Self-Compassion Test (SCS-SF)

Kristin Neff of the University of Texas at Austin published the 26-item Self-Compassion Scale in 2003 in Self and Identity. The short form came in 2011 from Filip Raes, Elizabeth Pommier, Neff and Dinska Van Gucht, in Clinical Psychology & Psychotherapy. The authors wanted a shorter version for settings where time is tight, such as survey research and tracking progress in therapy.

The 12 items come from the long scale, two from each of its six parts: self-kindness, self-judgment, common humanity, isolation, mindfulness and over-identification. The authors picked items that correlated highly with the long-scale total and with their own part. Self-judgment, isolation and over-identification are the negative parts, so their items are reverse-scored.

Raes and colleagues built and checked the short form in three samples: 271 first-year psychology students in Belgium, 185 Dutch-speaking adults recruited by e-mail (mean age 33), and 415 students at the University of Texas at Austin. Internal consistency (Cronbach's alpha) for the total score was at least .86 in all samples. The short-form total correlated .97 or higher with the long-form total in every sample. Factor analyses supported six related factors and one overall self-compassion factor.

The six subscales were less reliable, with alphas from .54 to .75 in the English sample, so Neff recommends the long scale if you want subscale scores. A 2019 study of the long scale in 20 samples (11,685 people) supported using a total score alongside six subscale scores.

Not everyone agrees on the total. A 2017 meta-analysis of 18 studies found that the positive items link to less distress and the negative items to more. Its authors warned that a total score, which reverses the negative items, may overstate the link with mental health problems.

Neff calls these cut-offs informal and says the scale has no clinical norms. Her information sheet and her FAQ differ by 0.01 at the 3.5 boundary, and some researchers split scores at the median of their own sample instead.

Two of the three validation samples were university students, and the third was an e-mail snowball sample of Dutch-speaking adults, so the short form has been studied less in older adults, clinical groups and other cultures. Neff's information sheet says there are no clinical norms that mark a person as high or low in self-compassion, and that scores are mainly used to compare groups. The test is self-report, so it reflects how you see your own habits.

Half the items describe harsh self-responding, so a low average can come from being hard on yourself, from being less kind to yourself, or both, and the total can't tell which. The subscale scores from this 12-item form are too unreliable to interpret on their own.

Use your result as a rough picture of how you treat yourself in hard moments. Neff calls the low, moderate and high labels informal, so treat them loosely. Retaking the test after a few weeks, or after a rough stretch, shows whether your answers change.

If harsh self-criticism or shame is weighing on your days, consider talking with a therapist. If you're thinking about suicide, call or text 988. Our articles on self-compassion and the inner critic cover the ideas behind the questions.

More on this topic: Self-compassion, Your inner critic, Self-esteem test, Shame vs. guilt.

Your answers are scored in your browser and aren't sent or saved anywhere. Close the page and they're gone.

• Raes, Pommier, Neff and Van Gucht, Clinical Psychology & Psychotherapy, 2011 (full text copy on Neff's site): Samples, alphas, correlations with the long form, factor structure, the 12 items and the original summed scoring. Read in full.
• Raes et al., 2011 (PubMed): Abstract only. Confirms alpha of .86 or more and r of .97 or more with the long form.
• Neff, Self-Compassion Scale Short Form information sheet: The permission letter, the scoring key, reverse-scored items, the informal 1.0 to 2.49 / 2.5 to 3.5 / 3.51 to 5.0 rubric, the statement that no clinical norms exist, and subscale reliability.
• Neff, Scales for Researchers: Lists the scales and says the long scale suits ages 14 and up. Says the short form correlates near-perfectly with the long form on total scores.
• Neff, FAQs for Researchers: Says the scales may be used for clinical purposes and gives the informal cut-offs, noting there are no formal norms.
• Neff, Test How Self-Compassionate You Are: Neff's own online test of the long scale. Says average overall scores tend to be around 3.0 and gives the rough low, moderate and high guide.
• Neff et al., Psychological Assessment, 2019 (PubMed): Abstract only. 20 samples, N = 11,685, long scale. Supports a total score and six subscale scores.
• Muris and Petrocchi, Clinical Psychology & Psychotherapy, 2017 (PubMed): Abstract only. Meta-analysis of 18 studies on positive and negative items and psychopathology.
• Kirby, Tellegen and Steindl, 2017, Behavior Therapy (PubMed): Abstract: 21 randomized trials, 1,285 participants; compassion-based interventions raised self-compassion (d = 0.70, 13 trials); the authors note the evidence rests mostly on small samples. Used for the sentence on the result page that self-compassion can be learned.