The explosive growth in the availability of mobile phones in societies around the world – even in some of the poorest, most remote communities – is increasingly leading many groups to explore how these devices might be used effectively as part of large scale data collection efforts in many sectors, including education. Utilizing small, portable electronic computing devices to help collect data is not new, of course. For over two decades, laptop computers and personal digital assistants (PDAs) have featured in initiatives to (e.g.) collect census information, interview consumers of various goods and services and poll potential voters. That said, such efforts often faced constraints related to, among other things: costs; the relative novelty of such devices among key segments of the population; the need to provide device-specific user training; and difficulties in exchanging data between these devices and other components of a larger system for data collection. If, as it has been argued, the best technology is often the one you already have, know how to use, can maintain and can afford, for most of the world, the mobile phone fits these criteria quite well. As of late 2013, rates of mobile phone penetration stood at 96% globally (128% in developed countries and 89% in developing countries). According to the International Telecommunications Union, “today there are almost as many almost as many mobile-cellular subscriptions as there are people in the world.”
Very simple, low-end ‘dumb phones’, for example, can utilize simple text messaging (or SMS) or voice to (e.g.) send out short queries by phone to a bank of phone numbers, prompting users to reply with a short response, which can be either predefined (‘text 1 for yes, 2 for no’) or open-ended. Smartphones can be used in much more sophisticated ways by presenting rich media survey questions directly to respondents or to help guide the actions of an ‘enumerator’ (someone who administers a survey in person) by presenting a user-friendly interface to help an enumerator input and transmit data in structured ways. Such phones may also contain help files and training aids for the enumerators. In between the high and low end, ‘feature phones’ (a catch-all category of sorts for phones which can do more than make basic voice calls and send and received text messages, but do not have the advanced functionality of smart phones) can make use of simple graphical forms (e.g) on screen as prompts for questions, and can store/transmit structured data as a result of responses.
Data input or captured into phones may be transmitted or shared in many ways (including SMS, MMS, USSD, Bluetooth, wireless Internet, or the exchange of physical memory cards). Where mobile connectivity is not available, data can be stored on the phone and transmitted later once a phone is within sufficient range of a cell tower.
be useful in large-scale data collection efforts,
and what comparative advantages might their use have when compared to other options?
Speed: Collecting data through the use of a mobile phone can greatly speed up the data collection process. Where network availability allows for near-instantaneous transmission of data to a central coordinating group, the reduced amount of time that elapses between local data collection and delivery can save weeks or even months of time in the overall data collection process. In addition, an early warning system of sorts can be established, allowing survey coordinators to quickly identify potential problems with data collection efforts and (potentially) correct them in almost real time.
Accuracy: Digital data capture at the source can greatly reduce transcription errors, and data transmission over mobile networks may ensure that no data are lost ‘in transit’. Capture and transmission of data digitally may also ensure that it is easier to store and access them at later dates, should this be required.
Ubiquity, familiarity and convenience: Enumerators and survey respondents may, at a general level, already be quite comfortable using a mobile phone (and indeed may be using their own personal device), even if they have not used it specifically as part of data collection efforts. Such devices may be widely available already in target populations, who may be accustomed to their use in a variety of contexts.
Training: Because people may already know how to use the devices for many purposes, less technical training may be necessary in some circumstances. In addition, on smart phones, and to a lesser extent with feature phones, help files and on-screen prompts may provide useful relevant supporting documentation and guidance that may reinforce messages from training that does occur, and potentially obviate the need for some sorts of training altogether.
Low power: Compared with devices such as laptops, mobile phones may be much easier to keep charged, as they require much less power and because many fast, low-cost charging options may be available in local communities because people are already utilizing such devices extensively for other purposes as part of their daily lives.
Combining with other data: Depending on the functionality of the phone used, textual data captured via mobile phone can be combined with data in other formats such as photographic images, audio, and video, as a way to substantiate the information provided by text. If a building being surveyed is noted as ‘damaged’, for example, an accompanying picture can provide further documentation. In addition, GPS or geo-locationary data can be passively collected and transmitted along with survey data. This can be used to help map the location of hospitals or schools, or to offer ‘proof’ that an enumerator actually visited a place that she has claimed to visit.
Low cost: All of these characteristics and affordances may mean that data collection enabled through the use of mobile phones can be done at substantially lower costs than is possible via traditional means.
It is worth noting that, in some circumstances, the comparison between data collection via traditional means and enabled through the use of mobile phones may be a false one. Indeed, in addition to improving the efficiency of data collection efforts when compared with traditional, largely paper-based practices, mobile data collection may also offer options for data collection that simply aren’t feasible, or even possible, using other tools or methods.
Issues and Challenges
While data collection via mobile phones may offer particular advantages when compared with traditional survey and data collection efforts, this is not to imply that such efforts are not without challenges or difficulties. Some common notable issues and challenges include those related to:
Technology: What technology should we use? What are the minimally viable specifications required for the devices used in mobile data collection efforts? What software applications are available, and what are their related advantages and disadvantages? These are often the first questions that many groups contemplating extending or complementing existing data collection efforts through the use of mobile phones often ask. While questions such as these will (or should) inevitably be posed at some point during the planning process, leading with them is typically a mistake. It is important that choices around a specific technology or device not define the initial scope or conception of the extent to which mobile data collection efforts might be beneficial and possible. No one technology may do everything that is required. Where one tool may appear to be a perfect fit for what is required, it may not work at the necessary scale. Vendors or partners may propose use of certain technologies or devices which may not be well suited for the specific data needs and/or data collection and sharing processes of the sponsoring group. Deciding which tools to use, and which partnerships to develop, may be critical pieces of the puzzle. First, however, it important to have a larger picture in place of what the objectives of a particular data collection might be, as well the relevant local context (including key stakeholder groups). Once that has been established, technology-related decisions can be made in the service of both higher order goals and on-the-ground realities.
Training: While in some instances less training and explicit instruction may be required, because the tool being used (a mobile phone) is familiar to respondents and enumerators alike, in some circumstances additional technology-related training and support may still be required. The adoption of user-centered design principles and techniques can help in some circumstances to avoid the need for certain types of training. (If, for example, an interface is easy or even ’intuitive’ to understand, users may not need to spend much time figuring it out.) That said, depending on the nature of the survey process and the methods of data transmission, additional training may well be required.
Cost: The costs of designing survey instruments delivered digitally may be considerably higher when constructing traditional paper-based questionnaires. In addition, new back-end technology infrastructure may need to be procured and put into place. Where it is not possible for enumerators to utilize their personal phones (whether because such use is not customary or permitted, because existing phones have insufficient functionality or because mechanisms for covering or reimbursing related data transmission costs are not in place), devices may need to be purchased and distributed (and potentially collected once they have served their purpose). Air time may need to be purchased. (More than a few data collection efforts by mobile phone to ‘benefit charities’ have foundered because sponsoring groups have wrongly assumed that mobile operators would be quick to donate free airtime ‘for a good cause’). Whether or not these costs in aggregate are cheaper than the way things are traditionally done or not – and they often are! – sponsoring groups may have difficulty estimating and budgeting for such costs on the front end, especially the first time they engage in efforts of this sort.
Data security: Digital collection and transmission of data as part of large scale survey efforts carries with it numerous potential risks and challenges related to data security and privacy that are often fundamentally different than those posed as a result of traditional paper-based survey efforts. If stored on devices, data can potentially be stolen or improperly accessed – the same holds true during data transmission. The use of encryption at both the device level and during transmission can greatly mitigate such risks, but the use of digital data security tools, protocols, and good practices, as well as related regulatory frameworks, laws and guidelines governing the use of such tools, may not be known widely known – or in some places, even be at hand. Where third party vendors or tools are utilized during the mobile data collection process – something that is usually the case – care needs to be taken to ensure that ownership, possession and utilization of data collected and transmitted are clearly articulated. Sufficient mechanisms need to be in place to audit related arrangements and agreements, and to ensure that penalties for non-compliance are clear and enforceable.
Speed of change: If mobile data collection efforts are expected to be repeated over time (for example, as a part of yearly census or annual planning activities), it should be expected that some of the enabling technologies may themselves change – as might the good practice guidelines and regulations related to their use. In some circumstances, this sort of change may even occur during the course of a specific data collection process itself! Technological change often outpaces the ability of planners and policy makers to anticipate and respond to these changes. Those involved in planning for and implementing mobile data collection activities would do well to keep this in mind, and should be prepared to monitor and respond to such changes over time. Planners should try to avoid making a ‘big bet’ on an unproven technology (especially one based on a closed or proprietary standard) or on single vendor, and should always consider how they may most effectively exit and transition from specific relationships, tools and standards.
[“source=worldbank”]
