Method of collecting data: paper-Based or electronic
The method of data collection affects many subsequent decisions and processes, including the budget, the timeline, hiring, training, logistics, equipment and supply needs, questionnaire development and the roles and responsibilities of data entry and management personnel. A decision on whether to collect data on paper or electronically needs to be made early on in the survey planning process.
Both systems of data collection have their security risks. These should be detailed and addressed in the protocol and the application for ethical approval. Risks include:
- Loss of information: Paper-based forms may be lost or damaged (during transport, due to theft or from fire/flood during transport or storage), while electronically completed forms could be lost due to a device malfunction or if the device is stolen prior to sharing with the supervisor and/or uploading to the selected storage-based system. In addition, electronically-collected data needs to be protected from possible accidental or deliberate deletion.
- Unauthorized access to information: Both paper-based and electronic forms require secure storage with authorized access only, whether this be a physical location for paper forms or a password-protected storage-based system (with a backup) for electronic data.
Electronic Data Collection
Electronic data collection is increasingly the default option. However, it is not always feasible or may not be the preferred option.
Electronic data collection has numerous advantages. A well-designed electronic data entry system greatly improves data quality, minimizes missing data, is more efficient to implement in the field, and saves on data entry and management time preparing for analyses. Inconsistencies in the interview can be prevented even before the data are entered into the device because the data entry programme can be designed to allow only the entry of unique values (for identification codes, for example), values within predetermined ranges, and responses that match with previously entered values. It can also automatically skip questions in accordance with previous responses. These features permit supervisors and team leaders in the field to focus on community relations and on supporting optimal enumerator and field technician performance, instead of investing time checking the quality of questionnaires as they are completed.
Uploading electronic data allows for checks on data frequency and quality to be shared with the Survey manager who will assess progress against data targets. The survey management team will be alerted to any unexpected results so that the associated cause can be explored, and ideally resolved. If needed, refresher training materials can be developed and delivered by survey coordinators or monitoring teams.
The availability and use of global positioning system (GPS) technology on mobile devices helps the survey management team ensure that the PSUs and households were visited in accordance to plan. GPS can also generate a map of all selected and visited PSUs for the survey report and archives. In addition, a GPS can provide data on altitude, which may be required for interpreting certain biomarkers, such as the haemoglobin levels for defining anaemia.
Some software can incorporate methods to assist with the sampling and selection of participants. System links are also usually set up to take photographs, for example, of a food brand or a salt iodine field test, and to scan household, individual, and sample identification barcodes into the questionnaire. Data will not be accepted if, for example, the individual part of the identifier does not match the household identification.
Budget and personnel resources required to design and set up a high-quality electronic data collection system can be substantial. Typically, electronic data collection requires technical expertise for programming and a longer lead time to develop and extensively test the electronic data entry platform. This expertise is also required during training, piloting and field work for troubleshooting.
It is also important to consider the capacity needed and available to handle hierarchical data (such as linking individuals with household data). This has previously been a weakness in some survey settings, however, working with a sampling statistician or data manager is advisable to appropriately manage hierarchical data.
Assessing national capacity for and feasibility of electronic data collection
The Technical committee should assess the preparedness and capacity of the survey team to undertake data collection using mobile devices, and verify the availability of equipment and network access. Factors to consider in the assessment are described in Box 4.4.
- Budget availability for purchasing electronic devices and accessories:
- This should be weighed against the savings that will be made on paper, printing, computer access, and personnel time required for paper-based data collection and entry. Purchase of the devices could also be considered an investment because the equipment could be used again in subsequent surveys or for other purposes.
- Availability of personnel with the specialized programming and data management skills necessary to set up and manage the system, or the ability of the Technical committee to work with an external individual or team who can provide these skills.
- Ease of locally obtaining spare devices, batteries, and SIM and memory cards.
- Any anticipated periods of time away from power supply. Enumerators would need to be able to recharge batteries regularly and the team should have spare batteries and/or devices for backup purposes. Take into account:
- The proportion of selected PSUs that are in remote areas and how long a team may spend there; and
- Whether in such areas, teams will have access to a car where devices could be charged.
- Infrastructure of the region, especially related to internet connections:
- There must be an expectation of reasonably frequent access to internet connections to upload data. Another option is the ability to purchase mobile wireless internet devices.
- Cost of internet connectivity if wireless internet devices for the survey must be purchased.
- Security of the team:
- How safe do enumerators feel carrying the tablets/phones in all survey locations?
- Where security is a concern in only a small number of PSUs, paper-based forms could be used instead of electronic-based forms for these locations.
- Any government data security concerns and approvals required.
Even when electronic data collection is the selected method, all teams should carry paper-based forms in case of 1) security issues, 2) community discomfort with the use of mobile devices and 3) the risk of device failure. Where, despite extensive pilot testing, an error in the electronic data entry form is discovered that would lead to loss of data, for example from an incorrect skip, paper-based forms can be used until the issue has been resolved and new data entry forms are downloaded into survey devices.
Any data from paper-based forms that are sent to a central location should be entered using the same data entry system as on the survey devices, so that data are in the same format and can be easily merged.
To fully understand the supply requirements and implementation process for electronic data collection, refer to the relevant sections in Module 8, Survey supervision and personnel, Module 9, Supplies and equipment, Module 10, Budget and timelines, Module 11, Data collection tools, field manual and database, Module 12, Training and pilot testing, Module 13, Field logistics, and Module 14, Data entry and cleaning.
Paper-Based Data Collection
Most national survey management teams are familiar with the use of paper-based data collection. Paper-based data collection is usually quicker to develop for field use and generally requires fewer resources in terms of budget and technological expertise. However, significantly more time is required at the “back end” to double-enter data and then compare, reconcile, and clean these data. Supervision and review of questionnaires in the field and a highly skilled data entry and cleaning manager are essential to efficiently produce the highest-quality information from the data generated.
Development of data entry forms, with checks and skips, is still required. It should be completed prior to the pilot test so that the data entry process can also be piloted and adjusted as needed. Data entry should begin as soon as paper questionnaires are transferred from the field, reducing the time required for post-completion of the survey. Data comparison and data cleaning can be done while waiting for laboratory results of sample and specimen analyses.
Paper-based collection does not enable all questionnaire-related errors and issues to be systematically identified and addressed in the same close-to-real-time manner as electronic data collection. However, if the survey management team decides to use paper-based data collection, it can be accompanied by the use of an electronic survey monitoring tool that can be run on any phone with SMS capability. This tool can provide information to track the PSU visited, the number of households with complete questionnaires and the number of different types of samples collected, so that fieldwork progress can be tracked and actual implementation numbers can be compared with survey target numbers.