{"id":4086,"date":"2020-10-14T14:49:09","date_gmt":"2020-10-14T14:49:09","guid":{"rendered":"https:\/\/data-science.gotoauthority.com\/2020\/10\/14\/goodharts-law-for-data-science-and-what-happens-when-a-measure-becomes-a-target\/"},"modified":"2020-10-14T14:49:09","modified_gmt":"2020-10-14T14:49:09","slug":"goodharts-law-for-data-science-and-what-happens-when-a-measure-becomes-a-target","status":"publish","type":"post","link":"https:\/\/wealthrevelation.com\/data-science\/2020\/10\/14\/goodharts-law-for-data-science-and-what-happens-when-a-measure-becomes-a-target\/","title":{"rendered":"Goodhart\u2019s Law for Data Science and what happens when a measure becomes a target?"},"content":{"rendered":"
\n

By Jamil Mirabito<\/a>, U. of Chicago & NYC Flatiron School<\/b>.<\/p>\n

In 2002, President Bush signed into law\u00a0No Child Left Behind<\/a>\u00a0(NCLB), which was an education policy stating that all schools receiving public funding must administer an annual standardized assessment to their students. One of the stipulations of the law required that schools make adequate yearly progress (AYP) on standardized assessments year over year (i.e., third grade students taking an assessment in the current year would have had to perform better than third grade students in the previous year\u2019s cohort). If schools were continuously unable to meet AYP requirements, there were drastic consequences, including school restructuring and school closure. As such, many district administrators developed internal policies requiring that teachers increase their students\u2019 test scores, using these scores as a metric for teacher quality. Eventually, with their jobs on the line, teachers began to \u201cteach to the test<\/a>.\u201d In fact, a policy of this sort inadvertently\u00a0incentivized<\/em>\u00a0cheating so that teachers and whole school systems could maintain necessary funding. One of the most prominent cases of alleged cheating was the\u00a0Atlanta Public Schools cheating scandal<\/a>.<\/p>\n

Unintended consequences of this sort are actually very common. Charles Goodhart, a British economist, once said, \u201cWhen a measure becomes a target, it ceases to be a good measure.\u201d This statement, known as\u00a0Goodhart\u2019s Law,<\/em>\u00a0can actually be applied to a number of real-world scenarios beyond just social policies and economics.<\/p>\n

<\/p>\n

Source: Jono Hey,\u00a0Sketchplanations<\/a>\u00a0(CC BY-NC 3.0).<\/em><\/p>\n

Another commonly cited example is a call center manager setting a target to increase the number of calls taken at the center each day. Eventually, call center employees increase their numbers at the cost of actual customer satisfaction. In observing employees\u2019 conversations, the manager notices that some employees are rushing to end the call without ensuring that the customer is fully satisfied. This example, as well as the accountability measures of No Child Left Behind, stresses one of the most important elements of Goodhart\u2019s Law \u2014\u00a0targets can and will be gamed.<\/strong><\/p>\n

<\/p>\n

Source: Szabo Viktor,\u00a0Unsplash<\/a>.<\/em><\/p>\n

The threat of gaming is much greater when considering how AI and machine learning models may be susceptible to gaming and\/or intrusion. A\u00a02019 analysis<\/a>\u00a0of 84,695 videos from YouTube found that a video by\u00a0Russia Today<\/em>, a state-owned media outlet, had been recommended by over 200 channels, far exceeding the number of recommendations that other videos on YouTube get on average. The findings from the analysis were suggestive that Russia, in some way, gamed YouTube\u2019s algorithm to propagate false information on the internet. The problem is further exacerbated by the platform\u2019s reliance on viewership as a metric for user satisfaction. This created the unintended consequence of\u00a0incentivizing conspiracy theories<\/em><\/a>\u00a0about the unreliability and dishonesty of major media institutions so that users would continue to source their information from YouTube.<\/p>\n

\n

\u201cThe question before us is the ethics of leading people down hateful rabbit holes full of misinformation and lies at scale just because it works to increase the time people spend on the site \u2014 and it does work.\u201d \u2014\u00a0Zeynep Tufekci<\/a><\/em><\/p>\n<\/blockquote>\n

So what can be done?<\/h3>\n

In this vein, it\u2019s important to think critically about how to effectively measure and achieve desired outcomes in a way that minimizes unintended consequences. A large part of this is not relying too heavily on a single metric. Rather, understanding how a\u00a0combination<\/em>\u00a0of variables can influence a target variable or outcome could help to better contextualize data.\u00a0Chris Wiggins<\/a>, Chief Data Scientist at the New York Times, provides\u00a0four useful steps<\/a>\u00a0for creating ethical computer algorithms to avoid harmful outcomes:<\/p>\n

    \n
  1. Start by defining your principles. I\u2019d suggest [five in particular<\/a>], which are informed by the collective research of the authors of the\u00a0Belmont<\/a>\u00a0and\u00a0Menlo<\/a>\u00a0reports on ethics in research, augmented by a concern for the safety of the users of a product. The choice is important, as is the choice to define, in advance, the principles which guide your company, from the high-level corporate goals to the individual product key performance indicators (KPIs) [or metrics].<\/em><\/li>\n
  2. Next: before optimizing a KPI, consider how this KPI would or would not align with your principles. Now document that and communicate, at least internally if not externally, to users or simply online.<\/em><\/li>\n
  3. Next: monitor user experience, both\u00a0quantitatively\u00a0<\/strong>and\u00a0qualitatively<\/strong>. Consider what unexpected user experiences you observe and how, irrespective of whether your KPIs are improving, your principles are challenged.<\/em><\/li>\n
  4. Repeat: these conflicts are opportunities to learn and grow as a company: how do we re-think our KPIs to align with our objectives and key results (OKRs), which should derive from our principles?\u00a0If you find yourself saying that one of your metrics is the \u201cde facto\u201d goal, then you\u2019re doing it wrong.<\/strong><\/em><\/li>\n<\/ol>\n

    Original<\/a>. Reposted with permission.<\/p>\n

    \u00a0<\/p>\n

    Bio:<\/strong> Jamil Mirabito<\/a> is a Project Manager, Poverty Lab at the University of Chicago and a Data Science Student at Flatiron School in NYC.<\/p>\n

    Related:<\/b><\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

    https:\/\/www.kdnuggets.com\/2020\/10\/goodharts-law-data-science-measure-target.html<\/p>\n","protected":false},"author":0,"featured_media":4087,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[2],"tags":[],"_links":{"self":[{"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/posts\/4086"}],"collection":[{"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/comments?post=4086"}],"version-history":[{"count":0,"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/posts\/4086\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/media\/4087"}],"wp:attachment":[{"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/media?parent=4086"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/categories?post=4086"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/wealthrevelation.com\/data-science\/wp-json\/wp\/v2\/tags?post=4086"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}