✦ScienexySCIENCE KNOWLEDGE LIBRARY
learning-resource · SCI-PSY-002-061

Prediction of Behaviour

Prediction is the psychological goal of estimating the likelihood of future behaviour or mental outcomes from established patterns, measured variables, and scientific evidence. It is probabilistic and conditional rather than a guarantee of what an individual will do.

Explanation

A fuller explanation is being developed. Verify this concept with cited sources before relying on it.

Example or application

A worked example has not yet been added.

Conceptual Coverage

Definition

Psychological prediction is the use of information about current or previous conditions to estimate the probability, timing, or form of future behaviour and psychological outcomes.

Purpose

Prediction helps psychologists anticipate patterns, evaluate theories, identify possible support needs, and make informed decisions in educational, clinical, organizational, and research settings.

Types

TypeDescription

Individual prediction

Estimating a future outcome for a particular person using relevant evidence and validated methods.

Group prediction

Estimating behavioural patterns or average outcomes in a defined population.

Short-term prediction

Estimating outcomes over a relatively brief period.

Long-term prediction

Estimating outcomes across an extended period, usually with greater uncertainty.

Conditional prediction

Estimating what may happen under specified circumstances.

Statistical prediction

Using mathematical models and measured variables to estimate outcomes.

Predictors And Outcomes

A predictor is a variable used to estimate an outcome. The outcome is the behaviour or psychological characteristic being predicted. Predictors can include previous behaviour, environmental conditions, assessment scores, and demographic or developmental information when ethically and scientifically appropriate.

Prediction And Explanation

Prediction asks what is likely to happen, whereas explanation asks why it happens. A model may predict accurately without identifying the true causal mechanism.

Scientific Foundation

Historical Development

Psychological prediction developed through psychometrics, individual-differences research, learning studies, educational assessment, and statistical modelling. Intelligence testing, personality measurement, and later predictive research expanded the use of quantitative evidence in estimating outcomes.

Major Contributions

ThinkerContribution

Francis Galton

Contributed to the study of individual differences and statistical relationships, while some of his historical work also reflected scientifically and ethically problematic eugenic assumptions.

Charles Spearman

Developed influential work on correlations and general intelligence, contributing to psychometric prediction.

Lewis Terman

Conducted longitudinal research on intellectually gifted children, illustrating the study of developmental outcomes.

Paul Meehl

Examined the relative performance of statistical and clinical prediction methods.

Research Methods

  • Longitudinal studies that track predictors and later outcomes.

  • Regression models that estimate relationships between variables.

  • Cross-validation that evaluates predictive performance on separate data.

  • Prospective studies that measure predictors before outcomes occur.

  • Classification methods that estimate membership in outcome categories.

Scientific Standard

A useful prediction model should be evaluated for accuracy, calibration, generalizability, fairness, and performance on data that were not used to develop it.

Deep Understanding

Probabilistic Nature

Human behaviour is influenced by changing internal states and external conditions. A prediction therefore expresses likelihood under specified circumstances rather than certainty.

Influencing Factors

  • Past behavioural patterns.

  • Current motivation and emotional state.

  • Biological and developmental characteristics.

  • Social relationships and cultural expectations.

  • Environmental opportunities and constraints.

  • Unexpected events and changing circumstances.

Example

Previous attendance, preparation habits, and access to learning support may help estimate a student's future academic engagement. These factors cannot determine the student's future with certainty.

Generalization

A prediction developed from one population may not remain accurate in another population because measurement, opportunities, culture, and environmental conditions can differ.

Prediction Versus Classification

Classification assigns a case to a category, while prediction estimates an unknown or future outcome. Classification models can be used for prediction, but the concepts are not identical.

Self Fulfilling And Self Defeating Predictions

A prediction can influence the behaviour of people who learn about it. Expectations may sometimes help produce the predicted outcome or encourage actions that prevent it.

Mathematical Coverage

Regression Equation

Ŷ = β₀ + β₁X₁ + β₂X₂ + ... + βₚXₚ

Variables

Ŷ

Predicted outcome.

Β₀

Intercept.

Β₁...βₚ

Estimated coefficients.

X₁...Xₚ

Predictor variables.

Probability

P(Y = 1 | X) represents the estimated probability of an outcome Y = 1 given predictor information X.

Classification Metrics

  • Accuracy = (True positives + True negatives) / Total predictions.

  • Precision = True positives / (True positives + False positives).

  • Recall = True positives / (True positives + False negatives).

Worked Example

If a model correctly classifies 84 of 100 observations, its observed accuracy is 84 / 100 = 0.84, or 84%. This does not establish that the model is well calibrated or equally accurate across different groups.

Validation

Cross-validation and independent test samples help determine whether predictive performance generalizes beyond the data used to fit a model.

Limitations

A high accuracy score can be misleading when outcomes are imbalanced. Prediction should also be assessed for false positives, false negatives, calibration, and subgroup performance.

Practical Learning

Education

Educational researchers may estimate which students could benefit from additional academic support using validated indicators, while avoiding fixed labels about ability or potential.

Clinical Psychology

Validated assessment information can contribute to estimating treatment outcomes or risk, but predictions must be interpreted by qualified professionals and should not replace individual assessment.

Organizational Psychology

Psychologists may study how training, work conditions, and prior experience relate to future performance.

Daily Life

A person can use previous study patterns and current circumstances to estimate whether a planned schedule is realistic.

Practice Activity

Identify three predictors of class attendance. Explain why each may be useful and identify at least two factors that could make a prediction inaccurate.

Critical Thinking

Common Misconceptions

  • A prediction is not a guaranteed statement about the future.

  • A strong predictor is not necessarily a cause.

  • Prediction accuracy for a group does not imply equal accuracy for every individual.

  • A model can be accurate in one population and biased in another.

Limitations

  • Unmeasured variables can reduce accuracy.

  • Human behaviour and environments change over time.

  • Historical datasets may contain social inequalities and measurement biases.

  • Models can become less accurate when applied to populations unlike their development samples.

Ethical Issues

Predictions about individuals can affect opportunities, privacy, treatment, and social expectations. Psychologists should avoid unjustified labelling, protect sensitive information, and explain uncertainty.

Critical Questions

  • Was the prediction tested on independent data?

  • How large and representative was the original sample?

  • What are the consequences of false positive and false negative predictions?

  • Does the model perform fairly across relevant groups?

Knowledge Assessment

Mcqs

QuestionOptionsAnswerExplanation

What does psychological prediction primarily estimate?

  • The probability of future outcomes

  • The exact cause of every action

  • A person's permanent identity

  • Only past behaviour

The probability of future outcomes

Prediction estimates likely outcomes based on available evidence.

Does a variable that predicts behaviour necessarily cause that behaviour?

  • Yes

  • No

No

Predictive relationships can exist without a direct causal relationship.

Analytical Questions

  • Why is psychological prediction probabilistic?

  • Explain the difference between prediction and explanation.

  • Why should predictive models be tested on independent data?

Worked Exercise

Problem

A prediction model correctly identifies 90 cases out of 120. Calculate its accuracy.

Solution

Accuracy = 90 / 120 = 0.75 = 75%.

Interpretation

The model correctly classified 75% of the cases in this sample. Further evaluation is needed to understand its errors and generalizability.

Further Learning

Prerequisites

  • Description of Behaviour

  • Explanation of Behaviour

  • Basic Statistics

Related Concepts

  • Psychometrics

  • Regression Analysis

  • Decision Theory

  • Risk Assessment

  • Individual Differences

Interdisciplinary Connections

  • Data Science

  • Statistics

  • Education

  • Medicine

  • Economics

Recommended Next Topics

  • Influence and Modification of Behaviour

  • Psychological Assessment and Measurement

  • General Principles and Individual Differences

Advanced Directions

Explore predictive validity, logistic regression, machine learning in psychology, calibration, fairness auditing, and longitudinal prediction.

Sources and verification

Sources are shown as provided in the project data. Not every source has yet been independently verified. Please report errors or missing citations.

  • Source verification pending; check an authoritative textbook, university, standards body, or peer-reviewed source.

Suggest a correction or source · More in Foundations of Psychology