Analysis of an Epidemiological Research Article Identity of the article Zambrana, I. M., Vollrath, M. E., Sengpiel, V., Jacobsson, B., & Ystrom, E. (2016). Preterm delivery and risk for early language delays: A sibling-control cohort study. International Journal of Epidemiology, 151–159.

 
Introduction

At the time of the study, the authors were leading researchers and professionals in various health institutions in Norway and Sweden.
The purpose of the study was to examine the relationship between early gestational age and language outcomes. To achieve this objective, the researchers used a control cohort study design, which they called “sibling-cohort” control design.

Methods
The researchers measured the outcomes of language comprehension and language production as the latent variables based on Confirmatory Factor Analysis (CFA).
The CFA model was chosen as it fits with high non-overlapping factor loadings for the descriptions of language construction, validation as well as the specific items included.

In addition, the researchers were measuring the outcome variables to determine the impact of early gestational age, which means that gestational age was the chief exposure that was being tested.
Nevertheless, gestational age as an exposure variable consists of a set of many variables, including gestational age at birth, malformations at birth, unplanned pregnancy, preeclampsia, urinary tract infections, alcohol intake by the mother during pregnancy, smoking by the mother during pregnancy and others.

The researchers used the Mplus version 7.2 as the statistical tool for data analysis. In addition, the statistical analysis was based on the item response theory approach, that is, the multi-level CFA for categorical data based on link function.

Although the researchers have not indicated he importance of using the Mplus statistical tool, it is evident that it was chosen because the system has features that relate to mediation analysis, latent class modeling and factor analysis