Table 7 Summary of matching outcome for each of the variables used in the matching procedure.

From: A quasi-experimental study of impacts of Tanzania’s wildlife management areas on rural livelihoods and wealth

Matching variable Tfm. Before/After WMA Non-WMA Std. Mean. Diff.
Distance from major town sqrt Before 37.8 17.4 1.25
   After 39.7 37.5 0.12
Distance from major road log Before 11.6 7.0 0.44
   After 11.6 10.1 0.14
Distance from wildlife corridor sqrt Before 53.6 105.5 −0.95
   After 50.5 63.6 -0.24
Distance from protected area sqrt Before 22.8 43.5 −1.21
   After 25.3 33.5 −0.47
Total annual precipitation log Before 809.3 783.4 0.11
   After 733.3 746.8 −0.06
Elevation sqrt Before 881.8 833.8 0.11
   After 912.4 892.7 0.07
Slope log Before 2.2 3.3 −0.79
   After 1.9 2.0 −0.04
Within elephant range none Before 0.9 0.1 3.57
   After 0.9 0.9 0.00
Within lion range none Before 1.0 0.9 NA
   After 1.0 1.0 0.00
Population density log Before 34.2 1151.9 −18.54
   After 42.3 64.0 −0.27
Proportion of forest none Before 0.15 0.27 −0.47
   After 0.13 0.10 0.12
Proportion of woodland and shrubs none Before 0.42 0.35 0.19
   After 0.36 0.39 −0.08
Proportion of grassland and crops none Before 0.38 0.30 0.21
   After 0.48 0.50 −0.04
  1. The column headed ‘Tfm.’ indicates the type of transformation applied to the variable prior to running the matching procedure to reduce the level of skew (sqrt=square root; ln=natural logarithm; none=no transformation applied). Mean values of matching variables within WMA and non-WMA villages included in the study, before and after matching are shown in the columns ‘WMA’ and ‘Non-WMA’. The standardised mean difference in the variable in WMA and non-WMA villages before and after matching is shown in the column ‘Std. Mean. Diff.’.