research

Statistical analysis of data in Life Sciences

Usually I collaborate in research projects conducting the statistical analysis of data (statistical inference, multivariate analysis, and methods of descriptive statistics) .   

• The problems that I have worked exhibit very different nature: physiology of fungi [1c, 2c], plant virology [4c], population ecology [3c, 5c], populations of cells [6c], geology [7c], botany [8c, 9c], vascular system in mammals [10c], diagnosis of depression through the analysis of laughter [11c, 12c].
  

• An important issue is the analysis of data from computer simulation experiments in artificial life and artificial intelligence. Obviously these data differ from those obtained in experiments performed in nature or in the laboratory. Statistical analysis of data from simulation experiments requires some special care: in [21] we reviewed some general protocols applied to evolutionary computation experiments.
  

A book about statistical methods in behavioral biology

In 1995 I wrote in collaboration with specialists in animal behavior the book entitled Statistical Methods in Behavioral Biology (Transl:. In spanish, 'Métodos Estadísticos en Biología del Comportamiento').

The book, eminently practical, deals with a simple and direct language, the theoretical foundation and application of statistical techniques in the field of Behavioral Biology, whether animal or human. The book is primarily aimed at students from Psychology, Biology, Medicine and, in general, all those who are interested in the statistical analysis. The book contains many examples that have been taken from these disciplines, to show the use of statistical methods. As innovation is noteworthy that is the first book in spanish incorporating a 'floppy disk' with the programs for the statistical tests described therein, and is the first to describe circular statistics techniques, useful in the study of animal orientation and biological rhythms.

The book is accompanied by 'PhiStat 1.0' program: a collection of programs in Turbo Pascal (MS-DOS). The program includes different techniques ranging from classical statistical tests (ANOVA, chi-square test, ANOVA and regression analysis, etc.) to circular statistics techniques (periodic regression, Hotelling test, the cosinor method, etc.). It also includes 'Dr Hypo' , a tool for interactive solving hypothesis testing and various utilities (T-test for paired observations, etc.) as well as routines for producing statistical tables (Z, F Fisher T-student, chi-square, etc.).

[HTML]  R. Lahoz-Beltra, J. Ortega, C. Fernández-Montraveta. 1995. Métodos Estadísticos en Biología del Comportamiento. Editorial Complutense (ISBN  84-7491-512-0), pp 1-232.

From statistical analysis to 'Big Data': a book with programs in R

In 2019 the book is published, the cover image of which is shown on the right. The book includes R programs on the elementary techniques of statistical inference with one and two populations as well as the usual statistical tests with percentages. The book introduces analysis of variance and linear regression. In a second part, the book introduces some Big Data statistical techniques: decision trees, vector support machines, cluster analysis and artificial neural networks.

[HTLM] R. Lahoz-Beltra. 2019. En las entrañas del big data. Una aproximación a la estadística. EMSE Publishing. Colección "Grandes Ideas de las Matemáticas" (ISBN 978-84-17811-471-1), pp. 1-144.

Statistical analysis of laughter and diagnosis of depression

There are no studies in psychiatry attempting the use of laughter as a diagnostic tool. In the present study [11c, 12c, 15c] we extracted nine temporal and acoustic variables from laughter episodes, looking for establishing better pattern-classification methods. These variables are number of bouts, voiced percentage, bout duration, mean energy, mean entropy, standard deviation of fundamental frequency, jitter, shimmer and harmonic to noise ratio. A library of records of laughter episodes was created and the values of the previous variables were statistically analyzed. We found that laughter may be applied as a diagnostic tool in the onset and evolution of depression and, potentially, of neuropsychiatric pathologies.

We registered laughs of depressed patients (n=30) and healthy controls (n=20), in total 934 laughs (517 from patients and 417 from controls). All patients were tested by the Hamilton Depression Rating Scale (HDRS). The processing was made in Matlab, with calculation of 8 variables per laugh plosive. General and discriminant statistical analysis distinguished patients, controls, gender, and the association between laughter and HDRS test. Depressed patients and healthy controls differed significantly on the type of laughter, with 88% efficacy. According to the Hamilton scale, 85.47% of the samples were correctly classified in males, and 66.17% in women, suggesting a tight relationship between laughter and the depressed condition.

Right Figure .- Sonogram of laughter recorded from (a) depressed patient and (b) healthy or normal control subject.