need 10 recordings which are less than 5 minutes


Final Reflection

The research process has been an enlightening experience as various articles have been selected and assessed to support the overall study on how AI can be applied in bettering agricultural practices in developing countries. With every literal data source assessed, a better appreciation of the benefits of the study outcomes were identified. The literature data also influenced the snowball sampling approach that is to be implemented in the interview process. The need to conduct interviews with various participants created the need to better understand the role of informed consents as well as risk assessment. The aforementioned documentations and requirements are some of the key necessities learned during the development phase of the study.

As identified in the study outline, the research will exploit the expertise of individuals from developing and low-income countries. As previously mentioned, the objective would be to understand the current systems and approaches used in agriculture with the intention of identifying how AI (artificial intelligence) would be applied to improve the practice. Nevertheless, the data collection process, the interviews, is set to be commenced from April to May 2022.

Each participant will undertake an interview process that will last not less than five minutes, as the data collected must be well documented for effective outcomes. The interview process will exploit the snowball sampling process as it exploits its attributes in assessing individuals difficult to locate. It is an exceptional approach for undertaking exploratory or descriptive studies with a tiny and difficult-to-identify or find group. Nevertheless, due to language barriers, the translation process will also be accounted for in the interview process, ensuring that each question is directly related to the research.

In order to conduct the interviews, attributes such as trust are essential for numerous reasons relating to the viability of the information collected (Geddes et al., 2017). In other words, trust is critical in any study involving human subjects, although it is more critical in a project involving the snowball sampling technique. To get consent from subjects to pinpoint other individuals of their community or tribe, the research must first establish positive rapport and a propensity for reliability. It can take some time, which is why it is necessary to exercise patience when utilizing the snowball sampling approach on recalcitrant groups of people (Berndt, 2020). The exploitation of the snowball sampling technique also contributes to the overall time allocated for the project.

In that, due to the need to develop a positive rapport, a significant period must be allocated. As identified, the process will take up to 62 days to interview the participants, in other words, 10 participants from each of the selected regions. The quality of the information collected outpaces the need for a large sample group. The research intends to understand the current practices and how AI can be implemented to improve the overall performance of the agricultural sector for developing and low-income countries. With such information, it creates the needed awareness for the allocated time as well as the number of participants to be exploited in the study.

Overall, as previously mentioned, the information to be collected will primarily address the current farming practices exploited in Nigeria, Africa, and the Yunnan area, China. Moreover, experienced farmers have higher consideration as they offer critical information that can simplify or diversify the research, thereby adding value to the study. The interview data will be accumulated with the literature data to identify the best approach to implement AI and agricultural practices.


Berndt, A. (2020). Sampling Methods. Journal Of Human Lactation36(2), 224-226.

Geddes, A., Parker, C., & Scott, S. (2017). When the snowball fails to roll and the use of ‘horizontal’ networking in qualitative social research. International Journal Of Social Research Methodology21(3), 347-358.