CB-SEM: see Covariance-based structural equation modeling. Coding: is the assignment of numbers to scales in a manner that facilitates measurement.
Confirmatory applications: aim at empirically testing theoretically developed models.
Constructs (also called latent variables): measure concepts that are ab stract, complex, and cannot be directly observed by means of (multiple) items. Constructs are represented in path models as circles or ovals. Covariance-based structural equation modeling: is used to confirm (or reject) theories. It does this by determining how well a proposed theo retical model can estimate the covariance matrix for a sample data set. Endogenous latent variables: serve only as dependent variables, or as both independent and dependent variables in a structural model. Equidistance: is given when the distance between data points of a scale is identical.
Error terms: capture the unexplained variance in constructs and indicators when path models are estimated.
Exogenous latent variables: are latent variables that serve only as independent variables in a structural model.
Exploratory: see Exploratory applications.
Exploratory applications: focus on exploring data patterns and identifying relationships.
First-generation techniques: are statistical methods traditionally used by researchers, such as regression and analysis of variance. Formative measurement model: is a type of measurement model setup in which the direction of the arrows is from the indicator vari ables to the construct, indicating the assumption that the indicator variables cause the measurement of the construct.
Indicators: are directly measured observations (raw data), generally referred to as either items or manifest variables� represented in path models as rectangles.
Inner model: see Structural model.
Interval scale: can be used to provide a rating of objects and has a constant unit of measurement so the distance between the scale points is equal.
Items: see Indicators.
Latent variable: see Constructs. Manifest variables: see Indicators.
Measurement: is the process of assigning numbers to a variable based on a set of rules.
Measurement error: is the difference between the true value of a variable and the value obtained by a measurement.
Measurement model: is an element of a path model that contains the indicators and their relationships with the constructs and is also called the outer model in PLS-SEM.
30 A Primer on Partial Least Squares
Measurement scale: is a tool with a predetermined number of closed ended responses that can be used to obtain an answer to a question. Measurement theory: specifies how the latent variables are measured. Multivariate analyses: are statistical methods that simultaneously analyze multiple variables.
Multivariate measurement: involves using several variables to indi rectly measure a concept.
Nominal scale: is a measurement scale where numbers are assigned that can be used to identify and classify objects (e.g., people, companies, products,
etc.).
Ordinal scale: is a measurement scale where numbers are assigned that indicate relative positions of objects in an ordered series. Outer models: see Measurement model.
Partial least squares path modeling: see Partial least squares struc tural equation modeling.
Partial least squares structural equation modeling: is a variance based method to estimate structural equation models. The goal is to maximize the explained variance of the endogenous latent variables. Path models: are diagrams that visually display the hypotheses and variable relationships that are examined when structural equation modeling is applied.
PLS regression: is an analysis technique that explores the linear relation ships between multiple independent variables and a single or multiple dependent variable(s). In developing the regression model, it con structs composites from both the multiple independent variables and the dependent variable(s) by means of principal component analysis. PLS-SEM: see Partial least squares structural equation modeling. Ratio scales: are the highest level of measurement because they have a constant unit of measurement and an absolute zero point; a ratio can be calculated using the scale points.
Recursive model: is a PLS path model that does not have a causal loop of relationships between latent variables in the structural mod el (i.e., no circular relationships).
Reflective measurement model: is a type of measurement model setup in which the direction of the arrows is from the construct to the indicator variables, indicating the assumption that the construct causes the mea surement (more precisely, the covariation) of the indicator variables. Second-generation techniques: overcome the limitations of first-generation techniques, for example, in terms of accounting for measurement error. SEM is the most prominent second-generation data analysis technique. SEM: see Structural equation modeling.
Single-item constructs: have only a single item to measure the construct. Statistical power: the probability to detect a significant relationship sig nificant when it is in fact significant in the population.
Structural equation modeling: is used to measure relationships between latent variables.
Structural model: is an element of a PLS path model that contains the constructs as well as the relationships between them. It is also called the inner model in PLS-SEM.
Structural theory: specifies how the latent variables are related to each other. That is, it shows the constructs and the paths between them. Theory: is a set of systematically related hypotheses developed fol lowing the scientific method that can be used to explain and predict outcomes and can be tested empirically.
Variance-based SEM: see Partial least squares structural equation modeling. Variate: is a linear combination of several variables.