Customized-Language Voice Survey on Mobile Devices for Text and Image Data Collection Among Ethnic Groups in Thailand: A Proof-of-Concept Study
Bibliographic record
Abstract
BACKGROUND: Public health surveys are often conducted using paper-based questionnaires. However, many problems are associated with this method, especially when collecting data among ethnic groups who speak a different language from the survey interviewer. The process can be time-consuming and there is the risk of missing important data due to incomplete surveys. OBJECTIVE: This study was conducted as a proof-of-concept to develop a new electronic tool for data collection, and compare it with standard paper-based questionnaire surveys using the research setting of assessing Knowledge Attitude and Practice (KAP) toward the Expanded Program on Immunization (EPI) among 6 ethnic groups in Chiang Rai Province, Thailand. The two data collection methods were compared on data quality in terms of data completeness and time consumed in collecting the information. In addition, the initiative assessed the participants' satisfaction toward the use of a smartphone customized-language voice-based questionnaire in terms of perceived ease of use and perceived usefulness. METHODS: Following a cross-over design, all study participants were interviewed using two data collection methods after a one-week washout period. Questions in the paper-based questionnaires in Thai language were translated to each ethnic language by the interviewer/translator when interviewing the study participant. The customized-language voice-based questionnaires were programmed to a smartphone tablet in six, selectable dialect languages and used by the trained interviewer when approaching participants. RESULTS: The study revealed positive data quality outcomes when using the smartphone, voice-based questionnaire survey compared with the paper-based questionnaire survey, both in terms of data completeness and time consumed in data collection process. Since the smartphone questionnaire survey was programmed to ask questions in sequence, no data was missing and there were no entry errors. Participants had positive attitudes toward answering the smartphone questionnaire; 69% (48/70) reported they understood the questions easily, 71% (50/70) found it convenient, and 66% (46/70) reported a reduced time in data collection. The smartphone data collection method was acceptable by both the interviewers and by the study participants of different ethnicities. CONCLUSIONS: To our knowledge, this is the first study showing that the application of specific features of mobile devices like smartphone tablets (including dropdown choices, capturing pictures, and voiced questions) can be successfully used for data collection. The mobile device can be effectively used for capturing photos of secondary data and collecting primary data with customized-language and voiced questionnaire survey. Using smartphone questionnaires can minimize or eliminate missing data and reduce the time consumed during the data collection process. Smartphone customized-language, voice-based questionnaires for data collection can be an alternative and better approach than standard translated paper-based questionnaires for public health surveys, especially when collecting data among ethnic and hard-to-reach groups residing in multilanguage-speaking settings.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.020 | 0.009 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".