Quantify precision of fits with SE or CI of parameters.Automatically graph curve over specified range of X values.Accept automatic initial estimated values or enter your own.Differentially weight points by several methods and assess how well your weighting method worked.Compare models using extra sum-of-squares F test or AICc.Automatic outlier identification or elimination.Global nonlinear regression вЂ“ share parameters between data sets.Enter different equations for different data sets.Enter differential or implicit equations.Now including family of growth equations: exponential growth, exponential plateau, Gompertz, logistic, and beta (growth and then decay). Fit one of our 105 built-in equations, or enter your own.Comparison of data from nested data tables using nested t test or nested one-way ANOVA (using mixed effects model).Analysis of repeated measures data (one-, two-, and three-way) using a mixed effects model (similar to repeated measures ANOVA, but capable of handling missing data).Three-way ANOVA (limited to two levels in two of the factors, and any number of levels in the third).Tukey, Newman-Keuls, Dunnett, Bonferroni, Holm-Sidak, or FisherвЂ™s LSD multiple comparisons testing main and simple effects. Two-way ANOVA, with repeated measures in one or both factors.Two-way ANOVA, even with missing values with some post tests.Calculate the relative risk and odds ratio with confidence intervals. Fisher's exact test or the chi-square test.Kruskal-Wallis or Friedman nonparametric one-way ANOVA with Dunn's post test.When this is chosen, multiple comparison tests also do not assume sphericity. Greenhouse-Geisser correction so repeated measures one-, two-, and three-way ANOVA do not have to assume sphericity.Many multiple comparisons test are accompanied by confidence intervals and multiplicity adjusted P values.One-way ANOVA without assuming populations with equal standard deviations using Brown-Forsythe and Welch ANOVA, followed by appropriate comparisons tests (Games-Howell, Tamhane T2, Dunnett T3).

Ordinary or repeated measures ANOVA followed by the Tukey, Newman-Keuls, Dunnett, Bonferroni or Holm-Sidak multiple comparison tests, the post-test for trend, or FisherвЂ™s Least Significant tests.

Perform many t tests at once, using False Discovery Rate (or Bonferroni multiple comparisons) to choose which comparisons are discoveries to study further.

Wilcoxon test with confidence interval of median.

Kolmogorov-Smirnov test to compare two groups.

Nonparametric Mann-Whitney test, including confidence interval of difference of medians.

Automatically generate volcano plot (difference vs.

Reports P values and confidence intervals. The corresponding confidence intervals are shown as 'very wide' with no numerical range (the range would be infinitely wide). When the best fit value is preceded by ~, so is the standard error. Other values would generate the same curve, or one that fits just as well. The parameter values preceded by ~ define the curve that Prism creates, but aren't actually useful in other ways. In other words, many combinations of parameter values lead to curves that fit equally well. But when the fit is 'ambiguous', changing other parameters can move the curve so it is near the data again. Changing the value of any parameter will always move the curve further from the data and increase the sum-of-squares. If you choose (on the Confidence tab) to use the older method of identifying some fits as 'ambiguous', then if a value is preceded by ~, it means the results are 'ambiguous'. Why is there a ~ symbol in front of some values? If you choose the recommended method, difficult parameters (and their confidence interval) will be marked as 'unstable'. On the confidence tab of the nonlinear regression dialog, you choose how Prism should deal with difficult fits.

Should you be fitting a family of datasets together using global nonlinear regression? Why does Prism say 'unstable' rather than reporting a best-fit value?.

Should you be constraining a parameter (or several) to constant values? If not, should you be constraining them to have a range of values (only positive numbers, perhaps)?.

Does your data actually define all the parameters?.

If a value doesn't make any sense, ask your self these questions:

What if a parameter value makes no biological or chemical sense? Next, make sure each value makes biological or scientific sense. First make sure you know what units each parameter is expressed in (if there are any units some parameters are unitless constants). Before trying to interpret the rest of the results, first look at the best-fit value of each parameter in your model.

Graphpad free trial