You are on page 1of 23

Statistics

Lecture and exercises (WS 2012/2013)

Dr. Olaf Lenz Institut fr Angewandte Geowissenschaften Angewandte Sedimentgeologie Technische Universitt Darmstadt

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 1

Structure
Basics (3 lectures with exercises)
17.10.2012 24.10.2012 31.10.2012 07.11.2012 14.11.2012 21.11.2012 28.11.2012 05.12.2012 12.12.2012 16.01.2013 23.01.2013 30.01.2013 Introduction on Statistics Data Presentation Requirements of Data for Statistical Analysis

Elementary Statistics (6 lectures with exercises)


t-tests and F-tests Analysis of Variance Correlation and Regression Chi-square Tests Non-parametric Tests Multivariate ANOVA/Repeated Measures

Analysis of Multivariate Data (3 lectures with exercises)


Cluster-Analysis Principal Component Analysis (Detrended) Correspondence Analysis

Time Series Analysis (1 lecture with exercises)


06.02.2013 13.02.2013 Analysis of stationary data: Spectral Analysis Analysis of non-stationary data: Wavelet Analysis

Final exam

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 2

PAST Software

http://folk.uio.no/ohammer/past/index.html or Google: PAST Hammer

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 3

Statistics - Definition

Statistics is the science of making effective use of numerical data relating to groups of individuals or experiments. It deals with all aspects of this, including not only the collection, analysis and interpretation of such data, but also the planning of the collection of data, in terms of the design of surveys and experiments. Classical statistical methods are methods which are concerned with the analysis of empirical (i.e. observed, measured) data.
(Dodge 2003: The Oxford Dictionary of Statistical Terms)

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 4

Statistics Definition II

We use Statistics to draw conclusions about very large groups of individuals (animate or inanimate) when we can only study small samples of them! The questions we are trying to answer: 1. If I assume that the sample of individuals I have studied is representative of the group they come from, what can I tell about the group as a whole? 2. How confident can I be that the sample of individuals I have studied was like the group as a whole?

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 5

Variability

Normal distribution

Frequency distribution
15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 6

Townend (2002)

Normal distribution

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 7

Davis (2002)

Measures of Location: Mean - Median - Mode


What is a typical member of a population?
The mode is the value that occurs with the greatest frequency.

The median is the value midway in the frequency distribution.

The mean is another word for the arithmetic average

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 8

Davis (2002)

Measures of Location: Mean Median - Mode


Data values: 8-12-10-7-7-11-8 n: 7 Sum: 63 Mean: 63/7 = 9

Mean:

Median: Data values: 7-7-8-8-10-11-12 n: 7 Sum: 63 Median = 8

Mode:

The mode is the value that occurs with the greatest frequency.
15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 9

Measures of Spread: Variance Standard deviation


How spread out are the values around the typical member of a population?

Oklahoma oil field

Texas oil field

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 10

Davis (2002)

Measures of Spread: Variance Standard deviation

The corrected mean of squared differences is 24/6 = 4.0 Variance The square root of variance Standard deviation
15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 11

Measures of shape: coefficient of variation


The dispersion in a variable is sometimes given by the coefficient of variation (CV), which is a dimensionless measure of variability expressed as a fraction of the mean. standard deviation CV = mean

Example: Ants standard deviation = 3 mm, mean length = 10 mm Dogs standard deviation = 20 cm, mean length = 100 cm Are ants or dogs more variable in their length? CVants = 3/10 = 0.3 = 30% CVdogs = 20/100 = 0.2 = 20%

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 12

Measures of shape: coefficient of skewness

positively skewed: long tail of high values to the right

skewness close to zero: histogram is approximately symmetric

negatively skewed: long tail of small values to the left

http://www.statistics4u.info/fundstat_eng/cc_skewness.html
15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 13

Measures of shape: Kurtosis


y>0 normal distribution y = 0 y<0

http://www.statistics4u.info/fundstat_eng/cc_kurtosis.html
15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 14

Summary Statistics
Mean Median Mode Quartiles

Measures of location:

location of the center of the distribution location of the other parts of the distribution

Measures of spread:

Variance Standard deviation Interquartile range

variability of the data values

Measures of shape:

Coefficent of skewness Coefficient of variation Kurtosis length of the tail

symmetry

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 15

Standard error and 95% confidence interval

population mean

sample mean

sample mean is an estimate of the population mean


15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 16

Townend (2002)

Standard error and 95% confidence interval


Temperature interval: 14.5 16 C Population mean: 15.2 C Margin of error: 16 - 15.2 C = 0.8 C

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 17

Townend (2002)

Standard error and 95% confidence interval

standard error

frequency distribution for a population

frequency distribution for the means of samples of 5 individuals

frequency distribution for the means of samples of 10 individuals

s.d. standard deviation

standard deviation of sample means standard error

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 18

Townend (2002)

Standard error and 95% confidence interval

Example:

There is approximately a 68% change that the true population mean lies in the range 9.0 +/- 0.76 = between 8.24 and 9.76
15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 19

Difference between standard deviation and standard error

Standard deviation: Standard deviation is a measure of how much deviation there is between individuals in a population.

Standard error: Standard error is a measure of the margin of error involved in estimating the mean of a population.

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 20

PAST

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 21

PAST Univariate Statistics


1 2

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 22

Next week
Basics (3 lectures with exercises)
17.10.2012 24.10.2012 31.10.2012 07.11.2012 14.11.2012 21.11.2012 28.11.2012 05.12.2012 12.12.2012 16.01.2013 23.01.2013 30.01.2013 Introduction on Statistics Data Presentation Requirements of Data for Statistical Analysis

Elementary Statistics (6 lectures with exercises)


t-tests and F-tests Analysis of Variance Correlation and Regression Chi-square Tests Non-parametric Tests Multivariate ANOVA/Repeated Measures

Analysis of Multivariate Data (3 lectures with exercises)


Cluster-Analysis Principal Component Analysis (Detrended) Correspondence Analysis

Time Series Analysis (1 lecture with exercises)


06.02.2013 13.02.2013 Analysis of stationary data: Spectral Analysis Analysis of non-stationary data: Wavelet Analysis

Final exam

15. Oktober 2012 | Fachbereich 11 | Angewandte Geowissenschaften | Dr. Olaf Lenz | 23

You might also like