Abstract
Purpose
The Social Difficulties Inventory (SDI-21) assesses everyday problems experienced by cancer patients, including difficulties
with self-care, work and relationships. Early development and psychometric evaluation studies have validated the SDI-21 for
computer administration. However, several recent studies have administered the SDI-21 on paper. We sought to test the score
equivalence of electronic and paper versions of the SDI-21.
with self-care, work and relationships. Early development and psychometric evaluation studies have validated the SDI-21 for
computer administration. However, several recent studies have administered the SDI-21 on paper. We sought to test the score
equivalence of electronic and paper versions of the SDI-21.
Methods
A randomised two-arm crossover trial in a sample of cancer patients with varied diagnoses. Patients completed electronic (via
the internet) and paper versions of the SDI-21, with half randomly assigned to complete the electronic version first (n = 51) and half the paper version first (n = 60). Patients were asked to complete both versions at home, within 2 weeks. Analyses were performed for the SDI-21 summary
score and three subscales.
the internet) and paper versions of the SDI-21, with half randomly assigned to complete the electronic version first (n = 51) and half the paper version first (n = 60). Patients were asked to complete both versions at home, within 2 weeks. Analyses were performed for the SDI-21 summary
score and three subscales.
Results
Score distributions and internal reliabilities for the paper and electronic versions were highly similar. There were no significant
differences between mean summary or subscale scores for the two administration modes. All mean score differences (all <0.25
of a scale point) were well below the SDI-21’s established minimally important differences, and all 95 % confidence intervals
were narrow and included zero. Intraclass correlations between paper and electronic scores were uniformly high and significant
(all ≥0.85) and above the standard acceptable level of reliability.
differences between mean summary or subscale scores for the two administration modes. All mean score differences (all <0.25
of a scale point) were well below the SDI-21’s established minimally important differences, and all 95 % confidence intervals
were narrow and included zero. Intraclass correlations between paper and electronic scores were uniformly high and significant
(all ≥0.85) and above the standard acceptable level of reliability.
- Content Type Journal Article
- Category Brief Communication
- Pages 1-6
- DOI 10.1007/s11136-012-0242-3
- Authors
- Laura Ashley, Psychosocial Oncology and Clinical Practice Research Group, University of Leeds, Level 3, Bexley Wing, St James’s Institute of Oncology, Leeds, LS9 7TF UK
- Ada Keding, Psychosocial Oncology and Clinical Practice Research Group, University of Leeds, Level 3, Bexley Wing, St James’s Institute of Oncology, Leeds, LS9 7TF UK
- Julia Brown, Clinical Trials Research Unit, University of Leeds, Leeds, LS2 9JT UK
- Galina Velikova, Psychosocial Oncology and Clinical Practice Research Group, University of Leeds, Level 3, Bexley Wing, St James’s Institute of Oncology, Leeds, LS9 7TF UK
- Penny Wright, Psychosocial Oncology and Clinical Practice Research Group, University of Leeds, Level 3, Bexley Wing, St James’s Institute of Oncology, Leeds, LS9 7TF UK
- Journal Quality of Life Research
- Online ISSN 1573-2649
- Print ISSN 0962-9343