Passing Judgement: The Use of Implicit Bias in Shaping Metadata Narratives
Updated: Mar 30
As a lateral thinker, I often imagine what our society would look like if we had to re-create it based solely on the metadata of our current state. While data has the ability to be proven through evidence-based methods; metadata, most often does not. In most instances, metadata is the metaphorical 'horse' before the 'cart' of data. As ambiguous and unreliable it has proven to be in the past decade, it remains the most widely used and accepted source of data for systems of accountability in the United States. To date, there is no proven scientific technique or method available to measure the reliability of metadata. The featured article in this month's blog, focuses on the dangers of using implicit-bias to shape metadata-driven narratives in an emerging technological society.
Data remains fixed as the anchor of quantitative evidentiary proof in support of an often skewed narrative. Opine-based metadata creates a context that gives meaning and dimensionality to fixed binary data. For those who have heard the phrase: "the devil is in the detail", well the 'detail' is undoubtedly ambiguous metadata. So what is metadata and why is it so powerful in the real world?
According to ASA-member Cathryn S. Dippo of the U. S. Bureau of Labor Statistics and Bo Sundgren, Statistics Sweden' s research in The Role of Metadata in Statistics; metadata provides the basis for human understanding of data. In recognizing the significance of the role metadata plays in the use of statistical information, Dippo and Sundgren postulated on the view of metadata management as an 'integral part of statistics production'. To provide a real-world picture of how meta-data is used and misused in today's society, I offer an example:
Annual Job Performance & Performance Improvement Plans (PIP)
Despite their hard work, impeccable attendance, and loyalty, every year thousands of employees are required to participate in job performance evaluations. While many employers apply the evaluative criteria within the structured annual assessment according to its intended use; there are others who wield this evaluative tool like a sword in King Arthur's Court. Smitten for the opportunity to cut a subordinate down, these tools for constructive feedback are instead transformed into weapons and used to create a segway for progressive discipline, termination, and withholding of well-deserved bonuses. One way in which metadata is widely misused, is during the appraisal reporting process in assessing how work is performed.
There was once a woman known as "Amazing" Grace Hopper and she served in the capacity of Rear Admiral for the U.S. Navy. In addition service to her Country, she is equally as famous for preparing the U.S. for technological advancement through use of a new form of communication in the language of business known as: 'common business-oriented language', or COBOL for short. Ms. Hopper; tired of the long, inefficient way she was told how she had to do things, came up with an easier one. As the "Mother of COBOL", Grace Hopper coined the phrase: The most dangerous phrase in the human language: "We've always done it this way." Click Here to Read More About Grace Hopper
Ms. Hopper's phrase means there is more than one way to get a job done and just because someone has done it one way (a million times), that does not make their way the only way or the 'best' way. In the management of human resources, I refer to these gray areas of how as the space mistakenly created for 'opine extra'. An 'opine extra' is the space of ambiguity in which implicitly biased metadata is inserted by the operator into an otherwise ethical system of reporting. Adverse metadata or the 'opine extra' in recordation of employee performance, is the biased data, applied to an employee record in order to show an adverse pattern of behavior over time. It is the 'monkey wrench', the 'cog in the wheel', or any other phrase commonly used that refers to the manipulation of an otherwise sound reporting system.
There are numerous areas within the employment process where implicit bias has crept into the framework of an otherwise sound system. The example below, illustrates a snapshot of how biased metadata was used against an employee to construct a false narrative regarding her job performance. The false performance narrative was then used to justify issuing a performance improvement plan (PIP) and subsequent proposed termination:
I have a friend who after working over a decade for a highly reputable employer, decided to transfer within her Company to another location. Upon moving to another part of the Country, she began experiencing mistreatment by her new colleagues. As a loyal and dedicated employee, she brought efficiency and profit to her employer for over a decade. Prior to her voluntary transfer, she was responsible annually for bringing nearly $1 million in profit to her employer. Unfortunately- despite her impressive 10+ year employment record, her confidence and professionalism was not welcome at the new location, and neither was her 'yankee' accent. Because she spoke with a strong northern accentuated emphasis on certain words, she was told that she sounded rude and needed to change her 'tone'. After several months at her new location, unable to change the way her voice sounded [as it sounded that way her entire life and was part of her identity], it came time for her annual performance appraisal. For the first-time in her 10+ years of employment, she was issued a poor performance rating in nearly every area based on her voice. At the conclusion of her Evaluation, she was placed on a PIP and threatened with termination.
While the story above is limited to one person, history is littered with examples of the use of false judgements and schadenfreude in the collection and application of metadata to manipulate otherwise sound reporting systems. People throughout time have used implicit- bias, wrapped in the falsehood of ambiguity to create metadata for use in the application of context. The false narrative that emerges from the 'opine extra', is not only damaging to its victims, it creates priori for business exposure to operational risks associated with complaints of discrimination, compliance violations, and economic loss.
Until adverse effects of the collection and use of metadata to drive false-narratives is fully realized, it will remain a danger to the legacy of future generations. Society should rethink the way in which it ethically collects, stores, and uses multi-dimensional data in decision-making. Who knows- maybe one-day, we will have to reconstruct our entire society based solely on Tweets.