Traditionally, survey data was collected by fieldworkers trekking across the country, knocking on doors and filling in paper questionnaires. As it sounds, it is possibly the slowest way of collecting data. However, recent advances in technology and data collection methodologies are changing this. More and more organisations are collecting data using innovative approaches that provide faster responses and quicker access to survey data.

In many developing countries, however, this traditional way of collecting survey data (i.e. face-to-face) still remains at large and the implication for timing, costs and data quality has become a topical debate amongst stakeholders. For example the United Nations Economic Commission for Africa (UNECA) held a workshop in October 2015 that aimed to strengthen the capacity of African countries to use computer assisted interviewing (CAI) and specifically mobile technologies to collect data for effective policy and decision-making.

The assumption is that technology, such as web and SMS-surveys, holds promise to deliver cleaner data quicker:

  • Technology can enable “plug and play” style surveys to account for respondents’ preferences, improving response rates.
  • Software can improve the quality of the data you get by sending notifications in real time when there are missing responses in the survey; ensuring that only applicable questions are asked; and identifying responses that are likely to be incorrect during the interview. This is done by comparing the results of different questions in the survey or identifying impossible responses.
  • Electronic submission of data can eliminate the need for transcribing paper based answers into a database and eliminates data capturing errors.
  • SMS and web surveys can collect data from much larger samples for a fraction of the cost of face-to-face, pen and paper surveys.

These are only a few examples of how technology can improve data collection. With costs continuously decreasing more examples will emerge as the use of technology in data collection becomes an even more viable alternative over the long term. However implementing these tech-based solutions to data collection do not come without their challenges in developing countries:

  • Representative sampling: Telephones (landline or cell phones) are not yet pervasive in all African countries. Of the 1.2 billion people living in Africa, only 367 million are unique mobile subscribers (2015); even less have access to broadband. If a telephone or SMS survey is undertaken, large parts of the population will be excluded. If online surveys are used even more will be. This creates a major problem as those that are excluded from the survey are also likely to be those who we most need information from.
  • Ability to answer: Even when people have access to technology it is important that literacy levels and language distributions are taken into account, especially when attempting to get a nationally representative sample of adults in a country. In 2011 the United Nations Education, Scientific and Cultural Organisation (UNESCO) found that only 59% of Sub Saharan African adults were literate compared to 80% in the rest of developing world. The UN defines literacy as the ability to read, write and comprehend a short, simple statement about one’s everyday life.

The implication is that when working in countries with these constraints there will still be a need for certain elements of a face-to-face survey, but we need to think innovatively about how we effectively leverage technology to get better results, faster.

The following are two innovative “plug and play” style demand-side survey approaches that we are keeping our eyes on that are improving the quality of data available in financial inclusion. They mix elements of the face-to-face survey methodology with new technology driven methodologies to:

  • Account for respondents’ preference by initiating the survey face-to-face to capture base-line information, but then introducing SMS-based or web-based modules depending on the willingness and capability of the respondent. This can increase overall response rates by allowing respondents to complete surveys in their own times, as well as improve the accuracy and response to sensitive questions, such as personal income. Further by not administering the questionnaire in one sitting you also reduce the chance of respondent fatigue. The trade-off is subtle – you are still restricted to those that are literate and have technology, but you only identify those respondents after capturing a base-line of information and assessing their capability. Not a perfect solution,but one that can increase responses and save costs on the margin.
  • Overcome literacy and language barriers in demand-side surveys by storing hundreds of visuals in different shapes and sizes on a tablet PC or smartphone to replace questions. These can be used while the survey is being administrated face-to-face and help in low-literacy environments or where language is a major barrier. We really liked how this was used for food consumption questions in Tanzania where images of food and the amount of food consumed where shown to respondents to collect survey data on food security in the country.

Technology has the potential to rapidly increase the speed at which data is available. But whether it is fully adapted to the developing world context is still up for debate. What we do know is technology has the potential to have a major impact on the margins and improve the quality of data that we have available. Over the next few years, i2i looks forward to working with partner countries and other interested parties to investigate how these innovations can be best applied to ensure their potential is recognized.

After all, we, like Marissa Mayer, want sooner to be better, but only if it really is better.

A Focus Note on innovations and best practices in the financial inclusion process will be available in the middle of July. Follow i2i on twitter, @i2ifacility to stay tuned!