6 ways to improve data quality in your organization

Posted by: Derrick Mwiti on

Not everything that can be counted counts, and not everything that counts can be counted. – Albert Einstein

Whether collected through surveys, assessments or routine monitoring and evaluation, high-quality data is important to the well-being of any development organization. Inaccurate data leads to false reports, poor decisions and, potentially, loss of time and money. On the other hand, good quality data promotes better decisions, optimal interventions and maximum impact. When planning for a malaria prevention campaign, for example, a firm understanding of the disease burden, community attitudes and socioeconomic characteristics is vital.

Data is considered to be of high quality when it is precise, reliable, complete and timely. Below are 6 steps that we recommend to help you improve the quality of data in your organization:

1. Determine the Metrics to Track

Different organizations need to track different aspects of their work. The critical indicators for the organization should inform the kind of data the organization collects. Therefore, the first step is for an organization to define the metrics they would like to track. For instance, an organization dispensing polio vaccine to children might track the number of children vaccinated in different locations. They might also track their age, sex and other relevant characteristics.

2. Designate a data champion

With huge amounts of data to manage, having a data champion in your organization is critical. This person’s job is to ensure that the data collected meets the standards laid out by the organization. He or she also ensures that the data supports the organization’s overall information demands. Designating a data champion ensures that data from different departments is gathered, stored and managed correctly, hence preventing any anomalies that may lead to wrong inferences.

3. Consolidate your data

Consolidation is the process of aggregating data from different sources into a single store. This is crucial for organizations that use multiple databases. The likelihood of having duplicate records is very high, particularly when disparate data stores are involved. Consolidation makes it easy for organizations to perform data analysis and presentation. It also reduces the cost involved in maintaining multiple databases.

4. Normalize your data

Normalization involves restructuring data in order to reduce redundancy and promote data integrity. Unnormalized data leads to false insights that can adversely affect the organization. For example, if you have a column for countries, Kenya and KE would be treated as two different entities when performing data analysis. Similarly, if you have a variable for currency, KSh and KES would be treated as different values, resulting in misleading conclusions. Normalizing your data prevents errors and saves time during the analysis process.

5. Profile your data

Data profiling is the process of routinely examining your data to ascertain that it meets your organization’s quality standards. This process may reveal inconsistencies with your datasets that you can address before they become problematic. Data profiling may involve detecting outliers and running basic summary statistics such as mean, mode, median, variance and standard deviation.

6. Use technology

Taking advantage of technology can help you implement the steps described here more easily and reliably. Although many organizations still rely on manual filing systems to manage their data, technology such as spreadsheets and specialized data collection applications are much better alternatives.

Here at Hoji, we use mobile technology to help organizations to collect better quality data more efficiently. Our integrated mobile data collection and analysis platform covers the entire decision-making value chain including data analysis and visualization. From small one-off surveys to country-wide Data Quality Assessments, our platform has helped many organizations super-charge their data management. Care to see what they like about us?

One Response

  1. Kingangi Gitau says:

    Great read ….. and just recently we had a problem with consolidating and normalizing our data after source became a bit more “diverse” …reading a bit more about this is good for reflecting & better planing

Leave a Reply