Related Experiment Videos
Evaluation of a risk factor survey with three assessment methods.
Beth Theis1, Jennifer Frood, Diane Nishri
1Division of Preventive Oncology, Cancer Care Ontario, 620 University Avenue, Toronto, Ontario M5G 2L7. Beth.theis@cancercare.on.ca
Chronic Diseases in Canada
|March 6, 2002
Summary
Evaluating cancer risk factor surveys using data patterns, qualitative feedback, and appraisal identified common and unique issues. These methods enhance survey data quality, particularly when resources for in-depth development are limited.
Area of Science:
- Epidemiology
- Health Survey Methodology
Background:
- Cancer risk factor surveys are crucial for public health research.
- Ensuring the quality of survey data is essential for accurate risk assessment.
- Limited resources often constrain comprehensive questionnaire development and validation.
Purpose of the Study:
- To evaluate the effectiveness of three distinct methods for assessing cancer risk factor survey questions.
- To identify common and unique issues revealed by each evaluation method.
- To demonstrate how these methods can improve survey data quality.
Main Methods:
- Dataset response pattern analysis to identify data anomalies and effects.
- Qualitative respondent feedback to uncover clarity and usability issues.
- Questionnaire appraisal by experts to assess technical and structural quality.
Main Results:
- Dataset analysis revealed missing data, out-of-range values, order effects, and coding inconsistencies.
- Qualitative feedback highlighted issues with clarity, sensitive topics, undefined terms, response option presentation, and recall.
- Questionnaire appraisal identified technical terms, complex syntax, hidden definitions, and ambiguous wording.
Conclusions:
- Multiple methods provide complementary insights into survey question quality.
- These evaluation techniques offer practical solutions for improving data accuracy in cancer risk surveys.
- The described methods are valuable for enhancing survey quality under resource constraints.