Data accuracy is the most fundamental parameter of your data quality. By data accuracy, we mean whether the current data are a correct and precise representation of the data you intended to collect. To what extent do they correspond to the truth and reality? Think of typographical errors in the data, one decimal too many or too few, or measurement scales that differ completely.

To provide high-quality data, the data must also be available to the right people. Think about the future as well. If one employee keeps all assessment scores locally, how will you gain access to them when that employee is on leave or even leaves the organisation?

Moreover, data must certainly not fall into the wrong hands. Think of employees’ personal information that is intended for HR, but could just as easily be visible to their manager, customers or the entire internet. This can lead to major losses if access to the data is not corrected quickly.

Data compatibility is the extent to which data from different sources are compatible with one another. When you want to combine data, it is important to ensure that there is a match between these datasets. Otherwise, this can cause errors when analysing the data. Ask yourself the following questions:

• Do my datasets have the same file format?
• Are there conflicts between the different systems from which I retrieve my data?
  For example, an automatically generated database of your website traffic cannot simply be added to an Excel file containing data about your job candidates.
• Are there common variables present in my datasets that allow me to link them together?
  For example, if you want to compare test scores from different years, they must measure the same variables.

Building on compatibility, you must also take data consistency into account. To what extent are the data the same? Suppose you would like to gain more insight into the job satisfaction of your employees, then you must always measure and collect this in the same way within a dataset. If satisfaction is scored for one employee on a 7-point scale, but for another employee is only assessed with a yes/no question, you will ultimately be able to do very little with these data. A first step towards consistent data is, for example, providing an overview of all measured variables together with their explanations.

Data relevance is the extent to which the data are suitable for your needs. Two key questions here are: “Is this data applicable to my situation?” and “Is this data sufficiently up to date?” Any data for which you cannot answer an unequivocal yes will not be able to help you at that moment. Irrelevant data only stand in the way of good decision-making.

Therefore, whenever possible, use data from your own company and with the most recent publication date available. Productivity data from your competitors cannot help you estimate how much profit your organisation will generate this quarter. And if you want to investigate the impact of a training programme, you will not be able to draw any conclusions without collecting new data after the training. Always try to make a trade-off based on the situation. Sometimes you want to examine as much data as possible to obtain an overall picture; at other times, you want to focus on very specific data to answer a specific question.

To what extent is your data, ultimately, usable as a guide for good decision-making? Do you have the right systems (or employees) that can read and process the data?

When it comes to data quality, you cannot cut corners on the basics. Inaccurate data lead to poor results. It is therefore essential to ensure that the data you rely on are worthy of that trust.

To improve the quality of your data, you can already take important aspects into account during data collection, such as accuracy, accessibility, compatibility, consistency, relevance and operability. Integrate these into your organisation and you will soon have an excellent, high-quality data programme.

Would you like even more inspiration to get started with HR analytics within your organisation? Then subscribe to our bi-monthly newsletter or read our book Reinventing HR-data.