Klara A.

The main objective of this project was to identify how personal characteristics and lifestyle habits influence the plasma retinol concentration of patients in the cohort.

I started with a detailed description of quantitative and categorical variables, which allowed for the detection of the intrinsic characteristics of the sample, including the assessment of central tendency measures and the detection of skewness. To test the associations between all pairs of variables, I used a wide range of inferential tests adapted to each data type (Pearson, Welch's t-test, ANOVA, Tukey's HSD, Chi-squared). Finally, I built multivariate models to isolate the effect of each predictor. Linear regression was used to model the continuous value of plasma retinol, and logistic regression, after transforming the dependent variable into a binary indicator, was used to model the probability of reaching a high level of plasma retinol. An exhaustive search for interactions between predictors was carried out for both types of models.

By providing complete R outputs in the appendix and summarizing the results in a clear report, I delivered a robust analysis allowing for a deep understanding of the complex mechanisms governing plasma retinol concentration in this cohort.

Study of the evolutionary and ecological dynamics of cancer, and its links with sexual reproduction

Teaching English up to post-baccalaureate level, in written and oral expression.