1/34
Vocabulary practice flashcards covering Null Hypothesis Significance Testing (NHST), sampling distributions, Central Limit Theorem, error types, statistical power, standard errors, and methods for assessing normality.
Hypothesis Testing
A statistical method used to evaluate whether sample data provides sufficient empirical evidence to infer that a specific condition, effect, or parameter is true for an entire target population.
Population
The entire group of individuals, objects, or measurements that a researcher is interested in studying and drawing conclusions about.
Sample
A smaller, selected subset of individuals chosen from the population that is directly observed, measured, and analyzed in a study.
Null Hypothesis (H0)
A default conservative assumption stating that there is no effect of an intervention, no difference between groups, or no association between variables in the population.
Alternative Hypothesis (H1 or Ha)
The research hypothesis or claim a researcher seeks evidence to support, proposing that an effect, relationship, or group difference exists in the population.
p-value
A probability value ranging from 0 to 1 indicating the likelihood of obtaining sample results as extreme as, or more extreme than, those observed, assuming the null hypothesis is true.
Significance Level (α)
A predetermined probability threshold (commonly set at 0.05 or 5%) specifying how unlikely an observed result must be under the null hypothesis before H0 can be rejected.
Statistical Significance
A formal determination made when the calculated p-value is less than or equal to the significance level (p≤α), indicating that the observed data is sufficiently incompatible with the null hypothesis to justify its rejection.
Type I Error
The mistake of rejecting a true null hypothesis (a false positive), the probability of which is directly set by the significance level (α).
Type II Error
The mistake of failing to reject a false null hypothesis when a real difference or effect actually exists (a false negative), denoted by the probability β.
Statistical Power (1−β)
The probability that a statistical test will correctly reject a false null hypothesis, which increases with larger sample size (n), larger effect size, or smaller sample variation (s).
Central Limit Theorem (CLT)
A fundamental theorem stating that as sample size (n) increases, the sampling distribution of the sample mean will approximate a normal distribution, regardless of the underlying shape of the population distribution.
Cauchy Distribution
A mathematical probability distribution that lacks a defined population mean, serving as a rare theoretical exception to the Central Limit Theorem.
Standard Deviation (SD)
A descriptive statistic that quantifies the typical dispersion or distance by which individual observed scores in a single sample vary around their sample mean.
Standard Error of the Mean (SEM)
A measure of precision that describes how far an estimated sample mean is expected to vary from the true population mean across repeated samples, calculated as SEM=nSD.
Standard Error of the Mean Difference (SEMD)
The standard deviation of the sampling distribution of the difference between two independent sample means, calculated as SEMD=nAsA2+nBsB2.
Effect Size
The quantitative magnitude of a difference or relationship between groups or variables, which is mathematically independent of the p-value and sample size.
Random Sampling Error
The natural fluctuation, chance, noise, or luck that causes a sample statistic to differ from the true population parameter when participants are randomly sampled.
Sample Distribution
The distribution of individual scores or data values collected directly from the participants within a single empirical sample.
Sampling Distribution of the Mean
The theoretical probability distribution formed by plotting the calculated means of an infinite number of independent random samples of identical size (n) drawn from the same population.
Law of Large Numbers
A mathematical law stating that as the sample size (n) increases, the sample mean (xˉ) will increasingly approximate the true population mean (μ).
Glivenko–Cantelli Theorem
Also called the Fundamental Theorem of Statistics, this theorem guarantees that an empirical sample distribution will approximate the true population distribution increasingly closely as sample size approaches infinity (n→∞).
Two-Tailed Test
A hypothesis test used for non-directional hypotheses (H1:μ1=μ2) where the critical region of significance is divided equally into both tails of the distribution (2.5% on each side at α=0.05, with critical values z=±1.96).
One-Tailed Test
A hypothesis test used for directional hypotheses (H1:μ1>μ2 or H1:μ1<μ2) where the entire critical rejection area (5% at α=0.05, with critical value z=+1.645 or −1.645) is allocated to a single specified tail.
Infinite Monkey Theorem
A probabilistic concept stating that an infinite number of monkeys randomly typing on typewriters for an infinite time will eventually produce every finite text, demonstrating that extreme outcomes naturally occur given enough chance attempts.
Skewness
A metric describing the asymmetry of a distribution, where positive values denote a long tail to the right and negative values denote a long tail to the left.
Kurtosis
A measure of the tailedness and peakedness of a probability distribution relative to a standard normal distribution.
Leptokurtic Distribution
A distribution possessing a sharper, higher peak and fatter tails than a normal distribution (positive excess kurtosis), indicating a higher frequency of outliers and extreme values.
Platykurtic Distribution
A distribution characterized by a flatter, broader peak and thinner tails than a normal distribution (negative excess kurtosis), reflecting fewer extreme values or outliers.
Mesokurtic Distribution
A distribution with moderate peakedness and tail weight characteristic of a standard normal distribution, having an excess kurtosis near 0 (or a raw kurtosis equal to 3).
Shapiro-Wilk Test
A statistical test evaluating the null hypothesis that sample data originated from a normal distribution, recommended as most powerful for small to moderate sample sizes (N<50).
Kolmogorov-Smirnov Test
A statistical test assessing whether a sample follows a specified distribution (such as normality) that is suited for larger samples (N≥50), though it is less powerful for small samples and overly sensitive when N>100.
Quantile-Quantile (Q-Q) Plot
A diagnostic probability plot that compares observed data quantiles against expected normal quantiles; points falling along the diagonal line indicate conformity to a normal distribution.
Box Plot
A standardized visual display of a dataset showing the median, interquartile range (25th to 75th percentiles), data spread, and individually flagged potential outliers.
Principle of Parsimony
The scientific guideline stating that researchers must default to the simplest plausible explanation (such as no effect or difference) unless compelling empirical data warrants a more complex explanation.