How to Make Recommendations for Educational Practice from Correlational Data Using Structural Equation Models
METADATA ONLY
Loading...
Author / Producer
Date
2023-06
Publication Type
Other Journal Item
ETH Bibliography
yes
Citations
Altmetric
METADATA ONLY
Data
Rights / License
Abstract
In this commentary, we outline a five-phase process by which recommendations for educational practice can be distilled from correlational data using structural equation modeling (SEM). First, meta-theoretical beliefs associated with latent variables-that mental attributes cause behavior and can therefore be measured indirectly by observing multiple indicators of that behavior-must be adopted and made explicit. Next, an SEM must be formulated with relevant pathways and covariates that exhaustively represent our theoretical knowledge and assumptions about the structure of the psychological phenomena being studied. Third, model-data-fit indices and estimated parameters associated with the SEM should be carefully interpreted. Fourth, the model should be replicated across educational contexts, and any necessary changes should be incorporated into the relevant psychological theory. Fifth, the results of multiple studies can then be interpreted together with other sources of evidence as a basis for communicating our current theoretical understanding and caveats to practitioners. We also point out that educational recommendations should likely never be entirely prescriptive, and instead lie on a continuum of specificity based on the strength of the evidence.
Permanent link
Publication status
published
External links
Editor
Book title
Journal / series
Volume
35 (2)
Pages / Article No.
48
Publisher
Springer
Event
Edition / version
Methods
Software
Geographic location
Date collected
Date created
Subject
Structural equation modeling; Recommendations for practice; Research design; Causation; Correlational Data
Organisational unit
Notes
Commentary