KNOWLEDGEBASE - ARTICLE #1701
The five kinds of data tables in Prism
To use GraphPad Prism effectively, you need to understand the differences between the five kinds of data tables it provides. Here is a summary of the differences between them:
XY Table
Data set columns: Different treatments
Rows: Different X values
Values: One column for X. Rest for Y.
Subcolumns: Replicates to compute error bars
Analyses: Linear or nonlinear regression.
Column Table -- One grouping variable
Data set columns: Different treatments (i.e. control vs. treated).
Rows: Replicate measurements
Values: Something you measured or assayed
Subcolumns: None; error bars computed from values stacked in a column
Analyses: t tests, one-way ANOVA
Grouped Table -- Two grouping variables.
Data set columns: Different levels of factor A (i.e. control vs. treated)
Rows: Different levels of factor B (i.e. male vs. female)
Values: Something you measured or assayed
Subcolumns: Replicates to compute error bars
Analysis: Two-way ANOVA
Contingency Table
Data set columns: Alternative outcomes
Rows: Alternative treatments or exposures
Values: Actual numbers (counts) of subjects or experiments.
Subcolumns: None
Analyses: Chi-square and Fisher's exact test
Survival Table
Data set columns: Alternative treatments
Rows: Different subjects.
Values: X is time; Y is codes for death (1) and censored (0)
Subcolumns: None; error bars can be computed from all the data
Analyses: Kaplan-Meier, Logrank test