← All apps

Sixty-two mice

Sixty-two cards, one per mouse. Thirteen are dealt to the E. coli group, and then twenty-seven mice develop a tumour, chosen completely at random with no regard for who was exposed. Exposure is fixed by the experiment; only the tumours are shuffled. That is the null hypothesis, made concrete.

Your shuffle

Germ-free, no tumour Germ-free, tumour E. coli, no tumour E. coli and tumour

–

of your 13 exposed mice developed a tumour

Your table

Response	E. coli	Germ-free	Total
Tumour	–	–	27
No tumour	–	–	35
Total	13	49	62

Draws

pooled so far

With T ≥ 8

as extreme as the real study

Class estimate

–

P(T ≥ 8) from these draws

Exact answer

0.124

hypergeometric

Distribution of T across the class

T < 8 T ≥ 8, at least as extreme as the real study

No draws yet. Bars appear here as you add them.

Running estimate of P(T ≥ 8)

class estimate exact value, 0.1239

The curve starts once you add the first draw.

Every draw

T	Students	Share