Teroka sains mengikut kaedah, bidang dan bukti.

Discriminant Analysis

Discriminant analysis finds linear combinations of predictor variables that best separate two or more known groups. It is used both to understand which predictors distinguish the groups and to classify new observations into those groups with minimum error.

2 sumber1936

field methods

Doctrinal Legal Research

Doctrinal legal research is the foundational methodology of legal scholarship. It systematically identifies, reads, and analyses authoritative legal sources — statutes, case law, constitutional texts, and regulations — to describe, explain, and critique the content and internal logic of legal doctrine. By working withi

2 sumber1860

Dolado-Lütkepohl Causality

The Dolado-Lütkepohl (DL) test, introduced by Dolado and Lütkepohl (1996), is a modified Wald procedure for testing Granger causality in vector autoregressive (VAR) systems whose variables may be integrated or cointegrated. By fitting a VAR of slightly higher order than necessary and restricting the Wald statistic to t

1 sumber1996

Double Bootstrap

The double bootstrap is a resampling method that calibrates a bootstrap confidence interval with a second, nested layer of bootstrap to bring its actual coverage closer to the nominal level. Introduced by Hall (1986) and Beran (1987), it is especially valuable for small samples and skewed distributions where a single-l

2 sumber1986

causal inference

Double Machine Learning

Double/Debiased Machine Learning (DML), introduced by Chernozhukov et al. (2018), is a semiparametric framework for estimating causal or structural parameters in the presence of high-dimensional controls. It uses flexible machine learning methods to model nuisance functions—the conditional expectations of the outcome a

1 sumber2018

experimental design

Double-blind A/B test

A double-blind A/B test is a randomized experiment that compares two variants — a control (A) and a treatment (B) — while concealing group assignment from both participants and those administering or assessing the experiment. Combining the causal isolation of randomized assignment with blinding on both sides eliminates

2 sumber1935

Driscoll-Kraay SE

Driscoll-Kraay standard errors provide a nonparametric, heteroskedasticity- and autocorrelation-consistent (HAC) covariance estimator for balanced and unbalanced panel datasets. Introduced by Driscoll and Kraay in 1998, the method corrects inference when residuals exhibit cross-sectional dependence, serial autocorrelat

1 sumber1998

DSGE Model

A DSGE model is a micro-founded macroeconomic general equilibrium model that combines the optimising decisions of households, firms, and government under rational expectations. Popularised for empirical policy work by Smets and Wouters (2007) and given its Bayesian estimation framework by An and Schorfheide (2007), it

3 sumber2007

time series

DTW Barycenter Averaging

DTW Barycenter Averaging (DBA) is a method for computing the average or representative sequence of a set of time series that respects temporal warping and elastic distance. Unlike Euclidean averaging which requires point-wise alignment, DBA minimizes the sum of Dynamic Time Warping (DTW) distances, producing a meaningf

3 sumber2011

Dumitrescu-Hurlin Causality

The Dumitrescu-Hurlin (DH) test, introduced by Elena-Ivona Dumitrescu and Christophe Hurlin in their 2012 Economic Modelling article, tests for Granger non-causality in heterogeneous panel datasets. Unlike standard panel causality approaches, it permits each cross-sectional unit to have its own distinct causal relation

1 sumber2012

Dunn Test

Dunn's test is a nonparametric post-hoc procedure introduced by Olive Jean Dunn in 1964 to identify which specific pairs of groups differ significantly after a Kruskal-Wallis test has returned a significant overall result. It compares groups pairwise using rank sums and applies a multiple-comparison correction — most c

1 sumber1964

finance

DuPont Analysis

DuPont Analysis is a financial performance framework that decomposes Return on Equity (ROE) into three multiplicative components: net profit margin, asset turnover, and the equity multiplier. Originally developed by engineers at DuPont Corporation in the early 1920s, the method gained renewed academic prominence throug

1 sumber2008

Durbin-Watson Test

The Durbin-Watson test, developed by James Durbin and Geoffrey Watson in 1950–1951, detects first-order serial correlation in the residuals of a linear regression. Its statistic ranges from 0 to 4, with a value near 2 indicating no autocorrelation, values toward 0 indicating positive autocorrelation, and values toward

2 sumber1950

Dynamic Bayesian Hierarchical Model

A Dynamic Bayesian Hierarchical Model combines the multilevel structure of Bayesian hierarchical models with an explicit time-evolution equation for the latent states. Observations at each time point are linked to unobserved dynamic states, which evolve according to a probabilistic transition law, while a shared hyperp

2 sumber1990

Dynamic Bayesian Inference

Dynamic Bayesian inference is a framework for performing Bayesian updating sequentially as new observations arrive over time. Rather than fitting a static model to a fixed dataset, it tracks how a posterior distribution over latent states or parameters evolves step by step, combining a prior with each new likelihood to

2 sumber1989

Dynamic Bayesian Model Averaging

Dynamic Bayesian Model Averaging (DMA) extends standard Bayesian model averaging to settings where the best predictive model may change over time. It maintains a probability distribution over a set of competing models and updates that distribution sequentially as new observations arrive, allowing model weights to evolv

2 sumber2010

Dynamic Bayesian Network

A Dynamic Bayesian Network (DBN) extends a standard Bayesian network over time by representing how a set of random variables evolve across discrete time steps. It captures both the conditional independence structure among variables at each instant and the probabilistic dependencies between consecutive time slices, enab

2 sumber1989

neuroimaging

Dynamic Causal Modeling

Dynamic Causal Modeling (DCM) is a Bayesian framework for specifying and inverting generative models of brain connectivity from neuroimaging data. Introduced by Karl Friston and colleagues in 2003, DCM treats brain regions as dynamical systems and estimates effective connectivity by fitting observed fMRI time series to

2 sumber2003

Dynamic Factor Model

A Dynamic Factor Model (DFM) extracts a small number of latent common factors from a large panel of economic time series and uses those factors to forecast or nowcast a target variable. Formalized for macroeconomic forecasting by James Stock and Mark Watson in their 2002 Journal of Business & Economic Statistics paper,

1 sumber2002

Dynamic Hamiltonian Monte Carlo

Dynamic Hamiltonian Monte Carlo — widely known as the No-U-Turn Sampler (NUTS) — is an adaptive extension of Hamiltonian Monte Carlo that automatically selects the number of leapfrog integration steps during each MCMC transition, removing the need to hand-tune the most sensitive tuning parameter of standard HMC. It is

2 sumber2014

materials science

Dynamic Light Scattering

Dynamic Light Scattering (DLS), also known as Photon Correlation Spectroscopy (PCS), is an analytical technique for determining the size and size distribution of particles suspended in fluids by analyzing the time-dependent intensity fluctuations of scattered laser light. Developed by Robert Pecora in 1964, DLS exploit

3 sumber1964

Dynamic Metropolis-Hastings Algorithm

The Dynamic Metropolis-Hastings (Dynamic MH) algorithm applies the Metropolis-Hastings MCMC sampler to Bayesian state-space and time-varying parameter models. At each time step, latent states or evolving parameters are updated via proposal-and-accept moves, yielding full posterior distributions over trajectories rather

2 sumber1970

Dynamic Monte Carlo Simulation

Dynamic Monte Carlo (DMC) simulation is a computational method that tracks the stochastic time evolution of a system by drawing random event sequences weighted by transition rates. Unlike static Monte Carlo sampling of equilibrium distributions, DMC explicitly advances a clock, making it suitable for kinetic, reaction,

2 sumber1975

Dynamic OLS

Dynamic OLS is a cointegrating-regression estimator introduced by Stock and Watson (1993) that recovers the long-run relationship between I(1) variables. It augments the static regression with leads and lags of the differenced regressors, correcting endogeneity bias parametrically so that the long-run coefficient can b

2 sumber1993

Dynamic Panel Data Model

The dynamic panel data model extends standard panel regression by including a lagged value of the outcome variable as a regressor, capturing persistence and adjustment dynamics. Because the lagged dependent variable is correlated with the unit-specific fixed effect, ordinary OLS or within estimators are biased; GMM-bas

2 sumber1988

Dynamic Particle Filter

A dynamic particle filter is a sequential Monte Carlo algorithm that tracks an evolving hidden state over time by maintaining a population of weighted random samples — particles — each representing a plausible trajectory. As new observations arrive, particle weights are updated via the likelihood and the population is

2 sumber1993

Dynamic Sequential Monte Carlo

Dynamic Sequential Monte Carlo (Dynamic SMC) is a Bayesian computational method that maintains and updates a population of weighted samples — particles — as new observations arrive over time. It propagates particles through a dynamic system model, reweights them by how well they match the observed data, and periodicall

2 sumber2006

Dynamic Variational Inference

Dynamic variational inference extends the variational inference framework to sequential and time-series settings by positing a structured approximate posterior that respects the temporal ordering of latent states. It jointly learns a generative model of how hidden states evolve over time and a recognition network that

2 sumber2014

epidemiology

Ecological Study

An ecological study is an observational epidemiological design in which the unit of analysis is a group or population — a country, region, city, or time period — rather than an individual. Exposures and outcomes are measured as aggregates (rates, proportions, or means) and then correlated across groups to generate or e

2 sumber1854

meteorology

Eddy Covariance

The eddy covariance method is a direct, micrometeorological technique that measures turbulent fluxes of momentum, heat, water vapor, and CO2 by computing the covariance between high-frequency fluctuations of wind velocity and scalar properties (temperature, humidity, concentration). It is the gold standard for measurin

2 sumber1951

EFA

Exploratory factor analysis reduces a large set of observed variables into a smaller number of latent common factors. It is widely used in scale development and psychometrics to uncover the dimensional structure that underlies a set of correlated items, without specifying that structure in advance.

2 sumber

psychometrics

EFA for Scale Development

Exploratory Factor Analysis for Scale Development is the psychometric application of EFA in which an item pool is administered and the resulting response data are analysed to discover the latent factor structure underlying the items. Originating with Spearman's (1904) factor theory and formalised for applied scale cons

2 sumber1904

research statistics

Effect Size

Effect size quantifies the magnitude of a research finding independent of sample size. While a p-value tells you whether a result is statistically significant, an effect size tells you how big the result is. Jacob Cohen formalized effect size measurement in behavioral sciences (1988), establishing standard benchmarks (

3 sumber1988

Effect size analysis

Effect size analysis quantifies the practical magnitude of a statistical result independently of sample size. Rather than asking only whether a difference or relationship is statistically significant, it asks how large it is, using standardized indices such as Cohen's d, eta-squared, omega-squared, or Pearson's r that

2 sumber1969

gerontology

EFS

The Edmonton Frail Scale (EFS) is a comprehensive, nine-domain assessment tool developed by Rolfson and colleagues in 2006 to systematically evaluate frailty across multiple physiological and functional dimensions in older adults. Combining clinical judgment with objective testing, the EFS assesses cognition, general h

3 sumber2006

EGARCH

EGARCH is an asymmetric GARCH variant, introduced by Nelson in 1991, that models the leverage effect in which bad news raises volatility more than good news of the same size. It captures the negative-shock asymmetry of financial return series by modelling the logarithm of the conditional variance.

2 sumber1991

EGARCH model

The Exponential GARCH (EGARCH) model, introduced by Nelson (1991), extends the standard GARCH framework by modelling the logarithm of conditional variance. This ensures variance is always positive without parameter constraints and, crucially, allows negative and positive shocks to have asymmetric effects on volatility

2 sumber1991

Elastic Net

Elastic Net is a regularized linear regression method introduced by Zou and Hastie in 2005 that blends the LASSO (L1) and Ridge (L2) penalties, so it performs variable selection and coefficient shrinkage at the same time. It is designed for predictive and explanatory modelling on data with many, possibly correlated, pr

1 sumber2005

Elastic Net Regression

Elastic net regression combines the L1 (lasso) and L2 (ridge) penalties into a single regularized regression framework. Controlled by a mixing parameter alpha and a shrinkage strength lambda, it can simultaneously select variables and handle correlated predictors — overcoming key limitations of pure lasso and pure ridg

2 sumber2005

EM Algorithm

The Expectation-Maximization (EM) algorithm is an iterative optimization procedure for finding maximum likelihood or maximum a posteriori estimates of parameters in statistical models with latent variables or missing data. Introduced by Dempster, Laird, and Rubin in their landmark 1977 paper, EM alternates between comp

1 sumber1977

pharmacometrics

Emax Model

The Emax model is a nonlinear pharmacodynamic model that describes the relationship between drug concentration and biological effect. Introduced by Holford and Sheiner in 1981, it characterizes dose-response curves using three fundamental parameters: the maximum achievable effect (Emax), the concentration producing hal

1 sumber1981

research design

Embedded Multilevel Mixed Methods

Embedded multilevel mixed methods design nests a secondary qualitative (or quantitative) strand within a primary study that spans hierarchically organized levels — such as students within classrooms, employees within organizations, or patients within clinics. The dominant strand addresses the research question at the s

2 sumber2000

Empirical Bayes

Empirical Bayes (EB) is an estimation strategy, introduced by Herbert Robbins in 1956 and developed into practical shrinkage estimators by Bradley Efron and Carl Morris in 1973, in which the hyperparameters of the prior distribution are estimated from the observed data via the marginal likelihood rather than specified

4 sumber

Engle-Granger Cointegration Test

The Engle-Granger two-step method tests whether two or more non-stationary I(1) time series share a common stochastic trend — that is, whether a linear combination of them is stationary. If cointegration is confirmed, an error-correction model (ECM) can be estimated to capture both short-run dynamics and long-run equil

2 sumber1987

Ensemble Linear Regression

Ensemble Linear Regression combines multiple ordinary least-squares models — each fitted on a different bootstrap sample or feature subset — and averages their predictions. The technique, grounded in Breiman's bagging framework (1996), reduces variance and improves predictive stability compared with a single linear reg

2 sumber1996

Ensemble Logistic Regression

Ensemble Logistic Regression trains multiple logistic regression classifiers on varied subsets or perturbations of the training data and combines their probability estimates by averaging or voting. The approach preserves logistic regression's probabilistic interpretability while reducing variance and improving predicti

2 sumber1996

research design

Equal-weight multilevel mixed methods

Equal-weight multilevel mixed methods is a mixed methods design in which quantitative and qualitative data strands are collected at two or more distinct levels of a social system — such as students, classrooms, and schools — and both strands carry equal analytic priority. The QUAN+QUAL notation (where '+' signals equal

2 sumber2000

Equivalence Test (TOST)

The equivalence test using the Two One-Sided Tests (TOST) procedure is a parametric hypothesis test designed to demonstrate that the difference between two group means falls within a pre-specified equivalence region ±Δ. Introduced by Schuirmann (1987) in the context of pharmaceutical bioequivalence, TOST reverses the l

2 sumber1987

ERS Point-Optimal Test

The Elliott-Rothenberg-Stock (ERS) Point-Optimal test, introduced in their landmark 1996 Econometrica paper, is a near-efficient parametric procedure for testing whether a univariate time series contains a unit root. By first applying GLS detrending at a carefully chosen local-to-unity value and then computing a likeli

1 sumber1996

ETS Model

ETS is a comprehensive exponential smoothing framework that automatically selects additive or multiplicative combinations of the error (E), trend (T) and seasonal (S) components of a time series. Formalised as an innovations state space model by Hyndman, Koehler, Ord and Snyder in 2008, it unifies and generalises the H

2 sumber2008

deep learning

ETSformer

ETSformer is a deep learning architecture for time-series forecasting introduced by Woo et al. in 2022. It integrates classical exponential smoothing principles directly into the Transformer framework by replacing standard self-attention with an exponential smoothing attention mechanism. The model decomposes a time ser

1 sumber2022

research design

Evaluation-oriented multilevel mixed methods

Evaluation-oriented multilevel mixed methods is a research design that combines quantitative and qualitative data across hierarchically nested levels of an organization or system — such as students within classrooms within schools — to evaluate a program, policy, or intervention. By capturing outcomes, processes, and c

2 sumber2000

finance

Event Study

The event study is a financial research method that measures the impact of a news release, policy change, or corporate event on asset prices through cumulative abnormal returns. Reviewed by MacKinlay (1997) and formalised econometrically by Kothari and Warner (2007), it is the standard tool for testing the efficient-ma

2 sumber1997

EWMA Chart

The exponentially weighted moving average (EWMA) control chart, introduced by S. W. Roberts in 1959, monitors a process using a weighted average that gives the most recent observation the greatest weight while letting older observations fade geometrically. Like CUSUM, this memory makes it highly effective at detecting

2 sumber1959

Expectation Propagation

Expectation Propagation (EP) is a deterministic message-passing algorithm for approximate posterior inference in Bayesian models, introduced by Thomas P. Minka at UAI 2001. It iteratively refines a set of local approximate factors — each drawn from the exponential family — so that their product closely matches the true

3 sumber2001

deep learning

Explainable BERT-based Classification

Explainable BERT-based Classification combines the predictive power of fine-tuned BERT transformers for text classification with post-hoc or intrinsic explainability techniques — such as SHAP, LIME, attention analysis, or integrated gradients — to reveal which words or tokens drove each prediction. The result is a clas

2 sumber2019

Explainable DBSCAN

Explainable DBSCAN pairs the DBSCAN density-based clustering algorithm with post-hoc interpretability methods — most commonly SHAP values or local surrogate models — to reveal which input features drive the algorithm's cluster and noise assignments. It enables analysts to understand why specific points were grouped tog

2 sumber1996

Explainable Extra Trees

Explainable Extra Trees combines the Extremely Randomized Trees (Extra Trees) ensemble algorithm with post-hoc explainability methods — most commonly SHAP values — to deliver both strong predictive performance and transparent, feature-level explanations. It extends the classic Extra Trees classifier or regressor so tha

2 sumber2006