Introduction
Self-compassion involves self-kindness, mindfulness and a recognition that inadequacies and suffering are shared human experiences (Neff and Germer, Reference Neff and Germer2013). Low levels are associated with higher levels of anxiety, depression, shame, rumination and perfectionism (for a review, see Barnard and Curry, Reference Barnard and Curry2011). Interventions targeting self-compassion have been developed to reduce psychopathology by increasing affiliative emotions toward the self and others. Mindful Self-Compassion (MSC; Neff and Germer, Reference Neff and Germer2013) is an 8-week therapist led group therapy program recently developed to enhance self-compassion in both clinical and non-clinical populations. Each session focuses on a specific topic and includes experiential exercises, discussions, and homework assignments. The MSC program yielded significant improvements in self-compassion, mindfulness, and wellbeing, with gains maintained at a 12-month follow-up (Neff and Germer, Reference Neff and Germer2013).
An open access psychoeducational website, entitled Self-Compassion, has been developed around content similar to MSC (Neff, Reference Neff2016). This website provides written, audio and video material including information on the conceptualization and measurement of self-compassion, and a number of self-compassion exercises such as meditation. No data on the efficacy of this website are available. Accordingly, the aim of the present study was to provide preliminary data regarding its efficacy in a non-clinical population. A secondary aim was to evaluate its impact on shame and rumination. People with high levels of shame can have difficulty being kind toward themselves and shame can become the focus of rumination, which further increases vulnerability to a number of mental health problems (Cheung, Gilbert and Irons, Reference Cheung, Gilbert and Irons2004).
Method
Design
A single-subject experimental design (AB) was used to provide an analysis of treatment impact within individual participants. Replication of treatment effects across subjects was provided through the use of multiple single-subject protocols.
Participants
Participants were recruited from a pool of 4179 university students and 629 university employees using posters, e-mails and information sessions during classes. The inclusion criteria were: being at least 18 years of age, reporting a “tendency to be hard on oneself”, and having Internet access. Of the 14 participants recruited (7 students; 7 employees), most were women (n = 11, 78.6%) and single (n = 10, 64.29%) with an average age of 31.8 years (SD = 11.5). Five participants (35.7%) withdrew from the study during the baseline phase due to self-reported lack of time, yielding a final sample of nine participants (N = 9). One participant could not complete the final outcome measures due to injury.
Measures
Self-Compassion Scale
The SCS is a 26-item measure evaluating three bipolar components of self-compassion: self-kindness as opposed to being overly critical or self-judgmental; common humanity (recognizing that all humans are imperfect) as opposed to isolation; mindfulness as opposed to avoidance or rumination over aspects of oneself or one's life that one dislikes. A total score was computed with higher scores representing higher levels of self-compassion.
The Other as Shamer Scale
The OAS is an18-item measure of external shame referring to how people think others perceive them. Scores are summed with higher scores corresponding to higher levels of shame.
Rumination-Reflection Questionnaire-Self-Rumination Subscale
The RRQ-SR is a12-item measure that assesses the frequency of 12 ruminative behaviors. The average item score was calculated with higher scores suggesting higher levels of rumination.
Treatment adherence
Participants reported in minutes on a weekly basis how much time they had spent, on average, navigating the self-compassion website.
Treatment benefits questionnaire
A 5-item questionnaire was developed to assess participants’ perceived benefits derived from the website, their intent to continue doing the exercises and if they would recommend this website to others.
Intervention
At the time of the study, the website Self-Compassion (Neff, Reference Neff2016) included seven self-compassion exercises: 1) How self-compassionate are you?; 2) Exploring self-compassion through writing; 3) The criticizer, the criticized, and the compassionate observer; 4) Changing your critical self-talk; 5) Self-compassion journal; 6) Identifying what we really want; and 7) Taking care of the caregiver. Video. Audio clips, including six guided meditation exercises, were also available to be downloaded and a workbook could be purchased.
Procedure
Single case AB protocols comprising a 4-week baseline and a 5-week treatment phase were used. Prior to the baseline phase, face-to-face meetings were held with individual participants to explain the study and to obtain informed consent. Outcome measures were administered online on a weekly basis using the SurveyMonkey software. When outcome measures were not completed within a day of the scheduled date, reminder e-mails were sent the following day. Although no clinical or therapeutic guidance was offered during the treatment phase, participants received weekly e-mails reminding them to complete outcome measures and measures of adherence. A posttreatment telephone interview was conducted to obtain participants’ feedback on the program and their overall experience.
Statistical analyses
In this study, the Critical Difference (CD) statistical method for analyzing data from single-case designs was used (for a description, see Nishith, Hearst, Mueser and Foa, Reference Nishith, Hearst, Mueser and Foa1995). To calculate CDs, all scores were first converted in ipsative z scores. The following formula for a one-tailed test was used, CD = 1.64[J(l-r)] ½, where J is the number of measurement points and r, the test-retest reliability of the measure. Each weekly score from the treatment phase was compared to the mean score at baseline. The absolute value of the difference was then compared to the CD and considered statistically significant when equal to or greater than the CD.
Results
Relative to the mean ipsative z scores at baseline for Weeks 1 to 4, seven of the nine participants (77%) showed statistically significant posttreatment improvements on the SCS (see Table 1). Changes were observed as early as after 2 weeks into the treatment phase. Similar results were found for shame. In fact, all but one of the participants who showed a statistically significant increase in self-compassion also showed a statistically significant reduction of their OAS scores compared to baseline levels. Few statistically significant changes were noted in rumination, with only two participants showing a reduction in RRQ-SR scores.
Table 1. Descriptive statistics for baseline, test-retest reliability, ipsative z-scores for baseline and treatment and Critical Difference score (CD) for each outcome measure
Notes: CS – Self-Compassion Scale. OAS = The Other as Shamer Scale. RRQ-SR =Rumination-Reflection Questionnaire-Self-Rumination subscale. P1 to P9 = Participants 1 to 9.
*p < .05
Treatment adherence
Participant 6 only spent a few minutes navigating the website at the beginning of the treatment and participant 8 chose to purchase the website's accompanying workbook instead of using the website. Self-reported total time spent on the website, excluding P6 and P8, averaged 199 minutes (SD =135; range 60 to 480 minutes).
Treatment benefits
Online questionnaire
Most participants reported having benefited from the self-compassion website. Elevated ratings of 3 (somewhat) or 4 (very much so) were given by a majority of the participants in relation to overall benefit (75%), benefit at school or work (50%), and improved quality of life (62.5%). Most (75%) also appreciated the self-compassion exercises, with 62.5%reporting that they intended to continue using the exercises. Overall, 89% of the participants said that they would recommend the website to others.
Telephone interview
All but one participant, who indicated he would prefer face-to-face therapy, gave positive feedback about the self-compassion website. Specific comments centered around themes of how the website helped them gain awareness of how they can be hard on themselves and helped them develop less self-defeating and critical perceptions of themselves. Increased self-confidence and patience, greater tolerance toward others, enhanced decatastrophizing of challenging situations, and a more positive outlook on life were also reported.
Discussion
This study provides preliminary evidence that a self-compassion psychoeducational website can produce improvements in both self-compassion and external shame within a relatively brief period of time. These results add further support to existing studies showing: 1) that self-compassion can be enhanced with targeted intervention; 2) that improvements can be observed early in treatment; and 3) that self-guided interventions can be an effective means of increasing self-compassion (e.g. Neff and Germer, Reference Neff and Germer2013). Participant satisfaction was high and almost all participants reported that they would recommend the website to others.The observed dropout rate was comparable to the general weighted average of 31% reported across internet-based psychological interventions (Melville, Casey and Kavanagh, Reference Melville, Casey and Kavanagh2010). Reported time each participant spent on the website varied, but the average was comparable to other similar studies (e.g. Clark et al., Reference Clark, Kelleher, Hornbrook, DeBar, Dickerson and Gullion2009). Given the small sample (N = 9), it is difficult to identify participant-factors that may underlie differences in time spent on the website. However, there was no apparent link between the time spent on the website and the magnitude of treatment gains. The extent to which participants are integrating the material in their day-to-day functioning is probably more relevant than absolute minutes at the website.
In addition to enhanced levels of self-compassion, the majority of participants also experienced an associated decrease in shame. Paradoxically, one participant reported greater self-compassion and higher levels of shame. It may be that some individuals with low self-esteem do not feel worthy of self-compassion when they first experience it, which increases levels of shame. Few significant changes of self-ruminative behaviors were observed. This could be due in part to the fact that none of the participants reported using the website's meditation exercises. Meditation has been found to reduce rumination (Jain et al., Reference Jain, Shapiro, Swanick, Roesch, Mills and Bell2007). It may also be that changes in self-rumination require a longer treatment than 5 weeks.
Limitations
Limitations to the present study include a low participation rate and the use of a single-subject experimental design in a small sample. Randomized controlled trials among well-defined diagnostic groups should be conducted and efforts to further encourage participation deployed as low rates of participation in clinical trials are commonly encountered. However, these encouraging preliminary results suggest that it may be worthwhile to develop a more structured Internet-based self-compassion course. Internet-based cognitive behaviour therapy has been found helpful for the treatment of anxiety and depression and would likely be equally effective for self-compassion. This type of first line approach to providing care may well increase access to effective treatment.
Acknowledgements
Conflict of interest: The authors have no conflicts of interest with respect to this publication.
Comments
No Comments have been published for this article.