The Four Types of Structural Equation Models

A several regression design could be conceptualized employing Structural Equation Model route diagrams. That’s the easiest SEM you can develop, but its authentic electrical power lies in growing on that regression design.  Here I will explore four styles of structural equation models.

Route Assessment

Extra interesting investigate queries could be requested and answered utilizing Path Investigation. Route Investigation is a form of structural equation modeling without latent variables.

1 of the strengths of route evaluation is the inclusion of interactions between variables that provide as predictors in one particular solitary product.

One certain and frequent instance is a mediation product.  Even even though it is not the only way of evaluating mediation, it is a very intuitive and successful a person.

This model here demonstrates a few various mediational relationships.  A single is the connection in between Exercising and Health issues.  You can see there is a immediate path from Workout to Illness, but there is also a mediated route that goes by means of Exercise.

This product is asserting that at the very least part of the effect of Work out on Health issues is that Work out has an effect on Exercise, and Exercise, in flip has an effect on Ailment.

Confirmatory Factor Assessment

Though we often consider of of Confirmatory Component Examination (CFA) as data reduction, it is really one particular of the main kinds of structural equation types. I uncovered about Exploratory Issue Investigation right before CFA, and even now use equally in most of the jobs I’m involved in.

CFA made me believe, at the beginning, of a whole “targeted” EFA. In essence, you only estimate the paths that website link every single indicator (aka noticed variable) to their corresponding issue (aka latent variable).

CFA is however a bit exploratory, the confirmatory actually speaks to the knowledge you have of the principle at the rear of the romantic relationship in between the constructs (represented by aspects) and their noticed indicators. Below is a simple path diagram of a two-variable CFA:

CFA is also acknowledged within SEM as the measurement product.  It is the action taken to figure out how the things (ε1 and ε1) are calculated by the indicators (x1 to x8).

Latent Variable Structural Model

The subsequent stage is to fit the structural product, which is what you possibly assume of when you listen to about SEM. It is generally making use of the calculated latent variables within just the path analysis framework.

The moment you have declared the latent variables you can hypothesize and take a look at their interactions. Here is an illustration of a entire latent variable structural equation model (observe the similarity with the instance of path analysis higher than):

Growth Curve Designs

An additional well known application of Structural Equation Modeling is longitudinal designs, usually referred to as Growth Curve Models. Presented you have several observations of the exact variable over time, you can declare an intercept and a slope for the subjects’ trajectories in excess of time as latent variables by constraining the route coefficients in a specific way (see diagram below).

Mainly because the paths are constrained, we are trying to estimate on development curve modeling the indicates of the latent variables.  These implies give us the all round intercept and the general slope across all topics.

Latent Progress Curve Products are linked to and an alternate to functioning Blended Models on longitudinal knowledge.  These combined models are normally termed Personal Progress Curve Versions.

By Manolo Romero Escobar

3 Disregarded Strengths of Structural Equation Modeling

Confirmatory component analysis (CFA) and route designs make up two core constructing blocks of SEM. Understand how these support you realize how SEM is employed.