Murtagh Fliss E M, Addington-Hall Julia M, Higginson Irene J
Department of Palliative Care and Policy, Kings College London, Weston Education Centre, London.
Palliat Med. 2007 Mar;21(2):87-93. doi: 10.1177/0269216306075367.
In answering questionnaires, research participants undertake complex cognitive processes, including understanding/interpreting questions, retrieval of information from memory, decision processes to estimate answers and response formulation. Cognitive interviewing techniques are widely used in large surveys, to improve questionnaires by understanding these processes, but their use is less familiar in other areas of palliative research.
This study applied cognitive interviewing techniques, alongside standard piloting, to refine a questionnaire for survey of symptoms in end-stage renal disease patients.
Ten consecutive renal patients were invited to undertake a cognitive interview, while completing a questionnaire comprised of the Memorial Symptom Assessment, Geriatric Depression and Palliative Care Outcomes Scales. Interviews were conducted using 'think-aloud' and concurrent probing techniques, and recorded, transcribed and analysed using content analysis.
Nine interviews were completed. A variety of cognitive problems were identified, including legibility/format, comprehension of specific words/phrases, inapplicability of some questions, response estimation and difficulties caused by combining instruments. These were categorized, using Tourangeau's information processing model, and used to refine the symptom questionnaire.
Cognitive interviewing was able to helpfully identify the range and depth of difficulties with questions, to a greater degree than with standard piloting. It may be of particular benefit when instruments are used in different combinations, or applied to new study populations. Wider use of these techniques in palliative research is recommended.
在回答问卷时，研究参与者要进行复杂的认知过程，包括理解/解释问题、从记忆中检索信息、估计答案的决策过程以及回答的表述。认知访谈技术在大型调查中被广泛使用，以通过了解这些过程来改进问卷，但在姑息治疗研究的其他领域，其应用并不那么为人熟知。
本研究应用认知访谈技术，结合标准预试验，完善一份用于调查终末期肾病患者症状的问卷。
连续邀请10名肾病患者进行认知访谈，同时填写一份由纪念症状评估量表、老年抑郁量表和姑息治疗结局量表组成的问卷。访谈采用“边想边说”和同步探查技术进行，并进行录音、转录和内容分析。
完成了9次访谈。发现了各种认知问题，包括清晰度/格式、特定单词/短语的理解、一些问题的不适用性、答案估计以及量表组合带来的困难。使用图兰乔的信息处理模型对这些问题进行了分类，并用于完善症状问卷。
与标准预试验相比，认知访谈能够更有效地识别问题的范围和深度。当量表以不同组合使用或应用于新的研究人群时，它可能特别有用。建议在姑息治疗研究中更广泛地使用这些技术。
