Research Design - Semantic 2

Definitions, facts, concepts

Control variable

A control variable is a variable that is held constant in an experiment to prevent it from affecting the outcome. This allows researchers to isolate the effect of the independent variable on the dependent variable, thus increasing internal validity.

Construct validity

The extent to which a dependent variable measures what it claims to measure, ensuring that the operational definition of a construct accurately reflects the theoretical concept.

Selection threats

are biases that occur when individuals or groups are selected for a study in a way that is not representative of the population, threatening external validity.

Probability sampling

A sampling technique where each member of a population has a known (or knowable) chance of being selected, allowing for the generalization of results.

Over-sampling

A sampling method in which certain groups within a population are deliberately selected at a higher rate than others, often in attempt to ensure adequate representation of minority groups in research. A common problem in stratified random sampling.

Q statistic

A statistical measure used to assess the degree of variation among independent studies or groups, often utilized in meta-analysis to determine heterogeneity. A significant Q statistic means the presence of a moderating variable is likely.

Observational design

A type of research design where the researcher observes subjects in their natural environment (status variables only) without manipulation or intervention, allowing for the assessment of behaviors or outcomes in real-world settings.

Inadequate explication of constructs

Refers to insufficient clarity or definition of the theoretical concepts being studied, which can lead to challenges in measurement and interpretation of research findings. The biggest threat to construct validity.

Solomon 4-group design

A research design that includes four groups, consisting of two experimental and two control groups, but only one of each is given a pre-test. This design aims to control for potential pre-test effects while allowing for comparison of the treatment's impact. .

Maturation effects

Changes in subjects' behavior or characteristics over time due to natural development or aging, which can influence the outcomes of a study. Threatens internal validity.

Construct

A theoretical concept that can only be examined latently and can only be defined through extrapolation from other measurements.

Moderator analysis

A statistical analysis of effect size used to examine how the effects of independent on dependent variables changes due to the presence of a moderator. This is common in meta-analysis and with the study of status independent variables.

Between-subjects independent variable

A type of independent variable where different groups of subjects are exposed to different levels or conditions of the variable, allowing for comparison across groups in an experiment. Any given participant is only in one level at any one time.

Treatment integrity

The degree to which an intervention is implemented as intended, ensuring that the treatment's effectiveness can be accurately assessed.

Hawthorne effect

The phenomenon where individuals alter their behavior in response to being observed or studied, potentially affecting the outcomes of research.

Randomized blocks design

A research design that involves dividing participants into subgroups or blocks based on a specific characteristic before randomly assigning them to different treatments, therefore controlling for confounds through matching.

History effects

External events that occur during a study that can influence participants' responses, potentially impacting the results. Threatens internal validity.

Non-probability sample

A sampling method where not all individuals have an equal chance of being selected, or where the probability of any given participant being selected is not known.

Internal validity

The degree to which a study accurately establishes a causal relationship between variables, ensuring that the results are not influenced by extraneous factors or confounds.

Cluster sampling

A sampling method where the population is divided into groups (clusters), and entire clusters are randomly selected for the study. Simple random sampling then proceeds from each cluster.

RABAB design

A case-study design where participants are randomly assigned to one of two conditions: measurement-treatment-withdrawal-reversal (ABAB), or treatment-withdrawal-reversal-measurement (BABA). Minimum of 2 cases (participants) needed.

Within-subjects independent variable

A variable in which each participant is exposed to all levels of the independent variable, allowing comparisons within the same subjects. Levels are different points in time as opposed to separate groups of people.

The File Drawer problem

Refers to the bias that occurs when studies with null or negative results remain unpublished, leading to an overrepresentation of positive outcomes in the literature.

Highly representative population-based dataset

A dataset that accurately reflects the diversity of a larger population, allowing for generalizable findings and robust statistical analysis.

Coding (meta-analysis)

The process of systematically categorizing and summarizing data from multiple studies in a meta-analysis to identify patterns, relationships, and overall effect sizes.

Statistical conclusion validity

The probability that conclusions drawn from statistical results may commit Type I or Type II errors.

True experimental design

A research design that includes manipulated independent variables, random assignment, and control groups.

Simple random sampling

A sampling method where each member of a population has an equal chance of being selected for the sample, ensuring randomness and minimizing bias.

Setting effects

Factors in a study's environment that can influence the outcomes and generalizability of the results. Threatens external validity.

Testing effects

Changes in participants' responses due to pre-testing, which can limit the reproducibility of results to conditions where the same pre-test is given. Threatens external validity.