Table 3. Comparison of socio-economic characteristics of Kenyan coastal households near and far from marine reserves, before and after matching.

Characteristic	Sample	Mean treatment (marine reserve)	Mean control	Difference in mean values	Mean eQQ differencea	Mean eCDF differenceb	t	t-test, p-value	KSc	KS, p-value
Household size, people	After matching	7.35	7.17	0.19	0.70	0.02	0.55	0.59	0.06	0.99
	Before matching	7.35	6.31	1.05	1.20	0.05	1.43	0.16	0.15	0.27
No. of jobs	After matching	2.67	2.46	0.20	0.20	0.03	0.95	0.34	0.11	0.37
	Before matching	2.67	2.14	0.53	0.57	0.08	2.73	0.01	0.21	0.02
Wealth	After matching	−0.97	−0.86	−0.10	0.23	0.04	−0.71	0.48	0.09	0.89
	Before matching	−0.97	0.88	−1.85	1.78	0.25	−4.94	<0.001	0.46	<0.001
Fortnightly expenditures, KES	After matching	4739.19	4633.85	105.33	321.37	0.03	0.97	0.34	0.13	0.72
	Before matching	4739.19	4597.40	141.79	281.48	0.02	0.28	0.78	0.08	0.97

^aMean difference in the empirical Q-Q plot of treatment and control groups.

^bMean difference in the empirical Cumulative Distribution Functions of treatment and control groups.

^cNon-parametric Kolmogorov-Smirnov test between treament and control groups.

Statistical matching (see Methods) identified pairs of similar households and reduced differences in key characteristics between households near and far from marine reserves to allow a relevant comparison of food security metrics using paired t-tests (Figure 2).