Table 3. Testing the effect of participant characteristics and behaviour on the proportion of participants who helped at the second opportunity, P(B).

Effect	Estimate	s.e.	z value	p value
intercept	−1.64	1.34	−1.23	0.22
cost	−0.23	0.05	−4.80	1.60 × 10−6
gender (male)	0.66	0.19	3.41	6.41 × 10−4
age	−0.06	0.01	−6.02	1.76 × 10−9
A occurs	0.99	0.19	5.16	2.52 × 10−7
training 1	0.11	0.13	0.85	0.40
training 2	0.32	0.18	1.80	7.17 × 10−2

Binomial GLM (logit link function). Response variable: Boolean indicating whether event B occurred. The table shows the estimate, standard error and statistical test results for the explanatory variables in the model. To give an example for interpreting this table: participants who helped at the first opportunity (A occurs) were more likely to help at the second opportunity (B occurs) and this effect was unlikely to have arisen by chance (p = 2.52 × 10⁻⁷).