Perceived social support, coping, and symptoms of distress in American and Spanish students

Abstract

The aim of this study was to test whether coping responses mediated the influence of perceived social support on symptoms of anxiety/depression, social withdrawal, and aggressive behavior in American (N=349) and Spanish students (N=437). Participants completed measures of perceived support, social stress, coping, and distress. Coping partially mediated relations between perceived support and distress, with coping mediation most evident in individuals facing high levels of social stress. Decreased use of disengagement coping by individuals with high perceived support appears to partially explain the protective value of perceived social support. Multiple group covariance structure analysis showed that models linking perceived support, coping, and distress were very similar across cultures, suggesting that the mechanisms underlying decreased risk for individuals with high perceived support may be relatively independent of cultural context, and that interventions designed to increase perceived support and decrease disengagement could be appropriate in both cultures.

Keywords:

• Perceived social support
• coping
• depression
• anxiety
• social withdrawal
• aggressive behavior
• cross-cultural psychology

Notes

1. Additional data about these analyses are available by contacting the authors.

2. Alpha coefficients for the perceived stressfulness of negative social events are not included because individuals only rate stressfulness for the events they actually experienced. Thus, it would only be possible to compute alpha coefficients for individuals who experienced all possible social stressors.

3. Equivalence of relations between social stress, perceived support, and symptoms across American and Spanish samples was tested using a structural model that included paths between social stress and perceived support, direct paths from social stress to psychological symptoms, and direct paths from perceived support to psychological symptoms. This model was a strong fit to the data and invariant across samples. It is available by contacting the authors.

4. Separate models were estimated for each psychological symptom because the sample size was insufficient to test a model including all symptoms simultaneously.

5. An alternative to using split-half indictors of perceived social support would be to use three MPSS scales assessing perceived support from family, friends, and significant others. The split-half indicators were used to limit the number of parameters in the model. A structural model based on three manifest variables failed to converge.

6. Low- and high-stress cut-off scores were set at the 30th and 70th percentile to balance the need to create clearly distinct groups with the need to maintain an adequate sample size for structural equation modeling. These percentiles corresponded to scores of ≤5 and ≥10 in the Spanish sample, and of ≤6 and ≥11 in the American sample.