Data Design, Data Collection, and Data Handling
Data Design
Social & Scientific Systems has been working for a number of years in topics relevant to data systems for epidemiologic studies. In the design phase, we consider the level of detail needed, the expected analytic use of the data, ease of administration, consistency of responses, and validity of the question content. For all topics, we research historical use of variables in order to create clear questions and resolve coding schemes before data are collected. We have subject matter expertise in many areas relevant to health and risk factors, including designing content for the following types of surveys:
- Ascertaining of health and disease histories, including family health
- Data collection of prescription and over-the-counter drug use, dietary supplements, use of complementary medicine, medical procedures, and medical record reviews
- Diet and nutrition surveys, including recall questionnaires, food diaries, and supervised dietary intake studies
- Lifestyle factors, such as residential and work histories, marital and family status, hobbies, exercise, and sleep
- Detailed surveys to assess occupational and environmental exposures
- Educational measurements, such as behavioral checklists, teacher assessments, and school record reviews
- Physical injury and impairment, such as head and neck injury, traumatic limb loss, and bone fractures
- Clinical data forms to record biomeasures, such as blood pressure, height and weight, body shape, coordination, and detailed clinical examinations
Data Collection
We use the following methods to collect data, including in-person and telephone interviews, home visits, clinical interviews, and mail and Web-based surveys:
- Computer-assisted telephone interviews (CATI), primarily programmed in Blaise, a software package developed by Statistics Netherlands
- Computer-assisted in-person interviews (CAPI)
- Optically marked questionnaires (scanned forms)
- Medical record abstracts and physician checklists
- Data from public and private databases
- Web-based data collection
Data Handling
We have strong expertise in data handling systems and our staff of programmers, analysts, and survey design experts are experienced in quality control procedures. Data may be collected via paper questionnaires, optically marked forms for electronic scanning, or computer-assisted interviews. We most typically use computerized data collection modes to streamline data collection and processing. For example, our computer-assisted interviews and scanned forms are pre-programmed to control for range and skip patterns, which helps reduce initial errors and greatly reduces time to edit the data. In consultation with the client, our staff develops coding schemes for classifying original data from source documents into codes for analysis. When pilot studies are conducted, provisional coding schemes are tested and changes are made to the questionnaire. As data are processed, edit procedures are in place to quickly identify missing data in case they can be retrieved and to identify problem responses. Documentation is updated as new values and issues are identified. Many studies conducted will include multiple forms and specimen collection variables; edit procedures include techniques to verify the match.