Contents - Previous - Next

Putting in place data quality checks

BOX 10. Some Examples of Purposeful Sampling young

The purpose of a trial hygiene evaluation study conducted in Tanzania was to carry out an exploratory assessment of hygiene practices among two selected rural populations. Observations conducted before the study began showed that there were more hygiene-related activities in households with young children than in those without, so only such house-holds were selected for further observation. Informal interviews with the mothers/caretakers of infants and young children were conducted in the same households. The number of interviews and observations in each site was twenty. This number was decided on the basis of the following considerations:

• Estimated number of households With young children.
• Total time allocated for interviews and observations in relation to group discussions conducted with participatory visual aids.

The participants of group discussion sessions were self-selected, that is, the groups consisted of those who were willing and able to attend meetings called by the study team. Special efforts were made by the team to invite different categories of people to attend group discussion meetings by visiting homes and explaining the purpose of the meetings to as many people as possible Participants of group discussions were sometimes divided d into sub-groups according to gender, age, and social status.

Processes of checking the quality or trustworthiness of information obtained by using qualitative methods differ from those applied to data obtained from quantitative surveys. Trustworthiness checks for qualitative information are essentially components of study design and execution which enhance the quality or goodness of the information gathered. They are not a set of tests to be applied to the data once collected (unlike statistical significance or goodness-of-fit tests), but in-built checks that are put in place before the start data collection and monitored throughout the conduct of investigation. Such checks include:

• Prolonged or intense engagement of the various participants. If you have ample time and other resources required for an extended study, you can check the quality of information you gather through prolonged engagement with the different stakeholders. If, however, the time allowed for a hygiene evaluation study is limited to a few weeks or months, as is often the case in water supply, sanitation and health/ hygiene education projects, you will opt for intensive interaction with the study participants. This requires you to build trust and rapport with the study population, to learn the particulars of the context relatively quickly, and to be open to multiple influences. Trust and rapport can only be established quickly if you already know the local language, understand the cultural nuances and have genuine respect for local people and their ways of life (see "Sensitizing the Study Team" in Chapter 3 for more detail on building rapport).

If your hygiene evaluation study is one of a series of intensive engagements with the study population, that is, one hygiene evaluation cycle to be followed by another (see "Hygiene Evaluation Cycle" in Chapter 4), then you may have both intense and prolonged engagement which will increase the trustworthiness of your findings.

• Triangulation of sources, methods, and investigators. Crosschecking of information on the same topic gathered from different sources, using different methods and/or by different investigators, is an integral part of qualitative investigation. The term triangulation derives from land surveying, where bearings are taken by drawing lines from at least two landmarks, in different directions, and finding their intersection point (Patton, 1990:187-9). A given problem can be thoroughly investigated only if information is collected from more than one source, when more than one method or tool of investigation is used, and/or when more than one investigator (with different perspectives) is involved. You can put in place means for triangulation of information by including people with different perspectives in your study team and by combining different methods/tools of investigation (see Table 3 at the end of Chapter 5, and Table 4 at the end of Chapter 6).

• Parallel investigations and team communications. If your study covers more than one location, and you have more than one study team, the teams can crosscheck the quality of each other's data sets by meeting regularly. If all teams are using the same methods, this will enable you to check how replicable the methods are. For parallel investigations to succeed, good communication between team members is important. This requires regular formal meetings and established group norms of behaviour.

• Diary of activities. Each member of the study team should keep a diary of activities throughout the period of preplanning, training, study design, and execution. These may or may not be revealed to others, but will help you at a later stage to remember the immediate reasons for methodological decisions and changes of direction.

• Participant checking. Periodic feedback sessions will enable you to present results of the investigation to members of the study population, and to test whether they agree with your understanding of what they are doing. This will enhance your rapport with them as it demonstrates your interest in their views and comments on your findings.

It will also set in motion ideas for the study participants to implement your findings.

Report(s) with working hypothesis(es), contextual descriptions and visualizations. Study reports which include sufficiently detailed or thick descriptions with visual materials and direct quotations capturing personal perspectives and experiences provide infinitely better checks of data quality than thin reports which present information that may be only partially set in context.

Peer review/checking. Peer reviews allow colleagues (not directly involved in your study) to explore important aspects of the study that might have been overlooked by the team of investigators. This will also help to keep members of the study team honest, by exposing them to searching questions which probe biases and explore meanings.

Impact on stakeholder's capacity to know and act. As a result of the study processes and outcomes, the study team and other stakeholders should have an increased awareness and appreciation of the issues addressed by the study. This should enable them to plan and execute follow-up action. If those who were involved in the study from the beginning remain unaffected or no wiser about the implications of the results at the end, then you have failed. The level of awareness and appreciation of the study results by those who participated in it ultimately depends on the level of their participation, not only in information gathering but also in the analysis, interpretation, reflection and judgement of the results (see Chapter 7).

Scheduling activities

In the next chapter, we describe a number of methods and tools you can use for gathering and analysing information. Some of these involve group activities, others are individual observations and interviews to be conducted during home visits. It is useful to plan activities in advance even if you end up having to make some changes when you start gathering information. A good activity plan allows for unforeseen events and includes substitute activities in case of mishap. Some flexibility will be necessary. It is necessary for the planned activity flow-chart to take account of the following considerations:

Ways of maintaining participants interest and the study team's interest and stamina. For example, by alternating group activities with individual interview and observations, information gathering with review (with and without participants), and feedback sessions. This can prevent fatigue and lack of motivation/interest while participants and investigators are engaged in intensive interaction. Qualitative investigations demand a lot of energy and stamina. Investigators need to keep alert almost all the time. Make sure you include periodic breaks and days off to allow members of the study team to rest, to refresh their minds, and/or to spend time with their families, especially if they have young children whom they have to leave behind while they travel to the study site(s). Try to correspond these with participant's holidays or busy days such as market days, Fridays or Sundays, and so on.

• Available time and resources. Project personnel often engage in several tasks at the same time, albeit for very short periods of time (a few days at a time), and they may commit themselves to the study, or be asked to do other jobs as well. It is important to think ahead and plan your study carefully to avoid unnecessary interruptions once you start. Interferences can put at risk the study team's motivation and its rapport with the study population(s).

(See Diagram 2. Arrows on the lines and loops indicate interconnections between the various decisions, statements, and activities/processes.)

DIAGRAM 2. Planning Flow Chart

Contents - Previous - Next