In a previous blog on 10 innovations in qualitative research, I wrote about the importance of using a greater variety of research designs, methods and data collection techniques when conducting financial inclusion research. Similarly, this blog discusses some innovative quantitative methods that could be used when conducting financial inclusion research. It is important to note that the specific design, method or data collection mode chosen should be the one the researcher believes will provide the greatest data quality given the problem being addressed. 

What is quantitative research?

Quantitative financial inclusion research allows us to attach numbers to financial inclusion phenomena and track these for changes over time. Quantifying various aspects of financial inclusion allows us to describe, explain relationships, to predict future outcomes, to monitor progress towards targets and to assess the extent of financial inclusion in a country. 
Quantitative research uses objective, rigorous and systematic strategies for generating and refining knowledge. They mainly use deductive reasoning and generalisation, where an established theory or framework is tested by gathering data to assess whether the theory or framework is supported.
The methods presented below can be used in conjunction with the typical large-scale demand-side surveys in order to enhance the quality of the quantitative research findings by adding dimensions to the data that cannot be obtained from survey data.

1. Behavioural experiments.  Experiments are used to establish cause and effect relationships between independent and dependent (outcome) variables. In an experiment, different groups of randomly selected and matched participants are administered different levels of an independent variable or condition. The independent variable is expected to have a particular causal effect on the outcome variable. An experiment is the only research design where the researcher manipulates a variable. 

A key aspect of an experiment is that it has at least two groups – an experimental group that receives the “treatment” and a control group that does not. Behavioural experiments become popular when it became clear that people do not always behave in a manner consistent with classical rational economic theory. In fact, people often behave in irrational and sub-optimal ways when making financial and other economic decisions. Experimental methods allow us to understand how various biases, heuristics (shortcuts to problem-solving) and contextual conditions affect decision-making. This allows us to develop “nudges” that can help people make better financial decisions.

As people often have limited insight into their own behaviour, it doesn’t always make sense to ask them to explain certain financial decisions. It is often more appropriate to conduct an experiment to reveal the true underlying decision-making process. It should be noted that these behavioural experiments could be true or quasi-experimental. 

2. Conjoint surveys help us understand how people make choices between products or services or a combination of these. That insight enables financial institutions to design new products or services that can better meet customers’ underlying needs. Conjoint analysis helps to determine the value consumers place on various product attributes, such as amount of cover, duration, price, credit limit, interest rate and so one. Within each attribute a different customer value is derived from changing levels of that attribute, for example, duration terms may be 1 year, 2 years, 3 years. 

A conjoint survey allows us to determine which attribute and level of an attribute is most valued by a consumer, for example, cover for a duration of three years. Conjoint surveys are useful for testing new financial products and being able to isolate the key value drivers of a product that will increase the likelihood of it being successful in the market.

More detail on conjoint surveys can be found here. []

3. Brand price trade off (BPTO) surveys are useful in situations where there is a strong interest in the interrelationship between the brand and the offer price. It is particularly suitable when comparing products and services that are similar to each other, and where brand may be a strong determining factor behind decision-making. In a market where banks, mobile network operators (MNOs) or government agencies have a strong brand, it might be useful to test the price premium each brand commands. Where the buying decision is more complex, for instance, where the specific features of a product or service are critical, a full conjoint modelling exercise is often more appropriate.

4. Implicit association test (IAT). When measuring sensitive issues, people can’t always articulate how they feel or they provide false responses that appear more socially acceptable. IATs can overcome these challenges and provide more accurate data as they are designed to detect the strength of a person's automatic association between mental representations of objects or concepts in memory. In a typical IAT, a person is shown two sets of stimuli simultaneously and has to respond as fast as they can. For example, they could be shown the word “savings” which is then paired with two other words such as  “easy” or “difficult”, etc. The faster the person pairs a particular word to “savings”, the greater the level of association between the two words. So, in the example above, if the person responds faster when the word “difficult” is offered over when the word “easy” is shown, then we can assume that they believe saving is difficult. 

Try an IAT for yourself here.

5. Survey gamification. Traditional surveys are often described as tedious and boring. Gamification is a great way to engage your respondents and get better quality data. Gamification applies gaming principles to non-game activities, like surveys. It is best suited to online platforms but can also be used on CAPI (computer-aided personal interview) devices. Examples of gamification include: slider bars, drag and drop ratings, video, audio, graphics, smileys, thumbs up/down and the use of chat bots to stimulate conversation. They make the survey more interactive, dynamic and relevant to the respondent, thus improving data quality.

6. Panel and longitudinal designs. Longitudinal designs allow us to understand change at an individual level over a specific span of time, which we don’t get from cross-sectional and repeated cross-sectional surveys. Longitudinal design can, for example, help us assess the impact of a financial inclusion programme on the lives of a specific community. With advances in technology, it is becoming increasingly easy and cost effective to establish and maintain a panel that can provide us with valuable longitudinal data. The mode of data collection can range from face-to-face to digital data collection, computer-aided telephone interviewing and online modalities. Telephone or online data collection is usually preferred as it is more cost effective. In some instances, financial diaries are employed as part of the data collection process.

7. Self-completion leave behind. Should it be necessary to collect a large amount of information that requires a significant amount of recall from the respondent, then it might be worthwhile considering a “leave behind” self-completion questionnaire. Using this method requires an easy-to-use questionnaire booklet being dropped off at a respondent’s household or place of work. They are shown how to complete it and instructed to complete it within a week or two. The interviewer then returns after a week, checks that it has been completed correctly and returns it to the office for capturing and further quality control.

What does this mean for financial inclusion research?

The examples above show how financial inclusion quantitative research has developed beyond the traditional survey to take our understanding of a research topic to another level and help us decipher consumer behaviour further. However, as always, it is critical that the methods chosen are appropriate to the problem at hand.

Explore our other blog on 10 innovations in qualitative researchas well as i2i’s online guide to demand-side survey implementation, which exposes the various methodological considerations when executing financial inclusion surveys.