introduction-eng

# introduction-eng - Biostatistics Prof dr Ann Vanreusel...

This preview shows pages 1–9. Sign up to view the full content.

Biostatistics - Descriptive - Experimental and sampling design - analysis of variances - correlation - regression - cluster - ordination Websites : www.statsoft.com Î electronic statistic textbook Biostatistical analysis JH Zar (75 EURO) Prof dr Ann Vanreusel

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Scientific research Objective Hypothesis Sampling or Experiment Processing data by statistics Î to test hypothesis Î data collection Interpretation Î discussion and comparison with literature Î presentation Identified based on base line or pilot study Nul hypothesis = mostly conservative Consider available statistics and their assumptions Correlation or manipulation Dataexploration Assumptions Objective measure for reliability Graphical presentation
Statistics is only a mean to interpret data (is never an objective) Impossible to investigate complete populations. Therefore samples are collected Not always possible to do in situ observations Therefore experiments are performed Statistics are a mean to see howfar observation is reliable (true for the whole population). Statistics are a mean to recognize and describe patterns. Parameters describing and characterising the data (greec letters) Estimation or statistics (latin letters)

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Examples Examples In situ - sampling community analysis population dynamics ….. - observations community analysis behavior functional morfology ……. - experiments -manipulations In situ Labo
- type data: variables Discrete or continuous ? e.g. counts versus measurements Nominal, Categorical e.g. Color, sex, .. - scales Ratio scale Î Constant interval size, 0 value Interval scale Î Constant interval size , no 0 value eg time scale eg 40° is not twice as warm as 20° Ordinal scale Î Ranking Î Less information Nominal scale Î Nominal data (eg 30 cm is half of 60 cm)

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Station 1 Station 2 Station 3 Station 4 Station 5 Station 6 spec 1 1 1 1 1 0 1 spec 2 1 1 1 1 0 1 spec 3 0 0 1 1 0 1 spec 4 1 1 1 1 1 0 Station 1 Station 2 Station 3 Station 4 Station 5 Station 6 spec 1 7 61 50 11 0 1 spec 2 4 13 155 6 0 4 spec 3 0 0 106 2 0 1 spec 4 5 42 100 13 1 0 Station 1 Station 2 Station 3 Station 4 Station 5 Station 6 spec 1 43,8 52,6 12,2 34,4 0,0 16,7 spec 2 25,0 11,2 37,7 18,8 0,0 66,7 spec 3 0,0 0,0 25,8 6,3 0,0 16,7 spec 4 31,3 36,2 24,3 40,6 100,0 0,0 Station 1 Station 2 Station 3 Station 4 Station 5 Station 6 spec 1 1 3 2 2 0 1 spec 2 1 2 4 1 0 1 spec 3 0 0 4 1 0 1 spec 4 1 2 3 2 1 0 Presence or Absence Counts : densities Relative abundances Coded abundances (classes) 1 : 1-10 2 : 11-50 3 : 51 - 100 4 : > 100 Presentation of data in a datamatrix: rows and columns
1st exploration datamatrix Common (undesirable) characteristics of a dataset : - “ noise” : variation due to measuring errors - overlap or redundancy: 2 or more variables give the same information - outliers : strongly deviating data Variation : - by measuring - genetical variation between organisms - influence of environmental factors Error Effect Variatie

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document