Why Schools Chase Test Scores Instead of Real Learning: The Surrogation Trap
Green Valley High School in Delhi had a problem. For five years running, their Class 10 board exam scores had been declining. Parents complained, the school board demanded improvement, and the principal’s job was on the line. The board exam scores—originally meant to measure how well students were learning—had become the only thing that mattered.
The principal called an emergency meeting. “We need to improve our board exam scores immediately,” he announced. Teachers understood: scores were the priority, not actual learning. What followed was predictable but tragic.
The school eliminated art, music, physical education, and project-based learning—anything not directly tested on board exams. Teachers stopped teaching concepts and started drilling test-taking strategies. “Don’t waste time understanding why,” one teacher told seventeen-year-old Priya’s math class. “Just memorize the steps that get you the right answer on the exam.”
Classes became endless practice tests. Teachers taught to the test format, covering only topics likely to appear, skipping “low-yield” material even when it was foundational. Students spent hours learning to identify which multiple-choice tricks examiners use and how to write answers in the exact format that maximized marks.
The strategy worked—board exam scores improved dramatically. The principal celebrated, parents were satisfied, and the school board praised the turnaround. But something fundamental had been lost.
Priya noticed the problem during a family discussion. “I got 95% on my board exams,” she said, “but I don’t actually understand most of what I was tested on. I memorized procedures and formats, but I couldn’t apply the knowledge to real problems or explain the concepts to someone else. The school was so focused on raising test scores that they forgot test scores are supposed to measure learning. We got better at taking tests, but we didn’t actually learn more—we might have learned less because we abandoned everything that makes education meaningful. The measure became more important than what it was supposed to measure.”
This phenomenon—losing sight of what a measure represents and treating the measure itself as the goal—affects education, business, healthcare, and countless domains where we use measurements to track important outcomes. Understanding surrogation reveals why focusing on metrics often undermines the very goals those metrics were created to achieve.
What Is Surrogation?
Surrogation, also called the “measure fixation” or “surrogacy trap,” is the cognitive error where people lose sight of the underlying construct that a measure is meant to represent and begin treating the measure itself as the actual goal. Originally, the measure was created to track progress toward a real objective. Through surrogation, the measure replaces the objective in people’s minds and behaviors, leading them to optimize the measure even when doing so undermines the original objective.
The phenomenon has been studied across organizational behavior and management. Research at Harvard Business School examining performance measurement systems found that organizations consistently fall into surrogation—metrics introduced to track strategic objectives gradually replace those objectives in decision-making, with people pursuing metric improvement even when it conflicts with strategic goals the metrics were supposed to serve.
According to studies from University of Michigan, surrogation operates through several mechanisms: cognitive substitution (complex constructs get replaced by simpler measurable proxies), accountability pressure (people are evaluated on measures rather than on underlying constructs), and loss of institutional memory (over time, people forget why measures were created and what they represent). These combine to transform measures from tools into goals.
Research from Stanford University demonstrates that surrogation is particularly likely when: (1) the underlying construct is abstract or difficult to observe directly, (2) the measure is concrete and easy to quantify, (3) there’s strong accountability or reward for improving the measure, and (4) time passes between measure creation and current use, allowing the original purpose to be forgotten. These conditions make surrogation nearly inevitable in organizations relying heavily on quantitative metrics.
The Farmer Who Counted Seeds Instead of Growing Crops
A parable tells of a farming village where harvest yields had been declining. Village elders decided to measure progress by counting seeds planted—reasoning that more seeds planted should lead to larger harvests. They began tracking “seeds planted per farmer” as their key metric and evaluating farmers based on this number.
The system initially made sense—farmers who planted more seeds generally produced more crops. But gradually, something changed. Farmers realized they were judged purely on seed count, not harvest quality or quantity. The measure—seeds planted—replaced the actual goal—successful harvest.
Clever farmers found ways to maximize their seed count: planting seeds more densely than optimal, planting in poor soil where seeds were wasted, planting late in season when seeds wouldn’t mature, even planting non-viable seeds that would never grow. Their seed counts skyrocketed. The elders celebrated the “improvement.”
But harvests continued declining. Despite dramatic increases in seeds planted, actual crop yields fell because farmers were optimizing the measure (seed count) rather than the underlying goal (harvest yield). They’d forgotten—or found it easier to ignore—that seed count was supposed to be a proxy for harvest success, not a goal in itself.
A wise farmer who’d resisted the seed-counting obsession explained the problem: “You created a measure to track progress toward good harvests. But the measure became the goal. Now we plant seeds not to grow crops but to raise our seed count. The measure has replaced what it was supposed to measure. We’re succeeding at the metric while failing at the actual objective. This is the danger of confusing the map with the territory, the measure with the goal.”
Buddhist philosophy addresses surrogation in teachings about attachment to concepts rather than reality. The Buddha taught that humans create concepts and measures to understand reality, then become attached to those concepts, treating them as reality itself rather than as imperfect representations. Surrogation represents this attachment—treating measures (concepts) as if they were the underlying reality those measures were created to track.
The Bhagavad Gita discusses this through Krishna’s teaching about acting for the right reasons versus acting for the rewards. Krishna warns that focusing on outcomes or measures (rewards, metrics) rather than on dharma (right action, true purpose) leads to distorted action. Surrogation represents exactly this distortion—pursuing measures rather than the purposes those measures should serve.
How Metric Fixation Undermines Real Goals
In education and standardized testing, surrogation explains “teaching to the test”—schools optimize test scores even when doing so undermines actual learning. Research shows that test score pressure leads schools to narrow curriculum to tested subjects, eliminate subjects not tested (art, music, physical education), reduce deep conceptual teaching in favor of test preparation, and sometimes cheat or manipulate testing conditions. Test scores—measures of learning—replace learning as the actual goal.
Studies from MIT tracking schools under high-stakes testing pressure found that increased focus on test scores correlated with decreased student outcomes in non-tested areas (creativity, critical thinking, problem-solving in novel contexts) and sometimes even in understanding of tested material beyond the specific test formats. Schools got better at producing high test scores while students learned less comprehensively—classic surrogation.
In business and corporate metrics, surrogation causes managers to optimize measured KPIs even when doing so undermines business strategy. Salespeople meet quota by pushing low-margin products or overselling customers in ways that damage long-term relationships. Executives cut R&D spending to boost quarterly earnings at expense of innovation. Customer service representatives minimize call time (a measured metric) by providing poor service that saves time but loses customers.
Research demonstrates that strong financial incentives tied to specific metrics increase surrogation—people optimize the measured metric increasingly divorced from the strategic goal it was meant to represent. Organizations sacrifice long-term strategic health to improve short-term measured metrics.
In healthcare and patient outcomes, surrogation makes hospitals and doctors optimize billing codes, procedure counts, and other measured metrics rather than actual patient health. Doctors order tests and procedures that increase billable activity without improving patient outcomes. Hospitals focus on metrics like “patient throughput” and “bed turnover” rather than on healing quality and patient wellbeing. The measures meant to track healthcare quality become the focus at expense of actual quality.
Studies show that healthcare organizations under strong metric-based accountability often game the system—selecting healthier patients to improve measured outcomes, documenting more severe diagnoses to justify lower outcome rates, focusing resources on measurable conditions while neglecting unmeasured problems. The measurement system drives behavior away from the goal (patient health) toward the measure (documented metrics).
In scientific research and publication metrics, surrogation makes researchers optimize publication counts, citation metrics, and impact factors rather than pursuing important questions and rigorous science. Scientists split results across multiple papers to increase publication count, pursue trendy topics likely to be cited rather than important but unfashionable questions, and prioritize getting published over getting things right. Measures of research productivity replace actual research quality as the goal.
Research on academic incentives shows clear surrogation—as publication metrics become more important for career advancement, research quality (replication rates, true discovery rates) declines while publication counts increase. Scientists optimize the measures while the underlying goal (scientific knowledge) suffers.
Remembering What Measures Are Measuring
The most important practice for avoiding surrogation is regularly asking: “What is this measure supposed to represent? Am I pursuing the measure or the underlying goal?” When you find yourself optimizing a metric, pause and reconnect with the actual strategic objective the metric was created to track. This conscious reconnection prevents the metric from replacing the goal.
Evaluate whether improvements in measures correspond to improvements in underlying goals. If test scores rise but students aren’t learning more, surrogation is occurring. If sales quotas are met but customer satisfaction falls, surrogation is occurring. If publication counts increase but research quality decreases, surrogation is occurring. Checking whether measure and goal move together reveals when surrogation has divorced them.
Use multiple diverse measures for important constructs rather than single metrics. Surrogation is harder when multiple different measures must all improve together. Student learning requires multiple assessments—tests, projects, demonstrations, applications. Using only one measure makes it easy to optimize that specific measure while neglecting the underlying construct. Multiple measures keep focus on the construct rather than any single measurement.
Build in qualitative assessment alongside quantitative metrics. Numbers enable surrogation by making optimization concrete and obvious. Combining quantitative metrics with qualitative judgment (expert review, case studies, narrative assessment) keeps focus on the underlying reality that numbers imperfectly represent. The numbers inform judgment; judgment prevents surrogation.
Create organizational memory about why measures exist and what they represent. Document the strategic purpose each measure serves. Train people on the distinction between improving measures and improving the underlying goals. Periodic review of whether current measures still serve original purposes prevents gradual drift into surrogation where measures outlive their usefulness but persist as goals.
Watch for perverse incentives and unintended consequences of measurement. When people are strongly incentivized to improve a measure, surrogation becomes likely—the measure becomes what people care about rather than what it represents. Designing incentive systems that reward underlying goals rather than just measured proxies reduces surrogation risk.
Remember Priya’s school that improved test scores while reducing actual learning, and the village farmers who planted more seeds while harvests declined. Both demonstrate surrogation—measures created to track important outcomes (learning, harvest) became goals themselves, optimized at the expense of what they were supposed to measure.
Surrogation isn’t malicious or stupid—it’s natural cognitive drift that occurs when dealing with abstract goals and concrete measures. The underlying construct (learning, health, research quality) is complex, abstract, hard to observe directly. The measure (test scores, procedure counts, publication numbers) is simple, concrete, easy to track. Under pressure to show improvement and accountability, people naturally focus on what’s measured because it’s tangible and clear. Through this natural process, the measure gradually replaces the construct, and optimization becomes divorced from purpose.
Breaking surrogation requires conscious effort to maintain connection between measures and meaning, to remember that metrics are tools for understanding goals rather than goals themselves, and to evaluate whether metric improvements represent real progress toward underlying objectives or just gaming of measurement systems. The number going up doesn’t mean you’re succeeding if you’ve lost track of what success actually means. Surrogation is succeeding at the measure while failing at what matters.
Frequently Asked Questions
How is surrogation different from just focusing on what’s measurable?
Focusing on what’s measurable is reasonable when measures accurately represent underlying goals. Surrogation is when the measure replaces the goal in people’s minds and behaviors, leading them to optimize measures even when doing so undermines the goals measures were supposed to represent. The difference is whether measures remain tools for tracking goals or become goals themselves that are pursued independently of whether they serve original purposes.
Why don’t organizations just measure the real goal directly instead of using proxies?
Because many important goals are abstract, long-term, or difficult to observe directly. “Student learning” is complex and multi-dimensional. “Research quality” requires expert judgment and long time horizons. “Customer satisfaction” involves subjective experience. Measures are created as proxies because direct measurement is impractical. The problem isn’t using measures—it’s losing sight of what they represent and treating them as goals themselves.
Can surrogation be prevented, or is it inevitable when using metrics?
It’s very common but can be reduced through: using multiple diverse measures, combining quantitative metrics with qualitative judgment, regularly reconnecting measures to strategic goals, avoiding strong incentives tied to single metrics, and maintaining organizational memory about what measures represent. Surrogation is hard to prevent entirely but awareness and careful measurement system design significantly reduce it.
Are some types of measures more prone to causing surrogation than others?
Yes—narrow quantitative measures of complex constructs are highest risk. “Test scores” for “learning,” “billable hours” for “legal service quality,” “lines of code” for “programming productivity”—these are narrow measures of broad constructs, making surrogation likely. Measures that more comprehensively capture underlying constructs (multiple diverse assessments, holistic evaluation) are less prone to surrogation because gaming one measure doesn’t optimize the full measure set.
What should I do if I notice surrogation in my organization or school?
Raise awareness by highlighting the gap between measure and goal—showing cases where the measure improved but the underlying goal suffered. Advocate for multiple diverse measures rather than single metrics. Push for qualitative assessment alongside quantitative metrics. Document and communicate what measures are supposed to represent. Question whether metric improvement represents real progress or gaming. Challenging surrogation requires courage because people invested in current measures resist acknowledging that “success” on measures doesn’t mean success on goals.
Observer Voice is the one stop site for National, International news, Sports, Editor’s Choice, Art/culture contents, Quotes and much more. We also cover historical contents. Historical contents includes World History, Indian History, and what happened today. The website also covers Entertainment across the India and World.
