Which statistical test should I use?

Made by Platelet

Which statistical test should I use?

Answer a few questions about your data and we'll point you to the right test.

Normalising plate readings: methods and when to use them

Made by Platelet

Normalising plate readings: methods and when to use them

From percent-of-control to Z-score: a practical guide to removing plate-to-plate variation without removing your signal

Choosing the right graph for your ELISA results

Made by Platelet

Choosing the right graph for your ELISA results

Bar charts, box plots, and dot plots: when each one is appropriate and how to avoid misleading your audience

What does a p-value actually mean?

Made by Platelet

What does a p-value actually mean?

What it is, what it isn't, and how to read p = 0.03 without overstating your result

How do I report effect size and confidence intervals?

Made by Platelet

How do I report effect size and confidence intervals?

Why the magnitude and precision of an effect matter more than the p-value alone

How many samples do I need?

Made by Platelet

How many samples do I need?

Power analysis explained, and why you calculate sample size before, not after, the experiment

When should I use a t-test?

Made by Platelet

When should I use a t-test?

One-sample, paired, and unpaired t-tests, and why Welch's is the safer default

When do I use ANOVA instead of a t-test?

Made by Platelet

When do I use ANOVA instead of a t-test?

Comparing three or more groups, choosing a post-hoc test, and reading interactions in a two-factor design

When should I use a non-parametric test?

Made by Platelet

When should I use a non-parametric test?

Mann–Whitney, Wilcoxon, and Kruskal–Wallis: the rank-based alternatives, what they actually test, and when a transform beats them

What's the difference between correlation and regression?

Made by Platelet

What's the difference between correlation and regression?

Pearson vs Spearman, interpreting R², and why correlation isn't causation

How do I compare proportions and categorical data?

Made by Platelet

How do I compare proportions and categorical data?

Chi-square, Fisher's exact, and McNemar's, and the effect sizes to report with them

How do I analyze survival data?

Made by Platelet

How do I analyze survival data?

Kaplan–Meier curves, the log-rank test, Cox regression, and what censoring means

What if my data aren't normally distributed?

Made by Platelet

What if my data aren't normally distributed?

Checking normality and variance with plots, and your options when assumptions fail

Do I need to correct for multiple comparisons?

Made by Platelet

Do I need to correct for multiple comparisons?

When to use Bonferroni and family-wise control versus FDR for large-scale screens

What counts as my n?

Made by Platelet

What counts as my n?

Technical vs biological replicates, and how pseudoreplication fakes statistical power

SD, SEM, or confidence interval: which error bar?

Made by Platelet

SD, SEM, or confidence interval: which error bar?

Choosing honest error bars and showing your data instead of hiding it behind a bar

How do I report statistics in a paper?

Made by Platelet

How do I report statistics in a paper?

What to put in methods and results so a reader knows exactly what you did and found

Which statistical test should I use for my ELISA data?

Made by Platelet

Which statistical test should I use for my ELISA data?

A practical decision framework for plate-based immunoassays

Importing and validating fluorescence readings from plate readers

Made by Platelet

Importing and validating fluorescence readings from plate readers

What to check before your raw OD or RFU values ever touch an analysis pipeline

Understanding Z′-factor in high-throughput screening

Made by Platelet

Understanding Z′-factor in high-throughput screening

The single most important quality metric for your assay: what it measures, how to calculate it, and what to do when it fails

96-well plate layout: design principles that protect your data

Made by Platelet

96-well plate layout: design principles that protect your data

How to arrange samples, controls, and replicates to minimise systematic bias before you pipette a single well