Brief COPE Inventory:
Scoring and Interpretation

The Brief COPE measures 14 distinct coping responses to stress. Carver (1997) abbreviated the longer COPE inventory so that applied studies could assess coping more efficiently without forcing all strategies into a single overall style score. Domains include active coping, planning, emotional and instrumental support, acceptance, avoidance-related strategies and others.

It is frequently used in health psychology, disaster recovery, chronic-illness and occupational stress studies. The instructional frame of the version matters: a response about coping with one health event need not represent a person’s typical coping in every situation. Researchers should model the subscales relevant to their hypotheses rather than create an arbitrary sum of all 28 items.

• Subscales: fourteen two-item subscales, each scored independently. In the standard 28-item order the item pairs are self distraction: 1 and 19; active coping: 2 and 7; denial: 3 and 8; substance use: 4 and 11; emotional support: 5 and 15; instrumental support: 10 and 23; behavioral disengagement: 6 and 16; venting: 9 and 21; positive reframing: 12 and 17; planning: 14 and 25; humor: 18 and 28; acceptance: 20 and 24; religion: 22 and 27; self blame: 13 and 26.
• Response scale: the original administered form commonly uses 1 to 4 (I haven’t been doing this at all, a little bit, a medium amount, I’ve been doing this a lot). Some electronic datasets store equivalent category ratings as 0 to 3, and the two conventions must not be mixed.
• Reverse-scored items: none in Carver’s subscale key.
• Unscored items and total: none unscored in the standard 28-item form, and no overall score was recommended by Carver.
• Recall period: depends on the instructions, for example coping with a specified event or an ongoing situation. Check the time frame in the participant instruction and the item numbering before computing scores.
• Versions: the 28-item Brief COPE, the longer COPE inventory and event-specific instructional adaptations. Item order may change in translations.

1. Verify that cope1 to cope28 follow Carver’s standard order and that the numeric response coding is 1 to 4. The syntax below assumes 1 to 4.
2. Compute the 14 distinct two-item sums using the pair list. Each subscale ranges from 2 to 8, and you can optionally divide by 2 to obtain a mean from 1 to 4.
3. Do not calculate or publish an overall Brief COPE total. If data are coded 0 to 3, the corresponding subscale sums range from 0 to 6 instead.
4. Require both items for each subscale in the sample. The original paper does not establish one global prorating rule for missing individual items, so describe any alternative method explicitly.

COMPUTE cope_self_distraction=SUM.2(cope1,cope19).
COMPUTE cope_active_coping=SUM.2(cope2,cope7).
COMPUTE cope_denial=SUM.2(cope3,cope8).
COMPUTE cope_substance_use=SUM.2(cope4,cope11).
COMPUTE cope_emotional_support=SUM.2(cope5,cope15).
COMPUTE cope_instrumental_support=SUM.2(cope10,cope23).
COMPUTE cope_behavioral_disengagement=SUM.2(cope6,cope16).
COMPUTE cope_venting=SUM.2(cope9,cope21).
COMPUTE cope_positive_reframing=SUM.2(cope12,cope17).
COMPUTE cope_planning=SUM.2(cope14,cope25).
COMPUTE cope_humor=SUM.2(cope18,cope28).
COMPUTE cope_acceptance=SUM.2(cope20,cope24).
COMPUTE cope_religion=SUM.2(cope22,cope27).
COMPUTE cope_self_blame=SUM.2(cope13,cope26).
EXECUTE.

pairs <- list(
  self_distraction=c(1,19),
  active_coping=c(2,7),
  denial=c(3,8),
  substance_use=c(4,11),
  emotional_support=c(5,15),
  instrumental_support=c(10,23),
  behavioral_disengagement=c(6,16),
  venting=c(9,21),
  positive_reframing=c(12,17),
  planning=c(14,25),
  humor=c(18,28),
  acceptance=c(20,24),
  religion=c(22,27),
  self_blame=c(13,26))
for (domain in names(pairs)) {
  data[[paste0("cope_",domain)]] <- rowSums(
    data[paste0("cope",pairs[[domain]])],na.rm=FALSE)
}

A higher score on a given domain indicates relatively greater endorsement or use of that particular coping strategy. It does not mean better or worse overall coping: the adaptiveness of a strategy depends on context, the outcome studied and the theoretical model.

Specify a small set of hypothesis-relevant coping subscales as primary outcomes or predictors. Correlated coping domains can enter a Multiple Linear Regression together, but check multicollinearity and adjust for multiple hypothesis tests when many subscales are compared. A two-item subscale has little room for internal-consistency estimation, so report the inter-item correlation, for instance with Correlation Analysis (Pearson & Spearman), and the Spearman-Brown coefficient where helpful.

Analyze repeated coping assessments with Linear Mixed Models (Multilevel Modeling) when the context remains consistent, because a change in the time frame or stressor can alter the construct assessed. For a factor analysis of all 28 items, such as Exploratory Factor Analysis (EFA), test the intended multidimensionality rather than assuming a common one-factor scale. No universal clinical minimal important difference is provided by Carver (1997).