# Interactive Graphs for Statistics for Social Science

### Martin Poulter, DeSTRESS Project

These graphs redraw themselves dynamically in response to user clicks (or touches, if you're using a tablet device). They are web pages with embedded programs, requiring no plug-ins or special software. They may be quite slow in some versions of Internet Explorer.

Licences: Creative Commons Attribution-ShareAlike (text content); MIT licence (Javascript code).

1. Effect on trendlines of censoring data
regression line; effect of choosing starting point

2. Evaluating the Rule of 70
exponential growth; percentage error

3. Mean versus median: film ratings
robustness of median relative to the mean

4. Pearson versus Spearman correlations
types of correlation; similarities and differences

5. Skew, mean and median
mode, mean, median of continuous (triangular) distribution

6. Skewness in a discrete distribution (counterpoint to above)
mode, mean, median of discrete (Poisson) distribution

absolute and percentage change; range, s.d. and variance

8. The power of compound interest
logarithmic axis

9. Logarithmic axes and income
measures of spread; absolute change versus percentage change

10. Arithmetic versus geometric mean
probability; expected return; geometric mean as long-term expected return

11. The shape of the sampling distribution
sampling from a skewed distribution; distinguishing the population, sample and sampling distribution

12. Distribution of sample standard deviations and why we use n-1 to calculate sample sd (incomplete)
sampling; Bessel's correction

13. Correlation of a sample (incomplete)

### University of Sussex

Kyle Siegrist, University of Alabama in Huntsville

This site aims "to provide free, high quality, interactive, web-based resources for students and teachers of probability and statistics" It covers probability, statistics and "special models" such as finite sampling models and the Poisson process - they are presented here with text and interactive Java tutorials. Each page suggests mathematical exercises and simulation exercises, with some external links. The site requires the Mozilla Firefox browser (version 1.5 or later), with the MathML fonts installed, and with the Java plug-in (version 1.5 or later).

Douglas Stirling, Massey University

This is a complete Java-based tutorial in introductory statistics, which can be used online or downloaded for offline use. It exposes the students to many data sets and uses interactive graphs which the student can alter by sliding controls. You must register to use this site.

Carnegie Mellon University

This is an open online course, available freely to independent learners but with a fee for students. The course makes much use of online text, diagrams and interactive assignments. It is comprised of four units: Exploratory Data Analysis, Producing Data, Probability and Inference. These are subdivided into a total of twelve modules and more than two hundred "pages" of material.

Creative Commons Attribution NonCommercial ShareAlike (CC-BY-NC-SA)
Carnegie Mellon University

This is an open online course, available freely to independent learners but with a fee for students who want feedback from an instructor. The course makes much use of online text, diagrams and interactive assignments. It is comprised of four units: Exploratory Data Analysis, Producing Data, Probability and Inference. These are subdivided into a total of twelve modules and more than two hundred "pages" of material. It is similar to the companion course on Statistical Reasoning, but with a more classical treatment of probability.

Creative Commons Attribution NonCommercial ShareAlike (CC-BY-NC-SA)
Richard Field, John Horton, Sue Cobb, Sandra Hill, University of Nottingham

An interactive tutorial combining graphs, text and interactive features. Twenty-five units cover topics from rounding and central tendency up to regression lines and Chi squared tests

Creative Commons Attribution NonCommercial ShareAlike (CC-BY-NC-SA)
Philip B. Stark, University of California, Berkeley

SticiGui: statistics tools for internet and classroom instruction with a graphical user interface is an on-line course, including a textbook and other statistical resources, and covers a spectrum of statistical applications, not just economics. It has been produced by Philip B. Stark of University of California, Berkeley. The site makes heavy use of Java / JavaScript to supply interactivity and the site recommends using the Firefox browser to work through the course.

Not specified

This is a set of online tools for performing statistical tests, generating random data or exact distributions, along with links to short explanations of statistical topics.

Stattucino

Stattucino is a collection of interactive applets and which illustrate some canonical distributions, statistical analysis and regression analysis. The site requires that a Java plugin is installed in your browser, and Javascript is enabled.

Webster West, University of South Carolina

On these pages are several Java applets with which to illustrate various statistical concepts such as regression, confidence limits and the Central Limit Theorem. The applets "provide interactive instruction on the Web and also to provide access to research procedures".