Being able to implement impactful interventions and get the intended results hinges upon making the informed decisions. The best way to make informed decisions is to collect and analyze data about development needs in a community; monitoring, evaluating, and continuously learning from the lessons data is bringing. For such decisions to be accurate, data must also be accurate. Therefore, it is of utmost importance to ensure that data being collected to support decision making in social development projects is of high quality.
In this article, I will outline different ways in which to ensure quality of data when the data is being collected.
Use correct data collection methods and tools
Collecting good data begins at using the appropriate data collection methods and tools.
Data collection methods include surveys, interviews, focus group discussions and observations, among others.
Data collection tools include structured and semi-structured questionnaires, checklists, and interview guides, among others.
If your study is more qualitative in nature, using close-ended questionnaires means that you won’t be able to capture long explanations required for a qualitative research.
For each data collection project, it is important to select the right tools and the right methods for collecting data.
Recruit the right data collection personnel
Different data collection projects require different sets of skills. Although some projects can be done by just about anyone who can be able to ask questions and record responses, some projects require that people with specialized knowledge be used.
For example, I you need to collect public health data and have to use specialized machines, you may need to recruit personnel who have at least some training in health.
Some projects may also require that the recruited individuals be experienced in asking certain kinds of questions. For example, where a questionnaire involves asking very sensitive information, beginner researchers may find it difficult to collect the right answers.
Train data collection team adequately
After having the right people for the data collection project, it is also important to train them adequately.
There are usually 2 parts to training data collection teams. First, they must be trained in the general aspects of administering a questionnaire such as sampling procedures, personal presentation, tone, probing and so on. Secondly, you need to orient the team on how they will ask and record responses on the data collection tools at hand.
It is important to not assume that the team you have, even if it is filled with experienced people, will be able to collect the right data without having them undergo some orientation.
Develop and document instructions
Even after having the data collection team undergo training, it is important that the data collection tools being used contain clear instructions on how to administer them.
Questions must be framed exactly the way they will be asked. Instructions must be included on what the enumerator is encouraged, allowed, or not allowed to do on the question, for example when probing is encouraged, and when reading out of available response options is allowed or not allowed.
Questions that require calculations must include a hint on how the calculations will be carried out. Questions whose response will be a number must include the unit of measurement to be used. In certain cases, the enumerator may have to convert a number provided into a different measurement unit – this information must be documented, and the enumerators trained on how to do this.
Pretest
Pretesting data collection tools before administering them to your targeted subjects is important for several reasons.
First, you will be able to find out if the tool you are using adequately covers your research objectives. Once the pretest data has been collected, you will be able to establish whether data analysis will yield the kind of results you intend to get from the data. With this information, you will be able to revise your tools accordingly.
Secondly, this is a hands-on orientation for your data collection team. It is also a way to get feedback from the data collection team in terms of whether the data collection instrument is adequately optimized for data collection. For example, some response options may need to be expanded, some questions may be repetitive and hence would have to be removed.
You will also gauge whether the tool is too long or just the right length. Issue to do with respondent fatigue may be uncovered from the pretest, and hence the tool may have to be revised accordingly.
Data quality control officer/team to do routine data collection checks
While in the field, data checks can be done after each day of data collection to ensure that the data being collected is of the intended quality.
This may involve assigning a data quality control officer or a team of data quality control officers. The team supervisor may also do this job. It is important to train these individuals from the beginning and lay out data quality guidelines to be followed and tracked by the data quality control officers.
Having the data checked frequently will ensure that you are not caught unawares on the final hour realizing that all the data collected is not of the optimal quality as you expected.
Length of data collection tools
Data collection tools must be of optima length to avoid respondent fatigue. To capture accurate responses, respondents must be in their right mind. When a data collection tool is too long, respondent devise a way to ensure that you do not waste any more of their time. This in most cases will result into poor quality data.
Summary
Data quality is of utmost importance for any data project. Your ability to trust the results from data hinges upon the quality and reliability of the data. As data professionals, it is our duty to ensure that we put in place measures for ensuring that data being collected to be used to inform decision making can be relied upon.