Why Your Memory Makes Everything Seem More Average Than It Was
When seventeen-year-old Rohan from Mumbai received his mathematics exam results, he was thrilled—he’d scored 94 out of 100, his highest score ever in a notoriously difficult subject. He immediately called his parents to share the exciting news, posted about it on social media, and celebrated with friends. The score felt exceptional, and it was—only three students in the entire class had scored above 90.
Three months later, Rohan’s cousin asked him how he’d done on that challenging math exam. Without hesitation, Rohan replied: “I did pretty well—got around 85 or 87, I think. It was a good score.”
His cousin looked confused. “Didn’t you tell me you got 94? You were really excited about it. You said it was your best score ever.”
Rohan paused, genuinely uncertain. “Did I say 94? That seems really high. Are you sure? I remember it being a good score, but 94 seems too high. I think it was more like mid-80s, maybe touching 90. That sounds more realistic.”
When Rohan checked his actual report card, he was surprised to find his cousin was right—he had indeed scored 94. But in his memory, the score had somehow shrunk to the mid-80s. The exceptional achievement had become merely good in his mind.
Meanwhile, Rohan’s friend Priya had scored 52 on the same exam—a failing grade that had devastated her at the time. When asked about it three months later, Priya said: “I did poorly, got around 58 or 60, I think. It was just below passing.”
Her memory had inflated her actual score from 52 to nearly 60—still bad, but not quite as catastrophic as the reality had been.
Their psychology teacher later explained what had happened: “You’re both experiencing conservatism bias, also called regressive bias. This is the tendency for memory to pull extreme values—whether very high or very low—toward the average or middle range. Rohan, your exceptional 94 got remembered as merely good 85-87. Priya, your very poor 52 got remembered as less catastrophically bad 58-60. Both of your memories ‘regressed toward the mean’—moved away from the extremes toward more moderate, average values.”
She continued: “This happens with numbers, likelihoods, frequencies, and experiences. If something was extremely good, memory softens it to merely good. If something was extremely bad, memory softens it to merely bad. If you experienced something very rarely, memory makes it seem less rare. If something happened very frequently, memory makes it seem less frequent. Memories are not extreme enough—they’re pulled toward the middle, making your past seem more average and moderate than it actually was. This affects how you remember prices, grades, frequencies, risks, and virtually any quantifiable aspect of your past.”
This cognitive bias—the tendency for remembered values to be less extreme than actual values, pulled toward average or moderate levels—affects financial decisions, risk assessment, learning from experience, and any domain where accurate memory of quantities matters. Understanding conservatism bias reveals why we misremember prices, underestimate actual risks we’ve faced, overestimate how often rare events occurred, and generally remember a more moderate past than we actually experienced.
What Is Conservatism or Regressive Bias?
Conservatism bias (also called regressive bias) is the memory phenomenon where extreme values—whether high or low—are remembered as less extreme than they actually were, pulled toward average or moderate levels. High test scores are remembered as lower, low scores as higher; high prices as lower, low prices as higher; very frequent events as less frequent, very rare events as more frequent; high probabilities as lower, low probabilities as higher. The bias is “regressive” because it operates like statistical regression toward the mean: extreme actual values get “regressed” toward more moderate remembered values. Memories are systematically not extreme enough.
The phenomenon was identified by memory researchers studying numerical recall. Research at University of Michigan demonstrated that when people experienced events with various frequencies or values and later tried to recall those frequencies/values, their memories systematically showed regression toward the mean: they overestimated low values (remembering them as higher than they were) and underestimated high values (remembering them as lower than they were). The pattern held across different types of quantities—prices, scores, frequencies, probabilities.
According to studies from Carnegie Mellon University, conservatism bias operates because memory for specific values is imperfect and uncertain. When reconstructing a forgotten value, people draw on both the specific memory (which is noisy and uncertain) and their general knowledge about typical values in that domain. This combination pulls remembered values toward what’s typical or average: if you can’t remember an exact score but know most scores are in the 70-80 range, a very high actual score gets pulled down toward that typical range in memory.
Research from University of Pennsylvania demonstrates that conservatism bias is particularly strong when: (1) time has passed since the original experience (making specific memory weaker), (2) the value was truly extreme compared to typical values (making regression more noticeable), (3) you have strong beliefs about what’s typical in that domain (providing a strong attractor for regression), and (4) the value wasn’t particularly meaningful emotionally (reducing motivation for accurate encoding). These conditions make memory pull extreme values toward moderate remembered values.
The Parable of the Merchant and the Fading Extremes
A teaching tale tells of a traveling merchant who visited villages throughout a kingdom, recording prices of rice in each location. In one wealthy village, rice cost 50 rupees per kilogram—an extraordinarily high price due to recent floods. In a distant farming village, rice cost only 8 rupees per kilogram—an extraordinarily low price due to local surplus.
The merchant noted these prices carefully in his ledger and continued his travels. Six months later, a trader asked him: “What were rice prices like in your travels? I’m trying to understand the market.”
The merchant thought back. “Well, prices varied. In that wealthy village affected by floods, rice was quite expensive—maybe around 35 or 40 rupees per kilogram. And in that farming village with surplus, it was quite cheap—around 15 or 20 rupees per kilogram.”
His apprentice, who had been present during the original visits, corrected him: “Master, your ledger shows 50 rupees in the wealthy village and 8 rupees in the farming village. Your memory has changed both numbers.”
The merchant checked his ledger and was surprised. “That can’t be right. 50 rupees? That seems impossibly high. And 8 rupees? That seems impossibly low. Are you sure I recorded correctly?”
A wise mathematician who overheard the conversation explained: “Your memory is demonstrating regression toward the mean. You experienced genuine extremes—genuinely very high price and genuinely very low price. But in memory, extreme values drift toward the middle. Your mind knows that rice prices typically range from 20-30 rupees, so when your memory of the specific prices faded, your brain reconstructed them pulling toward that typical range. The extremely high 50 became moderately high 35-40. The extremely low 8 became moderately low 15-20.”
He continued: “This is why merchants need written records and why human memory alone is unreliable for business. If you trusted your memory rather than your ledger, you’d think market conditions were more moderate and stable than they actually are. You’d fail to appreciate that prices sometimes reach genuine extremes. Your memory gives you a false picture of a more average, less volatile market than reality.”
The merchant protested: “But I was there! I experienced those prices! How can I misremember something I directly observed?”
The mathematician replied: “Memory doesn’t work like a ledger that preserves exact values. Memory works through reconstruction, and reconstruction is influenced by what you believe is typical or plausible. Extreme experiences get moderated in memory because extreme values feel less plausible than moderate ones. Your mind doesn’t believe rice reached 50 rupees or dropped to 8 rupees, so it adjusts your memory to more ‘believable’ moderate values. This happens unconsciously—you’re not lying or being careless. Your memory system is systematically pulling extremes toward the middle.”
Buddhist philosophy addresses conservatism bias in teachings about the impermanence and uniqueness of each moment. Buddhist practice emphasizes that each experience is distinct and unrepeatable—not interchangeable with other similar experiences or reducible to averages. Conservatism bias represents mental laziness: instead of holding the specific reality of each moment (this rice actually cost 50 rupees this day in this place), the mind substitutes average expectations (rice “normally” costs around 30 rupees). Mindfulness practice trains seeing what actually is rather than what typically is.
The Bhagavad Gita discusses this through Krishna’s teaching about seeing reality as it is rather than through mental constructs and expectations. Krishna teaches that maya (illusion) includes imposing mental patterns onto unique reality. Conservatism bias is a form of this: imposing average patterns (“prices are usually around X”) onto specific extremes (“this price was actually 50”). Wisdom requires recognizing when reality genuinely deviates from averages rather than mentally forcing extremes back toward comfortable averages.
How Memory Pulls Extremes Toward the Middle
In academic performance and grade recall, conservatism bias makes students remember very high and very low grades as more moderate than they actually were. Research shows that when students recall past exam scores months or years later, they systematically pull extremes toward the middle: exceptionally high scores (95+) are remembered as merely good (85-90), and exceptionally low scores (below 50) are remembered as less catastrophically bad (55-60). Students remember a more moderate academic history than they actually had.
Studies from Ohio State University found that students asked to recall their GPA from earlier semesters showed clear regression toward the mean: students with very high GPAs (3.8-4.0) remembered them as lower (3.5-3.7), and students with very low GPAs (below 2.5) remembered them as higher (2.7-2.9). The bias compressed the range, making academic performance seem more moderate in memory than in reality.
In price memory and consumer purchase decisions, conservatism bias makes people remember very high prices as lower and very low prices as higher than they actually were. Research shows that when people try to recall what they paid for purchases weeks or months ago, their memories regress toward typical prices: if you paid an unusually high price, memory lowers it; if you got an exceptional deal, memory raises the price. This affects whether you think you’re getting good deals or being overcharged.
Studies demonstrate that consumers shown products at sale prices and regular prices, then asked weeks later to recall which price was which, systematically moved both prices toward the middle: the high regular price was remembered as lower, and the low sale price was remembered as higher. This makes past deals seem less exceptional and past high prices seem less extreme, affecting future purchase decisions and value judgments.
In frequency and probability estimation from experience, conservatism bias makes people remember very frequent events as less frequent and very rare events as more frequent than they actually were. Research shows that when people experience events occurring at various frequencies (daily, weekly, monthly, yearly, once ever) and later try to recall those frequencies, their memories regress toward moderate frequencies: very frequent becomes moderately frequent, very rare becomes moderately rare.
Studies from Duke University tracking how people remember symptom frequencies found that patients experiencing symptoms very frequently (multiple times daily) remembered them as occurring less frequently (a few times per week), while patients experiencing symptoms very rarely (once a month) remembered them as occurring more frequently (weekly). This affects medical reporting and diagnosis—patients give doctors a more moderate frequency picture than reality.
In risk and safety assessment from past experiences, conservatism bias makes people remember very risky situations as less risky and very safe situations as less safe than they actually were. Research shows that when people experience situations with varying risk levels and later recall those risk levels, their memories pull toward moderate risk: extreme danger is remembered as merely concerning, and extreme safety is remembered as merely safe rather than exceptionally secure.
Studies found that people recalling past driving situations systematically moderated extremes: a near-miss accident (very dangerous) was remembered as a close call but not as dangerous as it was, while an empty highway (very safe) was remembered as safe but not as exceptionally risk-free as it actually was. This affects future risk-taking—people underestimate the genuine danger they’ve faced and overestimate the risk in safe situations, compressing their risk experience toward moderate remembered risk.
In weather and temperature recall, conservatism bias makes people remember very hot days as less hot and very cold days as less cold than they actually were. Research shows that when people recall temperatures from past days, weeks, or months, their memories regress toward typical seasonal temperatures: record heat is remembered as hot but not record-breaking, record cold is remembered as cold but not as extreme as it was.
Studies from University of Illinois tracking temperature memory found that people shown actual daily temperatures and asked to recall them weeks later systematically pulled extremes toward seasonal averages: a 42°C day got remembered as 38°C, a 2°C day got remembered as 8°C. Both memories moved toward the typical range, making weather seem more moderate in retrospect than it actually was.
In pain intensity and medical symptom memory, conservatism bias affects how patients remember symptom severity. Research shows that when patients experience varying pain levels and later try to recall peak pain intensity, their memories regress toward moderate pain levels: extreme pain is remembered as severe but less intense than it actually was, while mild discomfort is remembered as slightly more uncomfortable than it actually was.
Studies demonstrate that patients asked to recall their worst pain episodes weeks later systematically underestimated peak pain intensity, pulling extreme pain toward more moderate remembered pain. This affects treatment decisions—patients tell doctors about more moderate symptoms than they actually experienced, potentially leading to under-treatment of genuinely extreme symptoms that memory has moderated.
Recognizing When Your Memory Moderates Extremes
The most important practice for countering conservatism bias is recognizing that when you recall extreme values from memory, the actual values were probably even more extreme. If you remember something being quite expensive, it was probably very expensive. If you remember something being quite cheap, it was probably extremely cheap. Memory pulls extremes toward the middle, so adjust your recalled values outward when accuracy matters.
Keep written records of important quantifiable information rather than trusting memory. Conservatism bias shows that memory for numbers, frequencies, and magnitudes is systematically inaccurate in predictable ways. If exact values matter (financial records, academic history, medical symptoms, business data), write them down when they occur rather than relying on later recall that will moderate extremes.
Before making decisions based on remembered quantities, acknowledge that your memory probably moderates extremes. If you’re thinking “last time this cost around X” or “this happened about Y times before,” recognize that if X was extreme, it was probably more extreme than you remember, and if Y was very high or very low, it was probably even higher or lower than memory suggests.
Be especially skeptical of memories that seem suspiciously moderate or average. Conservatism bias makes everything seem more moderate in retrospect. If you remember past values clustering around typical or average levels without much variation, consider whether reality actually had more extremes that memory has pulled toward the middle. True experience usually includes more extremes and variation than memory preserves.
Recognize patterns in your memory distortions: are you consistently remembering excellent performances as merely good? Remembering terrible outcomes as merely bad? Remembering rare events as less rare? These patterns reveal conservatism bias operating, suggesting your actual experiences were more extreme than your moderate memories suggest.
Remember Rohan whose exceptional 94 became merely good 85-87 in memory, and the merchant whose extreme rice prices of 50 and 8 became moderate 35-40 and 15-20 in memory. Both illustrate how conservatism bias pulls extremes toward the middle, creating false memories of a more moderate and average past than was actually experienced.
Conservatism bias can’t be eliminated because memory genuinely is uncertain and genuinely does combine specific recall with general knowledge about typical values. But recognizing the bias allows correction: when recalling extreme values, adjust outward—the actual values were probably even more extreme. When memory says “quite high,” reality was probably “very high.” When memory says “pretty low,” reality was probably “extremely low.” Memory’s moderation serves psychological comfort (extremes are uncomfortable) but distorts reality. For accurate understanding of your past and for making good decisions based on past experience, recognize that your memories are systematically not extreme enough and that reality was more varied and less moderate than your moderate memories suggest.
Frequently Asked Questions
Why would my brain make extreme values seem more moderate? Wouldn’t accurate memory be better?
Memory’s moderation might serve psychological functions: extreme experiences (whether good or bad) are emotionally intense, and moderating them in memory reduces emotional impact. Also, moderate values seem more “normal” and believable, making them psychologically easier to hold in memory than extreme values that seem unusual or exceptional. Memory serves psychological comfort sometimes more than accuracy.
Does this mean I should always distrust my memory of numbers and quantities?
Not always distrust, but recognize the systematic pattern: memory pulls extremes toward averages. If you remember moderate values, those might be accurate. But if you experienced genuine extremes, your memory probably moderated them. The bias is strongest for true extremes, so the more extreme an actual value was, the more likely memory has moderated it. Recent memories and emotionally important values are somewhat more accurate.
If I write something down immediately, will that prevent the bias?
Yes—written records at the time of experience capture actual values before memory has a chance to moderate them. Conservatism bias operates on memory reconstruction, not on immediate recording. This is why financial records, written test scores, medical charts, and business ledgers are reliable while memory alone isn’t for quantitative information. Written records bypass memory’s moderating tendency.
Does this affect all types of memory or just numbers and quantities?
Conservatism bias primarily affects quantifiable information—numbers, frequencies, magnitudes, probabilities, amounts. Qualitative memories (what something looked like, how you felt, what happened) are subject to different biases. But anything expressible as a value on a scale (how good/bad, how often, how much) is vulnerable to regression toward moderate values in memory.
Can I train myself to remember extreme values more accurately?
Difficult to train away because the bias reflects fundamental features of how memory reconstruction works (combining noisy specific memory with knowledge of typical values). Best strategy isn’t trying to improve memory but rather recognizing the bias exists and compensating: when recalling values you know were extreme, deliberately adjust your memory outward. Or better yet, keep written records of important quantities rather than relying on memory.
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.
