Information Seeking Behavior

Abstract

Information seeking behavior is a type of behavior: the active pursuit of knowledge to close a gap between what a person knows and what they want or need to know. Cognitive psychology studies it as a motivated, goal-directed process driven by curiosity and by the value the mind places on reducing uncertainty, spanning everything from a scientist designing an experiment to a reader following a link to resolve a nagging question. Curiosity arises from an information gap and follows an inverted-U over prior knowledge; foragers allocate effort across information patches and leave when yield falls below the environment's average; the brain values information much as it values reward, which is why people sometimes avoid information they could freely obtain. A worked example illustrates optimal patch-leaving, and three interactive demonstrations let the reader manipulate the core effects.

Keywords: information seeking behavior, curiosity, information foraging, information value

Information seeking behavior is the purposive activity a person undertakes to satisfy a need for knowledge — the term Medical Subject Headings adopted to index the searching for, acquiring, and using of information (Wilson, 1999). It is not a single act but a process with phases: a need is recognized, a search is initiated and refined, sources are consulted and evaluated, and the result either satisfies the need or renews the search. Cognitive psychology takes an interest because the behavior sits at the junction of motivation, attention, memory, and decision making: curiosity supplies the drive, attention selects where to look, an answer's expected value governs how far to pursue it, and the result updates what the seeker knows. It is studied with the discipline's formal tools — optimal-foraging models and the economics of choice under uncertainty.

What Information Seeking Behavior Is

Information seeking behavior is any goal-directed activity aimed at acquiring information to meet a recognized need, and as a category it is both broad and structured. It covers the deliberate, effortful search of a researcher combing a literature, the quick lookup that resolves a passing question, and the exploratory browsing of someone who cannot yet specify what they are looking for. What unifies these is a sequence rather than a single response: an influential account describes information behavior as beginning with a need arising from a gap in knowledge, proceeding through active seeking and the evaluation of sources, and ending either in use or in a renewed search when the need is unmet (Wilson, 1999). The process is affective as well as cognitive — studies of searchers tracking a task over days find that uncertainty and anxiety are highest early, before a focus forms, and give way to confidence as the search converges (Kuhlthau, 1991).

That structure matters for cognition because different phases tax the mind differently. A known-item lookup is a targeted retrieval that loads memory and selective attention briefly; an open-ended inquiry is sustained exploratory search that unfolds over many steps, interleaving searching, browsing, and sense-making as understanding grows (Marchionini, 2006). The seeker must repeatedly decide whether the information so far is enough, whether the current source is worth continuing with, and where to look next — decisions that make information seeking a problem of allocating limited effort under uncertainty rather than a passive intake of facts.

Figure 1

The Information Seeking Cycle

Table 1

Frameworks Used to Explain Information Seeking Behavior

Curiosity and the Information Gap

The motive that initiates information seeking is curiosity, and the oldest cognitive account of it ties curiosity to arousal. Berlyne distinguished epistemic curiosity — the drive for knowledge — from perceptual curiosity, and argued that stimuli provoke exploration in proportion to their novelty, complexity, and surprise, variables that raise arousal to a level the organism is motivated to resolve (Berlyne, 1954). The influential modern reformulation is the information-gap theory: curiosity arises when attention fixes on a gap between what one knows and what one wants to know, and the intensity of curiosity depends on the size and salience of that gap (Loewenstein, 1994). The theory explains a characteristic non-monotonicity — people are least curious about things they know nothing about and about things they know completely, and most curious in between, where a gap is both present and sharply felt.

This inverted-U relation between curiosity and prior knowledge is the point at which the older arousal account and the information-gap account agree, and it has been borne out in studies that manipulate how much a person knows before posing a question (Kidd & Hayden, 2015). The practical reading is that curiosity is not a fixed trait discharged at random but a state the environment can induce by exposing a gap: a question whose answer the seeker can almost, but not quite, supply is the one most likely to drive a search. The demonstration below lets the reader move along that curve.

Foraging for Information

Once a search is under way, the seeker faces a continuous allocation problem: how long to stay with the current source before moving on. Information foraging theory borrows the mathematics of optimal foraging from behavioral ecology to answer it, treating documents, sites, and search results as patches that yield information at a diminishing rate the longer one works them, and treating the cost of moving between patches as a travel time (Pirolli & Card, 1999). The governing principle is the marginal value theorem: a rational forager leaves a patch at the moment its instantaneous yield drops to the average yield available in the environment as a whole, because beyond that point the time spent would be more productive elsewhere. A seeker following cues — the information scent of a link or heading that promises relevant content — is trading off the richness of the current patch against the cost of finding a better one.

The same logic scales up to how people search collections and the web, where exploratory search interleaves querying, browsing, and sense-making rather than retrieving a single known item (Marchionini, 2006). It also connects information seeking to a general control problem the brain must solve: the trade-off between exploration, gathering new information that may pay off later, and exploitation, using what is already known to secure a present gain (Cohen et al., 2007). Staying in a patch is exploitation; leaving to sample a new one is exploration, and the forager's departure rule is one solution to arbitrating between them. The demonstration below implements a patch with diminishing returns and a travel cost, and marks the departure time that maximizes the overall rate of gain.

The Value of Information

Why does the brain bother to seek information at all? The emerging neuroscientific answer is that it assigns information a subjective value and computes that value using machinery that overlaps with the valuation of primary rewards. Single-neuron recordings showed that midbrain dopamine neurons — the classic reward-prediction signal — also signal a preference for advance information about an upcoming reward, firing more when a cue allows the animal to learn its fate sooner even though the information does not change the outcome (Bromberg-Martin & Hikosaka, 2009). Human neuroimaging extended this: the subjective value of acquiring non-instrumental information and the value of a reward are encoded in a common region of the brain's valuation network, so that wanting to know and wanting to have draw on a shared currency (Kobayashi & Hsu, 2019). A synthesis of this work frames information seeking as the output of circuits that integrate the expected reduction in uncertainty with the attentional systems that select what to sample (Gottlieb et al., 2013; Gottlieb & Oudeyer, 2018).

This valuation is why the demand for information tracks uncertainty: the expected value of resolving a question is greatest when its outcome is most uncertain, and smallest when the answer is already nearly known or nearly impossible — the same intermediate-uncertainty peak that curiosity shows. But the value the mind places on information is not always positive. People sometimes decline information they could have for free, because the anticipated answer threatens a cherished belief, a hopeful expectation, or their mood — the phenomenon of information avoidance (Golman et al., 2017). Whether a person chooses to know, the evidence suggests, depends on how they expect the information to bear on their beliefs, their feelings, and their ability to act, so that the decision to seek or avoid is itself a value computation (Sharot & Sunstein, 2020). The demonstration below lets the reader set the stakes and see when the value of knowing turns negative.

Worked Example

Information foraging theory makes a quantitative prediction about when to abandon a source, and working it through shows how the marginal value theorem turns a vague intuition into a rule. Imagine a seeker working a single dense source — a long review article — that yields relevant facts quickly at first and then more slowly as the easy material is exhausted. Suppose the useful information gained in each successive minute is 10, 6, 4, 2, and 1 units, a typical diminishing-returns profile. Reaching any new source costs a fixed travel time of 2 minutes spent formulating a query and scanning results.

The cumulative information after each minute in the patch is the running total: 10, then 10 + 6 = 16, then 16 + 4 = 20, then 20 + 2 = 22, then 22 + 1 = 23 units. The quantity a rational forager maximizes is the overall rate of gain, which charges each patch visit with its travel cost: rate = cumulative gain ÷ (time in patch + travel time). Evaluating that for each possible departure gives 10 ÷ 3 ≈ 3.33, 16 ÷ 4 = 4.00, 20 ÷ 5 = 4.00, 22 ÷ 6 ≈ 3.67, and 23 ÷ 7 ≈ 3.29 units per minute.

The overall rate peaks at 4.00 units per minute, reached by leaving after either the second or the third minute. The marginal value theorem gives the same answer more directly: stay while the information gained in the next minute exceeds the environment's average rate, and leave when it falls to that rate. The marginal gains are 10, 6, 4, 2, 1; the average rate is about 4; so the forager stays through the minute that yields 4 and leaves after the third minute, when the next minute would return only 2 — below the 4 available elsewhere. The lesson generalizes beyond the numbers: the optimal moment to stop reading and search anew is not when the source runs dry but when its marginal yield drops below what a fresh source would offer, which is almost always earlier than the point of exhaustion.

Discussion

The cognitive psychology of information seeking converges on a single organizing idea: seeking is a value-guided allocation of limited effort, not a passive accumulation of facts. Curiosity supplies a motive that is itself tuned to the structure of what the seeker knows, peaking where a gap is present and salient (Loewenstein, 1994); foraging models specify how effort should be spread across sources and when to move on (Pirolli & Card, 1999); and valuation circuits supply the common currency in which the expected worth of an answer is weighed against its cost (Kobayashi & Hsu, 2019). Each of these is a specific instance of the general principle that the mind treats information as a resource with costs and benefits, and allocates attention and action to acquire it in roughly the way an optimizing agent would — the principle that links information seeking to decision making and reward processing rather than isolating it as a topic of library science.

What complicates the optimizing picture is that seeking is not the mind's only option. The same valuation that drives a person toward information can drive them away from it when the expected answer is unwelcome, so that information avoidance is not irrational noise but the predictable output of a value computation that happens to come out negative (Golman et al., 2017; Sharot & Sunstein, 2020). This reframes the applied question. If seeking depends on the anticipated hedonic and instrumental consequences of knowing, then getting people to seek useful but threatening information — a medical test result, a financial reality — is less a matter of improving access than of changing the expected value of the knowledge itself. The foraging and valuation accounts, taken together, predict both the diligent searcher and the deliberate avoider from one set of principles.

Current Directions

A central current question is how the brain computes the value of information before the information is in hand, and how that computation is shared with, or distinct from, the valuation of extrinsic reward. The synthesis of behavioral, computational, and neural work argues that information seeking is best understood as active sampling: the organism chooses what to look at so as to reduce uncertainty, and the machinery that assigns value to the anticipated reduction is now a tractable target for recording and modeling (Gottlieb & Oudeyer, 2018). A closely related strand treats curiosity itself as a computational quantity — the expected information gain of an action — and seeks to explain why intrinsically motivated exploration looks the way it does in children and in machines (Kidd & Hayden, 2015).

A second active front concerns the determinants of information avoidance and its consequences. Field and laboratory work is mapping when people choose not to know, showing that avoidance is widespread, systematic, and sensitive to whether the information bears on beliefs people are motivated to protect (Golman et al., 2017). Formalizing the decision of what to find out — as a function of expected effects on cognition, emotion, and action — is an attempt to bring seeking and avoiding under a single account rather than treating the second as a failure of the first (Sharot & Sunstein, 2020). Whether the valuation of information and the valuation of reward are ultimately the same computation or merely neighboring ones remains, on current evidence, genuinely open.

Common Misconceptions

People are most curious about topics they know nothing about.

Curiosity follows an inverted-U over prior knowledge: it is weakest when one knows nothing and when one knows everything, and strongest in between, where an information gap is both present and salient (Loewenstein, 1994).

The best time to stop searching a source is when it runs out of relevant material.

The marginal value theorem says to leave when a source's next-minute yield drops below the average yield available elsewhere, which is almost always before the source is exhausted (Pirolli & Card, 1999).

People always want more information when they can get it for free.

People often avoid freely available information when the expected answer threatens their beliefs, hopes, or mood; the decision to know is itself a value computation that can come out negative (Golman et al., 2017).

Glossary

Active Sampling.

The selection of what to look at or attend to so as to reduce uncertainty; a framing of information seeking as a choice the organism makes to gather evidence.

Curiosity.

The motivational state that drives information seeking; in gap theory it arises from a felt discrepancy between current and desired knowledge.

Decision Making.

The selection among options under uncertainty; information seeking is a sequence of decisions about whether to continue, which source to use, and where to look next.

Epistemic Curiosity.

Curiosity directed at knowledge and ideas, distinguished by Berlyne from perceptual curiosity directed at sensory stimuli.

Exploration-Exploitation Trade-off.

The tension between gathering new information that may pay off later and using current knowledge for a present gain; leaving a patch is exploration, staying is exploitation.

Exploratory Search.

Open-ended searching that interleaves querying, browsing, and sense-making, used when the seeker cannot yet specify a single known item.

Information Avoidance.

The deliberate choice not to acquire freely available information, typically when the expected answer threatens beliefs, hopes, or mood.

Information Foraging.

A theory that models information seeking on optimal foraging, treating sources as patches with diminishing yield and movement between them as costly travel.

Information Gap.

The discrepancy between what a person knows and what they want to know; its size and salience set the intensity of curiosity.

Information Scent.

The cues — a link, a heading, a snippet — by which a seeker judges whether a path leads to relevant information, guiding patch choice in foraging.

Marginal Value Theorem.

The optimal-foraging rule that an agent should leave a patch when its instantaneous yield falls to the average yield of the environment.

Non-Instrumental Information.

Information that carries subjective value even though it cannot change the outcome; the brain will pay for advance knowledge of a fixed result.

Reward Processing.

The valuation and motivational machinery that assigns worth to outcomes; it overlaps with the circuitry that values information.

Uncertainty.

The state of incomplete knowledge about an outcome; the expected value of information is greatest when uncertainty is high and the answer most resolvable.

Key Researchers

Daniel E. Berlyne

(1924-1976). Psychologist at the University of Toronto who founded the experimental study of curiosity, exploratory behavior, and arousal, and drew the enduring distinction between epistemic and perceptual curiosity.

Wikipedia

Jacqueline Gottlieb

. Neuroscientist at Columbia University's Zuckerman Institute; her work on attention and active sampling links curiosity and information seeking to the brain's mechanisms of valuation and selection.

ORCID - Google Scholar

Celeste Kidd

. Cognitive scientist at the University of California, Berkeley; she studies curiosity, attention, and learning, including how prior knowledge shapes what children and adults choose to explore.

Google Scholar - Wikipedia

George Loewenstein

. Professor of economics and psychology at Carnegie Mellon University; he formulated the information-gap theory of curiosity and, later, the analysis of information avoidance.

Google Scholar - Wikipedia

Peter Pirolli

. Cognitive scientist at the Institute for Human and Machine Cognition; with Stuart Card at Xerox PARC he developed information foraging theory, applying optimal-foraging mathematics to how people search for information.

Wikipedia

Tali Sharot

. Neuroscientist at University College London and MIT; her research on belief updating and the optimism bias extends to how people decide what they do and do not want to know.

ORCID - Google Scholar

Frequently Asked Questions

What is information seeking behavior in cognitive psychology?

It is the motivated, goal-directed pursuit of knowledge to close a gap between what a person knows and what they need or want to know, studied as a process of allocating limited attention and effort under uncertainty rather than as a passive intake of facts (Wilson, 1999).

What drives a person to seek information?

Curiosity does, and gap theory holds that curiosity arises when attention fixes on a discrepancy between current and desired knowledge; the larger and more salient the gap, the stronger the drive to resolve it (Loewenstein, 1994).

Why am I most curious about things I half-know?

Curiosity follows an inverted-U over prior knowledge: it is low when one knows nothing and low when one knows everything, and highest in between, where a gap is present and sharply felt (Kidd & Hayden, 2015).

What is information foraging theory?

It models information seeking on animal foraging, treating sources as patches that yield information at a diminishing rate and treating movement between them as costly, so a seeker should leave a source when its yield drops below the average available elsewhere (Pirolli & Card, 1999).

When should I stop reading one source and look for another?

The marginal value theorem says to leave when the information gained from the next minute in the current source falls below the average rate obtainable from a fresh source, which is typically before the source is exhausted (Pirolli & Card, 1999).

Does the brain treat information like a reward?

Largely yes: dopamine neurons signal a preference for advance information, and the subjective value of information and of reward are encoded in a shared region of the valuation network, so wanting to know and wanting to have draw on a common currency (Bromberg-Martin & Hikosaka, 2009; Kobayashi & Hsu, 2019).

Why do people sometimes avoid information they could get for free?

Because the value of knowing can be negative: when an expected answer threatens a belief, a hope, or one's mood, people deliberately avoid it, a systematic phenomenon called information avoidance (Golman et al., 2017).

How does information seeking relate to the exploration-exploitation trade-off?

Seeking new information is a form of exploration, and using what one already knows is exploitation; deciding whether to keep searching or act on current knowledge is one instance of the general control problem the brain must arbitrate (Cohen et al., 2007).

References