What becomes really interesting is the inter-play between variables, for example, in rural areas by gender, ethnicity or language and poverty.
Many public health research papers do exactly what Friedrich and colleagues warn for: they describe an empirical relationship between variables – for example, the correlation between availability of potable water and cases of diarrhea – and move directly towards an 'ought' statement – for example, that authorities ought to improve access to potable water.
There are multiple possibilities for integrating omic datasets, but these can be considered essentially to fall into two categories: statistical, that is, relying wholly on data-driven associations between variables (for example, [ 4, 72]); and knowledge-based, that is, using prior knowledge about biological organisation and pathways/networks (for example, [ 73, 74]).
Knowing the structure of such entities and the changes that they can undergo may provide a kind of mechanistic or process-understanding of biological events that goes beyond causal graphs, which merely express relations of causal relevance between variables (for examples see the next section).
The specificity of the effect is somewhat surprising given the moderate associations reported in previous studies between these variables, for example between trait EI and peer-reported aggression (Petrides et al., 2006), between peer reported bullying and the measures of adjustment used in this study (Viding et al., 2009).
Studies that assess the co-variation between two variables: for example, co-variation of functional or structural properties of the brain and a behavioural variable, such as reported stress.
The strength of linear associations between continuous variables, for example age at alcohol and sexual debut, was assessed using correlation coefficients (r), derived by simple linear regression.
The former represent algebraic relations between variables as in quantitative mathematics, for instance, addition, subtraction, and multiplication; the latter describes incomplete knowledge between two variables, for example, the monotonically increasing and decreasing relations, which state that one variable will monotonically increase with the increase/decrease of another.
One of the most widely applied causal inference approaches is MR. If the direction of the association is previously known between two variables (for example, a metabolite and a lipid in a SNP-MET-LIP set), MR can measure the extent of the unconfounded causal relationship using genetic variants as instrumental variables.
Alternatively, there may be an asymmetry in the relationship between the variables: for example, xi might represent the dose of an antihypertensive drug, and yi might represent the change in blood pressure in a group of subjects treated with various doses of the drug.
Multiple linear regression was used to assess the association between skewed continuous variables (for example, ICU length of stay, hospital length of stay) and sVARI while adjusting for a priori covariates.
