Conducting a Tracker Study: a Tough Nut to Crack

To enjoy a nut, you must first crack open the hard outer shell – the Malawi APRA research team have found one such nut and, as detailed in this blog, hope to have gathered the tools required to crack its shell.

The Malawi team intends to track members of households who were part of a 2007 study in the districts of Mchinji and Ntchisi, in Malawi. The government of Malawi and the UK Department for International Development (DfID) co-funded the study, titled the ‘Agricultural Input Subsidy Program Evaluation’. This study evaluated the impact of farm input subsidies and collected information about households and their members. It listed how many people were in a household, their ages and their main economic activities.

We want to go to the two districts that were involved in the 2007 study, to find the same households and enquire about household members who were listed in the study. If we find the households and their members, we want to ask them questions about their current livelihoods. We are interested in assessing the impact of agricultural commercialisation on the trajectory of their livelihoods. If some of the members have left their ‘original’ households, we will ask for their new addresses in order to ‘track’ them and interview them about their current living situation and economic activity. However, we have identified a number of challenges that will be faced in implementing our tracker study:

  • Data accuracy: the 2007 study made every effort to ensure that the data that was collected in 2007 was accurate. However, we still need to remain conscious of the possibility of errors in the original data, which may well effect the results of our tracker study.
  • Finding the households: we will conduct the study in Malawi where national citizens’ registration was not required until last year (2017). Even then, not many people have acquired national identity cards. People could have moved away from their ‘original’ households (in what we refer to as ‘branching’ households) or passed away. Very few locations in Malawi have easily identifiable addresses, especially in rural areas where there are no road names. In addition, not everyone has access to a phone, so we will need to visit households in person in order to confirm availability. With a substantial population that is illiterate, it is possible a respondent will give us an address or a phone number that is not correct.
  • Consent to conduct the tracker survey: research ethics demand that we conduct a study interview only when a respondent grants us consent. We can only hope that we will be granted consent to conduct studies by the households. It must be noted that over the past ten years, experience shows consistently very good response rates of over 90%.
  • Data collection: being granted consent to conduct the study does not mean that the respondent has to answer all of the questions. Some may even be sceptical or cynical of the motives behind the study. The respondents may therefore not give us any data. However, based on experience, we have not had any situations where a respondent chose not to answer any of the questions, or where an interview was terminated prematurely.
  • The data collectors: the success of the study will depend on how well we communicate the aims and approaches of the study to the data collectors. Failing to follow data collection procedures would compromise the study’s accuracy. Our training and data collection experience has shown, however, that data collectors have been able to grasp the aims of previous studies and we do not anticipate that they will not be able to understand what is expected of them in this exercise. We, the research team, have a pool of data collectors we have worked with for more than ten years and are confident that they will do as trained.

We are interested in assessing the impact of agricultural commercialisation on the trajectory of livelihoods, as defined in APRA’s methodology (stepping-up in agricultural activities, stepping-out of agricultural activities, hanging in to subsistence farming, dropping-out of agriculture). After the quantitative survey, we will identify households in different livelihood trajectories and conduct detailed qualitative interviews. This will involve interviews with ‘original’ households and ‘branching’ households. Where ‘branching’ households are sampled for detailed qualitative interviews, their ‘original’ households will also be interviewed to get a broader understanding of the role of agricultural commercialisation in livelihood development. It will not be easy, but we think by carefully thinking through the procedures, the science, and the context, we will do a good job.

With a thorough assessment of potential barriers to carrying out our study effectively, and identifying the ways in which we can circumvent these barriers, we believe that we can crack this nut.

Written by: Jacob Mazalale

Image credit: Ollivier Girard/EIF