Resilience Test (BRS)

The Brief Resilience Scale was published in 2008 in the International Journal of Behavioral Medicine by Bruce Smith and five colleagues at the University of New Mexico. They built it to measure one thing: the ability to bounce back or recover from stress. They argued that earlier "resilience" scales mostly measured resources that may promote resilience, such as optimism and coping styles, without measuring recovery itself.

The team chose the final items from a longer list, using feedback from the research team and piloting with undergraduates. Three items are worded positively and three negatively, to reduce the effects of social desirability and positive response bias. They tested the scale in four groups: 128 and 64 undergraduates, 112 cardiac rehabilitation patients, and 50 women, 20 with fibromyalgia and 30 healthy. All were recruited in Albuquerque, New Mexico.

In the four samples the items formed one factor that explained 55 to 67% of the variance. Cronbach's alpha ranged from .80 to .91. Test-retest reliability was .69 over one month in 48 people and .62 over three months in 61 people. Higher scores went with lower anxiety (correlations of -.46 to -.60), lower depression (-.41 to -.66) and lower perceived stress (-.60 to -.71). In the first student sample it correlated .59 with the Connor-Davidson Resilience Scale. Mean scores ran from 3.53 to 3.98 across the four samples.

The 2008 article gives no cut-offs. PsyToolkit credits the authors' 2013 book chapter, which described six samples totaling 844 adults, for these bands: below 3.00 is low, 3.00 to 4.30 is normal and above 4.30 is high. The sources we read don't test the bands against a diagnosis or an outcome.

A German study of 1,481 and 1,128 adults from the general population found an alpha of .85 in both samples. A model that accounted for the wording of positive and negative items fit better than a plain one-factor model, which points to a wording effect.

The 3.0 and 4.3 cut-offs are the developers' own, from a 2013 book chapter that we know only through PsyToolkit's summary. The 2008 article gives no cut-offs.

The scale was tested on undergraduates, cardiac rehabilitation patients and a small group of women, all recruited in one New Mexico city. The authors say it still needs longitudinal studies to show whether it predicts recovery, and comparison with physiological signs of recovery. It's self-report, so it reflects how you see your own recovery.

The six items ask about bouncing back from stress and setbacks. They don't ask how much stress you're under, how much support you have, or whether you're dealing with a mental health condition right now.

Read your average as a rough snapshot of how you usually recover. In the authors' samples, scores a month or three months apart correlated at .69 and .62, so some movement between tries is normal. If you retake the test, leave a few weeks between tries and note what was going on in your life.

If recovery has felt slow for weeks, or stress is affecting your sleep, work or mood, talk with a doctor or therapist and bring your result. If you're thinking about suicide, call or text 988. Our stress test and burnout recovery guide may help you see what's weighing on you.

More on this topic: Stress test (PSS-10), Burnout recovery, Self-compassion test, Depression, anxiety and stress test (DASS-21).

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

• Smith et al., International Journal of Behavioral Medicine, 2008 (PubMed): Abstract. Confirms the authors, journal, year and the four samples.
• Smith et al., 2008, full text (third-party copy of the published article): The six items, scoring, sample sizes, factor structure, alphas, test-retest values, correlations and limitations. Read in full. A third-party copy of the published article.
• Smith, Epstein, Ortiz, Christopher and Tooley, 2013 book chapter (abstract, Roger Williams University repository): Abstract only. Six samples and 844 adults. The chapter text, which PsyToolkit credits for the bands, wasn't opened.
• PsyToolkit, Resilience (BRS): A secondary source. Gives the 1.00 to 2.99 / 3.00 to 4.30 / 4.31 to 5.00 bands and credits the 2013 chapter for them.
• PhenX, Brief Resilience Scale protocol: A secondary source. Lists the items, scoring and the same bands, but prints the top band as 4.31 to 6.00, apparently a typo for 5.00.
• Chmitorz et al., PLoS ONE, 2018: German version in two general-population samples (n = 1,481 and n = 1,128). Alpha .85, method-factor model fit best.
• Joyce et al., 2018, BMJ Open (PubMed): Abstract: 11 randomized trials in the meta-analysis; a moderate positive effect of resilience interventions (0.44, 95% CI 0.23 to 0.64), with CBT-based, mindfulness and mixed programs effective. Used for the sentence on the result page about resilience training.