Skip to main content

Table 1 Sample description and comparison of agroecological (n = 61) and reference (n = 30) farmers on study variables

From: Promoting traditional foods for human and environmental health: lessons from agroecology and Indigenous communities in Ecuador

  Descriptive measurements Comparison by farmer category
Pooled sample Agroecological Reference
mean [SD] or percent mean [SD] or percent mean [SD] or percent
Traditional food (TF) practices
 TF production diversity (0–11 products) 4.7 [2.5] 5.7 [2.3]*** 2.8 [1.9]
 TF consumption diversity (0–12 products) 7.5 [2.0] 8.3 [1.7]*** 5.9 [1.6]
 TF consumption frequency (annual) 221 [182] 260 [193] 144 [129]
  median (interquartile range) 164 (82–301) 209 (130–351)*** 102 (56–180)
 Wild food consumption diversity (products) 7.5 [3.1] 7.7 [3.0] 7.0 [3.2]
Sociodemographics
 Age (years) 45 [13] 46 [13] 42 [13]
 Monthly income per capita (USD) 92 [89] 87 [81] 100 [105]
  median (interquartile range) 67 (37–110) 61 (37–110) 85 (40–109)
 Time to market (minutes) 47 [36] 49 [35] 43 [38]
  median (interquartile range) 40 (30–60) 38 (30–60) 43 (20–50)
 Education completed
  None or partial primary 44% 39% 53%
  Primary or partial secondary 38% 43% 30%
  Secondary or post-secondary 18% 18% 17%
Farm production diversity (products) 39 [16] 45 [15]*** 28 [14]
Share of total calories acquired from diverse sources
 Conventional markets (0–100%) 52 [27] 44 [23]*** 69 [25]
 Harvest (0–100%) 27 [24] 32 [24]*** 17 [19]
 Social economy (0–100%) 20 [24] 23 [24] 13 [23]
median (interquartile range) 12 (0.2–31) 17 (6–34)*** 0.3 (0.0–16)
  1. For continuous variables, mean is reported with standard deviation. For variables with non-parametric distributions, median and interquartile range are also reported. Frequency is reported for categorical variables. Share of total calories is based on an agroecological sample size of 60, due to missing information. Difference tested between agroecological and reference farmers with Student’s t-test, Mann-Whitney U-test or Chi-Squared test depending on variable distribution and type. *, ** and *** indicate significance at the 10, 5 and 1% levels, respectively.