These are similar to the standard tables seen in books. You can zoom these to the full size of you Macintosh window for easier viewing if desired. Additionally, you can use the scroll bar at right to view lower parts of the tables. To view other tables either use the submenu options under the Extras menu again, or click and hold the mouse on the title of a table (which acts as a pop up menu) and select the table you wish to see.
Below each title is a line of bold text which shows the meaning of each column in the table below. The available tables are the normal, inverse normal, chi-square, t and F distributions, and the non-parametric Wilcoxon sign-ranked matched-pairs and Mann-Whitney U tables.
Note that the t-distribution table shows only the one-tailed cumulative probability distribution, so the p values at the top must be carefully considered with respect to your specific problem, eg if you want a two tailed significance level of 0.95 then you should seek the critical value under the 0.975 column (at which there is 0.025 above the critical value and the same amount below the negative equivalent of this value). INF means infinity.
Other tables from the discrete probability distributions (binomial, poisson, hypergeometric and geometric) are also available by using the "Active Tables" option, typing in the parameters defining that specific exemplar of the distribution and then clicking the Generate Table button.
The Mann Whitney table is used in conjunction with the Mann Whitney U test analyses. The table gives the one-tailed critical values for the test statistic calculated by SchoolStat‚Ñ¢ for the two samples (for up to 10 data points in each). If the calculated value falls outside the critical values in the table for the 0.10 or 0.90 probability levels, then it is possible that the two independent samples are statistically significantly different. This is a discrete table of critical values so if your calculated value falls exactly on one of the table's critical values, you should accept the probabilty which is the next least significant.
The Wilcoxon Matched table is similar giving the discrete probability cut-off values for 5 to 25 paired samples.
The Kruskal Wallis table is in preparation for ANOVA.
