Language
Country/Area
No result found
How did Jöreskog revolutionize measurement in psychological research in 1969?
In 1969, a Swedish psychologist named Karl Jöreskog proposed a new research method—Confirmatory Factor Analysis (CFA). The method has quickly become an important tool in social science research because of its ability to test whether a set of measures is consistent with the researcher's understanding of a particular construct. CFA not only challenged the mainstream measurement methods at the time, but also opened up new ideas for psychological research.
The primary purpose of confirmatory factor analysis is to assess the fit between the data and a hypothesized measurement model based on theory and prior analytical research.
In the process of CFA, the researcher first establishes a hypothesis about the underlying factors, such as considering "depression" as the factor behind certain depression scales (such as the Beck Depression Scale and the Hamilton Depression Scale). By imposing constraints on the model, researchers are able to force the model to conform to their prior theoretical assumptions. If the fit between the data and the model is not good, the researcher may need to rethink the construct or consider another model.
For example, when researchers believe that there are two independent factors, they can create a model that constrains the correlation between the two factors to be zero. The resulting fit test will assist them in assessing whether the proposed model accurately captures the co-variation among all items.
"Mismatch between models may result from some items measuring multiple factors, or some items being more correlated than others."
Compared with Exploratory Factor Analysis (EFA), CFA is a more theory-driven analysis method. The goal of EFA is to automatically identify factors based on data, while CFA requires researchers to make assumptions about the number of factors and their correlations before analysis. This makes CFA critical in determining the validity of the measurement model.
It is worth noting that Jöreskog's CFA method replaces earlier methods, such as the multi-trait multi-method (MTMM) matrix, and provides a more reliable construct validity analysis in the case of spurious similarity. Such changes provide more precise and clear tools for psychological and social scientific research, further promoting the development of the discipline.
"The core of CFA is to test and adjust measurements based on theory."
Since the widespread use of CFA methods, researchers have realized that relying solely on one analytical method can lead to serious misunderstandings. For example, when dealing with non-normal data, the traditional maximum likelihood estimation (ML) method will experience bias, which has prompted the development of other alternative estimation methods, such as weighted least squares (WLS).
This series of developments does not stop at the evolution of methodology. The close connection between CFA and Structural Equation Modeling (SEM) also provides researchers with deeper insights. As a measurement model in SEM, CFA helps reveal the interaction between potential variables through clear causal relationships. This not only has great significance for basic research, but also provides a basis for the design of psychological intervention measures.
When evaluating model adaptation, researchers need to pay attention to a variety of indicators, such as chi-square test, root mean square error (RMSEA) and comparative fit index (CFI). These indicators jointly promote the optimization of the model and can Help researchers remain rational and objective in the process of diligent inquiry. A poorly fitting model indicates that researchers may need to reexamine or adjust their assumptions.
“Good model fit indicates that the model is reasonable, but this does not mean that the model is correct.”
Through Jöreskog's innovation in 1969, research methods in psychology changed dramatically. From then on, data analysis is no longer just a quantitative challenge, but a more complex process full of thinking and testing. As time goes by, CFA is not only respected by the academic community, but also gradually becomes an important tool in various related fields.
So, with the advancement of research technology, will more innovative methods appear in the future to subvert our understanding of psychological research?
