🚀 go-pugleaf

Dear stat-l readers,

I'd like to improve the effectivity of my postgraduate course entitled
"Experimental design and analysis by the methods of computer biometry".
I would be grateful for your comments and reflections on the new curriculum
of the course I've put together. This course is for those young researchers
(graduated in medicine or in biology) who study for the Ph.D. degree.
They work in different fields from morphology to psychophysiology.

The tentative curriculum:
____________________________________________________________________________

1.) Fundamental principles in experimental design and elements
    of probability theory
          - the random nature of the phenomena, experiments,
            random sampling, random variables
          - laws of probability
          - types of variables (scales)
          - frequencies,histograms,stem-and-leaf plots,important theoretical
            distributions (binomial,normal,Poisson)
          - descriptive statistics (first and second moments of the prob.
            distribution function as measures of location and dispersion)
          -association of variables (scattergram, covariance, correlation,
           distances)
          -uni- and multivariate experiments, explorative methods
           (graphical techniques: Chernoff faces and Voronoi tessellation,
           elements of clustering and multidimensional scaling)
          -multinomial experiments, contingency tables, Chi-square test,
           goodness of fit- and homogeneity tests

2.) Statistical inference: samples from two populations
          -expectation, estimation, criteria of "good" estimation
           (bias, effectivity, consistency)
          -least squares- and max. likelihood estimation
           (outlook: vector and matrix algebra)
          -sampling distributions, the central limit theorem
          -hypothesis testing, type of errors, power of the test,
           the concept of confidence intervals
          -inferences about population mean and variance
           (parametric and nonparametric tests, independent and correlated
           samples, Student's t and F distributions)
          -multivariate methods (Hotelling's T^2

3.) Statistical inference: samples from more than two populations
                           (and multiple factor experiments)
          -one-way analysis of variance
          -multiple comparisons
          -a posteriori comparisons (contrasts)
          -fixed- and random effects model
          -factorial designs (interaction terms)
          -other ANOVA designs (nested, randomized design, latin squares,
           split-plot design)
          -repeated measures ANOVA
          -testing the assumptions of ANOVA - nonparametric methods
          -MANOVA

4.) Association between variables
          -linear regression
          -inferences related to regression, and correlation
           (Pearson, Spearman tests)
          -nonlinear regressions: transformation of variables,
           curve fitting, decomposing of curves, dose-response curves,
           logit, probit models
          -ANCOVA, trend, orthogonal polynomials
          -time series analysis, prediction (ARIMA models)
          -multivariate regression, multiple and partial regression
          -canonical correlation

5.) Latent factors and structures in the experimental data
          -dimensionality determination and reduction:
           principal component analysis
          -discovering latent factors: factor analysis
          -classification of data: discriminant function analysis
_____________________________________________________________________________

Thanks for reading through (and for the comments in advance).
Even unfavourable critique would be worth for me.

Nora Szilagyi                                    neurocomp@ludens.elte.hu
Dept. of Comparative Physiology
Eotvos Lorand University
1088. Budapest, Muzeum krt. 4/A
Hungary