scale correlated .84 with the Session Impact Scale-Understanding Subscale (Elliott & Wexler, 1994). Because the 3-point scales allows for only limited variability, we used a 5-point scale (1 = no, 5 = a lot) in the present study. Internal consistency (alpha) for the present study was .81 for volunteer clients. Exploration-Insight Gains is a 7-item subscale, using a 9-point Likert scale (1 = strongly disagree, 9 = strongly agree), of the Gains for Dream Interpretation (Heaton, Hill, Petersen et al., 1998), which was developed from client responses to open-ended questions about the helpful aspects of dream interpretation sessions. Heaton et al. reported an internal consistency (Cronbach’s alpha) of .83 for the Exploration-Insight Gains. Zack and Hill (1998) reported that Exploration-Insight Gains was correlated .69 with Session Impact Scale-Understanding (Elliott & Wexler, 1994) and .53 with the Mastery-Insight Scale, suggesting evidence of concurrent validity. Internal consistency (Conbach’s alpha) of Exploration-Insight Gains for the present study was .88.

In the present study, the 3 measures of session outcome were highly intercorrelated (correlations ranged from .60 to .64) and the internal consistency of the 3 scores was .79. Hence, the 3 scores were standardized and averaged to represent session outcome in this study.

Insight into dream. Insight into the dreams that volunteer clients presented in the sessions was assessed using procedures developed by Falk and Hill (1993). Clients write an interpretation of their dreams in response to the following question: “What do you think this dream means to you? How would you interpret it?” Trained judges independently rate insight of the interpretation using a 9-point Likert scale (1 = no insight, 5 = moderate insight, 9 = high insight) using Hill et al.’s (1992) definition for insight (“Client expresses an understanding of something about himself or herself and can articulate patterns or reasons for behaviors, thoughts, or feelings. Insight usually involves an ‘aha’ experience, in which the client perceives self and world in a new way. The client takes appropriate responsibility rather than blaming others, using ‘shoulds’ imposed from the outside world, or rationalizing,” pp. 548-549). In previous studies, interrater reliability (Cronbach’s alpha) between judges ranged from .91 to .97 (Diemer et al., 1996; Falk & Hill, 1993; Hill, Nakayama, & Wonnell, 1998). Likewise, it was high for the two teams (each composed of 3 or 4 raters) in the present study (.86 and .90). Change in insight (post minus pre scores) was used as the measure of insight into the presented dream in the present study; change in insight was only moderately correlated with session outcome, r(103) = .24, p < .05, so it was considered to be an independent index of session outcome.

Measures of Adherence

Therapist self-rated adherence. The therapists rated the extent to which they competently and appropriately followed the structure of the Hill model using a measure developed by Heaton, Hill, Petersen et al. (1998). The items, rated on a 9-point Likert scale (1 = low, 9 = high), include “How completely did you do the exploration stage?”, “How completely did you do the insight stage?”, “How completely did you do the action stage?”, and “How competent did you feel doing the dream interpretation with the client?” In addition, we asked therapists to use the same scale to rate two additional items, “How much did you work to understand the dream according to what is going on currently in the client’s waking life?” “How much did you work to understand the dream according to how each of the parts of the dream might reflect a part of the client?” We expected no differences between the two conditions on the first four questions, but we expected that therapists in the waking life condition would score higher on the 1^st additional item and lower on the 2^nd additional item than therapists in the parts of self condition.

Judges’ Specification of Condition. To indicate which of the two insight conditions (current waking life or parts of self) therapists were implementing, judges used a 5-point Likert scale (1 = definitely waking life, 2 = maybe waking life, 3 = not sure, 4 = maybe parts of self, 5 = definitely parts of self) to rate the first 10-15 min of the insight stage. Interrater reliabilities (Cronbach’s alpha) for the two teams (each with two judges) were .87 and .99.

Procedures

Therapist recruitment/training. Therapists signed up for a seminar in dream interpretation being aware that one of the requirements was to participate as a therapist in a study. Students were not required to take this particular seminar, although they were required to take some seminar during their graduate program. All therapists signed consent forms indicating their willingness to have their data used in the research project. To increase the probability that therapists would be committed to their assigned condition and deliver the treatment effectively, they were allowed to select their preferred condition (waking life vs. parts of self). Therapists (6 in the waking life condition; 6 in the parts of self condition) conducted their chosen condition with all their assigned volunteer clients. All therapists had been trained previously in the Hill model and were trained further in the designated condition. All therapists conducted one practice session using the designated condition (waking life or parts of self). During the study, each therapist presented one case for group supervision and discussed other cases as needed.

Volunteer client recruitment. Volunteer clients were recruited from introductory psychology classes through sign-up sheets that advertised a study on personal problem-solving. They were asked to bring in a dream to use in problem solving. A total of 153 students signed up for the study, but 24 canceled their sessions ahead of time and 19 failed to show up for their appointments. Hence, 110 completed all assigned tasks of the study.

Pre-session testing. When volunteer clients first came in, they gave a written copy of their dream to a therapist. If the dream was too lengthy (i.e., more than a half page), the therapist asked the volunteer client to pick a segment to discuss in the session. Volunteer clients then read the selected segment of their dream to themselves and rated it on valence (using the Semantic Measure of Emotional States), vividness, and recency. Then volunteer clients wrote an initial interpretation of their dream. Finally, volunteer clients completed the Need for Cognition Scale and Attitudes toward Dreams-Revised (along with other measures not included in the present study) in a random order.

Dream interpretation sessions. Each therapist saw between 8-10 volunteer clients in single sessions that lasted from 35 to 145 min (M = 73.42, SD = 18.01). Sessions were audiotaped (with the client’s permission) through the insight stage so that judges could determine adherence to the assigned condition. Therapists informed volunteer clients that the sessions and tapes would be kept confidential. Also, therapists were not aware of volunteer client scores on any measures (e.g., need for cognition, attitudes toward dreams, ratings of dreams).

The therapist told the volunteer client that they would work collaboratively to figure out the meaning of the dream and briefly described the three-stage model. Volunteer clients were asked to tell their dream in the first person present tense and describe their feelings both during the dream and upon awakening. For each successive image, volunteer clients were asked to describe the image in great detail and then associate to the image by talking about whatever thoughts or images came to mind. Unlike the manner in which the exploration stage is typically conducted, therapists did not ask volunteer clients about triggers for the dream images from any current or past events in their lives because this might have favored the waking life condition. When volunteer clients spontaneously brought up triggers from waking life, therapists acknowledged the triggers but did not explore them. At the end of the exploration stage, which usually lasted about 30 to 45 min, the therapist summarized what the volunteer client had said about the images and stated clearly that they were moving on to the insight stage.

At the beginning of the insight stage, therapists gave a brief rationale for the assigned condition. For the waking life condition, therapists said something like, "What we would like to do now is try to understand your dream from the perspective of your current waking life. Research has shown that dreams are related to waking life events. So, what do you think the dream might mean in terms of your current waking life?" Therapists then worked with volunteer clients to understand how the dream fit with what was going on in the client's life. If volunteer clients tried to shift focus (e.g., to parts of self), therapists stated that although other types of interpretation might be helpful, they were going to focus only on waking life. For the parts of self condition, therapists said something like, "Let’s try to understand your dream from the perspective that each of the images of the dream represents a part of yourself. Theory suggests that we incorporate (take in) important people, ideas, events, or feelings and they become parts of ourselves and get enacted in the dream as different characters or objects. By understanding the different parts of ourselves, we can begin to integrate all of the parts and develop a more cohesive sense of self. So, what do you think the dream might mean if you think of the images as representing different parts of yourself?” If volunteer clients tried to shift focus (e.g., to waking life), therapists stated that although other types of interpretation might be helpful, they were going to focus only on parts of self. At the end of the insight stage, which lasted at least 15 min, therapists in both conditions asked volunteer clients to summarize what they had learned about themselves.

Therapists in both conditions then said that they were moving to the action stage. They asked volunteer clients how they would change the dream, translate fantasized changes in dreams to changes in their lives, continue working on the dream, or develop a ritual to honor the dream. To maintain the distinction between the two conditions, therapists focused on changes in waking life or parts of self respectively. At the end of the action stage, which lasted at least 10 min, therapists asked the volunteer clients to summarize what they had learned about the dream and what they were going to do about what they had learned.

Post-session measures. Immediately after the session, volunteer clients completed (in random order) the post-session outcome measures (Mastery-Insight Scale, Gains from Dream Interpretation, Depth) and wrote another interpretation of the dream. Concurrently, therapists completed the adherence measure. After the measures were completed, therapists provided volunteer clients with a written debriefing statement about the study, references for dream interpretation, and information about the campus counseling center. They also answered any questions that volunteer clients had about the study.

Insight Ratings. Seven judges were trained to rate insight into dreams using data from previous studies until they reached high interrater reliability (Cronbach’s alpha > .70). After training using dreams from previous studies, judges were given randomly-ordered copies of the dreams and the corresponding pre- and post-session interpretations (judges were not aware of condition or whether interpretations were pre- or post-session). Judges independently rated the amount of overall insight in the pre- and post-session interpretations considering them at the same time so that they could compare levels of insight. Three or four judges rated each dream. Scores were averaged across judges.

Distortion Ratings. Three different judges rated the dreams for distortion. They were trained using data from previous studies until they reached high interrater reliability (Cronbach’s alpha > .70). They then rated the dreams from the present study in random order.

Adherence Ratings. Therapists rewound the tapes to the point immediately after they described what would happen in the insight stage so that judges would not hear what condition was being conducted. Two judges (different people than those who judged insight and distortion) listened to the first 10 to 15 min of the insight stage for each session and rated therapist adherence to condition.

Determining whether sessions fit the criteria for the assigned conditions. We established a priori the criterion that therapists had to judge themselves as adhering at least 7 on a 9-point scale to their own condition and 4 or less on the other condition. Furthermore, we a priori set the criterion that judges had to rate therapists at least 4 on the 5-point scale, indicating that therapists adhered to the assigned condition. For two sessions in which the tapes were inaudible, we examined the therapists’ adherence ratings and found them to be very acceptable (8 or 9 on the assigned condition and 1 on the other condition). In addition, these two therapists were judged as adhering highly in all their other sessions, so we assumed that their adherence ratings for the inaudible sessions were accurate; hence, we retained these two sessions for the data analyses. Five sessions (1 in the waking life condition, 4 in the parts of self condition) were dropped for not adhering to the assigned condition, leaving a total of 105 sessions for the data analyses.

Results

Preliminary Analyses

Table 1 shows the means and standard deviations for all variables. Scores for Depth and Exporation-Insight were within one standard deviation of scores in other dream interpretation studies (e.g., Diemer et al., 1996; Heaton et al., 1998; Rochlen et al., 1999), indicating that therapists were comparable to therapists in previous dream interpretation studies (note we could not compare scores on the Master-Insight Scale because we changed the rating scale from a 3- to 5-point scale). Table 2 shows the intercorrelations (using Pearson Product Moment correlations) among variables used in the analyses.

Therapists in the waking life condition had higher ratings of their beliefs in the waking life focus (M = 4.83, SD = .41) than did therapists in the parts of self condition, M = 4.00, SD = .63, t(10) = 2.71, p < .05, although therapist belief in the parts of self focus did not differ between conditions. Furthermore, no differences were found (using t-tests) between conditions for other therapist variables (age, years of experience, number of dream interpretation sessions conducted previously, and self-ratings of theoretical orientation).

No differences between conditions were found for therapist-rated adherence to the stages (exploration, insight, and action) or for overall competence. For this study, therapist adherence ratings on the exploration stage (M = 8.12, SD = 0.88), insight stage (M = 7.81, SD = 0.96), and action stage (M = 7.01, SD = 1.24) were within one standard deviation of scores (with one exception) reported for therapists in the Hill et al. (1997) dream interpretation study (exploration M = 7.25, SD = 0.97; insight M = 6.47, SD = 1.19, and action M = 6.08, SD = 1.33) and the Rochlen et al. (1999) dream interpretation study (exploration M = 7.46, SD = 1.52; insight M = 7.11, SD = 1.56, and action M = 6.49, SD = 1.53). (The exception was that therapist adherence ratings for the insight stage for this study were a standard deviation higher than in Hill et al., 1997). Hence, therapists adhered to the three stages about the same as previous therapists.

As expected, therapists in the waking life condition rated themselves as adhering more to waking life interpretations than did therapists in the parts of self condition, t(96.15 for unequal variances) = -.29.36, p < .001. Likewise, therapists in the parts of self condition rated themselves as adhering more to parts of self interpretations than did therapists in the waking life condition, t(103) = 35.11, p < .001. In conjunction with the evidence presented earlier (i.e., that judges who listened to the tapes of the insight stage were able to accurately determine who did which condition), we can assert that therapists adhered to their appropriate conditions.

Therapists (using dummy codes) were entered into regression equations for session outcome and change in insight. No overall (or individual) therapist effects were found for the total sample, waking life condition, or parts of self condition for any of the regression analyses. Hence, even though the professor had more experience than the graduate student therapists, no differences were evident in the data. Furthermore, using t-tests, no effects were found for volunteer client gender or race on session outcome or change in insight. Hence, therapist effects, volunteer gender and race were not considered in subsequent analyses.

Because there was such a range in terms of length of session, we were concerned that session length might have been related to the session outcome measures. However, correlations revealed that session length was not related to session outcome or insight change.

Predictors of Session Outcome and Insight Change

To determine whether relationships between variables were linear, quadratic, or cubic, plots were drawn and curves estimations calculated for all the linear, quadratic, and cubic trends for all of the predictor variables paired with each session outcome variable. For attitudes toward dreams and valence in relation to session outcome and for valence in relation to insight change, the linear, quadratic, and cubic relationships were all significant, but the linear relationships accounted for almost all of the variance. Hence, only the linear relationships were considered in subsequent analyses. To test the simple relationships between the predictor variables and session outcome and changes in insight, we first examined the Pearson correlations. We also did regression analyses to examine the unique relationships among predictors and session outcome variables.

Session Outcome. Pearson correlations (see Table 2) revealed that Attitudes Towards Dreams-Revised and valence of dreams were both significantly related to the composite index session outcome (composed of scores on Depth, Mastery-Insight Scale, Exploration-Insight Gains), such that better session outcome was attained when volunteer clients had positive attitudes and presented pleasant dreams. A simultaneous multiple regression analysis with the predictor variables of condition (waking life or parts of self focus in the insight stage), need for cognition, attitudes toward dreams, valence of the dream, distortion of the dream, vividness of the dream, and recency of the dream and the criterion of session outcome was significant, F(9,97) = 4.58, p < .001, R² = .25, adjusted R² = .19. Attitudes toward dreams was a significant unique predictor, part correlation (partialing out variance from all other predictors and the criterion) = .27, t = 3.07, p < .01. In addition, valence of dreams was a significant unique predictor, part correlation = .29, t = 3.25, p < .01. Another hierarchical regression analysis was conducted including all the above variables in the first block and the interaction between need for cognition and distortion and the interaction between need for cognition and condition in the second block. Neither of the interaction effects was significant. Hence, volunteers clients who had the best session outcome had positive attitudes toward dreams and presented pleasant dreams.

Changes in insight. Pearson correlations (see Table 2) revealed that valence of dreams was significantly related to changes in insight, such that more change in insight occurred when the volunteer clients presented pleasant dreams. Using the same predictors as in the analysis for session outcome, a simultaneous multiple regression analysis was not significant. An additional regression including the interaction between need for cognition and distortion and the interaction between need for cognition and condition was not significant (nor were any of the individual predictors). Thus, volunteer clients who gained the most insight into their dreams presented pleasant dreams.

Discussion

Volunteer client attitudes toward dreams and the valence (pleasantness) of dreams were the best predictors of a composite index of session outcome for volunteer clients in single sessions of dream interpretation. In addition, valence of dreams was the best predictor of gains in insight into dreams (based on changes in ratings of written dream interpretations), although this finding must be considered as tentative given that no control group was used and changes could have been due to effects of interpreting the dream twice. Hence, volunteer clients gained most from dream interpretation when they had positive attitudes toward dreams and worked with pleasant dreams. In contrast, the type of interpretation (waking life vs. parts of self), need for cognition, and other characteristics of the presented dream (distortion from waking life, vividness, and recency) were not significantly associated with session outcome or insight change. We discuss each of the findings in the following sections.

Volunteer Client Attitudes Toward Dreams

Attitudes toward dreams was positively related to session outcome. Attitudes toward dreams probably reflects volunteer clients’ willingness to consider that dreams have something to do with what is going on in their lives or in their personality. If volunteer clients value dreams, they probably feel that they can learn a lot from working with them in sessions. On the other hand, if volunteer clients do not think that dreams are valuable, they may be unwilling to consider any interpretation. They may think that dream interpretation is akin to palm reading or astrology and hence might not pay much attention to the session.

The importance of attitudes toward dreams in predicting session outcome is consistent with the Hill et al. (1997) findings that students who were willing to participate in dream interpretation sessions had positive attitudes. The findings differ somewhat from those of Zack and Hill (1998), who found a quadratic but not a linear relationship between attitudes and session outcome. It should be noted that the quadratic and cubic relationships were also significant in the present study but were minuscule compared to the amount of variance accounted for by the linear effect. In both studies, volunteer clients with more positive attitudes toward dreams had better session outcomes than did students with more negative attitudes. Differences occurred in the middle range. We would suggest that the larger sample size in the present study allowed for more stable relationships to emerge, although these results need to be replicated.

The results for attitudes should be considered in light of the Rochlen et al. (1999) results. They found that volunteer clients with below average attitudes toward dreams and dream recall could be trained in the skills needed for dream interpretation (e.g., description, association, interpretation, action). Once trained, their outcome for dream interpretation sessions was equivalent to that of previous studies of samples with a normal range of scores on attitudes and dream recall. These results suggested that training may be effective for helping people with negative attitudes recognize the benefits of working with their dreams. In addition, therapists should not dismiss the idea of working with dreams with people who have negative attitudes; instead they could train clients in the skills needed to do dream interpretation (e.g., description, association) and in what to expect from the dream interpretation process.

Valence of the Presented Dream

Volunteer clients had better session outcome and gained more insight into their dreams when they presented pleasant rather than unpleasant dreams. For example, a volunteer client who had a positive session outcome brought in the following pleasant dream (before the dream, she wrote, “About four years ago I had this dream that I really did not know what to make of. It was when I was pregnant with my daughter X. I remember being upset with her father at the time. When I hung up on him, I was crying and very upset. I then called on Jesus to help me. At this time I remember my tears drying and me suddenly falling asleep”):

“I was walking in the most beautiful green meadow that I have ever seen. The skies were very blue and everything seemed to be at rest. Every little creature you can imagine was at peace. I remember suddenly that a white baby lamb came up to me. When I looked down to see the lamb, I was dressed in a long white robe. The lamb rubbed my leg and then we started walking together and playing. When I woke up all I could remember was that I felt loved and peaceful.”

In contrast, a volunteer client who had a negative session outcome brought in the following unpleasant dream,

“My dream started out with a black sky when I realized I am in the woods. I notice the two figures in all white outfits coming towards me. I started to run for my life. I started to run as far as I could and stumbled a lot but never got caught. I was being pursued relentless. I saw a black pond and decided to jump into it to avoid being caught. Regardless, the two white figures jumped into the pool in pursuit after me. I was panicking so much that I almost drowned. I got out of the pool and saw that the two figures were still running after me. I immediately woke up.”

It may have been easier for the first volunteer client to discuss her dream because it was soothing, affirming, hopeful, and provided a resolution to her painful life situation. In contrast, it may have been difficult for the second volunteer client to discuss her dream because it suggested threat and conflict, without much of a sense of hope or resolution. Therapists may need to work harder when volunteer clients present unpleasant dreams to help them figure out how to solve problems suggested by such dreams.

An important caveat in understanding these results about the pleasantness of dreams is that these were single sessions. It may be that volunteer clients felt vulnerable discussing unpleasant dreams in single sessions because they would not have an ongoing relationship with their therapists. Perhaps therapists in ongoing therapy should consider working first with pleasant dreams to help build the alliance, and then only venture into working with more unpleasant dreams once they have a solid alliance. Given that clients in ongoing therapy probably bring in more unpleasant than pleasant dreams, researchers should investigate the factors that allow long-term clients to feel safe and profit from discussing unpleasant dreams.

The results for valence are somewhat consistent with those of Zack and Hill (1998). In both studies, volunteer clients with more pleasant dreams had better session outcomes than did those with less pleasant attitudes. Differences occurred in the middle range. It should be noted that quadratic and cubic relationships were also significant in the present study (as was true in the Zack and Hill study) but they were minuscule compared to the amount of variance accounted for by the linear effect. As we suggested for the results for attitudes and session outcome, the larger sample size in the present study may have allowed for more stable relationships to emerge, although these results need to be replicated.

Need for Cognition

Contrary to our hypothesis, need for cognition did not predict client-rated session outcome either by itself or in interaction with the distortion level of the dream or the type of interpretation. These findings are similar to those of previous studies that found that various measures of psychological-mindedness did not predict session outcome (Cogar & Hill, 1992; Diemer et al., 1996; Hill et al., 1997). Hence, even though therapists may like to work with clients who have high need for cognition and who are psychologically-minded, we find no empirical evidence that such clients benefit more from dream interpretation.

We should note, however, that need for cognition was related to attitudes toward dreams. People who enjoy thinking in general also like to think about their dreams. However, it appears that it is the more specific attitudes toward dreams rather than the more general enjoyment of thinking that is related to session outcome. Need for cognition was also related to distortion of dreams, suggesting that people who are high in need for cognition have more unusual dreams, which makes sense because they are interested in their thought processes and hence might be more open to fantastical dreams.

Other Characteristics of the Presented Dream

We were surprised that distortion level, vividness, and recency of dreams did not predict how much volunteer clients benefitted from dream interpretation sessions. Perhaps distant dreams were just as or maybe were even more important to work with than more recent dreams because the volunteer clients persisted in remembering them. In addition, volunteer clients probably only brought in dreams that were vivid enough to make them salient (this is supported by the mean ratings of almost 7 on a 9-point scale, indicating that the ratings were at the high end of the scale). A limitation of the present study in terms of studying dream characteristics is that volunteer clients were allowed to bring in dreams of their choosing, which may have skewed the selection to more interesting dreams. Probably the only way to test whether variables such as distortion, recency, and vividness make a difference in terms of session outcome would be to require participants to bring in randomly-chosen dreams. At this point, we do not know why volunteer clients selected particular dreams to discuss in sessions.

Waking Life Vs. Parts of Self Interpretation

Session depth and insight were equally high for both waking life and parts of self conditions. These session outcomes were consistent with those of previous dream interpretation studies and higher than regular therapy (Cogar & Hill, 1992; Diemer et al., 1996; Heaton, Hill, Hess et al., 1998; Heaton, Hill, Petersen et al., 1998; Hill et al., 1993; Hill et al., 1997; Hill et al., 2000; Rochlen et al., 1999; Wonnell & Hill, 2000; Zack & Hill, 1998). Hence, both types of interpretation were as effective as when therapists used their clinical judgment about what type of interpretation to use.

Although no differences were found between conditions in terms of volunteer client-rated session outcome or insight gains, therapists reported subjective differences in implementing the conditions. Therapists in the waking life condition thought that it was relatively easy to do the sessions because volunteer clients could readily relate their dreams to waking life and hence needed minimal justification for participating. Therapists in the parts of self condition, however, reported that they had more difficulty explaining the rationale and convincing volunteer clients that parts of dreams reflected parts of themselves (which may have been reflected in the differential therapist non-adherence in that we had to discard four parts of self sessions and only one waking life session). With particular volunteer clients, therapists enjoyed the parts of self condition because they could be creative and playful, but they had more difficulty with some reluctant volunteer clients and with dreams that were obviously related to waking life. Obviously, therapists were able to provide convincing rationales because the session outcomes were equivalent for the two conditions, but the subjective difficulty in implementing the parts of self condition with some volunteer clients is an important consideration. Therapists may need to be trained in how to provide convincing rationales for interventions that lack immediate face validity.

Limitations

The results of this study must be considered in the context that most of the therapists were graduate students (although several had extensive experience with dream interpretation), clients were volunteers receiving course credit and may not have been motivated for dream interpretation, only single sessions were used, and session outcome was assessed only immediately post-session. Furthermore, results of this study might not generalize to dream interpretation conducted within ongoing therapy. This limitation is particularly important in light of the consideration that clients in ongoing therapy might be more willing to work with unpleasant dreams if they have a strong therapeutic alliance.

Implications

One implication of these findings is that therapists should be cautious about using dream interpretation with clients who do not value dreams and with unpleasant dreams, especially in single or initial sessions. We would not suggest that therapists avoid using dream interpretation in these situations, but we would suggest that therapists educate clients about the value of working with dreams, be supportive with clients who present unpleasant dreams, and spend more time in the action stage to help resolve problems reflected in unpleasant dreams.

On the basis of these results, therapists can feel justified using their clinical intuition to decide whether to use waking life or parts of self interpretations. Both seem to be effective, although therapists may need to give better rationales and explanations to help clients understand the parts of self focus. Further research should examine other types of interpretation, such as the experience itself or early childhood experiences (Hill, 1996), interpersonal dynamics (Kolchakian & Hill, 2000), or spirituality (Davis & Hill, in press).

Given that the amount of variance accounted for by attitudes toward dreams and valence of dreams was small, we suggest that researchers investigate other factors (e.g., the therapeutic alliance, therapist techniques, client involvement) that might influence the outcome of dream interpretation sessions. We need to know more about when, with whom, and how to use this intervention. Moreover, research is needed about the manner of interpretation. In this study, therapists worked collaboratively with clients to construct interpretations, as Hill (1996) suggested. One can easily imagine, however, that we would have obtained different results if therapists had provided the interpretations authoritatively or had encouraged clients to construct their own interpretations. In conclusion, we recommend that researchers continue looking for variables that influence the process and outcome of dream interpretation sessions.

References

Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of Personality and Social Psychology, 42, 116-131.

Cacioppo, J. T., Petty, R. E., & Kao, C. F. (1984). The efficient assessment of need for cognition. Journal of Personality Assessment, 48, 306-307.

Cernovsky, Z. Z. (1984). Dream recall and attitudes towards dreams. Perceptual and Motor Skills, 58, 911-914.

Cogar, M. M., & Hill, C. E. (1992). Examining the effects of brief individual dream interpretation. Dreaming, 2, 239-248.

Costa, P. T. Jr., & McCrae, R. R. (1992). Revised NEO-Personality Inventory (NEO PI-R) and NEO Five-Factor Inventory (NEO-FFI): Professional manual. Odessa FL: Psychological Assessment Resources.

Davis, T. L., & Hill, C. E. (In press). Including spirituality into the Hill cognitive-experiential model of dream interpretation. In L. VandeCreek & T. L. Jackson (Ed.), Innovations in Clinical Practice: A source book. (Vol. 19, pp. ). Sarasota, FL: Professional Resource Press.

Diemer, R. A., Lobell, L. K., Vivino, B. L, & Hill, C. E. (1996). Comparison of dream interpretation, event interpretation, and unstructured sessions in brief therapy. Journal of Counseling Psychology, 43, 99-112.

Elliott, R., & Wexler, M. M. (1994). Measuring the impact of sessions: in process-experiential therapy of depression: The Session Impacts Scale. Journal of Counseling Psychology, 41, 166-174.

Falk, D. R., & Hill, C. E. (1995). The process and outcome of dream interpretation groups for divorcing women. Dreaming, 5, 29-42.

Heaton, K. J., Hill, C. E., Hess, S., Hoffman, M. A., & Leotta, C. (1998). Assimilation in therapy involving interpretation of recurrent and nonrecurrent dreams. Psychotherapy, 35, 147-162.

Heaton, K. J., Hill, C. E., & Petersen, D., Rochlen, A. B., & Zack, J. (1998). A comparison of therapist-facilitated and self-guided dream interpretation sessions. Journal of Counseling Psychology, 115-121.

Hill, C. E. (1996). Working with dreams in psychotherapy. New York: Guilford Press.

Hill, C. E., Corbett, M. M., Kanitz, B., Rios, P., Lightsey, R., & Gomez, M. (1992). Client behavior in counseling and therapy sessions: Development of a pantheoretical measure. Journal of Counseling Psychology, 39, 539-549.

Hill, C. E., Diemer, R., & Heaton, K. J. (1997). Dream interpretation sessions: Who volunteers, who benefits, and what volunteer clients view as most and least helpful. Journal of Counseling Psychology, 44, 53-62.

Hill, C. E., Diemer, R., Hess, S., Hillyer, A., & Seeman, R. (1993). Are the effects of dream interpretation on session quality, insight, and emotions due to the dream itself, to projection, or to the interpretation process? Dreaming, 3, 269-280.

Hill, C. E., Nakayama, E. Y., & Wonnell, T. L. (1998). A comparison of description, association, and combined description/association in exploring dream images. Dreaming, 8, 1-13.

Hill, C. E., Zack, J., Wonnell, T., Hoffman, M. A., Rochlen, A., Goldberg, J., Nakayama, E., Heaton, K. J., Kelley, F., Eiche, K., Tomlinson, M., & Hess, S. (2000). Structured brief therapy with a focus on dreams or loss for clients with troubling dreams and recent loss. Journal of Counseling Psychology, 47, 90-101.

Jung, C. G. (1974). Dreams (R. F. C. Hull, Trans.). Princeton, NJ: Princeton Univ Press.

Kolchakian, M., & Hill, C. E. (2000). A cognitive-experiential model of dream interpretation for couples. In L. VandeCreek & T. L. Jackson (Ed.), Innovations in Clinical Practice, Vol. 18, 85-101.

Kolden, G. C. (1991). The generic model of psychotherapy: An empirical investigation of patterns of process and outcome relationships. Psychotherapy Research, 1, 62-73.

Lang, P. J., Greenwald, M. K., Bradley, M. M., & Hamm, A. O. (1993). Looking at pictures: Affective, facial, visceral, and behavioral reactions. Psychophysiology, 30, 261-273.

Mehrabian, A., & Russell, J. A. (1974). An approach to environmental psychology. Cambridge, MA: The MIT Press.

Orlinsky, D. E., & Howard, K. I. (1975). Varieties of psychotherapy experience. New York: Teachers College Press.

Perls, F. (1970). Dream seminars. In J. F. Shepherd and I. L. Shepherd (Ed.), Gestalt therapy now. New York: Harper Colophon.

Robbins, P. R., & Tanck, R. H. (1988). Interest in dreams and dream recall. Perceptual and Motor Skills, 66, 291-294.

Rochlen, A., Ligiero, D., Hill, C. E., & Heaton, K. Preparation for dreamwork: Training for dream recall and dream interpretation. (In press). Journal of Counseling Psychology,.

Russell, J. A. (1978). Evidence of convergent validity on the dimensions of affect. Journal of Personality and Social psychology, 36, 1152-1168.

Stiles, W. B., & Snow, J. S. (1984). Dimensions of psychotherapy impact as viewed by novice counselors and their clients. Journal of Counseling Psychology, 31, 3-12.

Wonnell, T., & Hill, C. E. (2000). Effects of including the action stage in dream interpretation. Journal of Counseling Psychology, 47, 372-379.

Zack, J., & Hill, C. E. (1998). Predicting dream interpretation outcome by attitudes, stress, and emotion. Dreaming, 8, 169-185.

Author Notes

Clara E. Hill, Rachel E. Crook, Maria A. Turkson, Teresa L. Wonnell, Jason S. Zack, Aaron B. Rochlen, Misty R. Kolchakian, and Jamila N. Codrington were all in the Department of Psychology, University of Maryland at the time of the study. Aaron Rochlen is now in the Department of Educational Psychology at the University of Texas. Jason Zack is with etherapy.com. Frances A. Kelley, Timothy L. Davis, Leslie E. Laldanado, and Suthakaran Veeresamy were all in the Counseling and Personnel Services Department, University of Maryland at the time of the study. The fourth through twelfth authors are listed in random order because they contributed equally to the study.

We wish to express our gratitude to Gregory Hancock for statistical consultation and to Emilie Y. Nakayama for reading a draft of this paper.

This paper was presented at the Association for the Study of Dreams, Washington DC, July, 2000 and at the American Psychological Association, Washington DC, August 2000.

Correspondence concerning this article should be addressed to Clara E. Hill, Department of Psychology, University of Maryland, College Park, MD 20742. Electronic mail may be sent to hill@psyc.umd.edu.

Table 1. Means and standard deviations of all Measures Completed by Clients

______________________________________________________________________________

M SD

Predictor Variables

Need for Cognition Scale 63.51 11.68

Attitudes Toward Dreams-Revised 3.66 0.80

Valence of Dream 4.54 1.80

Distortion of Dream 5.39 2.24

Vividness of Dream 6.76 1.93

Recency of Dream 2.47 1.00

Session Outcome Variables

Mastery Insight Scale 3.79 0.69

Exploration-Insight Gains 7.77 1.20

Depth Scale 6.07 0.81

Pre Insight into Dream 3.34 1.32

Post Insight into Dream 3.87 1.54

______________________________________________________________________________

Note. N = 105. Pre and post insight were judges’ ratings on insight in client written dream interpretation from before and after the session. High scores indicate high levels on all variables.

Table 2. Correlation Matrices of all Predictor and Session Outcome Variables Used in the Analyses

______________________________________________________________________________

Session

NCS ATD-R Valence Distortion Vivid Recency Outcome

Predictor Variables

ATD-R .23*

Valence -.15 .12

Distortion .22* .13 .02

Vividness .04 .15 .16 .12

Recency .03 -.10 -.06 .07 .22*

Outcome Variables

Session Outcome .15 .37** .33** .11 .18 -.07

Insight Change -.08 .05 .25** .07 .09 .09 .24*

______________________________________________________________________________

Note. N = 105. Session Outcome = a composite of the Depth Scale, the Mastery-Insight Scale, and the Exploration-Insight Gains Scale. Insight Change = post - pre scores on judge -rated insight in written interpretations of dreams, NCS = Need for Cognition Scale, ATD-R = Attitudes Toward Dreams, Valence = pleasantness of the dream as measured by the Semantic Differential Measure of Emotional States, Distortion = a judge-rated measure of how close the dream was to waking life, Vividness = client-rated index of vividness of the dream, recency = client-rated index of how recent the dream was. High scores indicate high levels on all variables.

* = .05, ** = .01, *** = .001

September 22, 2000

Degrees of Authors: The first author has a Ph.D., the rest have M.A.s
Affiliation of First Author: Department of Psychology, University of Maryland, College Park, MD 20742,
email: hill@psyc.umd.edu