“Saturar” means in this context “to load” (here: loaded), and “saturación” is “loading”.

I was actually expecting to see the term “factor” present in your excerpt, and it does appear once in it. Hence I infer that your text deals with the notion of factor analysis, a technique employed in statistics to limit the number of variables into a reduced number of factors to manage statistical data more easily.

The fact that another very technical term, “varianza” (= variance), is also used here, confirms my suspicion that your texts is about statistical analysis and factor analysis.

Here is one example with presence of the key terms “factor analysis”, “items” and “loading”:
https://core.ac.uk/download/pdf/301346386.pdf
Measuring Organizational Level IS Usage and Its Impact on Manufacturing Performance
(…)
Construct-Level Exploratory Factor Analysis. In this step, all the remaining 22 ISU items were submitted to construct-level exploratory factor analysis to check for discriminant validity of the measurement instrument. Four factors emerged from the factor analysis with all factor loadings above 0.50 and most above 0.60. Serious cross-loading occurred on item INI7. Hence item INI7 was dropped. The remaining 21 items were again put into construct-level exploratory factor analysis. This time four clear factors emerged with all items loaded correctly on the expected dimensions. Most factor loadings were above 0.60. No cross-loading was observed. The KMO measure of 0.93 indicated outstanding sampling adequacy. The final set of measurement items for the ISU construct organized by factor loadings are shown in Table 1.

I am not sure if you will be able to open this link, but here it is:
https://www.raco.cat › article › download
Saturaciones factoriales e índices de discriminación en la Teoría clásica del test y en la Teoría de respuesta a los items
(…)
Se presenta un estudio relacional que trata de mostrar en forma integrada las equivalencias existentes entre las saturaciones en el modelo de un factor común, los índices de discriminación en el modelo de dos parámetros en la Teoria de Respuesta a los ítems y las correlaciones item-total en el análisis de reactivos basado en el modelo clásico. Las relaciones obtenidas se ilustran con un ejemplo numérico.
Palabras clave: Análisis factorial, teoria de respuesta a los items, analisis de reactivos, unidimensionalidad, fiabilidad.

The aim of this paper is to show the relationships existent among the loadings in the one common factor model, the discrimination indexes in the two parameter Item Response Model and the item-total (biserial and point biserial) correlations in the classical item analysis. A numerical study which illustrate the equivalences is also presented.
Key words: Factor analysis, Item Response Theory, Item Analysis, Unidimensionality, Reliability.


--------------------------------------------------
Note added at 57 mins (2022-12-13 18:38:07 GMT)
--------------------------------------------------

Correction: (...) that your text is about (...)