TL;DR
This study explores the feasibility and acceptability of using handheld computers for data collection (HCDC) in public health research in China, a region where this technology is underutilized. The findings suggest that HCDC is a viable, preferred, and efficient method for collecting data from hard-to-reach populations in China, despite some initial technical challenges.
What Problem Does This Paper Address?
Collecting accurate and timely data is crucial for effective public health research and interventions. Traditionally, paper-based questionnaires and manual data entry have been the norm, especially in developing countries. However, this method is labor-intensive, prone to errors, and can be inefficient, particularly when dealing with large sample sizes or hard-to-reach populations. While handheld computer data collection (HCDC) has gained traction in developed countries for its accuracy, speed, and data management capabilities, its adoption in public health research within China has been limited. This paper addresses this gap by investigating the acceptability and adoption of HCDC among interviewers conducting a tobacco use survey among rural-to-urban migrant women in China, a population that is often difficult to access.
Methods
The study employed a mixed-methods approach, combining HCDC implementation with post-training and post-data collection surveys to evaluate interviewer experiences.
- HCDC System: Hewlett-Packard iPaqs running Windows Mobile 5 and Questionnaire Development Studio (QDS) software (version 2.6) were used. QDS facilitated the creation of electronic questionnaires with features like handwriting recognition for Chinese characters, data validation, and error checking.
- Training: A comprehensive training program was delivered to survey supervisors and interviewers, covering HCDC functions, data entry, transfer, and management.
- Data Collection: A cross-sectional survey on tobacco use was conducted using the programmed HCDC. Data was managed centrally using the QDS Warehouse system.
- Evaluation:
- Post-training survey: Assessed trainees' self-assessed skills, knowledge, and acceptability of HCDC.
- Quality-control evaluation: Supervisors observed interviewers' performance during the study.
- Post-data collection survey: Interviewers provided feedback on their experience with HCDC, including ease of use, problem-solving confidence, and preference for future surveys.
Datasets
Not discussed in the paper. The study utilized a custom-designed questionnaire for a specific tobacco use survey.
Key Insights and Ideas
- Bridging the Digital Divide: The study demonstrates that even in contexts with limited prior experience, HCDC can be successfully implemented and adopted by local staff, fostering capacity building.
- Language is Key: The use of Chinese characters within the HCDC system, facilitated by QDS software, was crucial for interviewer comprehension and data accuracy, overcoming a significant barrier in previous Chinese HCDC studies.
- Preference for Efficiency: Interviewers, after experiencing HCDC, showed a strong preference for it over traditional paper-and-pencil methods, citing ease of use and confidence in data management.
- Technical Proficiency Matters: The study suggests that a certain level of computer proficiency among interviewers is beneficial for HCDC adoption, potentially influencing selection criteria for future studies.
Key Results
- High Confidence Post-Training: After training, over 90% of trainees felt confident in their ability to collect data, transfer it, and encrypt it.
- Positive Post-Collection Experience: 80% of interviewers found data collection and management easy, and 60% felt confident in solving encountered problems.
- Preference for HCDC: Nearly 70% of interviewers preferred using handheld computers for future surveys, with 55% considering HCDC a particularly useful tool for China.
- Efficiency Gains: While initial pilot testing revealed time-consuming aspects (e.g., consent form reading, character entry), continuous practice reduced survey completion time significantly.
- Minimal Refusals: Only two potential interviewees refused to participate due to confidentiality concerns, indicating general acceptance by the target population.
- Technical Issues: The most common technical issues were system freezing and SD card reading errors, which were usually resolved by rebooting. Persistent difficulty with Chinese character entry was also noted.
Limitations and Future Work
- No Pre-Training Assessment: The study lacked a pre-training assessment, making it difficult to quantify the exact skill improvement from training alone.
- Potential Observer Bias: The quality control evaluation, involving direct observation, might have influenced interviewer behavior.
- Software Limitations: QDS software's programming environment is primarily in English, posing challenges for non-English language support, which requires further development.
- Future Work: The authors suggest focusing on capacity building of local staff, further software development for non-English languages, and exploring the use of HCDC for other applications beyond surveys, such as therapeutic interventions. They also recommend considering computer proficiency as a selection criterion for interviewers.
Summary updated