Data Gathering Practices

The term “big data” refers to substantial datasets that are analyzed using computer algorithms to extract trends and patterns. This cybernated means of processing data was intended to handle the massive amounts of data that have arisen in our modern, digitized world; however, as time has passed, it has become clear that having a lot of data does not necessarily mean the data is representative or that it is adequately utilized by government or organizational entities. 
Whether intentional or not, big data by its very architecture is designed to reproduce existing biases, allowing systemic prejudice to enter into large datasets. 
Marginalized groups generate less data for a variety of reasons, including unequal access to the resources necessary to engage in data-gathering practices. Therefore, the minimal amount of data on these underrepresented populations is not reliable, yet it continues to be circulated within big data analysis, giving rise to new manifestations of inequality that are difficult to recognize. Over the long term, this can lead to data outputs that are outright discriminatory. In particular, racial minorities and women are at risk due to persistent underrepresentation; this unreliable information originates from historical data that has not been adjusted to account for our modern understanding of data-gathering flaws and awareness of bias. Conclusions based on inequitable or partial data don’t just paint an undependable picture, but also have enduring consequences; this is particularly impactful in healthcare, where misguided data analysis can prompt inaccurate decision-making by clinicians and healthcare policymakers.  This leads to increased medical costs, patient mortality, decreased funding, and health worker shortages, along with other unquantified drawbacks.
“Hidden data gaps” is a term to describe these datasets that are routinely utilized for policymaking, but contain misrepresentation or missing data, without governments being aware.
An algorithm is only as good as the data it works with and inappropriate modeling produces inaccurate predictions. Part of our mission is to conduct and disseminate research on AMTCs in order to advance their role in the global health workforce, and we believe this is only possible when we are collecting and employing valid, current, and inclusive data, as well as staying aware of the pitfalls of big data and avoiding hidden data gaps. We are dedicated to recognizing the individual humans that numbers and figures represent, and using a multitude of sources to ensure our understanding of a situation is true to life and represented in our data.
We embrace traditional statistical techniques and small data curation approaches that have been seen to significantly improve healthcare outcomes; we believe that by applying this methodology, we can bring accurate and effective awareness to the often overlooked or misunderstood AMTC workforce. We value the safety of our members and strive to remain informed on the benefits and drawbacks of current data practices, so that we can maintain responsible data gathering methods.

​

​

​