Research Design and Statistics

Comprehensive vocabulary flashcards covering research design types, sampling methods, threats to validity, and statistical concepts based on the lecture notes.

Variable

An attribute, behavior, event, or phenomenon that is capable of varying or having two states, conditions, or levels.

Constant

A characteristic that is restricted to one state or condition.

Independent variable

A variable that is intentionally changed (often represented as $X$) and is believed to affect another variable; it must have at least two levels.

Dependent variable

The event studied and measured in an experiment (often represented as $Y$) that is expected to change based on the independent variable; it is not manipulated.

Operationalization

The process of strictly defining variables into measurable factors defined in terms of the method by which they will be measured.

Content analysis

A measurement method that involves organizing data into categories.

Protocol analysis

A type of content analysis where subjects think aloud while solving problems.

Interval recording/sampling

A behavior sampling method where behavior is observed for a period of time divided into intervals, and it is recorded whether the behavior occurs in each interval.

Event sampling

Observing and recording behavior as it occurs, often using a pre-coded checklist or recording the start and end times.

Situational Sampling

Observing behavior in a number of settings to increase the generalizability of the study.

Sequential analysis

A method of coding behavioral sequences rather than separate behaviors.

True Experiments

A research arrangement that permits maximum control over independent variables through random assignment and controlling for bias.

Randomized control clinical trial

A true experiment conducted within the context of an intervention.

Quasi-experiment

A design where conditions of true experiments are approximated but there is no random assignment; subjects are selected based on varying characteristics.

Random Selection

A sampling method where each member of the population has an equal probability of being selected, enhancing external validity and reducing bias.

Stratified Random Sampling

Dividing population members into homogeneous subgroups (strata) before randomly selecting subjects from each stratum to ensure all subpopulations are represented.

Cluster Sampling

A sampling method where groups of individuals (clusters) are selected first, followed by individuals from within those clusters.

Extraneous (confounding) variable

A source of systematic error that is irrelevant to the study but correlates with and has a systematic effect on the dependent variable.

Blocking

The process of building an extraneous variable into the study as an independent variable, where subjects in each block are randomly assigned to the independent variable.

Statistical control

Using statistical methods to remove the influence of an extraneous variable and equalize subjects on that variable.

Maturation

A threat to internal validity involving processes that change over time, such as growing older, stronger, wiser, tired, or bored.

History

A threat to internal validity involving any event other than the independent variable that occurs inside or outside the experiment and may account for results.

Testing (internal validity threat)

The effects that taking a test one time may have on subsequent performance on the test.

Instrumentation

A threat to internal validity involving changes in the measuring instrument, measuring device, or measurement procedures over time.

Statistical Regression

The tendency for extreme scores to revert toward the mean of a distribution when measurement is re-administered.

Selection

Systematic differences between groups before any manipulation, occurring on the basis of the assignment of subjects to groups.

Attrition

The loss of subjects in an investigation, occurring when studies last longer than one session.

External Validity

The extent to which investigation results can be generalized to other populations (population validity), settings (ecological validity), and circumstances.

Analogue study

A study that examines the relationship between variables in a laboratory setting.

Solomon four-group design

A design used to control for the interaction between testing and treatment by including groups that do not receive a pretest.

Evaluation apprehension

A form of reactivity where subjects act in ways to avoid negative evaluation.

Demand characteristics

Cues in the experimental setting that inform subjects of the purpose of the study and how they are supposed to behave.

Experimenter expectancy

When the experimenter unintentionally provides cues to subjects that may impact the results.

Multiple treatment interference

A threat to external validity occurring when subjects receive two or more treatments, where one treatment influences the reaction to subsequent ones.

Counterbalanced design

A solution used to control for carry-over effects in within-subject or repeated measures designs.

Between group Design

A design where different groups of subjects are administered different conditions, such as the simple two group design.

Factorial design

A design involving more than two independent variables, allowing for the simultaneous investigation of multiple variables.

Main effect

The effect of an independent variable on the dependent variable, disregarding the effect of all other independent variables.

Interaction effect

Occurs when the effect of an independent variable is different at different levels of another independent variable.

Within Subject Design

A design where all participants are exposed to every treatment or condition, allowing for comparison within the same subject.

Autocorrelation

A problem in within-subject designs where performance on the posttest correlates with performance on the pretest, increasing Type I error probability.

Type I error

Rejecting the null hypothesis when it is true, also known as a false positive.

Mixed Design

A design that combines at least one between-subject and one within-subject independent variable.

Baseline phase (A phase)

The initial phase of a single subject design where no intervention is applied and repeated measurements are taken.

Reversal Design

A single subject design (e.g., ABAB) where treatment is withdrawn after a phase to determine if behavior returns to baseline levels.

Multiple Baseline Design

A design where treatment is introduced sequentially across different subjects, behaviors, settings, or tasks without withdrawing treatment.

Continuous variable

A variable with an infinite number of values on the measurement scale, such as time or age.

Discrete variables

Variables with a countable number of values between any two values.

Dichotomous variable

A variable that has only two values, such as boy or girl.

Nominal Scale

Categorical data using numbers only as identifiers or labels for unordered categories.

Ordinal Scale

A scale that divides observations into categories and also rank orders them.

Interval Scale

A scale representing quantity with equal units where zero is simply a point of measurement and does not indicate a lack of characteristic.

Ratio Scale

A scale with equal quantities, rank order, and an absolute zero where no numbers exist below zero.

Kurtosis

The sharpness of the peak of a distribution curve.

Platykurtic

A distribution that is flatter than the normal distribution.

Leptokurtic

A distribution that is more peaked than the normal distribution.

Positively skewed distribution

A distribution where most scores are on the negative side with few high scores.

Negatively skewed distribution

A distribution where most scores are on the positive side with few low scores.

Mode

The score that occurs most frequently in a distribution.

Median

The score that divides a distribution perfectly in half when data is ordered from low to high.

Mean

The arithmetic average of all scores in a distribution.

Range

The difference between the largest and smallest values in a data set.

Variance

A thorough measure of variability that includes all scores of the distribution.

Standard deviation

The square root of the variance, used to compare the variability of distributions.

Standard error

Calculated by dividing the standard deviation by the square root of the sample size ($n$); it represents the standard deviation of the sampling distribution.

Central limit theorem

States that as sample size increases, the sampling distribution of the mean will become closer to a bell shape.

Alpha (level of significance)

The size of the rejection region (typically set at $0.5$ or $0.1$) determining the probability of a Type I error.

Type II error

Retaining a false null hypothesis, also known as a miss; the probability is represented by beta ($\beta$).

Statistical power

The ability of a statistical test to correctly reject a false null hypothesis.

Parametric Tests

Tests like the t-test that evaluate population parameters and assume the variables are normally distributed with equal variances (homoscedasticity).

Homoscedasticity

The assumption that the variances in compared groups are equal, or that the range of $Y$ scores is the same as the range of $X$ scores.

Nonparametric Tests

Distribution-free tests used for nominal or ordinal scale variables that are generally less powerful than parametric tests.

Mann-Whitney U test

A nonparametric test used to compare the medians of two independent samples with rank-ordered values.

Kruskal-Wallis test

A nonparametric test similar to Mann-Whitney but used for two or more independent groups.

Wilcoxon matched-paired signed rank test

A nonparametric test for correlated groups where the dependent variable measure is not normally distributed.

One way ANOVA

An analysis of variance involving one independent variable and two or more groups.

Effect Size

The magnitude of the difference between conditions, often expressed in standard deviation units (e.g., Cohen's d).

Pearson correlation coefficient (r)

A measure ranging from $-1.00$ to $+1.00$ indicating the direction and strength of a linear relationship between two variables.

Multiple regression

A technique using two or more predictors to predict one criterion, resulting in a multiple correlation coefficient ($R$).

Canonical Correlation

An extension of multiple regression involving two or more predictors and two or more criteria.

Factor analysis

A technique used to analyze the structure of variables and reduce many data points to a few factors that explain intercorrelation.

Cluster Analysis

A method used to group people or data into subgroups based on similarities to achieve within-group homogeneity and between-group heterogeneity.