1/25
Vocabulary flashcards covering core concepts of inferential statistics, sampling error, probability models (classical vs. frequentist), binomial and normal distributions, Z-score standardization, and the Central Limit Theorem.
Descriptive statistics
Statistical procedures used to organise and summarise collected sample data (answering "What does this sample look like?"), including measures such as means, standard deviations, medians, interquartile ranges, counts, and graphs.
Inferential statistics
Statistical techniques that use sample data to make estimates, draw inferences, and test hypotheses about a wider population (answering "What can we conclude about the population based on the data from the sample?").
Sampling error
The naturally occurring, random variation and discrepancy between a sample statistic and the true population parameter that arises purely by chance because a sample is only an incomplete subset of the population.
Measurement error
Variability arising because data collection instruments or tools are imperfect, representing a source of error that is not accounted for or addressed by inferential statistics.
Proof-of-concept studies
Early-stage experimental studies (such as Phase 1 and Phase 2 clinical trials) designed to test whether a proposed mechanism is biologically or psychologically possible under ideal circumstances, where researchers are solely interested in the specific sample rather than broad population inference.
Complete enumeration
A research context (such as a census or closed population) where the sample comprises the entire population of interest, eliminating the distinction between sample statistics and population parameters.
Nominal level of measurement
A scale of measurement consisting of mutually exclusive named categories that cannot be ordered, lack consistent numerical intervals, and have no true zero point.
Classical probability
An approach that derives probability theoretically through logical reasoning and symmetry, calculated as the ratio of favourable outcomes to the total number of equally likely, mutually exclusive, and exhaustive outcomes.
Frequentist probability
An empirical approach that defines probability as the long-run relative frequency of an event occurring across an indefinitely large series of repeated, identical trials.
Law of Large Numbers
The mathematical principle stating that as the number of experimental repetitions or trials increases, the observed empirical relative frequency of an outcome converges toward its true theoretical probability.
Binomial distribution
A discrete probability distribution that models the number of successes (k) across a fixed number of independent trials (n), where each trial has only two possible outcomes and an identical probability of success (p).
Binomial coefficient
The algebraic term denoted as (kn)=k!(n−k)!n!, which calculates the total number of distinct ways to choose k successes out of n trials.
Normal distribution
A symmetrical, unimodal, bell-shaped continuous probability distribution (also known as a Gaussian distribution) defined entirely by its mean (M) and standard deviation (SD), under which the total area is exactly 1.
Skewness
A measure of the asymmetry of a distribution, indicating whether the majority of scores cluster to one side while the tail stretches toward the opposite side.
Floor effect
A measurement limitation occurring when an assessment is excessively difficult, causing scores to pile up at the lower boundary of the scale and contributing to a positive skew.
Ceiling effect
A measurement limitation occurring when an assessment is excessively easy, causing scores to cluster at the upper boundary of the scale and contributing to a negative skew.
Kurtosis
A measure of the peakedness of a distribution curve and the heaviness or thickness of its tails relative to a normal distribution.
Platykurtic
A distribution profile that is broad, relatively flat, and characterized by lighter tails than a normal distribution, with a kurtosis value of <3.
Leptokurtic
A distribution profile featuring a narrow, acute peak and heavier tails than a normal distribution, with a kurtosis value of >3.
Mesokurtic
A distribution profile exhibiting baseline peakedness and tail weight identical to that of a normal distribution, characterized by a kurtosis value of =3.
Z-score (Standard score)
A standardized metric calculated as zi=sXi−Xˉ, expressing the signed distance of a raw score from the sample mean in units of standard deviation.
Standard normal distribution
A specific normal distribution obtained by converting raw scores into Z-scores, resulting in a mean of 0 and a standard deviation of 1.
Area under the curve (AUC)
The mathematical integral beneath a continuous probability density function between two values, representing the exact probability or p-value of an observation falling within that range.
Central Limit Theorem (CLT)
A foundational theorem stating that as sample size (n) increases (typically n≥30), the sampling distribution of sample means approaches normality regardless of the shape of the underlying population distribution.
Sampling distribution of the mean
The theoretical probability distribution consisting of the calculated means from all possible random samples of a fixed size (n) drawn repeatedly from a population.
Theorem
A mathematical proposition that is proven true through step-by-step logical reasoning starting from accepted fundamental axioms.