Connor Davidson Resilience Scale: A Closer Look at Its Impact

The Connor-Davidson Resilience Scale, usually called the CD-RISC, is one of the most widely used psychological questionnaires for measuring how well a person bounces back from stress, adversity, and trauma. Developed in 2003, the original version contains 25 items scored on a 0-to-4 scale, with higher totals indicating greater resilience.¹ Since its introduction it has been translated into dozens of languages, shortened into briefer versions, and applied in settings ranging from military recruitment to cancer wards to college campuses. Its influence on how researchers and clinicians think about resilience is hard to overstate, yet the scale has also attracted pointed criticism about what it actually captures and whether its internal structure holds up across different populations.

What the Scale Measures and How It Was Built

Kathryn Connor and Jonathan Davidson designed the CD-RISC to capture qualities they believed protected people against the psychological fallout of hardship. The 25 items tap into themes like personal competence, tolerance of negative emotions, acceptance of change, trust in one’s instincts, perceived control, and the role of spiritual or social support. You rate each statement from “not true at all” (0) to “true nearly all of the time” (4), so total scores can range from 0 to 100. In the original validation work, the scale showed strong internal consistency, good test-retest reliability, and the ability to distinguish people under significant stress from those coping well.¹

Connor and Davidson’s initial factor analysis suggested five underlying factors, grouping items into clusters related to personal competence, perseverance, spiritual influences, trust in one’s instincts, and tolerance of negative affect. That five-factor structure became the standard description of the scale for years. It has since come under serious scrutiny, which matters for anyone using or interpreting the scale today.

The Five-Factor Problem

When independent teams took the CD-RISC into new populations, the original five-factor structure often failed to replicate. One study in post-9/11 U.S. military veterans ran a full exploratory and confirmatory factor analysis and found no support for the five-factor model. The best statistical fit was a two-factor solution, with factors organized around adaptability and self-efficacy, and only the adaptability factor behaved the way resilience theory would predict, acting as a buffer against psychopathology after trauma exposure.² A Swedish validation similarly found high discriminant and predictive validity but could not reproduce the original factor structure.³

This pattern has repeated across countries and populations. The practical takeaway is that the CD-RISC works well as a single overall score of resilience, but interpreting the five subscale scores as though they reflect five distinct psychological traits is on shaky ground. Most researchers today treat the total score as the reliable metric and handle subscale results cautiously, if they report them at all.

Shorter Versions and Why They Exist

Twenty-five items is not a long questionnaire by research standards, but in busy clinical settings or large-scale surveys, even a few minutes of respondent time matters. That practical pressure led to the development of shorter versions. The CD-RISC 10 strips the scale down to ten items and was validated as a single-factor measure with strong internal consistency and construct validity.⁴ It has since been tested in a wide range of cultural contexts. A Russian youth validation, for instance, confirmed the single-factor structure with reliability coefficients around 0.85 and showed the expected positive links to well-being and self-esteem and negative links to depressive symptoms.⁵

The 10-item version does have its own wrinkles. An Australian study using Rasch modeling found that two of the ten items did not perform well statistically, suggesting an eight-item version might be more precise.⁶ That finding has not led to widespread adoption of an eight-item version, but it flags that even refined scales can carry items that behave differently across samples.

At the extreme short end, a two-item version (the CD-RISC 2) focuses on the ability to bounce back and to adapt to change. It was developed partly for use in large pharmacological trials where questionnaire space is at a premium.⁷ The trade-off is obvious: with only two items, you gain speed but lose the ability to pick up subtleties. In head-to-head comparisons within clinical trials, though, the 2-item, 10-item, and 25-item versions have produced broadly comparable effect sizes when tracking treatment response, which says something about how much information is concentrated in those core items.⁸

How CD-RISC Scores Behave in Clinical Populations

One of the scale’s most consistent findings is that people dealing with depression, anxiety disorders, and PTSD tend to score lower than the general population. In the original validation work, the mean score for a general population sample was around 80 out of 100, while people with PTSD posted the lowest scores of any clinical group studied.⁹ That gradient, general population scoring higher than depressed or anxious groups and those groups scoring higher than PTSD groups, has been replicated in many countries.

In oncology, the CD-RISC has been used to understand what helps cancer patients cope. A principal component analysis in cancer patients found the scale accounted for roughly two-thirds of the variance in resilience, but about a third remained unexplained, hinting at determinants the scale does not capture. The authors noted that interventions targeting coping strategies and ensuring adequate social and spiritual support could improve resilience in at-risk patients.¹⁰ That unexplained third is a useful reminder: no questionnaire captures everything about a complex trait.

Predicting Outcomes in the Military

Military organizations have a strong practical interest in resilience. If you could identify who is likely to struggle before they are exposed to the stressors of service, you could intervene early or adjust training. The CD-RISC has been tested in exactly this context. A study of U.S. Air Force trainees found that resilience scores measured at the start of military service significantly predicted both attrition from service and the likelihood of receiving a mental health diagnosis within the first six months.¹¹ Most trainees reported relatively high resilience, which makes intuitive sense for a self-selected group entering a demanding career. But even within that high-scoring pool, the scale still differentiated who was more likely to leave early or develop problems.

Separately, pretreatment resilience scores on the CD-RISC have been shown to predict treatment response in PTSD patients. In a trial of venlafaxine (a commonly prescribed antidepressant), patients with higher baseline resilience were more likely to achieve remission of PTSD symptoms by week 12, even after controlling for which treatment group they were in and how severe their symptoms were at the start.¹² The implication is that resilience is not just a feel-good concept: it predicts how likely someone is to benefit from a specific pharmacological treatment.

Resilience Among Healthcare Workers

Burnout and secondary trauma are occupational hazards for clinicians, and the CD-RISC has been widely used to study these relationships. Among ICU nurses, those who met diagnostic criteria for PTSD scored significantly lower on resilience (a mean of about 75) compared to those without a PTSD diagnosis (a mean of about 83). The gap was consistent across subscales measuring personal competence, perseverance, and leadership.¹³

A study of Canadian palliative care physicians found an even more striking pattern. Their median resilience score was 74, placing them at just the 25th percentile of the normative population. Those with high burnout averaged around 67.5, compared to roughly 77 for those with low burnout.¹⁴ These numbers paint a sobering picture: the physicians tasked with end-of-life care scored, on average, well below general population norms for psychological resilience.

Whether interventions can meaningfully raise those scores is a more complicated question. A randomized trial of mindfulness-based cognitive therapy in psychiatric nurses found that the intervention group showed higher resilience and lower burnout after eight weeks compared to controls.¹⁵ A similar mindfulness-based trial in patients with schizophrenia also found significant increases in CD-RISC total scores after eight weeks.¹⁶ But results are not universal. A randomized trial of mindfulness-based cognitive therapy in critical care nurses found no difference in CD-RISC scores between intervention and control groups.¹⁷ The mixed evidence suggests that mindfulness interventions can shift resilience in some populations and contexts, but it is not a reliable across-the-board effect.

How the CD-RISC Compares to Other Resilience Measures

The CD-RISC is not the only resilience questionnaire in use. The Brief Resilience Scale (BRS) is a common alternative that frames its questions differently, focusing more narrowly on the ability to recover from stress. You might assume two scales measuring “resilience” would produce interchangeable results, but they don’t quite. A confirmatory factor analysis across multiple samples found the two scales were highly correlated yet statistically distinct, with each having unique predictive effects on depression and life satisfaction.¹⁸

A study comparing the two scales specifically in adolescents recovering from concussions found only a modest correlation between them. The BRS showed a small inverse relationship with anxiety symptoms, while the CD-RISC did not, suggesting the two instruments are tapping into somewhat different slices of what we casually call resilience.¹⁹ This matters practically: a clinician choosing a resilience measure for a particular purpose should consider which aspect of resilience is most relevant, not just grab whichever scale is most familiar.

The Overlap with Personality

A persistent concern in resilience research is how much the CD-RISC simply measures personality traits that have been studied for decades under different names. Multiple studies have documented a substantial overlap between resilience scores and the personality dimensions of neuroticism and extraversion. People who score low on neuroticism and high on extraversion tend to score high on resilience, which raises the question of whether the CD-RISC is capturing a genuinely distinct psychological construct or repackaging personality. The evidence suggests both neuroticism and extraversion are reliable predictors of resilience, so any resilience score should be interpreted with an awareness that it partly reflects temperament.

This does not make the scale useless. Even if resilience overlaps with personality, it may capture patterns of coping and adaptation that personality inventories miss. But it does mean that claiming someone “lacks resilience” based on a low CD-RISC score might be partly restating that they tend to experience negative emotions intensely, something a neuroticism scale would already tell you.

Subjective Versus Objective Resilience

Another layer of complexity involves what kind of resilience a self-report questionnaire can capture. The CD-RISC measures how resilient you think you are. Objective measures of resilience, by contrast, look at actual outcomes after adversity: did the person maintain functioning, stay employed, avoid clinical depression? A study in adolescents found no significant correlation between a subjective resilience measure and an objective one, and concluded the two should not be used interchangeably.²⁰ In other words, believing you are resilient and actually demonstrating resilience under pressure appear to be separate things, at least in younger populations.

This is not unique to the CD-RISC. All self-report questionnaires face the same limitation. But it is worth keeping in mind when interpreting scores, especially in high-stakes contexts like military screening or clinical risk assessment.

What Brain Imaging Studies Have Found

Some researchers have tried to connect CD-RISC scores to what is happening in the brain. A study of at-risk children and adolescents using resting-state functional brain imaging found that youth with higher resilience scores spent less time in a brain connectivity pattern characterized by heightened communication between default-mode and executive-control networks. These effects were independent of depressive symptoms and adversity exposure, suggesting resilience has its own neural signature rather than just being the absence of distress.²¹

In adults, higher resilience on the CD-RISC has been associated with greater gray matter volume in the inferior frontal gyrus, a region involved in emotional regulation and impulse control, and with structural differences in the insula, which plays a role in empathy and interoception. These structural correlates were in turn associated with better quality of life and lower depression.²² Work linking CD-RISC scores to heart rate variability has produced more tentative results, with some associations appearing only when measured at the same time point and differing between men and women.²³

Using the Scale with Adolescents

The CD-RISC was originally developed and validated in adults, but it has increasingly been used with younger populations. A scoping review of adolescent resilience self-report scales found that in teen samples, the original five-factor model again did not hold up well. The best fit was typically for the 10-item single-factor version.²⁴ A 2025 study of adolescent and young adult psychiatric outpatients confirmed the 10-item version’s single-factor structure and reported good internal consistency and test-retest reliability over one month. However, it flagged that items related to versatility, humor, and persistence did not perform as well under item-response-theory analysis.²⁵

Whether those weaker-performing items reflect genuine developmental differences in how teenagers think about resilience, or simply awkward wording for younger respondents, remains an open question. For now, the 10-item version appears workable in youth settings, but clinicians should be cautious about treating every item as equally informative for a 15-year-old as for a 45-year-old.

Cross-Cultural Adaptations

The CD-RISC has been translated and validated in dozens of languages. A Danish validation of the 10-item version in hospital staff confirmed sound psychometric properties and adequate internal consistency.²⁶ An Iranian validation of the full 25-item scale found good model fit and measurement invariance between men and women, meaning the scale appeared to function the same way regardless of the respondent’s sex.²⁷ Korean and Russian validations have produced similar findings.

The consistency is encouraging, but a recurring theme across these studies is that the original five-factor structure rarely replicates outside the original sample. Most cross-cultural studies end up endorsing either a single-factor or two-factor model. This is not just a statistical curiosity. If the five factors do not hold, then any research comparing, say, the “spiritual influences” subscale between cultures is comparing something that may not represent the same psychological construct in both groups. Using the total score or the 10-item version sidesteps this issue, which is likely part of why the shorter forms have become more popular than the full 25-item scale in international research.

Resilience and Academic Persistence in College Students

Beyond clinical and occupational settings, the CD-RISC has been applied in higher education to understand what keeps students enrolled and performing. A study of undergraduate students found that intrapersonal resilience factors (qualities like self-regulation and optimism, the kind the CD-RISC is designed to capture) contributed to explaining variance in cumulative GPA even after accounting for aptitude and prior achievement. There was also a strong correlation between resilience factors and mental health, suggesting these are not independent domains but feed into each other.²⁸ For university counseling centers interested in identifying students at risk of dropping out, resilience screening could add something useful beyond traditional academic metrics, though it is worth remembering that a self-report score is a single data point, not a diagnosis.